,̂ ">? ^& ^l̂ ssy%«^

:.. 77UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

WASHINGTON, D.C. 20460

OFFICE OF

SUBJECT: Exposure Ana lys i s of Vertac Fac i l i t y - J a c k s o n v i l l e , Arkansas

FROM: John Schaum, Environmental Engineer 'Exposure Assessment Group (RD-689)

Steven Bayard, StatisticianCarcinogen Assessment Group (RD-689)

TO: Richard Mays, Attorney (UH-527)

Mike Kilpatrick. Environmental EngineerOffice of Waste Program Enforcement (WH-527-M)

T H R U : James W. Fa1co, DirectorExposure Assessment Group (RD-689)

Per your request, we have analyzed the amount of off-site* human exposureto 2,3,7,8-TCDD** occuring as a result of the operations at Ver tac 's facil ityin Jacksonvi l le, Arkansas. Due to time constraints we decided to evaluateonly the exposure caused by eating fish caught from Bayou Meto. As explainedbelow, this exposure route appears most significant. The results of ouranalysis are summarized below and the details of the exposure analysis aredescribed in the attached report.

Fish Exposure Route Appears Most Significant

TCOH has a very low vapor pressure, low solubility in water, and a strongtendency to be adsorbed by soil particles. This means that TCDD is mostreadily released from the site via transport on suspended solids in overlandrun-off or on airborne dust and less readily released via leaching into thegroundwater or direct volatilization into the air. The run-off transportroute is of particular concern since this type of release will accumulate inthe sediment of nearby waters and bioaccumulate in fish. The monitoringdata which indicate that the highest off-site concentrations of TCDD have beenfound in the fish and sediment of Bayou Meto support these arguments. As aresult, we concluded that human exposure caused by eating fish caught fromBayou Meto represents the most significant potential health threat andwarranted first consideration. However, the other exposure routes could alsobe significant, and if time/resources allow we recommend further analyses ofthem as well.

*Per our discussion with you we decided not to evaluate worker exposure.

**2,3,7,8-TCOD has been abbreviated to simply TCDD in remainder of memo.

The Exposure Analysis Supports the Monitoring Data

We calculated that the human exposure levels could be as high as 1 1 2 ng/kgof body weight/yr for fish caught near the confluence of Bayou Meto and RockyBranch to as low as 0.09 ng/kg of body weight/yr for fish caught near themouth of Bayou Meto. This calculation involved the fol lowing steps:

o The monitoring data suggest that the concentration of TCDD in thesediment of Rocky Branch averages to about 500 ppt. Using this valueas a starting point, we calculated the TCOO load to Bayou Meto, and theresulting distribution in the sediment from the confluence of RockyBranch and the Bayou Meto to the mouth of Bayou Meto.

o We calculated the concentration of TCOO in the fish by multiplying thesediment concentrations by a fish-sediment distribution coefficient.We derived this coeff icient in two ways giving values of 5.4 and 555.2,and by using both values, expressed the results as a range.

o Finally we computed the individual exposure levels according to thefollowing equation: (TCnn concentrat ion in f ish x average consumptionrate)/average hodyweight. In keeping with current Agency practices weused 5.2 Ib/yr (2.37 kg/yr) as the national average fish consumptionrate and 154.2 Ib (70 kg) as the average body weight.

AH of the exposure estimates and corresponding monitoring data aresurnnarized in Table 1. Table 1 shows that the monitoring data generallyagree with the exposure predictions. We believe that this gives credibilityto both the monitoring data and the exposure estimates.

The only significant discrepancy between the values based on calculationsand those based on monitoring occur at river mile 132 (the mouth of RockyBranch). The calculated levels assume that the sediment exits the Branch at500 ppt TCDD and is immediately reduced to 5.25-5.96 ppt TCDO due to dilutionfrom the sediment in the Bayou from upstream of the confluence. Themonitoring indicates, however, that the TCDD levels remain at 500 ppt justoutside the mouth of Rocky Branch. Apparently, this is due to the lack ofmixing between the Branch and Bayou sediments at the sampling point.

The Individual Cancer Risk Appears High Over the Length of the River

Rased on the linearized multistage model, the EPA Carcinogen AssessmentGroup has calculated the following upper-limit risk estimate associated withingestion of TCDD:

P(d) ' l-exp(-4.25 x IQSd)

where d is exposure i n un i t s of mg/kg/day assumed cont inuous for 70 years.

Estimates of l i fetime individual r isk, shown in Table 2 vary by distancefrom the pollutant source, but remain high over the entire length of theriver. In fact, the risk is greater than 1 per 10,000 at the mouth of the

river which is approximately 130 miles fron the Vertac facil ity. Estimatedrisks have been calculated basen on both modeled exposure and actual f ishmonitoring deta where they exist. These results appear consistent since therisks estimated from the monitoring data fall within the range of r isksestimated from the modeled exposure data. Consumption considerably higherthan 5.2 1b/yr of these contaminated f ish would lead to upper limit individualcancer risk on the order of 10-2 to 10-1. Although this individual riskappears high, it must be evaluated in conjunction with the size of the exposedpopulation (discussed below).

Me Cannot Precisely Estimate the Number of People Exposed

Since fishing on Bayou Meto is currently banned, exposure via the fishroute should not be occuring. However, Figure 1 provides a means ofestimating the number of people who could be exposed under normalcircumstances. The exposure estimates were obtained by dividing the size ofthe fish catch from Bayou Meto by the individual consumption rate. Since,both of these parameters are difficult to assess accurately, we decided to useFigure 1 to show how thp estimated exposed population s ize will vary accordingto assumptions made regarding catch size and consumption rate. In evaluatingthe data in Figure 1 , the reader should consider the following:

o The national average for consumption of non-marine fish is ^.2 1b/yr.However, surveys have shown that individual total f ish consumptionrates vary as much as three-fold between the average and 95thpercentile. These statistics represent an individual's total f ishconsumption which is likely to involve f ish caught from more than onesource. Thus, an individuals consumption of f ish from Bayou Meto onlyis probably less than that suggested by the statistics for totalconsumption.

o Precise estimates of fish catch size require the acquisition of sitespecific data which was beyond the scope of this project. However, weobtained an estimate of the mean catch rate from the U.S. Fish andWildl i fe Service. This estimate (62 Ib/acre/yr) was based on ananalysis of a Bayou Meto offstream reservoir which may differ from theBayou itself. The total catch was estimated by multiplying this figureby an estimate of the surface area of the river.

o Since the fish are likely to be consumed locally, the number of peopleexposed is probably less than the local population size (about476,000).

This Study Improves Our Understanding of the Situation, but Much UncertaintyRemains

One important benefit of this analysis is that it provides exposureestiiates in situations where the TCOO leve ls in f ish or sediment were be lowdetectable limits. Such exposures can be significant in cases involving TCODdue to its extreme toxicity. Secondly, this study provides a means ofbounding the possible range of exposures. The limited amount of data are not

sufficient for producing a statistically valid estimate of the error range.Thus, the calculated exposures provide a range which otherwise could not havebeen calculated.

It must be emphasized that these estimates are based on a very limitedamount of data. As the range of exposure indicates, the analysis must heviewed as a preliminary estimate. A far greater amount of data would herequired to precisely define the magnitude and extent of TCDD exposure in theBayou Meto area.

He Recommend Further Investigation

The fish-sediment distribution coeff icient (Kps) contributes most to theuncertainty in the calculated exposure levels. Kps could be calculated moreaccurately with definitive data on the organic carbon content in the BayouMeto sediment or with additional data on TCDD levels in fish. Since the Kpsis the most sensitive parameter in these calculations, we recommend that anyfuture monitoring efforts give first priority to collecting the above data.

The uncertainty in the est imates of f ish catch s ize and f ish consumptionrates a lso contributes to the uncertainty in the exposure calculations. Theaccuracy of these parameters could be improved by conducting local surveys.

Finally, we also recommend further investigation of several areas whichwere beyond the scope of this study:

o Surface run-off could result in the deposition of TCnn on residentialproperties located downgradient of the site. He recommend furthermonitoring of these areas, particularly after periods of heavy rainfall.

o TCnP releases could also occur via leaching into groundwater ortransport with dust into the air. We recommend more thoroughassessments of the exposure which may occur via these routes.

o Releases of harmful substances other than TCDD may also be occurring.For example, the groundwater monitoring indicates that phenols may beleaking from the site. We recommend further investigation of thehazards which these substances may pose.

Attachment

TABLE 1. TCDD CONCENTRATION IN SEDIMENT, TCDO CONCENTRATION IN FISH,AND HUMAN EXPOSURE BY RIVER MILE

River Cs (ppt)Mile Based on Based on

Calculations Monitoring

(2)CF (PPt) Exposure (ng/kg/yr)

Based on Based on Based on Based onCalculations Monitoring Calculations^) Monitoring(^)

132100

75543416

D

5.25-5.963.06-3.482.61-2.970.83-0.940.71-0.800.62-0.700.49-0.56

500<70<85/on\ou<30<20

28.35-330916.52-193214.09-16494.48-5223.83-4443.35-3892.65-311

300112

30<30<25<25

0.95-1120.56-650.48-560.15-180.13-150.11-130.09-10

10.13.791.0

• . " ;

;•-".Notes:

1. Cs = concentration of TCnn in sediment 0== TC"n load/total sediment loadLow values assume trappinp eff iciency = 0.71High va lues assume trapping efficiency = 0.41

2. Cp •=• concentration of TCno in fish= Cs x Kps where Kps = sediment fish distribution coefficient

Low values assume Kps = 5.4High values assume Kps = 555.2

3. Calculated Exposure = (Cs x Kps x consumption rate)/70 kg bodyweightConsuinption rate assumed = 5.2 Ib/yr (2.37 kg/yr)Low values assume Kps = 5.4 and use low Cs valuesHigh values assume Kps = 555.2 and use high Cs values

4. Exposure based on monitoring = (Cp x consumption rate)/70 kg bodyweightConsumption rate assumed = 5.2 Ib/yr (2.37 kg/yr)

5. Less than values equal detection limits.

6. All Cc and Cp monitoring data was gathered by Arkansas Department of PollutionControl and Ecology in 1981. The Cp data were derived from catf ish samples.

TABLE 2. HUMAN EXPOSURE AMD CANCER RISK BY R I V E K MILE

River ExposureMile Rased on

CalculationsO)

(ng/kg/yr)Rased on

Momtoring(7)

Cancer Probabi1 i ty (3)Based on I-iased onCalculations Monitoring

U2inn

75543416

n

O.q6-l12O.'i6-650.48-560.15-180.13-150.11-13n - n Q - i n

10.13.711.0

0.0012 -o.ooofi5 -0.00056 -0.00017 -0.00015 -n.00013 -n - n n m n -

0.17?0.0730.063O.o?l0.0170.015n . n i ?

".010.00140.0012

r-

Notes:

1. Calculated Exposure = (Cs x Kps x consumption rate)/70 kg bodyweightwhere Cs = concentration of TCOD in sediment, and 0Kps = sediment f ish distributuion coefficientConsumption rate assumed = 5.2 1b / y r (2 .37 kg/yr)Low values assume Kps = 5 . 4High values assume Kps ^ 555.2

2. Exposures based on monitoring = (Cp x consumption rate)/70 kg bodyweightwhere C(: = measured concentration of TCDD in fish (only available at river miles 132 ,100 and 75)Consumption rate assumed = 5.2 1b/y r (2.37 kg/yr)

3. Cancer Probability = 1 - e t -1 .25 x 105 x exposure)where exposure is in units of mg/kg/day

Figure 1

lAmber of People Exposed vs. Consumption Rate and Fish Catch Rate

10 20Consmption Rate (ib/yr)

0

~1. Plot A is based on a fish catch rate of 90 Ib/acre/yr which is 150t of theestimated mean.

2. Plot B is based on a fish catch rate of 62 Ib/acre which is the estimated mean.3. Plot C is based on a fish catch rate of 30 Ib/acre which is SOt of the

estimated mean.4. Plot D is based on a fish-catch rate of 6 Ib/acre/yr which is 104 of the

estimated mean.5. These plots were confuted using the equation: Exposed Population =

[Fish Catch Rate x Stream Area)/Consumption Rate.

EXPOSURE ANALYSIS OF VERTAC FACILITY

February 26, 1981

By John SchaumJames Fa1co

Exposure Assessment Gro.up

Office of Health and Environmental Assessment

Office of Research and Development

U.S. Environmental Protection Agency

TABLE OF CONTENTS

1.0 INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Overview of Problem. . . . . . . . . . . . . . . . . . . . . . . . 21.2 Data Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2.0 SUMMARY OF CONCLUSIONS AND RECOMMENDATIONS. . . . . . . . . . . . . . . 4

3.0 POSSIBLE ROUTES OF EXPOSURE . . . . . . . . . . . . . . . . . . . . . . 1 1

4.0 LEVELS OF EXPOSURE. . . . . . . . . . . . . . . . . . . . . . . . . . . 1 3

4.1 Calculation of TCDD Load to Bayou. . . . . . . . . . . . . . . . . 1 34.2 Calculation of TCOO Concentration in Fish

and Resulting Exposure Levels. . . . . . . . . . . . . . . . . . .15

5.0 SIZE OF EXPOSED POPULATION. . . . . . . . . . . . . . . . . . . . . . .20

fi.n APPENDICES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24

A. Calculation of Sediment Loads . . . . . . . . . . . . . . . . . . .25B. Calculation of Fish - Sediment Distribution Coefficient . . . . . .30C. Calculation of Pond Trapping Efficiency,. . . . . . . . . . . . . .33

7.0 REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37

1.0 INTRODUCTION

The objective of this report is to analyze the amount of off-site hunan

exposure to 2,3,7,8-TCDD* occurring as a result of the operations at Ver tac 's

pesticide manufacturing plant in Jacksonvil le, Arkansas. Due to time

constraints, only the exposure caused by eating fish is covered in depth. We

prepared the report in response to a request from the Office of Waste Program

Enforcement who is considering enforcement actions against the company.

* ? , 3 , 7 , R - T C n n has been abbreviated to simply TCDD in remainder of memo

1.1 O V E R V I E W OF THE P R O B L E M

2,4,5-T was manufactured at this site by a variety of different companies

from 1948-1979. During the late 60 's Agent Urange was also produced for the

Air Force. These herbicides contained TCDD as an impurity. Waste disposal,

cooling water handling and other plant operations have apparently resulted in

the release of TCDD. TCOD has been detected in the disposal areas and soi ls

surrounding the plant. The highest off-site concentrations have been found in

a small stream (Rocky Branch) located immediately adjacent to the property.

About two miles fron Vertac, Rocky Branch flows into the Bayou Meto which is a

tributary of the Missouri River. TCDD has also been detected in fish caught

fron the Rayou Meto (EPA, Case Fi le).

Arkansas of f ic ia ls have currently banned f ishing on the Bayou Meto, so

human exposure via the fish consumption route sho'uld not be occurring.

However, we believe that this route would represent the most serious human

health threat under normal circumstances and decided to focus our analysis on

exposure via f ish consumption.

This assessment involved determining the mechanism by which TCDO was

transported to the stream, computing the levels of TCDD at various points

downstream from the facility, estimating the amount which would bioaccumulate

in fish, and finally estimating human consumption rates and the resulting

exposure levels. Additionally, we estimated the number of people exposed.

1.2 DATA SOURCES

We obtained the monitoring data used in this report from the Case File

maintained by EPA 's Office of Waste Program Enforcement. Although access

to this file is restricted, interested people should contact Mike Kilpatrick

(EPA, Off ice of Waste Program Enforcement, 202-3^2 -3111) . The monitoring and

analyses were actually conducted by the Arkansas Department of Pollution

Control and Ecology.

The watershed area and sediment load data were derived frcm reports by the

U.S. Geological Survey and Arkansas Soil and Water Conservation Commission

(see reference list for further details). The river flow rate data were

obtained from EPA 's Hater Quality Control Information System (STORET) . This a

computerized file maintained by EPA and updated on a weekly basis. Anyone

wishing to use this data should contact the STORET User Assistance Section at

(202)426-7792.

The estimate of the mean fish catch rate is based on a computer analysis

done by Robert Jenkins of the Fish and Wildl i fe Service, Fayettevil le,

Arkansas. For more information about this analysis contact John Schaum (EPA,

Office of Research and Development, 202-382-2931).

The above discussion covers the principal sources of data used in this

study. Several other minor pieces of data were obtained from var ious sources

as cited. Anyone who desires should be able to consult these sources from the

information provided in the list of references.

2.0 SUMMARY OF CONCLUSIONS AND R E C O M M E M D A T I O H S

This report analyzes the amount of off-site exposure to ?,3,7,S-TCrin

occurring as a result of the operations at the Ver tac 's pesticide

manufacturing plant in Jacksonvil le, Arkansas.

Fish Exposure Route Appears Most S i g n i f i c a n t

TCDD has a very low vapor pressure, low so lub i l i t y in water , and a strong ;

^ ''•.

tendency to be adsorbed by soil par t ic les (Esposi to , et a1. 1980, pp. 5, 241 :".

and 247) . This means ttiat TCDD is most readily released from the site v i a ,̂

transport on suspended sol ids in overland run-off or on airborne dust and less °

readi ly released v i a l e a c h i n g in to the g r o u n d w a t e r or d i rec t v o l a t i l i z a t i o n

into the a i r . The run-off transport route is of p a r t i c u l a r concern since th i s

type of release w i l l accumulate in the sediment of nearby waters and

bioaccuitulate in f i s h . The moni to r ing data which indicate that the highest

off-s i te concentrations of TCnR have been found in the f i s h and sediment of

Bayou Meto support these arguments. As a result , we concluded that human

exposure caused by ea t ing f i s h caught from Bayou Meto represents the most

s i g n i f i c a n t potential hea l th threat and warranted f i r s t cons idera t ion .

However, the other exposure routes could also be s i g n i f i c a n t , and if

time/resources a l l ow we recommend further analyses of them as wel l .

The Exposure Analysis Supports the Moni tor ing Data

We calculated that the human exposure levels could be as h igh as 112 n g / k g

of body weight/yr for f ish caught near the confluence of Bayou Meto and Rocky

Branch to as low as 0.09 ng/kg of body weight/yr for f i s h caught near the

mouth of Bayou Meto. This c a l cu l a t i on involved the f o l l o w i n g steps:

o The monitoring data suggest that the concentration of TCU3 in the

sediment of Rocky Branch averages to about 500 ppt. Using this value

as a starting point, we calculated the TCDD load to Bayou Meto and the

resulting distribution in the sediment from the confluence of Rocky

Branch and the Bayou Meto to the mouth of Bayou Meto.

o We calculated the concentration of TCDD in the fish by multiplying the

sediment concentrations by a fish-sediment distribution coefficient.

We derived this coefficient In two ways giving values of 5.4 and 555.2,

and by using both values, expressed the results as a range.

o Finally we computed the individual exposure,levels according to the

following equation: (TCDO concentration in fish x average consumption

rate)/average bodyweight. In keeping with current Agency practices we

used 5.2 Ib/yr (2.37 kg/yr) as the national average fish consumption

rate and 154.2 1 b (70 kg) as the average body weight.

AH of the exposure estimates and corresponding monitoring data are

summarized in Table 1. Table 1 shows that the monitoring data generally

agree with the exposure predictions. We believe that this gives credibility

to both the monitoring data and the exposure estimates.

The only significant discrepancy between the values based on calculations

and those based on monitoring occur at river mile 132 (the mouth of Rocky

Branch). The calculated levels assume that the sediment exits the Branch at

500 ppt TCDD and is immediately reduced to 5.25-5.96 ppt TCDD due to dilution

TABLE 1. TCDO COflCENTRATION IN SEDIMENT, TCDD CONCENTRATION IN FISH,AND HUMAN EXPOSURE BY RIVER MILE

River Cg (ppt)Mile Rased on RaserI on

Calculat ions Monitoring

CF (PPt)^Rased on Based onCalculations Monitoring

Exposure (ng/kg/yr)Rased on Based onCalculations^) Monitoring (4)

132100

7554^d«3^160

5.25-5.963.06-3.482.61-2.970.83-0.940.71-0.800.62-0.700.49-0.56

500<70<fi5/on\ulJ

<30<20

28.35-330916.52-193214.09-16494.48-522•a R ^ m.J•OJ—L t l t t+'3 '31; '3QQJ» JO-Jo-7

2.65-311

300112

30<30<25<25

0.96-1120.56-650.48-560.15-180.13-150.11-130.09-10

10.13.791.0

^.~----- CM

'r.Notes:

1. Cs = concentration of TCDD in sediment= TCDD load/total sediment load

Low values assume trapping efficiency = 0.71High values assume trapping efficiency » 0.41

C

2. CF = concentration of TCDD in f ish= Cc x Kps where Kps e sediment f ish distribution, coefficient

Low values assume Kps = 5.4High values assume Kps s 555.2

3. Calculated Exposure = (Cg x Kpg x consumption rate)/70 kg bodyweightConsumption rate assumed = 5.2 Ib/yr (2 .37 kg/yr)Low values assume Kpg = 5.4 and use low C; valuesHigh values assume Kp; = 555.2 and use high Cs values

4. Exposure based on monitoring = (Cp x consumption rate)/70 kg bodyweightConsumption rate assumed = 5.2 1b/yr (2.37 kg/yr)

5. Less than values equal detection limits.

6. A 1 I C$ and Cp monitoring data was gathered by Arkansas Department of PollutionControl and Ecology in 1981. The Cp data were derived from catf ish samples.

from the sediment in the Bayou from upstream of the confluence. The

monitoring indicates, however, that the TCDD levels remain at 500 ppt just

outside the mouth of Rocky Branch. Apparently, this is due to the lack of

mixing between the Branch and Bayou sediments at the sampling point.

We Cannot Precisely Estimate the Number of People Exposed

.Since fishing on Bayou Meto is currently banned, exposure via the fish

route should not be occurring. However, Figure 1 provides a means of

estimating the number of people who could be exposed under normal

circumstances. The exposure estimates were obtained by dividing the size of

the fish catch from Bayou Meto by the individual consumption rate. Since,

both of these parameters are di f f icult to assess accurately, we decided to use

Figure 1 to show how the estimated exposed population size will vary according

to assumptions made regarding catch size and consumption rate. In evaluating

the data in Figure 1 , the reader should consider the fol lowing:

o The national average for consumption of non-marine fish is 5.2 Ib/yr

(2.37 kg/yr) (Stephan 19SO). However, surveys have shown that

individual total fish consumption rates vary as much as three-fold

between the average and 95th percentile (SRI 1980). These stat is t ics

represent an individual's total f ish consumption which is likely to

involve f ish caught from more than one source. Thus, an individual's

consumption of fish from Bayou Meto only is probably less than that

suggested by the statistics for total consumption.

o Precise estimates of fish catch size require the acquisition of site

specific data which was beyond the scope of this project. However, the

Number

of Pe

ople Exp

osed/Y

ear

e»B>

nwrt

rt

r+

S.K-

S.K'

S'&K

-

§

§§

§

B> »

?! s- s & 3 i? S, o> »-• ^ 8 .̂ ^

P'? ss-s

.SS &S

.S3!?

?0

0

Hi i

-h as ^ ^Hi

^ S & 5 rt <t 0 Ml

§ •-;

^ s ^̂§> I •a i' !̂? ? ^̂ /

1-1- p

- n

^ «, «

b-

" gg

. K-

»-' <p

(»0

<»

i-h (n g-i'

°B

P

W

,

£ ?

mean catch rate has been estimated at 62 Ib/acre/yr (Personal

communication, John Schaum to Bob Jenkins, U.S. Fish and Wi ld l i fe

Service, Fayettevi He, Arkansas February 24, 19R2) . This estimate was

based on an analysis of a Bayou Meto offstrea.n reservoir which may

differ from the Bayou itself. The total catch was estimated by

Tiuitiplying this figure by an estimate of the surface area of the

river.

o Since the fish are likely to be consumed locally, the number of people

exposed is likely to he less than the local population size (about

476,000 - see Section 5 ) .

This Study Improves Our Understanding of the Situation, but Much Uncertainty

Remains

One important benefit of this analysis is that it provides exposure

estimates in situations where the TCDD levels in fish or sediment were below

detectable limits. Such exposures can be significant in cases involving TCDO

due to its extreme toxicity. Secondly, this study provides a means of

bounding the possible range of exposures. The limited amount of data are not

sufficient for producing a statistically valid estimate of the error range.

Thus, the calculated exposures provide a range which otherwise could not have

been calculated.

It must be emphasized that these estimates are based on a very limited

amount of data. As the range of exposure indicate-s, the analysis must be

viewed as a preliminary estimate. A far greater amount of data would be

required to precisely define the magnitude and extent of TCPD exposure in the

Bayou Meto area.

He Recommend Further Investigation

The fish-sediment distribution coefficient (Kps) contributes most to the

uncertainty in the calculated exposure levels. Kps cuu1d be calculated more

accurately with definit ive data on the organic carbon content in the Bayou

Meto sediment or with additional data on TCDD levels in fish. Since the Kf^

is the most sensitive parameter in these calculations, we recommend that any

future monitoring efforts give first priority to collecting the above data.

The uncertainty in the estimates of fish catch size and fish consumption

rates also contributes to the uncertainty in the exposure calculations. The

accuracy of these parameters could be improved by conducting local surveys.

Finally, we also recommend further investigation of several areas which

were beyond the scope of this study:

o Surface run-off could result in the deposition of TCDD on residential

properties located downgradient of the site. We recommend further

monitoring of these areas, particularly after periods of heavy

rainfall.

o TCDD releases could also occur v ia leaching into groundwater or

transport with dust into the air. We recommend more thorough

assessments of the exposure which may occur via the routes.

o Releases of harmful substances other than TCDD may also be occurring.

For example, the groundwater monitoring indicates that phenols may be

leaking from the site. We recommend further investigation of the

hazards which these substances may pose.

10

3.0 POSSIBLE ROUTES OF EXPOSURE

According to the monitoring data, 2,3,7,8-TCDD is present in the disposal

areas at Vertac and has been detected in the surface soi ls surrounding the

plant (EPA, Case File). Potential routes of releases from the site are as

fo l lows:

o Transport by run-off - Run-off can carry significant amounts of soil as

suspended solids. This is the most likely route which caused the

sediment contamination detected in Rocky Branch. Although the remedial

actions should have reduced such releases, this probably represents the

most significant way in which TCPO could continue to be released from

the site. The significance of this exposure route is increased hy the

fact that several investigators have shown'that TCnn is very stable in

aquatic sediment and that it is likely to bioaccynulate in aquatic

organisms (Perwak et a 1 . 1980, p. 48). Accordingly, we chose this

exposure route to analyze in depth.

Run-off could also carry TCDD contaminated soil to residential

properties located downgradient of the site. We did not analyze this

possibility due to difficulties in calculation methods and lack of

time. However, we recommend further monitoring of these areas,

particularly after periods of heavy rainfall.

o Leaching into groundwater - TCDD has a very low solubility in water

(0.2 ppb) and a high octanol-water partition coefficient (1.38 x 107)

which means that TCOD will leach at very t o w levels and be readily

11

adsorbed by the soil (Versar 1978). Additionally, we understand that

no one now uses the groundwater for drinking purposes within a couple

of miles fron the plant. Accordingly, since it is uncertain whether

any significant human exposure occurs via the groundwater route at

present, we decided not to evaluate it further. However, the

groundwater route nay represent a long-tenn threat and should be

further analyzed. We also recommend further analysis of exposure to

phenols, which are much more mobile in the subsurface than TCOD as the

monitoring data indicates, and nay represent a more serious current

threat.

o Volat i l izat ion and transport with dust into the air - TCDD has a very

low vapor pressure (estimated at 10-6 - in-7 mm Hg-Perwak, et

a1 . 1980, p. 57) and is, therefore, unlikely to be measurable in the

ambient air as vapor at detectable levels off the site. Since TCDO is

strongly adsorbed to soil particles and has been found in the surface

soils we would expect to find it in the dust particles. However, the

remedial actions taken to date (covering disposal areas and blow-out

area) should greatly diminish dust contamination. Accordingly, we

decided not to evaluate it further, but recommend, if possible, to

analyze it further in the future.

The remainder of this paper discusses exposure to TCDD resulting from

transport by run-off.

12

4.0 LEVELS OF EXPOSURE

4.1 CALCULATION OF TCDP LOAD TO BAYOU

The Vertac property borders Rocky Branch at a point about 2 miles from the

Bayou Meto. EPA and Arkansas state officials have analyzed the sediment in

the Rocky Branch for TCDD and found it to average approximately 500 ppt near

the mouth of the Branch. Using this as a starting point, this paper attempts

to model how the TCDD is transported through the environment.

The Nonpoint Source Pollution Assessment Summaries for Arkansas River

Basin estimate the sediment yield rate for the Upper Bayou Meto Water Shed

(No. 1 1 0 1 ) at 8?,000 tons/yr which averages to 0.63 tons/yr/acre (Arkansas

Soil and Water Conservation Commission, 1979). Assuming that this rate is

constant throughout the water shed, the sediment load from Rocky Branch can be

calculated by multiplying the Rocky Branch drainage area by 0.63. This

assumption should be val id since the factors affecting erosion (i.e.,

rainfall, slope steepness, cover, etc.) appear roughly equal over the area.

The size of the Rocky Branch drainage area was measured directly from a

USGS contour map and found equal to 502 acres above the cooling pond and 721

acres below. We calculated that the pond has a trapping efficiency of 0.46 to

0.71 as explained in Appendix C. Using these values, the sediment load from

Rocky Branch to Bayou Meto can be calculated as follows:

Sediment load = [amount of sediment entering pond x (1-trapping

efficiency)] + amount of sediment entering Branch below pond

Max load » 502 acre (0.63 tons/acre/yr)(1-0.46) + 721 acres

(0.63 tons/acre/yr) « 625 tons/yr

Min load ' 502 acres (0.63 tons/acre/yr)(1-0.71) + 721 acres

(0.53 tons/acres/yr) = 546 tons/yr

13

According to the most recent monitoring data, SCO ppt of TCon is present

in the sediment near the mouth of Rocky Branch. Using this value, we

calculated the annual TCDD load to the Bayou as follows:

Max TCDD load to Bayou = 625 tons/yr (500 x 10 -12 )

= 3.1 x 10-7 tons/yr

Min TCDD load to Bayou = 546 tons/yr (500 x 10-12)

' 2.73 x 10-7 tons/yr

0On the basis of a TCDD mass balance, we computed the concentration of TCDD ,̂

in the sediment (Cg) at a number of points downstream of Rocky Branch: c

TCDD load to Bayou = Cg (sediment load from upstream of point)

Max Cs = Max TCDD Load = 3.1 x 10-7Sediment load Sediment load

Min Cg = Min TCDD Load = 2.73 x 10-7

Sediment load sediment load

The sediment loads were derived from a combination of USGS data and the

Nonpoint Source Pollution Assessment Summaries for Arkansas River Basin as

explained in Appendix A. The sediment loads and TCDD concentrations in the

sediment (Cs) are summarized below for 7 different points of the Bayou:

14

Loi

Rt

Rt

Rt

Rt

Rt

Rt

Moi

sat ion

161

31

13

79

152

11

i th

R ive r M i l e

13?

ion75

54

34

16

0

Sediment Load( tons /yr)

52013

89080

1044411

3?q?34

385554

441874

551444

GS ((Max

5. PR

3.48

2.97

0.94

0.80

0.70

0.56

a p t )M i n

5.25

3.06

7.61

0.83

0.71

0.62

0.49

4.2 CALCULATION OF TCOO CONCENTRATION IN FISH AND RESULTING EXPOSURE LEVELS

Assuming that the levels of TCDD in the sediment reach an equilibrium with

the levels of TCDD in fish, we can calculate a fish sediment distribution

coefficient. This situation is most likely to occur in nearby Lake Oupree*

where TCDD has also been detected. Equilibrium conditions are more likely to

he reached in the lake, since it would probably have a more stable fish

population than the Bayou. A few fish samples fron Lake Dupree have been

collected over the past few years. The most recent samples show 810 ppt in

the fish and 150 ppt in the sediment. Using these values a distribution

coefficient between sediment and f ish TCDD concentrations can be calculated as

fol lows:

*Lake Dupree is located about 1 mile from Vertac and close to RockyBranch. Although, the Lake and Branch are not normally connected, the lowterrain suggests that they would connect during floods which is probably howit became contaminated with TCOD.

15

KF-S s CF/C-S

= 810 ppt150 ppt

= 5.4

where Cp = concentration of TCDD in fish and

Cs = concentration of TCnn in sediment.

The fish found to contain RID ppt from Lake Dupree was a bass (Personal

communication from John Schaum to Dick Cassat, Department of Pollution Control

and Ecology, Little Rock, Arkansas, February 24, 19S2). Fish which feed

primarily on the bottom such as catfish, have been found to accumulate more

TCDD than f ish which do not, such as bass (Harles, R. and Lewis R. 1980).

Since catfish is a common species in Bayou Meto, the Kp-s derived above on

the basis of a bass, 1 s probably lower than the KF.S representative of the

entire population. We also calculated a value for KF.S froni the biota-water

and organic matter-water distribution coefficients as reported in the

literature (as explained in Appendix B). This method predicts that the Kp.g

could be as high as 555.2.

The concentration of TCDP in fish (Cp) is now easily calculated as

fol1ows:

CF ' Cs Kp.s

Min Cp = 5.4 Cs

Max CF ' 555.2 Cs

In order to complete the exposure level analyses, it is necessary to

estimate the amount of fish a person may eat from the Bayou Meto. Ideally,

such estimates are based on local surveys, but unfortunately, such surveys

16

have not been conducted. Consequently, we decided to base our estimates on

national statistics for fish consumption. It is important to understand that

this approach introduces a number of uncertainties and requires careful

interpretation of the results.

The fish consumption rate statistics used in this report, were derived

from the following studies:

o Puring 1973-74, NPD Research Inc. conducted a survey and estimated that ^ _

the average per capita consumption rate in the U.S. is IS.6 g/day ( 1 5 ;

Ib/yr) and 54.7 g/day (44 Ib/yr) for the 95th percentile. In iqp,R, SRI oC-

International using NPD's survey data recalculated the average as 14 .3 ^

g/day ( 1 1 . 5 Ib/yr) and the 95th percentile as 4 1 . 7 g/day (33.5 Ib/yr).

These figures cover a 1 1 types of marine, estuarine, and freshwater f ish

and shellfish (SRI, 1980).

o Using the SRI data, EPA has derived a national average of 6.5 g/day

(5.2 Ib/yr) for the consumption of freshwater fish only (Stephan, C.E.

1980) It is now Agency policy to use this estimate in setting ambient

water quality criteria.

The NPD/SRI estimates are probably not applicable in this situation since

they include the marine fish portion of a person's diet. Therefore, EPA's fi.l

g/day (5.2 Ib/yr) estimate (based on freshwater fish only) is probably more

representative of individual consumption of fish taken from Bayou Meto.

However, the possibility that people may eat freshwater fish from other

sources, suggests that their actual consumption of contaminated f ish would

17

he lower than the average consumption rate. On the other hand, the Npn survey

shows that some segments of the population eat fish at a rate three times

greater than the average. In keeping with current Agency practices, we

decided to compute the exposure levels on the basis of the average freshwater

fish consumption rate (i.e., 6.5 g/day). The exposure level was calculated

using the following formula:

Annual Exposure = (Cg x Kpc; x Consumption Rate)/70 kg bodyweight

The exposure levels at a number of locations on the river are summarized in

Table 1.

18

TABLE 1. TCDD CONCENTRATION IN SEDIMENT, TCDD CONCENTRATION IN FISH,AND HUMAN EXPOSURE BY RIVER MILE

(1) (2 )River Cs (ppt) CF (ppt ) Exposure ( n g / k g / y r )M i l e Based on Based on Based on Based on Based on Based on

Calcu la t ions Moni tor ing Ca lcu la t ions M o n i t o r i n g C a 1 c u l a t i o n s ( 3 ) M o n i t o r i n g ( 4 )

132inn75513416

nTloft

1.

5.25-5.963.06-3.48?.61-2.97n.83-0.140.71-0.po0.62-0.700.19-0.56

•s:

Cg = concentration= TCDD load/tot

Low va lues assumeH i g h v a l u e s assume

500<70<85/finsou<3n<20

of TCDOa1 sedimet r a p p i n g

t r a p p i n g

28.35-330916.52-193214.09-1649a aa,!;??*+.tO"* 3L£

3.P3-4443.35-3R92.65-311

in sedimentnt loade f f i c i ency = 0.71

e f f i c i e n c y = 0.41

30011230

<30<25<25

0.96-1120.56-650.48-560.15-1R0.13-150.11-130.09-10

10.13.79l.n

1 '"'i1 ̂ -.0

2. Cf » concentration of TCDD i n f i s h= Cc x Kps where Kps = sediinent f ish d i s t r ibu t ion" coeff ic ient

Low vaTues assume Kps = 5.4High values assume Kps = 555.2

3. Calculated Exposure = (Cg x Kps x consumption rate)/70 kg bodyweightConsumption rate assumed = 5.2 Ib/yr (2.37 kg/yr)Low va lues assume Kps = 5.4 and use low Cs va luesH i g h va lues assume Kps = 555.2 and use h i g h Cg va lues

4. Exposure based on mon i to r ing = (Cp x consumption rate)/70 kg bodyweightConsumption rate assumed = 5.2 Ib/yr (2.37 kg/yr)

5. Less than values equal detection l imi t s .

6. AH Cc and Cp moni tor ing data was gathered by Arkansas Department of P o l l u t i o nControl and Ecology in 19R1. The Cp data were derived from ca t f i sh samples.

19

5.0 SIZE OF EXPOSED POPULATION

Since f i sh ing on Bayou Meto is current ly banned , exposure v ia the f i s h

route should not be occurr ing. However, rough estimates of the number of

people exposed under normal circumstances can be obtained by d i v i d i n g the size

of the f ish catch from Bayou Meto by the i n d i v i d u a l f i sh consumption rate.

The method used to estimate catch size is described below.

The size of a f i s h catch is most accurately estimated from surveys

conducted in the area of interest. According to the Arkansas F i sh and Game

Coinmission such surveys have not been conducted on the Bayou Meto (Personal

communicat ion from John Schaum to Barry Beaver, Arkansas Fish and Game

Conmiiss ion, Lonoke, A r k a n s a s , Noveriber 10, 19R1). Consequen t ly , the f i s h

catch estimates must be based on less precise methods.^

The mean catch per un i t area has been estimated at 62 Ib/acre/yr (personal

communica t ion from John Schaum to Robert J enk ins , U.S. Fish and W i l d l i f e

Service, Fayettevil le, A r k a n s a s , February 24, 1982). This estimate inc ludes

the cont r ibut ions from both sport and commercial f i sh ing . It was based on an

ana lys i s of a Bayou Meto offstream reservoir wh ich may d i f f e r from the Bayou

i tse l f .

The other parameter which must be estimated in order to determine the size

of the f ish catch is the surface area of the river. Bayou Meto is

approximately 300 ft wide at its mouth and 150 ft wide i n its mid section

(Personal communication from John Schaum to John Giese, Arkansas Department of

Pol lut ion Control and Ecology, Lit t le Rock, Arkansas , November 3, 1981).

Assuming that the average width is 150 ft , the surface area of the river can

be computed as shown:

Area = River length x average width

» (144 m i ) ( 1 5 0 f t ) (640 ac rea /mi2) / (528D f t / m i )

20

== ?600 acres - after founding.

Finally, the s ize of the total fish catch can he estimated as shown helow:

Total catch = (62 1b/acre/yr)(2f i00 acres)

= 161,?nn 1b/yr

Since the estimates of both fish consumption and catch size are difficult

to assess accurately we decided to use Figure 1 to show how the estimated

exposed population size will vary according to assumptions made regarding

these parameters. In evaluating the data in Figure 1 , the reader should

consider the following:

o The national average for consumption of non-marine fish is 5.2 1b/yr.

However, surveys have shown that individual total f ish consumption

rates vary as much as three-fold between the average and 95th

percentile (SRI, 1980). These statistics represent an individual 's

total f ish consumption which is likely to involve fish caught from more

than one source. Thus, an individual 's consumption of f ish from Rayou

Meto only is probably less than that suggested by the stat ist ics for

total consumption.

o Althouth the catch rate has been estimated to average 62 Ib/acre/yr, it

could vary due to differences between the Bayou and reservoir upon

which the estimate is based.

o Since most fish from Bayou Meto are likely to be consumed locally, the

number of people exposed is likely to be less than the size of the

local population. The total number of people inhabiting the four

21

Figure 1

htiinber. of People Exposed vs. Consunption Rate and Fish Catch Rate

Consunption. Bate Clb/yr)0" " I . Plot A is based on a fish catch rate of 90 Ib/acre/yr which is 1504 of the

estimated mean.2. Plot B is based on a fish catch rate of 62 Ib/acre which is the estimated mean.3. Plot C is based on a fish catch rate of 30 Ib/acre which is 501 of theestimated mean.4. Plot D is based on a fish catch rate of 6 Ib/acre/yr which is lOt of theestimated mean.5. These plots were computed using the equation: Exposed Population »

(Fish Catch Rate x Stream Area)/Consmiptiop Rate.

counties that are at least partially drained by Bayou rieto is

approximately 47fi,000 (Personal communication from John Schaum to

Johanna Barten, U.^. Census Bureau, Washington, ri.c. on February 1 1 ,

19S?) as detailed below;

Pulaski 310,613

Lonoke 34.51R

Praire 10 ,140

Jefferson 90,718475,990

23

6 . 0 APPENDICES

A - CALCULATION OF SEDIMENT LOAOSR - CALCULATION OF FISH-SEDIMENT niSTRIBDTION COEFFICIEMTC - CALCULATION OF POND TRAPPING EFFICIENCY

24

A P P E N D I X A

C A L C U L A T I O N OF S E D I M E N T LOADS

The sediment lodds used in th i s report were derived from two sources:

1) S u l l i v a n , N . N . and J.E. Terry. 1970. U.S. Geological Survey, Dra inage

Areas of Streams in Arkansas.

2) Arkansas Soil and Water Conservation Commiss ion . 1979. Nonpoin t . Source

P o l l u t i o n Assessment Suirxnaries for Arkansas River Bas in .

The (JSGS study a l l o w s one to accurately est imate d ra inage areas, so we

used it for th is purposs. The Arkansas study provides good erosion/sediment

yie ld da ta , so we used it to obta in these va lues . The d r a inage area data

d i f f e r s l i gh t l y between the studies. We judged that the USGS data was

probably more accurate i n t h i s respect and adjusted the A r k a n s a s numbers to

correspond to the USGS numbers (explained below).'

The Arkansas study d i v i d e s the Bayou Meto d r a i n a g e area into four water

sheds (Figure A-l) and gives size and sediment y i e ld estimates for each (Table

A-l) .

Table A-l

Water Shed water Shed Size Sediment YieldName Number (acres) (tons/yr) (tons/yr/acre)

Upper Bayou MetoBayou MetoBayou Two PrairieMi 11 Bayou

1101 130970 82000 0.631102 232761 187500 0.81103 149717 177900 1 . 1 91104 96739 82400 0.85

We decided to compute the sediment load at each point of interest on the

Bayou by m u l t i p l y i n g the per acre sediment y i e ld s by the size of the d r a i n a g e

25

It Diaqraw. o-T bo^nu Mdo Drainage Arfia

0 0 0 ^ .•: 6

area above each point. Since the per acre sediment yield is different for

each water shed, we needed to determine how much of each water shed made up

the drainage area. The first step in this determination was to obtain the

total drainage area above each point of interest on the Bayou from the USGS

data (Table A-2).

Table A-2

River Mile(mile from mouth)

Drainage Area(mi1e2) (acres)

132 129 82560100 220 14080075 250 16000054 574 36736034 684 43776016 794 5081600 998 638720

Next, we adjusted the water shed size estimates given in the Arkansas study tocorrespond to the USGS data. The USGS study estimated the total drainage areaof Bayou Meto as 638,720 acres and the Arkansas study computed it as 610,187acres. We were able to positively identify that the Arkansas estimate forWater Shed 1103 was 1323 acres less than the USGS estimate, so we adjusted theArkansas estimate from 149,717 acres to 151,040 acres. The data did not allowus to identify exactly how much of the remaining 27,210 acre discrepancy wasdistributed among sheds 1101, 1102, and 1104. Accordingly, we simplyincreased the Arkansas estimates for these sheds by the fraction of 27,210acres proportional to their size. These adjustments are summarized below(Table A - 3 ) .

27

Table A-3

Water ShedNumber

1101

1102

1103

1104

Arkansas estimateof size (acres)

130970

232761

149717

96739

Adjustment(acres)

+0.28(27210)

+0.51(27210)

+1323

+0.21(27210)

Adjustedsize (acres)

1385S9

?46638

151040

102453

Total 610187 +28533 638720

The above in format ion a l lows one to compute the amount of each water shed

c o n t r i b u t i n g to the total d ra inage area ups t ream of a p a r t i c u l a r p o i i n t on the

Bayou. Where a water shed is located ent i re ly upstream of a poin t , obviously^

its entire area is included in the drainge area. Conversely, if a water shed

is located entirely downstream of a po in t , none of it is inc luded i n the

dra inage area. In s i tuat ions where only a portion of a shed was upstream of a

point , we ca lcula ted the portion of its area contr ibut ing to the total

dra inage area by subtracting the size of any water sheds located ent i re ly

above the point from the total dra inage areas estimated from the USGS data

(Table A-2). These computations are given below:

1) Portion of Shed 1101 above River M i l e 132 (no entire sheds above th is

p o i n t ) : 82,560 - 0 - 82,560

2) Portion of Shed 1102 above R ive r M i l e 100 (only Shed 1101 above th i s

po in t ) : 140 ,800-138 ,589=2211

3) Portion of Shed 1102 above River M i l e 75 (only Shed 1101 above this

poin t ) : 160,000 - 138,589 = 21,411

4) Port ion of Shed 1102 above River M i l e 54 (Sheds 1101 and 1103 above th i s

28

poin t ) : 367,360 - 138,589 - 151,040 = 77,731

5) Portion of Shed 1102 above R i v e r M i l e 34 (Sheds 1101 and 1102 above t h i s

po in t ) : 437,760 - 138,589 - 151,040 = 148,131

6) Portion of Shed 1102 above R i v e r M i l e 16 (Sheds 1101 and 1102 above t h i s

p o i n t ) : 508,160 - 138,589 - 151,040 = 218,531

F i n a l l y , the sediment y ie ld is ca lcu la ted by m u l t i p l y i n g the yield/acre

for each shed (Table A-l) by the area of the shed inc luded i n the d ra inage

area. The y ie ld data is summarized below i n Table A-4.

D r a i n a g e Area (acres )R i v e rM i l e 1101 1102 1103

Table A-4Sediment Yield (tons/yr)

1104 1101 1102 1103 1104 Total

132

100

75

54

34

16

0

82560

138589

138589

138589

138589

138589

13R5R9

0

2211

21411

77731

148131

218531

246638

0

0

0

151040

151040

151040

151040

0

0

0

0

0

0

102453

52013

87311

87311

87311

87311

87311

87311

0

1769

17129

62185

118505

174825

197310

0

0

0

179738

179738

179738

179738

0

0

0

0

n0

87085

52013

89080

104440

329234

385554

441874

551444

29

APPENDIX B

CALCULATION OP FISH-SEUIMENT DISTRIBUTION COEFFlCI£r,T

Previously in this study, a fish-sediment distribution coefficient for

TCDD (KF-S) was calculated from the measured levels of TCUD in the fish and

sediment of Lake Dupree. Alternatively, Kp.s can be calculated on the basis

of the organic matter-water distribution coefficient (KQM-W) and biota-water

distribution coefficient (Kp.y)*:

1 ) According to Perwak et a 1 . (l^n),

^g ^M-K = d.36

KnM.w = 2.29 x inA

log KB.H ^ ^^

KB-W = 1.096 x lO5

2) By definition,

KOM-W s COM/CW and

KB-W = CB/CW

where CQ^ == concentration of TCDD in organic matter

Cy = concentration of TCDD in water

CR = concentration of TCDD in biota

*As used here, Kp.y represents a bioconcentration factor defined as theratio of contaminant concentration in the biota and in the water and isassumed equal to Kp.y (fish-water distribution coefficient). Since KB-Wis used in the above derivation of Kp.s, it should be dear that Kp.s andthe hioconcentration factor are closely related. The exact relationship isdescribed by the equation: Kg.y = Kp-s KS.K (where KS.K equals thesediment-water distribution coefficient). This relationship is deriveddirectly from fundamentals of equilibrium thermodynamics and says that Kp.yand Kp.5 are linearly related.

30

3) T h u s ,

KB-M/KOM - w - (CB/Cri)/(Con/Cri)1.096 x 105/2.29 x m4 = Kg - OH

KB-OM = 4.786

4) According to Har-iaker (1978):

KOM-M = ^'OC-W x I-7?4 . where Kpr-u = organic carbon-waterd i s t r i b u t i o n coef f i c ien t

s". Kp.oc = KR-nM/l - 7 ? 1

= 4 .7RS/1.724

= 2.776

Fi) According to the U.S. Soil Conservation Service (Personal communicationfrom John Schaum to Charles Fu1tz , U.S. Soil Conservation Service,Little Rock, Arkansas November 10, 1981) the'organic carbon in this arearanges from 4 to 0.5%.

Thus, at 4%:

Cs = 0.04 CQC. where Cs = concentration of TCDD in sediment,

CQC = concentration of TCDO in organic carbon

and, KB.S - CB/CS

assuming Kp.^ = ^F-S' then

KF.S = CB/n.04 coc= KR-OC/"-04

= 2.77R/0.04

= 69.4

if organic carbon is 0.5*, Kf.s = 2.776/0.005 = 555.2

The fish-sediment distribution coefficient derived in this study on the

basis of the monitoring data was 5.4. Thus, it is about one order of

31

magnitude lower than the 69.4 value and two orders of magnitude lower than the

555.? value. This discrepancy could be explained in several ways:

o The organic carbon content of the sediment could be much higher than

0.5 to 4% estimates we received from the U.S. Soil Conservation

Service.

o The fish and/or TCDD may not have been in place long enough to reach an

equilibrium. ; . ,

o The rate of water replacement in the Rayou may have not provided c

adequate time for reaching equilibrium conditions.

Obviously, the organic carbon content is an extremely sensitive parameter,

and the accuracy of this study could be greatly improved with a more

definitive estimate. We strongly recommend that any future monitoring

act iv i t ies give high priority to accurately determining the organic carbon

content of Bayou Meto sediment.

32

APPENDIX C

CALCULATION OF POND TRAPPING EFFICIENCY

Trapping efficiency is a function of residence time. Thus, we first

estimated the pond volume and flow rate as explained below.

The pond on Rocky Branch has a dam at one end. The maximum width and

depth is at the dam and measures 400 ft and 6 ft, respectively (Personal

communication from John Schaum to Imre Szekelyhide, Ecology and Environmental

Consultants, Dallas. Texas, November 10, 1981). Both the width and depth

dimensions taper gradually toward a minimum at the point where the Branch

enters the pond. We assumed that the cross sectional profiles resembled

elliptic curves with the maximum depth at the center. Accordingly, we

approximated the shape of the pond as diagramed below:

tTop View: 400'

ISide View: 6

^ f c - — — ' ( n o t to scale)

3-D view showing ell iptic cross-sections:

The volume of the pond was calculated as follows:

1. Using similar triangles, the depth (D) and width (W) can be expressed as a

function of the length (L) of the pond:

33

D L~5 ° T?no

n 6L L/200" = T2TO =

and,

W L4TO " 1200

"-^-^

2. The area of an ell ipse equals TTah where a and b represent the dimensions

illustrated (Eshbach 1975, p. 258) :

Thus, the cross-sectional area of tne pond (Ac) is calculated as

fol lows;

Ac = 1/2 (-IT)(D)( l /2)(W) = TTDW/4

3. Finally the volume (V) of the pond is calculated:

V = ^\: dL

= ^CVW/^di

TT/4 ^ ( L / 2 0 0 ) ( L / 3 ) d L

1200ir/2400\ 1.2 dL

• J O

(ir/2400) (1/3) (12003)

34

7.'iA x in^ ft3

In addition to volume, the flow rate through the pond must he determined in

order to calculate the residence time. The flow rate computation involved two

steps:

1. Assuming that the rainfall and run-off characteristics are constant over

the upper Bayou Meto water shed (i.e., area above Lonoke or River Mile 106),

an average f low contribution (Fc) per acre can be calculated as fol lows:

F(; = average flow rate/size of water shed

- 3?1 ft^sec (3fi00 sec/hr)(24 hr/day)(3P5 day/yr)/(207 mi2^^ acres/mi)

« 76412 ft^/yr/acre

The average flow rate value used ahove was obtained from EPA 's STORET data

base for Bayou Meto (EPA, Water Quality Control Information System).

2. The size of the drainage area of the pond was measured directly as 503

acres. Using this value the flow through the pond (Fp) can be calculated:

Fp =s Fc (size of pond drainge area)

' 76412 ftS/yr/acre (502 acres)

= 3.84 x 107 ft3/yr

35

Now the average residence time is easily calculated as follows:

residence time » Pond Volume/flow rate

» 7.54 x 105 ft3/3.S4 x lO? ft3/yr

= 0.02 yr

Using the average residence time of 0.02 yr the trapping efficiency can be

determined from established charts to range from 0.46 to 0.71 (EPA 1979).

36

.001 .002 .005 .01 .02 .05 .10 .20 .50T'Meon Hydraulic Residence Time (years)

8.0 R E F E R E N C E S

1. U.S. Environmental Protection Agency, Office of Uaste Program Enforcement.Case File.

2. Esposito, M.P., T.O. Tiernam, and F.E. Dryden. 1980. Oioxins.EPA-600/2-80-197, U.S. Environmental Protection Agency, Washington, n.C.pp. 5, 241. and 247.

3. Stephan, C.E. 1980. Memorandum to J. Stara, U.S. Environmental •Protection Agency. July 3, 1980.

4. S R I , I n t e rna t iona l . 1980. Seafood Consumption Data A n a l y s i s - F ina lReport.

5. Perwak, J., A. Eschenroeder, et a1. 1980. An Exposure and R i s kAssessment for 2,3,7,8-TCDO - In ter im Draf t Report. EPA Contract f68-01-3857, U.S. Environmental Protection Agency, Wash ing ton , D.C. , p. 48.

6. Versar. 197B. Statement of Probable Fate - TCno. Prepared for EPAOff ice of Water Planning and Standards, U.S. Environmental ProtectionAgency.

7. Arkansas Soil and Water Conservation Commission. June 1979. NonpointSource Pollution Assessment Summaries for the Arkansas River Rasin.

s. Sullivan, N.N., and J.E. Terry. 1970. Drainage Areas of Streams inArkansas. U.S. Geological Survey

9. Hamaker, John W. 1978. Interpretation of Soil Leaching Experiments,republished in Volume I of Chemicals, Human Health and the Environment,p. 24.

10. Eshbach. 1975. Handbooks of Engineering Fundamentals, p. 258.

1 1 . EPA, Water Quality Control Information System (STORET), STORET UserAssistance Section (202) 426-7792.

12. EPA. August 1979. Costs and Water Quality Impacts of ReducingAgricultural Non-Point Source Pollution. EPA-600/5-79-009, p. 225.

13. Harless, R., and R. Lewis. 1980. Paper presented at Workshop on Impactof Chlorinated Dioxins and Related Compounds on the Environment. Rone,Italy, October 22-24, 1980.

37