Recommended Screening Limits for Contaminated Surface Soil and Review of Factors Relevant to Site-Specific Studies: Recommendations of the National ... Protection and Measurements

NCRP REPORT No. 129

RECOMMENDED SCREENING LIMITS FOR CONTAMINATED SURFACE SOIL AND REVIEW OF FACTORS RELEVANT TO SITE-SPECIFIC STUDIES Recommendations of the NATIONAL COUNCIL O N RADIATION PROTECTION AND MEASUREMENTS

Issued January 29,1999

National Council on Radiation Protection and Measurements 7910 Woodmont Avenue / Bethesda, M D 20814-3095

LEGAL NOTICE This Report was prepared by the National Council on Radiation Protection and Measurements (NCRP). The Council strives to provide accurate, complete and useful information in its documents. However, neither the NCRP, the members of NCRP, other persons contributing to or assisting in the preparation of this Report, nor any person acting on the behalf of any of these parties: (a) makes any warranty or representation, express or implied, with respect to the accuracy, completeness or usefulness of the information contained in this Report, or that the use of any information, method or process disclosed in this Report may not infringe on privately owned rights; or (b)assumes any liability with respect to the use of, or for damages resulting from the use of any information, method or process disclosed in this Report, under the Civil Rights Act of 1964, Section 701 et seq. as amended 42 U.S.C. Section 2000e et seq. (Title VZZJ or any other statutory or common law theory gouerning liability.

Library of Congress Cataloging-in-PublieationData Recommended screening limits for contaminated surface soil and review of fadors relevant to site-specific studies. p. cm. - (NCRP report ; no. 129) "January 1999." Includes bibliographical references and index. "SC a-20." ISBN 0-929600-61-4 1. Radioactive pollution of soils -Standards. 2. RadioisotopesEnvironmental aspeds. 3. Radiation dosimetry. I. National Council on Radiation Protection and Measuremenb. 11. Series. TD879.R34R43 1998 98-48643 628.5'5-dc21 CIP

Copyright O National Council on Radiation Protection and Measurements 1999 rights reserved. This publication is protected by copyright. NO part of this publication may be reproduced in any form or by any means, including photocopying, or utilized by any information storage and retrieval system without written permission from the copyrightowner, except for brief quotation in critical articles or reviews.

Preface The decision regarding the need to cleanup surface soil contaminated with radionuclides can be complicated. The use of this Report is intended to assist in this decision. This Report provides screening limits which can be applied to sites where the surface soil is known to be contaminated with radionuclides. The screening limits are calculated using methods which are chosen to be conservative under most conditions. Their use will allow reasonable judgments to be made regarding whether additional action is needed. Further action will generally not be required if the surface soil concentration is below the suggested limits. If the soil concentration is above the screening limit, a site-specific dose assessment is recommended. It is emphasized that doses calculated with the dose factors in this Report da not represent estimates of doses to particular or typical individuals or thresholds for possible adverse effects. Thus, the screening doses calculated using the methods of this Report are inappropriate for use in calculating population exposures or to estimate health effects. The calculation of doses to actual individuals requires the use of site-specific and individual-specific parameters in the formulas used for the calculations. This Report was prepared by Scientific Committee 64-20 on Contaminated Soil. Serving on the Committee were:

Harold L. Beck,Chairman Environmental Measurements Laboratory New York, New York Members

David G Baker Richland, Washington

William L. Robison Lawrence Livermore National Laboratory Livermore, California

Andre Bouville

Joseph H. Shinn

National Cancer Institute Bethesda, Maryland

Lawrence Livermore National Laboratory Livermore, California

F. Owen Hofhnan SENES Oak Ridge, Inc. Oak Ridge, Tennessee

Steven L. Simon National Academy of Sciences Washington, D.C.

NCRP Secretariat E. Ivan White, Senior Staff Scientist Cindy L. O'Brien, Managing Editor The Council wishes to express its appreciation to the Committee members for the time and effort devoted to the preparation of this Report and to the U.S.Department of Energy for the financial assistance in developing this Report.

Charles B. Meinhold President, NCRP

Contents ..................................................................................... 1. Introduction ........................................................................ 1.1 Scope of Report ............................................................. 1.2 Approach ......................................................................... 1.3 Land-Use Scenarios ........................................................ 1.3.1 Agricultural (AG).................................................. 1.3.2 Heavily Vegetated Pasture (PV).......................... 1.3.3 Sparsely Vegetated Pasture (PS) ......................... 13.4 Heavily Vegetated Rural (RV) ............................. 1.3.5 Sparsely Vegetated Rural (RS)............................ 1.3.6 Suburban (SU) ...................................................... 13.7 No Food Suburban (SN) .......................................

Preface

1.3.8 Construction. Commercial. Industrial (CC)

iii

........

. Soil .........................................................................................

2 Recommended Screening Limits for Contaminated

2.1 Recommended Screening Limits ................................... 2.1.1 Multiple Nuclide Contamination ......................... 2.1.2 Alternate Limiting Doses ..................................... 2.2 Application of Screening Limits .................................... 2.2.1 Applicability of Tabulated Land-Use Scenarios ............................................................... 2.2.2 Special Situations ............................................... 23.3 Nonapplicable Contamination Scenarios ............ 2.2.4 Site-Specific Dose Assessments ........................... 2.3 Scientific Basic for Screening Limits ............................

. 3.1 Dose Model .................................................................... .........................................

3 Dose from External Exposure

34 34 36 36 37

3.2 Discussion of Model Parameters ................................... 3.2.1 Concentration in Soil ............................................ 3.2.2 Effective Dose Factor ............................................ 3.2.2.1 Effective Dose versus Effective DoseEquivalent. Organ Dose .......................... 38 3.2.2.2 Exposure Rate. Free-in-Air Kerma versus Effective Dose .............................. 39

vi

/

CONTENTS 3.2.2.3 Dependence of Dose Factor on

Orientation ............................................... 3.2.2.4 Dependence of Dose Factor on Nuclide Depth Distribution .................................. 3.2.2.5 Dependence of Dose Factor on Soil Composition .............................................. 3.2.2.6 Dependence of Dose Factor on Areal Extent of Contamination ......................... 3.2.2.7 Accuracy of Dose Factor Calculations .... 3.2.2.8 Dose Factors Chosen for Screening Dose Calculations ............................................. 3.2.3 Dependence on Soil Moisture and Bulk Density 3.2.4 Shielding by Dwellings ......................................... 3.2.5 Indoor and Outdoor Exposure Times .................. 3.2.6 Age. Sex Correction .............................................. 3.3 Summary of Parameter Values for Screening Calculations .................................................................... 3.4 Calculated Screening Doses ...........................................

.

4 Dose from Inhaled Radionuclides .................................. 4.1 Introduction to the Resuspension-Migration Pathway 4.1.1 General Classes of Resuspension ........................ 4.1.1.1 Wind-Driven Open-Environment

Resuspension

...........................................

4.1.1.2 Localized and Direct Contamination

.....

4.1.2 Parameters of Contaminants that Affect Open-

Air Resuspension .................................................. 4.1.2.1 Particle Size ............................................. 4.1.2.2 Physical-Chemical Form Meeting Availability ............................................... 4.1.2.3 Availability with Time ............................ 4.2 Resuspension and Dose Models ..................................... 4.2.1 Dose Model ............................................................ 4.2.2 Resuspension Models ............................................ 4.2.2.1 Method 1, Estimation of C , by Modified Mass Loading ........................................... 4.2.2.2 Method 2, Estimation of C h by Resuspension Factor ................................ 4.2.2.3 Method 3, Estimation of C& by MassLoading-Derived Resuspension Fadors ... 4.2.3 Estimating the Uncertainty in Values of Air Concentration .......................................................

1

CONTENTS 46.4 Derivation of Radionuclide Concentrations in

Air for Screening

................................................... ...

4.2.5 Example of Calculation of Air Concentrations

4.3 Discussion of Dose Model Parameters .......................... 4.3.1 Average Nuclide Concentration in Soil ............... 4.3.1.1 Concentration in Soil for a Site-Specific

Assessment

...............................................

4.3.1.2 Concentration in Soil for Screening

.......

4.3.2 Indoor versus Outdoor Concentrations in Air .... 4.3.3 Inhalation Dose Factor ......................................... 4.3.3.1 Dependence of Dose Factor on Lung-

Absorption and Particle Size

..................

4.3.3.2 Dependence of Dose Factor on Age ........ 4.3.3.3 Organ Doses versus Effective Dose ........ 4.3.3.4 Uncertainty and Variability in Dose

Fadors ...................................................... 4.3.3.5 Recommended Dose Factors for

Screening

..................................................

Dose Calculations

.................................................

4.3.4 Usage Factors ....................................................... 4.3.4.1 General ..................................................... 4.3.4.2 Screening Values ..................................... 4.3.5 Age Dependence of Dose.Chi1d. infant Screening 4.3.6 Dose from Inhalation of Airborne Radon and

Progeny

..................................................................

4.4 Summary of Recommended Parameter Values for

Inhalation Dose Estimation

...........................................

4.5 Calculated Screening Inhalation Doses ........................

.

5 Dose from Ingested Radionuclides ................................ 5.1 Dose Model ...................................................................... 5.2 ~ k u s s i o nof Model Parameters ................................... 5.2.1 Eladionuclide Concentration in the Soil :,............ 5.2.2 Human Diets ......................................................... 5.2.2.1 Variability in Human Diet ...................... 5.2.2.2 Screening Values for Human Diet ......... 5.2.3 Soil to Vegetation Transfer Factors .................... 5.2.3.1 Effect of Soil Depth Profile on Soil to

Vegetation Transfer Factor

.....................

5.2.3.2 Soil to Vegetation Transfer Factor

Values for Screening

...............................

5.2.4 Animal Diets ......................................................... 5.2.5 Meat, Milk Transfer Factors ................................

~ i i

viii

1

CONTENTS

5.2.6 Decay Correction for Delay from Harvest to Consumption ......................................................... 5.2.7 Committed Effective Dose Factors ...................... 5.2.7.1 Dependence of Dose Factor on Age ........ 5.2.7.2 Dependence of Dose Factor on Gastrointestinal Uptake ......................... 5.2.7.3 Uncertainty in Biokinetic Models .......... 5.2.7.4 Recommended Ingestion Dose Factor Values for Screening ............................... 5.2.7.5 Dose Factors for Site-Specific Dose Assessments ............................................. 5.2.7.6 Organ Dose versus Effective Dose ......... 5.2.8 Uncertainty in Ingestion Dose Factors Chosen for Screening Calculations ................................... 5.3 Direct Ingestion of Soil ................................................... 5.3.1 Factors of Inadvertent Ingestion and the Etiology of Purposeful Intake .............................. 5.3.1.1 Inadvertent Intake .................................. 5.3.1.2 Geophagia ................................................. 5.3.2 Review of Literature on Soil Intake .................... 5.3.3 Recommended Ingestion Rates ............................ 5.3.4 Use of Soil Ingestion-Rate Data in Screening Calculations ........................................................... 5.3.5 Calculation of Screening Doses ........................... 5.4 Dependence of Committed Effective Dose on Age ....... 5.5 Summary of Recommended Parameter Values for Screening (Ingestion Pathway) ...................................... 5.6 Calculated Screening Doses ...........................................

.

6 Determination of Radionuclide Concentration in Soil ......................................................................................... 6.1 Factors to Consider in the Design of Screening and Sampling Programs for Radionuclide Concentrations in Contaminated Soil ..................................................... 6.2 Instrumental Measurement Techniques ....................... 6.2.1 In Situ Gamma-Ray Spectrometry ...................... 6.2.2 Exposure Rate Measurements ............................. 6.3 Soil Sampling .................................................................. 6.3.1 Soil Sampling Methodology .................................. 6.3.2 Sample Preparation and Analysis ....................... 6.4 Air Sampling ................................................................... 6.5 Strategy of Determining Radiouclide Concentrations for Screening ...................................................................

6.5.1 Estimating Soil Concentrations by In Situ

Spectrometry or Exposure Rate Measurements

...

6.5.2 Sampling Soil for Screening .................................. 6.5.3 Site-Specific Soil Sampling ....................................

.

7 Calculation of Screening Doses........................................ 7.1 Distribution of Individual Doses ..................................... 7.2 "Maximum" Dose Estimates for Screening .................... 7.3 Site-Specific Dose Assessments .......................................

Appendix k Calculated Screening Doses ..........................

. Appendix C. Dose Factors. Shielding Factors .................. Appendix D. Transfer Factors ..............................................

Appendix B Radionuclide Decay Data...............................

Glossary ....................................................................................... References .................................................................................. The NCRP ................................................................................... NCRP Publications................................................................... Index .......................................................................................... 352

1. Introduction Surface soil can become contaminated with radionuclides through many different mechanisms such as airborne deposition, spills and leaching from contaminated material stored above ground. Current activities associated with the cleanup of contaminated weapons production and storage facilities and the future decommissioning of nuclear facilities might result in additional soil contamination as well as in the discovery of past contamination. Even after cleanup of known contaminated land, some residual contamination will remain. No matter how the contamination occurred, the issue becomes whether the level is ~ ~ c i e n thigh, l y either before or after cleanup, to warrant action or restriction on the use of the contaminated site. The primary purpose of this Report is to provide screening limits (in Bq kg-') which can be applied to sites where the surface soil is determined to be contaminated with one or more radionuclides. These screening limits, which can be defined a s a conservative method of relating an effective dose ( E ) limit for a critical group to a corresponding soil contamination level (EPA, 1990a), can be used to allow reasonable judgments to be made regarding the need for possible (further) action based on present soil radionuclide levels. Such judgments need to be consistent with the recommendations of the National Council on Radiation Protection and Measurements (NCRP) regarding radiation exposure to members of the general population, with applicable regulatory limits, and with the principle of ALARA (as low as reasonably achievable). If the surface soil concentration is below the suggested limits, then no further action will generally be required. If the concentration is above the suggested limit, a site-specific dose assessment should be conducted. I t is emphasized that the doses given in this Report are strictly for comparison with a limiting value to establish a screening level and do not represent estimates of doses to particular or typical individuals or threshold values for possible adverse effects. The calculated doses are deliberately designed to conservatively represent the maximum dose to any individual. Thus,these doses are inappropriate for use in calculatingpopulation exposures or to estimate health effects. The calculation of doses to actual individuals requires the use of site-specificand individual-specificparameters in the formulas used for the calculations.

2

1

1.

INTRODUCTION

To justify the guidance provided, this Report reviews in some detail the scientific basis for estimating both site-specific and generic doses to individuals from all pathways that could result from direct or indirect exposure to the contaminated soil. This review includes discussions of the uncertainty and variability in all the important parameters included in the calculational models. The Report also provides guidance on how to determine the site average radionuclide concentration in surface soil to be used for applying these generic screening criteria. Factors important in site-specific studies are cited in many of the cases discussed.

1.1 Scope of Report Only surface soil contamination is considered. Buried wastes are not considered in this Report. Surface soil refers only to depths comprising the plow layer, i.e., down to a depth of about 30 cm. The guidance herein is not intended to be used for evaluating the implications of an ongoing contamination episode such as a continuing airborne deposition, which is treated in NCRP Commentary No. 3 (NCRP, 1989) and NCRP Report No. 123 (NCRP, 1996). Only radionuclides whose half-lives are longer than 30 d (or are supported by a precursor with a half-life >30 d) are considered. Ground water contamination is not calculated explicitly, although it is recognized that the latter could be an important dose pathway for some sites (see Section 2.2.2).The area contaminated is considered to be relatively extensive (hundreds of square meters or more). All other important dose pathways are considered including external radiation exposure, beta-ray skin dose, ingestion of contaminated foodstuffs, direct and indirect ingestion of soil by humans and animals, and both indoor and outdoor inhalation of resuspended material. Examples of contamination scenarios for which these limits are applicable include widespread contamination from fallout from weapons tests and nuclear facility accidents (such as occurred a t Chernobyl) as well as more localized contamination resulting from nuclear facility operations andlor decontamination and decommissioning. An example of the latter would be the plutonium contamination of soils downwind from the Rocky Flats plant (Krey and Hardy, 1970).

1.2

Approach

For screening purposes, it is appropriate to establish as a goal the limitation of the maximum annual E to a member of the most

1.2 APPROACH

/

3

critically exposed group, i.e., to insure that no individual is likely to receive a dose that exceeds some recommended limit. The approach used to calculate the screening doses in this Report is to first review the current models for estimating the dose to individuals for each pathway. Then, based on an extensive literature review, recommended parameter values are presented for eight different land-use scenarios describing the present or intended use of the contaminated site. These chosen parameter values are used to calculate the highest annual E from external exposure or committed effective dose [E(.s)l from inhalation or ingestion that would be delivered by both the nuclide and its progeny for each radionuclide considered for each land-use scenario.' Prudently conservative values are suggested for the uncertainty in each parameter value used in the models and scenarios. This uncertainty reflects both true variability due to biological and environmental variations, human lifestyle differences, etc., and lack of knowledge as t o the correct mean or central tendency. If possible, this uncertainty is characterized by a distribution parameter such as a standard deviation or geometric standard deviation. Sometimes,only a range can be estimated for a particular parameter because of a lack of adequate information regarding the true distribution of potential values. In those cases, a triangular distribution is assumed, encompassing the estimated 5 to 95 percent limits. To estimate the likely median and maximum dose to an individual in the critical group, the distribution of potential individual doses from each pathway and the total dose from all relevant pathways is calculated stochastically using the estimated uncertainty for each chosen parameter value. The 95th percentile of the calculated total dose distribution was then used as an estimate of the maximum dose to any individual in the critical group. Section 7 of this Report describes the methodology used for these stochastic calculations. These "maximum" dose estimates and a postulated "acceptable" maximum annual dose to any individual from exposure to any single contaminated site is used to estimate the recommended soil concentration screening limits presented in this Report. Each of the calculated screening doses can be thought of as a conservative estimate of the maximum annual E(T)to a representative member of the most 'Usually the maximum annual dose occurs in the first year of exposure. However, for a few very long-lived nuclides, the ingrowth of progeny may result in a higher annual dose sometime in the future. For those nuclides, the highest annual dose in any year over the next 1,000 y is calculated, and this value is used for setting screening criteria. A 1,000 y limit is consistent with delays used in the past for screening guidance (NCRP, 1996). In any case, only the dose for a few very long-lived radionuclides, primarily members of the naturally occumng uranium and thorium series, will continue to increase beyond 1,000 y (see Appendix A).

4

/

1. INTRODUCTION

exposed population group. The most exposed population will depend not only on the land-use scenario, but also on the particular radionuclide. Separate calculations are made for children and for adults and the highest dose is used for the guidance presented in this Report. The screening limits and total screening doses are tabulated in Section 2. The calculated E(z) for each pathway and land-use scenario are tabulated in Appendix A Using specific guidance provided in this Report, the reader can use the individual pathway doses for site-specific assessments and for the summation of doses from sites contaminated with more than one nuclide. By following this approach, the calculated annual E(z)and screening limits provided in this Report are expected to be much more realistic than previous screening models, yet still conservative. Previous NCRP screening recommendations (NCRP, 1989; 1996) assigned a single safety factor of 10 to the calculated total screening dose which was already a conservative value. The more extensive uncertainty analysis used for this Report provides separate land use, pathway and nuclide dependent screening limits. This approach eliminates the need to assume a single overly conservative overall safety factor to cover the range of all possible exposures and thus avoids undue conservatism in recommending screening soil concentration limits for particular sites and nuclides. NCRP Commentary No. 8 (NCRP, 1993a), for example, states that the screening limits given in Commentary No. 3 (NCRP, 1989) for airborne deposition scenarios were likely to be too conservative for external exposure while perhaps not conservative enough for ingestion of food grown on nutrient-depleted soil. Thus, by providing only a single screening limit for all land uses, the screening guidance in Commentary No. 3 is probably too conservative for certain land-use scenarios, for example where no food is grown. Finally, the approach taken here assumes that if the measured nuclide concentration in soil can result in an annual E(z) higher than the recommended screening limit, a site-specific dose assessment will be performed. Such an assessment might consider not only annual doses but also the lifetime risk to potentially exposed populations.

1.3 Land-Use Scenarios Separate calculations have been made for eight different land-use scenarios. The intent is to differentiate among uses where different dose pathways might dominate, the most exposed population group

1.3 LAND-USE SCENARIOS

/

5

might differ, or the range of a particular critical parameter might be more limited, while keeping the number of different scenarios considered to a manageable level. This results in a narrower range of possible individual doses for each projected land use. Thus, the screening limits for particular land uses are lower than would be required to maintain a comparable level of conservatism if all possible land uses were lumped together. The land-use scenarios chosen are described below.

1.3.1 Agricultural (AG) This scenario is intended for sites used primarily for food production for human consumption, e.g., vegetables, fruit, grain, etc. It is assumed that there are no dwellings on the contaminated site itself in order to more clearly distinguish this type of land-use from a rural land-use scenario. Thus, only adults are assumed to be exposed via inhalation and/or external radiation, although children and infants, in addition to adults, may be exposed via ingestion of contaminated food produced on the site. The most exposed individuals would depend on the dose pathway and radionuclide. If external exposure and/or inhalation is the most important dose pathway, the most exposed population would be farmers working on the site. If the ingestion pathway is significant for the nuclide in question, infants and children might be the most exposed population, even if they do not live on the site. Separate calculations of the maximum dose are made for adults and for children and infants and the highest dose used for the screening limits presented for this scenario.

1.3.2

Heavily Vegetated Pasture (PV)

This scenario is for sites used primarily for milk or meat production. Again, it is assumed that there are no dwellings and thus no direct external or inhalation exposure to children or infants. The most exposed population for the external radiation and inhalation dose pathways would be adults working on the land. For the ingestion dose pathway, the most exposed population group would be individuals who obtain most of their meat and milk from animals ingesting fodder grown on the site. Thus, the most exposed population group for the ingestion pathway will depend on the particular radionuclide since adults ingest more meat than children or infants but the dose per unit intake for many radionuclides is higher for infants and children (see Section 5). As for the agricultural scenario, separate

6

1

1. INTRODUCTION

calculations of the maximum dose from all relevant pathways are made for adults and for children and infants and the highest total dose used for the screening limits presented for this scenario. This scenario should also be used for agricultural sites used to grow feed for meat or milk production.

1.3.3 Sparsely Vegetated Pasture (PS) This scenario is identical to the previous except that the site is assumed to be located in an arid area. Resuspension of surface soil is assumed to be higher at these sites. A typical site might consist of open range land. Grazing animals are assumed to get less of their total diet from the site than for more heavily vegetated sites. This scenario should be used for sites where one might expect higher than average resuspension. It should also be used for sites with a higher than average potential for ingestion of soil by both animals and humans, either directly or via resuspension onto vegetation.

1.3.4 Heavily Vegetated Rural (RV) This scenario is intended to include open fields and forested sites. Some ingestion of contaminated food is assumed from gardens, wild game, or fruits and mushrooms fmm forests. It is assumed that there may be dwellings directly on the contaminated site. The most exposed population group will depend on the particular radionuclide but would most likely be children and infants living on the site that ingest milk from a backyard cow. This scenario can also be used for farms where persons live on site, as contrasted to the agricultural land-use scenario where it is assumed that no people live on site.

1.3.5 Sparsely Vegetated Rural (RS) The rural-sparsely vegetated scenario is similar to RV except, as for the sparsely vegetated pasture scenario, the site is likely to be in an arid area. Less food is assumed to be produced on these types of sites. However, resuspension of surface soil is assumed to be higher than for more heavily vegetated sites.

1.3.6 Suburban (SU) This scenario is for residential properties. Some minor food production such as that from vegetable gardens is assumed to occur. The

1.3 LAND-USE SCENARIOS

/

7

most exposed population would likely be children who live on the site, play outdoors, and ingest vegetables grown on the site and perhaps some contaminated soil.

1.3.7

No Food Suburban ( S N )

This scenario is identical to SU except that no food is produced from the site. Parks, schools, developed recreational sites and residential lawns would be included in this land-use scenario. The most exposed individuals would likely be children playing outdoors, inhaling resuspended soil and possibly ingesting contaminated soil.

1.3.8

Construction, Commercial, Industrial (CC)

This scenario is for sites where the soil is likely to be disturbed due to present or future construction activities or activities involving earth moving or for sites used for industrial or commercial purposes. It is assumed that there are no dwellings on the site and no children are exposed. The critically exposed group consists of adult workers from external radiation exposure andlor inhalation and ingestion of suspended contaminated soil. The doses from construction and earth moving activities are likely to be short term and thus the screening limits will be somewhat more conservative than for long-term exposures.

2. Recommended Screening Limits for Contaminated Soil This Section lists screening guidance for over 200 radionuclides with half-lives >30 d. This guidance is based on the screening dose calculations described in Sections 3 through 7. If the exposure is from a single site, as for the doses calculated in this Report, the NCRP recommends that the dose to the maximally exposed individual from any single set of sources, e g . , at a particular site, should not exceed 0.26 mSv y-I. This is intended to insure that the total dose rate to that individual from all man-made sources other than medical exposures does not exceed 1mSv y-I. Thus, to remain consistent with these recommendations, the screening limits recommended in this Report are based on limiting the maximum E(z) rate to any individual to 0.25 mSv y-l. Dose limits that appear to be lower than that used here are currently under consideration by regulatory bodies such as the U.S. Nuclear Regulatory Commission (NRC) and U.S.Environmental Protection Agency (EPA). However, the limits proposed by NRC and EPA, which are intended for cleanup of contaminated sites, are based on the median dose to an individual in the most critically exposed population rather than the maximum dose to any individual as used in this Report. Thus, the screening limits proposed here will generally be considerably higher (i.e.,more conservative) than the values proposed for regulatory purposes. It is again emphasized that the guidance proposed in this Report is for use in screening and is not intended for use a s cleanup criteria, since the conservative nature of the guidance given here could result in greater amounts of soil being removed than would be necessary with realistic, site-specific calculations.

2.1 Recommended Screening Limits

Table 2.1 lists the recommended screening limits for each nuclide considered in this Report. The limits were calculated by dividing 0.25 mSv by the calculated "maximum"screening total dose per unit

Isotope

T,,,(d)

TABLE2.1 -Recommended screening limits (Bq

Isotope

AG PV

PS

RV

RS

SU

TABLE 2.1 -Recornmended screening limits (Bq kg-').a (continued) SN

Isotope

.-Recommended screening limits (Bq kg-')." (continued)

Gd-153 Eu-154 Eu-155 Tb-157 Tb-158 Dy-159 Tb-160 Ho-166m Yb-169 Tm-170 Tm-171 Hf-172 Lu-173 Lu-174m Lu-174' Hf-175 Lu-176 Lu-177m Hf-178m Ta-179 Ta-180 Hf-181 W-181 Hf-182 Ta-182' Re-184m Re- 184' 0s-185 W-185 Re-186m

Isotope

TABLE 2.1 -Recommended screening limits (Bq kg-').' (continued)

9.4E + 04 1.9E + 02 5.3E t- 02 2.OE + 01 2.1E + 02

3.2E + 02 1.3E + 05 (4.8E+ 03 2.5E + 06 3.33 + 05

Bk-249 Cf-249 C f-25W Cm-250 Cf-251

RV 9.8E + 04 2.OE + 02 7;2E+B2 3.2E+01 3BE + 02

PS 1.2E + 05 2.4E + 02 6.73 + 02 3.2E+ 01 2.8E + 02

W

2.OE + 06 3.9E + 03 8.7E + 03 3.OE + 02 4.OE + 03 5.3E + 04 l . l E + 02 2.8E + 02 1.4E + 01 1.3E + 02

RS l . l E + 05 2.3E + 02 1.OE + 03 4.7E + 01 3.7E + 02

SU

1.3E + 05 2.7E + 02 2.OE + 03 7.7E +01 5.5E + 02

SN

8.8E + 04 1.8E+ 02 4.6E + 02 2.4E + 01 2.2E + 02

CC

"Average Bq kg-' dry soil measured over top 5 cm of soil (see Section 6). bSee Section 2.2.3 for additional guidance regarding the screening limits for naturally occurring radionuclides. 'A decay product of another radionuclide with half-life >30 d. If this nuclide is present only as a result of parent decay, its dose is included in that of i t s parent and only .the screening limit for the parent need be applied.

AG

Tvz (d)

Isotope

TABLE2.1 -Recommended screening limits (Bq kg-').a (continued)

2.1 RECOMMENDED SCREENING LIMITS

1

17

soil concentration in Sv (Bq kg-')-' from Table 2 Z 2The doses given in Table 2.2 represent the median and maximum annual dose (calculated using the median dose and safety factor) over the next 1,000 y from all pathways from the listed parent nuclide and all of its progeny for the most critically exposed population group. The dose distributions were calculated using a Monte Carlo technique as described in Section 7. The calculated doses for each dose pathway and landuse scenario are given in Appendix A along with the delay indicating the year of maximum exposure and whether the most exposed population consists of children or of adults. Progeny with half-lives >30 d are also listed separately in Tables 2.1 and 2.2 since the progeny may be present at the time the contamination is discovered even though the parent has already decayed away or the contamination scenario may have been such that the particular decay product was separated from its parent prior to release to the soil. However, if a radionuclide is present in the soil only as a result of decay of a precursor also present in the soil, only the screening limit for the parent should be used because the dose from the daughter product is included in that of the precursor. As described earlier, the screening limits given in Table 2.1 are designed to restrict the total annual dose to any exposed person from a single contaminated site to I00 keV) with dose factors exhibiting the smallest differenceswith assumed depth profile, i.e., 4) but as discussed earlier, the larger median-diameter distribution affects mostly the upper respiratory deposition. ICRP Publications 66 and 71 (ICRP, 1994a; 1995a) should be consulted for calculating sitespecific dose endpoints from estimated particle-size distributions. Method 1has been shown to hold for many sites of aged deposits and provides more site-specific information than the alternative Method 2 described below. Before considering the alternative methods of estimation of C,,, it is important to emphasize the requirement for representative measurement of S . The soil radionuclide concentration will have natural variability that cannot be characterized by a few samples. This natural variability is due to the complex nature of the original deposition, surface effects, and variability of the weathering process after initial deposition. Experience has shown that within a larger region of relatively homogeneous deposition there will be spatial variability depending on the scale of the soil-surface sampled. That is, the area of the soil-surface that is sampled is the critical dimension. Since the air sampled for estimates of C& contains particles whose trajectories came from many upwind locations, it is best to use a large area (tens of meters in radius) to determine an average of S based on many soil samples or from a n in situ gamma spectroscopic measurement that represents a large view of the soil surface. More detail on soil variability and methods for determinings are presented in Section 6. 4.2.2.2 Method 2, Estimation of C,,, by Resuspension Factor. The second method of estimation of C& is known as the resuspension factor method:

where:

Sr D

= the = the

resuspension fador (m-') total decay-corrected soil deposition (inventory) in Bq m-2

4.2 RESUSPENSION AND DOSE MODELS

1

69

By comparison with Equation 4.3, one sees that less information is and, in fact, by rearranging available about factors that influence Cd,, Equation 4.5 one obtains the definition of the "resuspension factor," which is really an air concentration normalized to D and has units of inverse length (m-I). Method 2 is more difficult to relate to site properties and thus less specific than Method 1.The deposition, D, can be related to soil surface concentration, S. Experimental evidence abounds that indicates the vertical distribution of radionuclides in the soil within a few months after a surface contaminating event will have a n approximately exponential distribution with depth: A = A, exp ( - 0 2 )

(4.6)

where: a

= known as the inverse relaxation depth (1RL-')

z

=

A,

= the extrapolated value of soil surface concentration

the depth below ground surface level (Bq kg- '1

Actually, the value of A at the surface is probably uniform with depth over some shallow depth, zl.The surface is likely to be wellmixed in the first few centimeters, because of geophysical and biological processes such as rain drop impact, shrinking-swelling cycles, wind saltation, root growth, or freezing-thawingthat are very active in the span of a only a few years (Shinn, 1992). Furthermore, in practice, soil is sampled from the top down in thin layers, usually at least 2.5 cm in depth, and, in fact, it becomes very difficult and sometimes arbitrary to decide what is the appropriate "zero" depth a t very shallow depths, simply because of natural microtopography. Let us define A,, in terms of S, the average concentration in the shallow layer of depth 21, assuming for practical reasons that the radionuclide is distributed uniformly through the layer. Then from Equation 4.6:

& = Sexp(azl/2)

(4.7)

By definition, the total deposition (inventory)D (Bq m-2)is obtained by summing over all soil layers:

D

= pX

(Ai &i)

(4.8)

where: P

Ai

= the soil bulk density (kg m-7 = the average activity in soil layer i in Bq kg-'

Thus, for site-specific assessments, soil sampling should provide the vertical profile distribution A, in order to determine D, and to solve

Equation 4.6 for the inverse relaxation depth. Analyzing a large number of depth increments from a single sampled site may be prohibitive, so a curve-fitting least squares analysis of three to four depth increments is usually employed to solve Equation 4.6. D can be expressed in terms of the inverse relaxation depth by integrating Equation 4.6 to great depth, and substituting Equation 4.7: D

=

pShe

(4.9)

where h,, which is defined as the "characteristic" or effective depth of deposition contributing to resuspension, has the properties: h,

=

(1 1 a)exp (az,1 2 )

(4.10)

he can be estimated easily for the reasonable values of z , (-2.5 cm), and for relaxation depths commonly observed a few months or more after deposition, 0.2 cm-' < a < 0.7 cm-' (1.4 cm < RL < 5 crn); he -5 k 1.5 cm (4.11) Note that heis not equal to the RL, i.e., l / a ,because of the presence of the well-mixed upper layer with average concentration S. Thus, estimates of deposition, D, can be made by applying Equation 4.9, with knowledge of the average surface-soil concentration, S, a reasonable judgment of penetration of the radionuclide so that h from Equation 4.11 applies, a n d using t h e soil bulk density p 1.6 g ~ m - One ~ . is now in a position to utilize the resuspension factor method of Equation 4.5 to estimate Cak,by applying available estimates of Sf.Resuspension factors for the first two months can be derived from the experiments by Garland (1982), those from 2 to 42 months from the observations following the Chernobyl event summarized by Garland et al. (1992), and for long times, from the results of Anspaugh et al. (1975). The latter authors also show a method for estimating resuspension factors for all times, but the more recent evidence from the Chernobyl accident shows a more rapid decline in resuspension factors with time in the short to intermediate times than their formula predicts. Based on these data, where no measurements of this resuspension factor are available, one should use the following estimates:

-

Sf = 1 0 - V t for the first 1,000 d

(4.12a)

Sf =

(4.12b)

for longer times, out to many years,

where:

Sr

=

t

=

has units of m-I the time in days since the contaminating event

4.2 RESUSPENSION AND DOSE MODELS

1

71

4.2.2.3 Method 3, Estimation of C,,, by Mass-Loading-Derived Resuspension Factors. A combination of Methods 1and 2 makes it possible to estimate C,.,, the concentration of radionuclides in air by means of a site-specific resuspension factor derived from the modified mass loading approach. This method results in much less uncertainty for purposes of comparing scenarios in human health risk assessments. The uncertainty due to variations in surface soil concentration S is eliminated, and more conservative, less uncertain variables are used in the estimation of the resuspension factor Sf. Recalling Equation 4.5 and Equation 4.9,

C,i,

=

Sr D

(4.5)

where:

Sr = the site-specific resuspension factor now defined by inserting Equations 4.5 and 4.9 into Equation 4.3 as follows:

sr

= (Er M ) I (p x h,)

(4.13)

To apply this method, Cai, i s estimated from Equation 4.5, with D from Equation 4.9, but with Sr estimated by Equation 4.13. This method is recommended for undisturbed soil surface conditions only, and the uncertainty of estimating CG,in that case is usually lower than for Methods 1or 2. 4.2.3

Estimating the Uncertainty in Values of Air concentration

The Methods 1, 2 and 3 are products, so that an estimate of the expected relative error in C~ can be obtained by the square root of the sum of the squares of the relative errors of each factor. One can provide an estimate for each of the factors under certain conditions. However, these factors are usually the median value of some distribution. Thus, estimates of the relative errors of the above factors are provided by estimating the coefficients of variation (CV), defined as the ratio of the sample S.D. to the mean. I n the case where there is empirical evidence of a logarithmic distribution, the parameters of the log-distribution were converted to CV (Gilbert, 1987):

cv = [exp (ln2s,) where: Sg

In

= =

the GSD the natural logarithm

-

Illn

(4.14)

72

/

4. DOSE FROM INHALED RADIONUCLIDES

Values of CV determined empirically for the modified mass loading and resuspension methods, respectively, are provided in Tables 4.2 and 4.3. The CV values of the enhancement factor, Ef, in Table 4.2, parts a and b, are calculated from sample GSDs of Elvalues reported by the authors cited. The CV values of S for soil sampling were taken from a study where Shinn et al. (1993a) compared 63 soil profiles after the data had been normalized to the deposition D, so that the samples were naturally statistically stratified, and there was no difference in analytic precision between sample locations. The CV values of S for in situ gamma spectroscopy were estimated by Shinn et al. (1993b1, where the CV was shown to be dependent on the combined uncertainty in the instrument calibration and the variability in bulk density, p, inverse relaxation depth, a,and characteristic depth, he. Even the uncertainty in the isotope ratios commonly needed by the method, such as the plutoniumlamericium ratio, contributes to the total CV. The CV values of mass loading, M, i n Table 4.2 came from the sample GSDs of the cited authors' data. Thus, it can be seen from Table 4.2 that the expected minimum error in Ca using the mass loading method can be expressed as a coefficient of variation approximately equal to 0.77.This CV is a combination of the smallest relative errors in Er,S and M shown in the available literature, in either the undisturbed or disturbed case. Furthermore, the expected maximum error in Cai,would be represented by a coefficient of variation of 3.4 or greater if soil sampling, rather than in situ gamma spectroscopy is used, or if mass loading is more uncertain, such as in the case of highly disturbed soil conditions. If an alternative method is used, the expected minimum error (CV)in Cabusing the resuspension factor method is 0.86 soon after a contaminating event, and increases in time (see Table 4.3). At later times, the CV value of the resuspension factor dominates over the CV of deposition (which has nearly the same expected error as the soil measurement). The reason for the differences between the CV values of long-term resuspension factors that were estimated for Chemobyl fallout and for studies in Nevada is likely due to the differences in horizontal representativeness of the Cai,measurements. For Nevada, the site studied was limited in size and the air samplers sometimes were not perfectly located with respect to the wind. It should be expected that this was not a problem where the fallout was more nearly horizontally homogeneous as in the Chernoby1 case. From Chemobyl data, the expected maximum error in Cai, using the resuspension factor method corresponds to a coefficient of variation of about 3.3. This value, obtained when soil sampling is used, is based on the long-term resuspension factor estimates under optimum conditions. However, the error increases drastically if the

CV

0.40 0.40 2.6 0.53 0.36 10.5 0.77 11.1

0.43 2.0 2.6 0.53 0.36 0.52 1.02 0.77 3.4 Disturbed soil Agriculture tractor Soil sampling In situ gamma Urban samplers Agriculture tractor Expected minimum Expected maximum

Undisturbed soil Nuclear event site By soil sampling By in situ gamma Urban samplers Rural samplers Replicate sampler Expected minimum Expected maximum

Description

"n is the number of measurements. bValues obtained by combining several error terms.

Cair

M M cot

S

Er Er S

Disturbed Soil Case

CBir

c~,

M

Er Ef S S M M

Undisturbed Soil Case

Variable s,

(GSD)

Method 1-Modified Mass Loading n'

Reference

Shinn (1992) Loshchilov et al. (1992) Shinn (1993a) Shinn (1993b) Shah et al. (1986) Loshchilov et al. (1992) Calculated Calculated

Shinn (1992) Shinn (1992) Shinn et al. (1993a) Shinn et al. (1993b) Shah et al. (1986) Shah et al.(1986) Luna et al.(1971) Calculated Calculated

TABLE 4.2 -Empirical CV for variables used in estimating resuspension.

cv 0.43 2.0 0.36 0.52 1.02 0.64 2.3

Er

12 14 14 14 20 115

1.83 1.98 2.24 2.77 3.7 10.3 4.2 1.76 2.1 4.9 10.4

-

-

-

nu

s, (GSD)

Undisturbed soil Nuclear event site Urban samplers Rural samplers Replicate sampler Expected minimum Expected maximum

Description S,

6 2 46 28 137

1.51 3.5 1.42 1.63 2.32 1.8 3.8

-

nn

(GSD)

8

P

1

Shinn (1992) Shinn (1992) Shah et al. (1986) Shah et al. (1986) Luna et al. (1971) Calculated Calculated

Reference

ul

'3

2c

i

i

Garland (1982) a Garland et al. (1992) m Garland et al. (1992) Garland et al. (1992) Garland et al. (1992) Kercher and Anspaugh (1993)b From S and h errors From S and h errors m Expected minimum Long term expected Expected maximum

Reference

"n is the number of independent sample periods or sites. bKercher, J.A. and Anspaugh, L.R. (1993). Personal communication (Health and Ecological Division, Lawrence National Laboratory, Livermore, California).

Cair

cair

M

3M

First 2 months At 12 months At 24 months At 42 months Long t e d c h e r n o b y l Long termNevada Soil sampling I n situ gamma First 2 months Long terdchernobyl Long term/Nevada

Description

Method 3 -Mass-Loading-Derived Resuspension Factor Method. Undisturbed Case Only.

0.61 0.77 0.96 1.35 2.1 15.1 2.6 0.61 0.86 3.3 15.3

CV

Variable

c,

cam c&

D

D

Sr Sr

s f

Sr Sr Sr

Variable

used in estimating resuspension.

Method 2-Resuspension Factor Method. Undisturbed Case Only.

TABLE4.3 -Empirical CV for variables

4.2 RESUSPENSION

AND DOSE MODELS

1

75

measurement is confounded by wind direction problems (e.g., for Nevada data). In summary, the uncertainty in the estimation of C& depends highly on several key variables. The surface soil concentration, S, contributes uncertainty to all methods of estimation. S has a large variance if it is derived from a limited number of soil samples and a smaller variance if it is obtained from measurement by in situ gamma spectroscopy. If measurements are made of the mass loading or the resuspension factor, then one must be concerned about the representativeness of the air sampler location. If no measurements are available, then estimates may become highly uncertain in the cases where either long-term resuspension or disturbed soil conditions is the subject in question. The latter is particularly true if C ~ , is being estimated for a soil cleanup exercise. The CV values provided here are based on the empirical evidence obtained from available contaminated sites. More experience in the future should assist in providing a better estimate. 4.2.4

Derivation of Radionuclide Concentrations in Air for Screening

The concentrations C,L for screening for the land-use scenarios discussed in Section 1were estimated from the resuspension factor Sfand Equations 4.5, 4.9 and 4.13. The combination of variables found in Table 4.4 were chosen as parameters for these estimates. The values for M, Efand Sf are annual averages. At a minimum, the mean concentration S, over the top 5 cm is assumed to be measured (Bq kg-'). The other parameters are estimated. The importance of obtaining a representative and areaaveraged value of S cannot be overemphasized (see Section 6.5, Strategy of Determining Radionuclide Concentrations for Screening). The values for Slare derived from Equation 4.6, values for Ef are from observations of resuspension from normal and disturbed soils and values of M from monitoring data in the literature. The bulk density, p, is assumed, conservatively, to have a common value of TABLE 4.4-Selected parameter values for screening calculations. Land-Use Scenario

AG

PV

PS

RV

RS

SUtSN

CC

76

/

4.

DOSE FROM LNHALED RADIONUCLIDES

1.6 g c m 3for topsoil, although the surface soil density is often less than this. The effective depth of deposition, h, can thus be estimated from Equation 4.9 using a sampling depth of the surface soil, 0.025 to 0.05 m. The value of a for most soils will fall in the range 10 to 100 m-l, so that h calculated from Equation 4.10 will have a value on the order of 0.05 to 0.1 m. For sandy soils, soils that have been tilled, or soils subject to construction after the contaminating event, h has a value closer to 0.1 m. The values for h in Table 4.4 should thus be prudently conservative. The diameters of airborne radionuclides have been determined primarily by the diameters of the host soil particles to which the radionuclides have become firmly attached. As discussed earlier, observations have shown that the diameters of airborne radionuclide particles originating from the soil are distributed approximately lognormally for the range of aerodynamic diameters between 0.3 and 10 pm. These distributions are broadly dispersed (GSD > 4), and have a median aerodynamic diameter usually between 2 and 6 pm. For screening purposes, a value of 1 pm is used. This favors the conservative estimate of pulmonary (deep lung) retention and corresponds to the value generally recommended for calculating inhalation dose factors for members of the general public (see Section 4.3.3;ICRP, 1994a; 1996a). The values given in Table 4.4 have been observed a t enough sites to permit determination of the uncertainty in each parameter for these selected scenarios. The major source of uncertainty is usually the combined natural variability of the observed parameters and not measurement imprecision. Thus, most of the parameters have an approximately lognormal frequency distribution, since concentrations of contaminants in soil and air also have lognormal frequency distributions. Observed estimates of the CV derived from these distributions are given in Table 4.5. The uncertainty estimates chosen TABLE4.5-Uncertainty in screening parameter values (CV. Parameter

CV

"Observed in soil contamination zones stratified by in situ gamma spectroscopy measurements.

4.2 RESUSPENSION AND DOSE MODELS

1

77

for use in screening refer to the variability in average annual values of the parameters. Thus the values are closer to the minimum values observed and smaller than would be appropriate for calculating exposures over shorter time intervals. Example of Calculation of Air Concentrations The first step in the estimation of concentrations of radionuclides in air is to calculate the effective deposition (inventory),D, given the radionuclide concentration, S, in the topsoil (representative, areaaveraged soil samples). Equation 4.9 is used, and Table 4.4 provides selected values for h. For example, assume the topsoil concentration, S, is 1Bq kg-' a t a site that is slated for development of commercial buildings. Using the construction land-use scenario, because the soil surface will be stirred by earth-moving and traffic activity, the estimate of D from Equation 4.9, using values from Table 4.4, is D = (1 Bq kg-') X (160 kg m-2) = 160 Bq m-2. The next step is to estimate radionuclide concentrations in air, C&, from Equation 4.5. The value of D is as calculated above and values of the site-specific Sfare calculated, from the method shown in Equation 4.13, or from values given in Table 4.4. Given the value of S = 1Bq kg-', the estimate of D = 160 Bq m-2, and site-specific SF= 4 x m-l, the estimate of C& is C&, = 6 x Bq m-3. = Sf x D = (1.6 x lo2) x 4 x The uncertainty of C& can be estimated by determining the coefficient of variation of C& as a figure of merit. The CV is calculated from the square root of the sum of squares of each of the CV in Sf, ph and S given in Table 4.5: CV = [(0.78)2 + (0.41)2 + (0.5)21* = 1.0. (4.15) The average annual outdoor concentrations in air were estimated in this manner for each of the land-use scenarios considered. These values, in turn, were used to estimate the annual effective committed doses using Equation 4.2. These estimated average annual air concentrations and associated uncertaintylvariability are given in Table 4.6. Note that the values are per unit radionuclide concentration 4.2.5

TABLE 4.6-Radionuclide concentrations in air used in screening dose ~alculations.~ Land-Use Scenario

AG

PV

'Per 1 Bq kg-' average in top 5 cm.

PSRS

RV

SUISN

CC

78

1

4. DOSE FROM INHALED RADIONUCLIDES

in soil. The CV value does not include the uncertainty in the soil concentration. The uncertainty estimates in Table 4.6 are used in Section 7 to estimate the overall uncertainty (range) of inhalation dose to the most exposed populations.

4.3

Discussion of Dose Model Parameters

Once the average annual radionuclide concentration in air has been estimated, one can use Equation 4.2 to estimate the annual E(T).Each of the remaining required parameters is discussed below.

4.3.1 Auerage Nuclide Concentration in Soil In order to calculate the annual dose commitment, one needs to determine the concentration of the particular nuclide at the time of inhalation, and the integral activity inhaled. The activity inhaled is in t u n directly proportional to the surface soil nuclide concentration, S . 4.3.1.1 Concentration in Soil for a Site-SpecificAssessment. For a site-specific dose assessment one should determine the local average depth distribution (see Section 6) and, if possible, the average concentration in the top 2.5 cm. Preferably, one should consider all the factors discussed in Section 4.1 and, if at all possible, measure the average daily airborne radionuclide concentrations as well as concentration in soil. 4.3.1.2 Concentration in Soil for Screening. For screening purposes, the median and maximum annual committed doses to a member of the most exposed population group must be estimated. It is assumed that the average concentration of each contaminant in the top 5 cm has been measured a t some time ,t in a manner sufficient (see Section 6.5) to achieve an acceptably small CV. It is assumed for screening that the radionuclide is, and remains, uniformly distributed over the top 5 cm. As discussed in Section 3, however, the depth distribution soon after the original deposition may be such that the concentration near the surface is somewhat higher and the concentration in the surface soil will decrease with time due to percolation to deeper layers or due to erosion. The additional uncertainty resulting from this assumption is likely to be minor, since, as discussed in Section 4.2.4, a conservative estimate for the effective depth of penetration is assumed. For most situations, as

4.3 DISCUSSION OF DOSE MODEL PARAMETERS

1

79

discussed in Section 3.2.2.4, the radionuclide will likely have penetrated significantly into the ground over the course of the first year. As discussed earlier for the external dose pathway (see Section 3.2.1.1), the concentrations will also change with time due to decay and buildup of progeny. Thus the soil concentrations used for calculating screening doses were corrected to provide the maximum decay series dose from all pathways in any year over a 1,000 y interval. 4.3.2

Indoor versus Outdoor Concentrations in Air

Even when indoors, one may still be subjected to contaminated air. Studies have shown that indoor/outdoor concentration ratios can vary widely depending on the tightness of the dwelling (air exchange rate), the type of structure, etc. (DOE, 1990; Roed and Camell, 1987). Radionuclides can also be brought into the building on shoes and clothing and then resuspended and inhaled. For site-specific studies, it may be useful to actually compare indoor versus outdoor airborne radionuclide concentration measurements under various meteorological and living conditions. For screening studies, however, an average indoor to outdoor ratio of 0.3 (GSD = 1.45) was chosen as a reasonable yet prudently conservative value for screening. The estimated uncertainty corresponds to an estimated range of from about 0.1 to 0.9. Both the ratio and range are consistent with results from a number of studies (NCRP, 1993a). Note that although rac2lonuclide concentration levels are generally lower indoors than out, total resuspendable dust levels indoors may be comparable to or even higher than outdoors due to indoor activities (Thatcher and Layton, 1995). 4.3.3

Inhalation Dose Factor

The E(z) to a reference human due to inhalation of a given amount of radionuclide has been reported for a large number of radionuclides (EPA, 1988; ICRP, 1979-82; 1996b; Phipps et al., 1991) using biokinetic models developed by the ICRP (1977; 1994a). These calculations account for the buildup and decay of daughter products in the body. The biokinetic models consider the particle size distribution, transfer from lung to other body organs, and uptake from the GI tract. The calculated individual organ and effective doses are for one or more particular lung absorption types based on the likely chemical form of the inhaled radionuclide. The ICRP inhalation dose model has recently been revised (ICRP, 1994a) and the ICRP has published calculated E(z) for both adults

80

1

4. DOSE FROM INHALED RADIONUCLIDES

and children as members of the general public of various ages for over 800 radionuclides (ICRP, 1995a; 1996b).E(T)for workers have also been published based on the new inhalation dose model (ICRP, 199410). Adult ICRP organ dose and dose-equivalent estimates for most radionuclides, as a function of clearance class based on the ICRP Publication 30 lung model and using ICRP Publication 26 organ weighting factors, are reported in Federal Guidance Report No. 11 (EPA, 1988). The National Radiological Protection Board (NRPB) of the United Kingdom has also published calculations of effective dose factors based on the ICRP Publication 30 lung model (Phipps et al., 1991). The NRPB,however, used ICRP Publication 60 organ weighting factors and also provided estimates as a function of age for a large number of nuclides. 4.3.3.1 Dependence of Dose Factor on Lung Absorption and Particle Size. The physical-chemicalform of the radionuclide andlor matrix material and size of the particle upon which it is imbedded determines the absorption and clearance properties of the material in the various compartments of the respiratory system after it is either inhaled or ingested. For example, if plutonium is in the form of h 0 2 , it is insoluble and less available for transfer across the gut wall and the lung-blood barrier than if it is in an ionized or other more chemically soluble form or bound to small particles or organic complexes of some type (ICRP, 1979-82; 1994a; 199413). Often, however, it is not possibleto determine the chemical form. Thus, a conservative approach, used in previous NCRP screening models (NCRP, 1996), is to assume that the inhaled material estimated from resuspension measurements is of the absorption type that results in the highest E(.r). The estimated dose for some nuclides varies significantly depending on the assumed absorption type as shown in Table 4.7. The new ICRP Publication 66 inhalation dose model no longer uses only three discrete clearance-rate classes (ICRP, 1994a); however, the three absorption types generally listed, slow, moderate and fast, roughly correspond to the ICRP Publication 30 day, week and year clearance rates, respectively. The ICRP (1996133lists recommended absorption types to be used in calculating E(z) to the general public for various chemical compounds likely to be encountered in the environment as well as for unspecified compounds. When the dose factor has been calculated for more than one absorption type or particle size, the dose factor corresponding to the recommended absorption type clearance rate should be used for site-specificstudies if the chemical form of the nuclide in the soil is known. Similarly, if the size distribution is known to be significantly different from that assumed for screening

4.3 DISCUSSION OF DOSE MODEL PARAMETERS

81

1

TABLE4.7-Comparison of adult inhalation E(T)factors (Sv Bq') versus lung absorption type for selected radionuclides." Absorption Type Nuclide

F (fast)

M (moderate)

S (slow)

"Data from ICRP Publication 72 (ICRP, 199613)

purposes, one should use a dose factor calculated for particles nearer that size rather than the ICRP default values. Table 4.8 compares the calculated dose factors for an adult for 1Fm versus 5 km median particle sizes. Note that for absorption types S and M, the use of dose factors for 1 krn AMAD particles is generally conservative but may underestimate the dose for fast absorption. However, since the recommended absorption type used in this Report for the nuclides where inhalation is likely to be an important dose pathway is usually type M (see Appendix C), the dose factors used should be conservative. The ICRP recommends and uses a median particle size of 1 km TABLE4.8-Variation of inhalation dose factor (Sv Bq-') with median particle sizea Absorption Type-F

Absorption Type-M

Nuclide

1 pm

5 pm

1pm

5 pm

Co-60 Sr-90 Zr-95 Ru-103 Ce-144 U-238 Pu-239

5.23-9 2.43-8 2.53-9 4.93-10

-b

9.63-9

7.1E-9

3.03-8 3.OE-9 6.83-10

-b

-b

4.53-9 2.33-9 3.43-8 2.63-6 4.73-5

3.63-9 1.93-9 2.33-8 1.63-6 3.23-5

-b

-b

4.93-7

5.83-7

-b

-b

"Data from ICRP Publication 68 (ICRP, 1994b). bNodata.

Absorption Type-S 1pm

5pm

82

1

4.

DOSE FROM INHALED RADIONUCLIDES

for members of the general public and 5 p m for adult workers (ICRP, 1996b),although, as discussed earlier, resuspended material is likely to have an AMAD higher than 1 pm. Dependence of Dose Factor on Age. The dose factor also varies significantly with age, as shown in Table 4.9 which gives the ratio of the infant (1y) and child (10 y) dose factors to the corresponding adult dose factor. Thus, a site-specific dose assessment should calculate separate inhalation doses for infants, children and adults. This requires age-specific breathing rates, which are discussed in Section 4.3.4.A site-specific assessment should also be based on realistic exposure times for each age group.

4.3.3.2

Organ Doses versus Effective Dose. For site-specific dose assessments it may also be appropriate to estimate the doses to specific organs rather than E. For some nuclides, the highest organ doses differ significantly from E. Table 4.10compares the dose factor for the most critical organ with E(T).

4.3.3.3

Uncertainty and Variability in Dose Factors. The NCRP (1998)has estimated the uncertainty in the biokinetic models used to determine inhalation dose factors. Table 4.11 summarizes this uncertainty for some selected nuclides of interest. Because of the number and complexity of the assumptions needed to calculate the dose factors and the lack of adequate biokinetic data for many nuclides, the uncertainty estimates given in Table 4.11 are based TABLE 4.9 -Variation of inhalation dose factor with age for selected radi~nuclides.~

4.3.3.4

Nuclide

Sr-90 Se-75 Tc-99 1-129 CS-134 CS-137 Ra-226 U-238 Pu-238 Pu-239 Am-24 1

Absorption Type

M F M F F F

M M M M M

Infant I Adult

Child / Adult

3.1 6.0 3.3 2.4 1.1 1.2 3.1 3.2 1.6 1.5 1.6

1.4 2.5 1.4 1.9

0.8 0.8 1.4 1.4 1.O 1.0 1.0

LDatafrom ICRP Publication 72 (ICRP, 1996b). The infant values are for age 1 y while the child values are for age 10 y. See Appendix C for a complete listing for all covered radionuclides.

4.3 DISCUSSION OF DOSE MODEL PARAMETERS

1

83

TABLE 4.10-Ratio of dose to critical (most exposed) organ to effective committed inhalation dose for selected radionuclides." Absorption

Nuclide

Type

Critical Organ

Sr-90 Zr-95 Tc-99 1-129 Ce-144 Cs-137 U-238 Pu-239

M M M F M

Lung Lung Lung Thyroid Lung Respiratory tract Lug Bone surface

F M M

Ratio

'For adults, data from ICRP Publication 71 (ICRP, 1995a).

TABLE 4.11-Estimated uncertainty in inhalation dose factors for selected radwnuclides." Estimated Range Nuclide

Male Adult

Othersb

10 (female teen) 10 10

5 20 10 20 20 5 10 10 5 10 10 10 10 10 10 20 "dapted from NCRP (1998). The estimated range R can be interpreted as indicating that the dose factor for some individuals may be a s much as a factor of R higher or lower than the dose factor recommended by the ICRP. bSpecial population groups generally consisting of diseased people, or of infants or children.

84

1

4.

DOSE FROM INHALED RADIONUCLIDES

on expert judgment rather than a mathematical calculation. The uncertainty ranges given use ICRP Publication 30 values as a starting point and include uncertainty in the appropriate gut transfer factor as well as the uncertainty in the biokinetic factors used in the ICRP Lung Model. They reflect the degree of reliability that a particular group of experts has in the ICRP Publication 30 values and thus may include an element ofbias. The column labeled "others" reflects the larger uncertainty in model parameters for infants or children or for groups suffering from certain diseases. Although the uncertainty estimates are based on the ICRP Publication 30 model rather than the newer ICRP Publication 66 model, we have utilized them for estimating the uncertainty in the inhalation dose factors used for calculating the screening doses in this Report since no similar estimates have been made for the newer model. 4.3.3.5

Recommended Dose Factors for Screening. In this Report,

E(T)factors published by the ICRP (1996b) for members of the general public are used for calculating screening doses. These dose factors are listed in Appendix C. If a value for a given nuclide was calculated for more than one clearance rate or median particle size, the value recommended for use by the ICRP for unspecified compounds was used. It was also assumed that all the airborne material was in the respirable size range covered by the dose model which assumed a 1p m median diameter particle size. As discussed earlier, this is generally conservative since the median diameter of suspended material is likely to be higher than 1 bm. GSDs were chosen to reflect the estimated range for the male adult for use with the adult dose factors discussed above, assuming the range about the median corresponds to about 3 GSD. A conservative estimate of uncertainty (GSD = 1.4 to 2.2) inferred from the uncertainty ranges presented in Table 4.11 was assigned to E(T) factor. A GSD of 1.7 was used for all nuclides listed in Table 4.11 except that for the well-studied nuclides, 90Sr,=9pU and 137Cs.For '"Sr, ='Pu and L37Cs a GSD of 1.4 was used. A GSD of 2.2 was assigned for all other radionuclides. As indicated in Table 4.9, the mean dose factors for some nuclides for infants or children may be much higher than those for adults. However, as will be shown later in Section 4.3.5, this higher dose factor is compensated for by a much lower inhalation rate. 4.3.4

Usage Factors

4.3.4.1 General. The true inhalation dose depends on the time a person is exposed to resuspended soil and the amount inhaled over

85

/

4.3 DISCUSSION OF DOSE MODEL PARAMETERS

the course of the exposure. The time exposed, both indoors and out, is assumed to be the same as the time exposed to external radiation (see Table 3.12). The amount inhaled, however, will depend on the breathing rate, which will vary with the type of activity and with age. Table 4.12 gives reported average breathing rates for various types of activity versus age. Screening Values. Table 4.13 gives the chosen values for breathing rates and associated variability used in the screening calculation for both indoors and outdoors. These values represent averages for outdoor and indoor exposure over the course of a year. The outdoor rate is based on the assumption that the individuals making up the critical population are active outdoor workers. The inhalation rates are for adults to correspond to the use of adult dose factors.

4.3.4.2

TABLE 4.12-Average breathing rates (m3d-I) for various types of activity. Reference

Adult

Child (10y)

Infant(1 y)

UNSCEAR (1993)-all activities

22

15

4

15 7 27 43-53

5 4 8

ICRP Publication 66 (1994~) -all activities 16-20 Resting 8-11 Light activity 30-36 Heavy activity 65-72

-

ICRP Publication 66 ( 1 9 9 4 ~-average ) annual indoor and out Housewife 18 Sedentary male worker 22 Sedentary female worker 18 Outdoor worker 25

TABLE 4.13 -Inhalation rates used for calculating adult screening doses for land-use scenarios considered (m3d-3. Land-Use Scenario

Outdoor-median GSD Indoor-median GSD

35 1.2

35 1.2

35 1.2

30 1.2

30 1.2

25 1.2

35 1.2

-

-

-

20 1.2

20 1.2

20 1.2

-

86

1

4. DOSE FROM INHALED RADIONUCLIDES

Table 3.12 gave the chosen indoor and outdoor exposure times and associated uncertainty assumed for the screening calculations for each land-use scenario. These times apply to both external exposure and inhalation exposure. Based on information reviewed in NCRP Report No. 76 (NCRP, 1984b), minute breathing rates tend to be distributed lognormally with a GSD of about 1.3 to 1.5. Since the screening criteria in this Report are based on average annual doses, and the variability in inhalation rate over the course of a year will tend to average out, the GSD of 1.2 assigned to the chosen estimates is believed to be reasonable and conservative. It allows for the possibility t h a t a member of the critical group will spend a significant fraction of hisher time in heavy activity, both while indoors or outdoors. 4.3.5 Age Dependence of Dose-Child, Infant Screening Dose Calculations

Using the age variation in the inhalation dose factor given in ICRP Publication 72 (ICRP, 1996b) (see Table 4.9 and Appendix C) and the differences in average (all activities) inhalation rate given in Table 4.12 for infants (1y old) and children (10 y old) versus a d ~ l t s , ~ the committed effective inhalation dose was also calculated for children (age 10 y) and infants (age l y) for each nuclide for the landuse scenario sites where children a r e allowed to be exposed. Table 4.14 shows the ratio of the inhalation dose for children and infants to adults corrected for breathing rate for some selected radionuclides. As can be seen, the ratio of infant or child dose to adult dose ranges from about 0.2 to 1.7. The child (age 10 y) to adult dose ratio, in particular, is sometimes >1 for some nuclides despite the much lower breathing rates. Because the adult-child-infant differences are small compared to the uncertaintylvariability in the biokinetic factors, applicable clearance class, and usage factors, the same uncertainty estimates for breathing rate and dose factor are used for the calculations of the childlinfant doses as for the adult calculations. Since the calculated screening doses use an adult outdoor breathing rate based on an active worker, the maximum dose calculated based on these uncertainty estimates should still include the dose to even the most exposed infant or child. However, for site-specific assessments where children or infants might be exposed, it is recommended that whenever possible, appropriate age-specific parameters and 'A ratio of 0.25 was used for the infant to adult (VA) breathing rate and 0.8 for the child to adult (CIA) rate.

4.3

DISCUSSION OF DOSE MODEL PARAMETERS

1

87

TABLE4.14-Ratio of child to adult a n d infant to adult committed effective inhalation dose for selected radionuclides." Nuclide

Se-75 Sr-90

Tc-99 1-129 (3-134 CS-137 Ra-226 Pu-238 Pu-239 Am-241 U-238

Child/Adult

InfanVAdult

1.7 1.0 1.0 1.3 0.6 0.6 1.0 0.7 0.7 0.7 1.0

1.1 0.6 0.6 0.4 0.2 0.2 0.6 0.3 0.3 0.3 0.6

"Estimated using child (age = 10 y), infant (age = 1 y), adult dose factors from ICRP Publication 72 (ICRP, 1996b), and relative average breathing rates from Table 4.12. related uncertainty be used. Note that ICRP (199613) provides dose factors for a wider range of ages than the three age groupings used for this Report. For the land-use scenarios where children can be present, the separate inhalation, ingestion and external exposure doses calculated for them were summed stochastically, as described in Section 7, and the highest dose (for either children or adults) used to calculate the screening guidance given in Section 2. Note that in Appendix A, separate calculations for infants and children are not tabulated since only the highest inhalation dose (generally that for children) was conservatively combined with the childlinfant doses calculated for the other dose pathways. 4.3.6

Dose from Inhalation of Airborne Radon a n d Progeny

The potential doses due to exposure to radon and its progeny were reviewed in NCRP Report No. 78 (NCRP, 1 9 8 4 ~and ) by UNSCEAR (1988; 1993). Most of the inhalation dose ascribed to 226Rain the surface soil will be from inhalation of radon progeny. The screening dose calculations for 226Raused the values given in UNSCEAR (1993) relating E(T)from radon and its progeny to a given outdoor air radon concentration, i-e., the average 226Ra concentration measured in soil of 40 Bq k g 1 to the average outdoor radon gas air concentration of

88

/

4. DOSE FROM INHALED RADIONUCLIDES

10 Bq me3,resulting in an annual E ( d of 650 pSv y-I from radon and its progeny if exposed 100 percent of the time (UNSCEAR, 1993). Thus, an average 226Ra soil concentration of 1Bq kg-' would correspond to an inhalation dose of about 15 pSv y-l. For sites where children might be the most exposed population, UNSCEAR recommends using a dose 1.5 to 2 times this value. The inhalation screening doses for a26Ragiven in this Report thus include a contribution of 15 pSv y-I (Bq kg-')-' for adults and 25 pSv y-' (Bq for children (for sites with dwellings) for the fraction of time spent outdoors on the contaminated site, and 8 pSv y-l (Bq kg-')-' for adults and 13 pSv y-I (Bqkg-')-' for children for time spent indoors.'

4.4 Summary of Recommended Parameter Values for

Inhalation Dose Estimation Land-Use Dependent Outdoor air concentrations Indoor/outdoor ratio Breathing rates Indoor/outdoor occupancy Land-Use Independent Dose factors Age correction factor

see Table 4.6 0.3 (GSD = 1.45) see Table 4.12 see Table 3.12 see text, Appendix C (GSD = 1.4 to 2.2) see text, Table 4.14, Appendix C

4.5 Calculated Screening Inhalation Doses

Table 4.15 lists the calculated median annual E(z) from inhalation for an adult member of the most exposed population group for each land-use scenario for a number of important radionuclides likely to be found in contaminated soils. Estimates of the likely range (5 to 95 percentiles), calculated stochastically based on the uncertainty analysis discussed in Section 7, are also given. The complete listing of inhalation screening doses for all nuclides considered is given in Appendix A. 'For screening, the indoor gas concentration is arreumed to be equal to the outdoor level but since the equilibrium equivalent radon concentration indoors is about 40 percent compared ta 80 percent outdoors (UNSCEAR,1993) the dose when indoors L halved.

AG PV

PS

RV

RS SU/SN

CC

"Median dose to representative adult member of critically exposed population. Ranges in parentheses represent 5 to 95 percent percentiles. Doses include contributions of decay products and are maximum annual doses over a 1,000 y interval. Nuclide concentration is per kilogram of dry soil assumed uniform over top 5 cm. See Appendix A for complete listing of inhalation doses. Doses are also given for 226Rawithout the contribution from 2a2Rnto illustrate contribution from radon.

Nuclide

Land-Use Scenario

TABLE 4.15-Annual E(T.from inhalation of selected radionuclides [Sv (Bq &-')-'I."

m w

.

2

90

1

4.

DOSE FROM INHALED RADIONUCLIDES

The inhalation doses vary more with land-use scenario than do the external pathway doses, and exhibit a much larger likely range. However, the inhalation doses generally control the screening limits only for nuclides such as the transuranics that emit nonpenetrating radiations and generally only for relatively dusty sites such a s sparsely vegetated and construction sites.

5. Dose from Ingested Radionuclides If contaminated land is used for the production of food for either human or animal ingestion, the ingestion of this food, or of the by-products (meat, milk) from animals eating forage grown on the contaminated land may result in a radiation exposure and E(z). Ingestion of contaminated soil may also result in E(z). This Section discusses models which can be used for estimating such ingestion doses both for screening purposes and for subsequent site-specific assessments, the relevant parameters used in these models, and the calculated screening doses.

5.1 Dose Model The annual committed effective dose, incurred from the ingestion of a given quantity of food containing a particular radionuclide can be estimated from the following expression: E,,, = Z,(Ci x Ri) x fi

x exp[- A(ti - to)] x Dfi,,

(5.1)

where: of a particular radionuclide in foodstuff i at harvest =the average annual intake of foodstuffi (kg) Ri = t h e fraction of R i derived from the contami6 nated site exp[ - A(ti- to)]= a correction factor to account for radioactive decay between harvest (to)and ingestion (t) = the committed effective dose for ingestion (Sv per Dfi, Bq) that would result from an intake of 1 Bq of this nuclide

Ci

= the concentration in Bq kg-'

Foodstuff i can be vegetation grown in the contaminated soil or milk or meat from animals which have ingested vegetation grown on the contaminated soil. The foodstuff can also take the form of contaminated soil ingested by humans either directly or as a result of eating

92

1

5. DOSE FROM INGESTED RADIONUCLIDES

unwashed contaminated vegetables or fruit. The total dose from ingestion is obtained by summing the individual contributions from each nuclide for each foodstuff and from soil itself. The concentration in a given foodstuff, Ci,depends on the foodstuff and the particular pathway by which the radionuclides in the soil were transferred to the foodstuff. For edible vegetation, eg., vegetables, grains and fruits directly ingested by humans, the contamination can result from direct root uptake by t h e plant or from contamination due to resuspension of contaminated soil with subsequent deposition and adhesion onto the plant surfaces and also possible translocation into its edible portions. The concentration (Bq kg-') in a given type of vegetation due to root uptake and resuspension processes can be estimated for screening calculations from the following expression:

B,

= an empirically determined soil to vegetation transfer factor

for root uptake usually expressed as Bq kg-' of wet vegetation or dry animal fodder per Bq kg-' of dry soil B,, = a similar transfer factor representing the net effect of all resuspension processes S = the concentration Bq kg-' (dry) at harvest of the soil to which the vegetation is exposed (for screening S is assumed to be the surface soil concentration) In the case of contamination by resuspended soil, the mechanisms can be quite complex. Not only can airborne resuspension and subsequent redeposit contaminate the vegetation, but also phenomena such as rain splash, saltation and mechanical disturbances during harvest. Thus, B,, must also be determined empirically for the particular site or type of site and the type of vegetation. Previous NCRP screening models [e.g., NCRP Commentary No. 3 (NCRP, 1989; 1996)lfocused on contamination of vegetation by direct airborne deposition. Thus, the contamination was calculated using an estimated interception fraction and deposition velocity. Such an approach is not valid for contamination dominated by resuspension. In this case, the major source of contamination is due primarily to contaminated soil being deposited onto, and adhering to, the vegetation or to root uptake. The major dose pathways are thus direct ingestion of the contaminated soil adhering to the vegetation by humans or animals or ingestion of edible portions contaminated by root uptake or translocation of radionuclides from plant surfaces to edible portions. Incorporation of the radionuclide into the plant itself via resuspension is generally of much smaller importance than for

5.1 DOSE MODEL

1

93

the case of direct deposition. Thus the use of the airborne deposition screening model formulation for contamination by resuspension will generally significantly underestimate the concentration in or on unwashed vegetation. Vegetation for human consumption is usually washed, although not always thoroughly enough to remove all attached soil. However, soil adhesion can be particularly important for forage or pasture vegetation eaten by animals because, unlike vegetation consumed by humans, it is unwashed. The concentration CmilLWt of a particular nuclide in contaminated animal products depends on several additional factors. C& ,, can be calculated from the following expression: C*meat

=

Cfodder

X Qmilk, meat X

T Q ~ Imeat I I ,X F d ,meat

(5.3)

where: = the nuclide concentration (Bq kg-') in the dry fodder

Cfodder

Qrmlk, meat

2Qw

,,

F d ,meat

eaten by the animal, nuclide concentrations in fodder can be calculated from Equation 5.2 = the average daily intake (kg d-') of the animal of dry contaminated fodder = the fraction of the animal's total feed derived from the site = an empirically determined transfer factor representing the equilibrium concentration in meat or milk (d kg-' or d L-') resulting from a given daily intake of radionuclide by the animal (C* ,, is thus given in Bq kg-' or Bq L-' of meat or milk, respectively.)

If the animal consumes different types of fodder, i-e., grain, grass, etc., one first uses Equation 5.2 to calculate the nuclide concentration in each type of vegetation. The average fodder concentration is calculated by summing the different radionuclide concentrations, weighting by the fraction of the animals diet from each fodder type. The dose incurred by an individual ingesting foodstuffs or soil from a contaminated site will, of course, depend on the type and amount of food grown on the site (the land-use scenario). The dose will also depend on the fraction of hidher total diet that originates from the contaminated site. The type of site, the type of soil at the site, the particular vegetation, and the distribution of the radionuclides in the soil will influence the amount of radionuclide available for root uptake or that may be resuspended and deposited onto the vegetation. The amount of soil ingested will also depend on the type of site and the particular lifestyles and habits of the inhabitants (see Section 5.3). The soil to vegetation and feed to milklmeat transfer factors in the expressions given above also differ considerably from

94

1

5. DOSE FROM INGESTED RADIONUCLIDES

nuclide to nuclide and vary depending on the nuclide's chemical form in the soil. For the screening models used in this Report, the effect of many physical and chemical processes are often grouped together into a single transfer factor for which a steady-state annual average is assumed. One may wish to consider using more sophisticated foodchain pathway models such as PATHWAY (Whicker et al., 1990) for site-specific assessments. Dynamic models such as PATHWAY simulate the time-dependence of root uptake and soil adhesion during the course of plant growth and harvest and more readily allow for inclusion of site-specific conditions. Models such as PATHWAY also allow one to treat temporal variations in animal feed practices, pasture season, etc., in much more detail than do the screening models described in this Report.

5.2 Discussion of Model Parameters Each of the important parameters needed to estimate E(T)from ingestion of radionuclides, including the recommended values for both screening and site-specific assessments, uncertainties and variability, is discussed in the following paragraphs.

5.2.1 tiadionuclide Concentration in the Soil To calculate the dose from Equations 5.1 through 5.3, one needs to know the relevant nuclide concentrations in the soil. For sitespecific assessments one should use a suitably spatially averaged nuclide concentration a t the time of interest, i.e., the concentration in the surface soil during above-ground vegetation exposure for contamination via resuspension or the concentration in the root zone at harvest for contamination via root uptake. The decay and ingrowth of progeny during the growing period should also be considered. The NRC (Kennedy and Strenge, 1993) has provided guidance on how to calculate the concentration of each nuclide and its progeny in growing vegetation from measurements in soil. For the purposes of the screening calculations described in this Report, it is assumed that the average concentration of each radionuclide of concern in the top 5 cm of soil has been measured a t some time to (see Section 6). The screening calculations assume a mean concentration of 1 Bq kg-' over the entire root zone. As discussed in Sections 3 and 4 for the external and inhalation dose pathways,

5.2 DISCUSSION OF MODEL PARAMETERS

1

95

the average annual concentration in soil of that nuclide and each of its progeny for each year over the succeeding 1,000 y period were then calculated from the decay data given in Appendix B. Screening doses were computed for each radionuclide using Equations 5.1 through 5.3, and the highest series sum used for the screening criteria reported in Section 2. Thus, for calculating ingestion screening doses from Equations 5.1 and 5.2, the average concentration over the entire year of highest dose was used for the concentration in soil a t harvest, under the assumption that crops are grown and ingested throughout the year. The average concentration in the top 5 cm is assumed to have been measured according to the guidance given in Section 6, resulting in an uncertainty in the mean concentration within certain bounds (i.e., with CV 5 0.5).

5.2.2 Human Diets

In order to calculate the average annual dose from ingestion, it is necessary to determine the average quantity of vegetables, fruits and grains grown in the contaminated soil that were consumed during the year. One must also determine the quantities of meat and milk consumed annually from animals that ingested fodder or other vegetation grown on the site. Human diets vary considerably, both with locale as well as with age. For site-specific studies, it may be possible to determine the specific diet of individuals who may be impacted. For these screening calculations, however, generic diets were used. 5.2.2.1 Variability in Human Diet. Reported per caput annual consumptions of major food groups as a function of age for the United States are listed in Table 5.1. Also listed in the table are some estimates of the annual consumption of each food group as a function of age. Site-specific dose assessments may require a survey of the exposed population to determine their dietary habits and the fraction of their diet that is derived from the contaminated site. It is also important to consider the potential exposure to populations who do not live on or near the contaminated site but who may be exposed as a result of ingesting food produced on the site. On average, beef is generally consumed in slightly greater quantities than other types of red meat (about 28 kg y-I versus 22 for pork and 5 y: 100 mg d-I

M

E

?

U

U

Calabrese et al. (1989)

Calabrese et al. (1990)

(m)

Wong (19881, Wong et al. (1988)

Reference

Urban

Rural or

A

C

C

I

I/G

IIG

Study Summary/ Analysis Methods

6 adults studied as part of a validation program in a child soil ingestion study (Calabrease et al., 1989)

Fecal analysis for soil trace elements during 8 d period for 65 children, 1-4 y of age in greater Amherst, MA area; one soil pica child identified

- 50 mg d-'

non-pica, 5-8 g d-I for pica child

9 to 40 mg d-I for

M

M

M

Basis for Ingestion-Rate Estimates Ingestion-Rate Estimates (A, E, IM, M. LRY

25% > 100 mg d-I Observations made for 4 14% > 200 mg d-I months in 2 government 4% > 300 mg d - L C supervised homes in Jamaica for children awaiting foster home placement, analysis by fecal sampling for silica

Geophagia or Child Inadvertent or Adult Ingestion (CIA) (GA)

geophagia), ordered by publication date (Simon, 1998). (continued)

TABLE 5.18-Selected literature reporting quantitative estimates of soil ingestion-rate (inadvertent and

u

m

el

2 m

2

8

z

5: m

VI

ZI

P

E3

R/U

U

van Wijnen et al. (1990)

Calabrese et at. (1991) C

C

G

0 to 90 mg d-I (geometric mean) for day care groups, 30 to 200 mg d-' for camping groups

10 g d-I for all age groups (noted as unrealistic for 3 mo old infants)

200 mg d-I for children 100 mg d-' for adult

39 mg d-I (Al) 82 mg d-I (Si) 245 mg d-I (Ti)

Fecal analysis for soil trace 10 to 13 g d-I during elements for one child (3.5 2nd of 2-week y, female) of 64 (see observation period Calabrese, 1989) who displayed pica during 2nd of two week observation period

Fecal analysis for titanium of two different groups in the Netherlands during summer compared to hospitalized children

I

C/A

R

Haywood and Smith (1990)

Based on observation of aboriginal population and estimates of consumption of soil contaminated water and other sources

Regulatory guidance values

I

CIA

R/U

EPA (1989a; 1989b; 1989c; 1990aj

I/G

Fecal analysis for soil trace elements of 104 normal school children randomly selected, 2-7 y of age in S.E. Washington state

I

C

R/U

Davis et aE. (1990)

M

M

E, A

LR, A, E

M

U

Finley and Paustenbach (1994)

C/A

C

U

de Silva (1994)

I

I

I/G

C/A

I

C

R/U

U

(R/U)

ASTDR (1992)

Thompson and Burmaster (1991)

Reference

Rural or Urban Study Summary1 Analysis Methods

Fitted child data of Calabrese and Stanek (1992) to parametric (lognormal) distribution

Blood levels of lead from Barltrop (1979) and Barltrop et al. (1975) used to determine ingestion-rate of lead contaminated soil

Regulatory guidance values

Re-analysis of data of Binder et al. (1986) using actual stool weights, fitted parametric (lognormal) distribution

Geophagia or Child Inadvertent or Adult Ingestion (CIA) (GO)

LR

LR, A, E 50-100 mg d-I for non-geophagic child, 50 mg d-' for adult, 5-10 g d-I for geophagic child

Arithmetic mean value of 16 mg d-', 90 mg d-' S.D.

LR

59 mg d-' (geometric mean), 91 mg d-I (arithmetic mean), 126 mg d-I (S.DJd

Basis for Ingestion-Rate Estimates Ingestion-Rate Estimates (A, E, IM,M, LRY

geophagia), ordered by publication date (Simon, 1998). (continued)

TABLE5.18-Selected literature reporting quantitative estimates of soil ingestion-rate (inadvertent and

2

i4

i2

m

9 $

g

2

$ m

?'

\

Q, to

+'

C/A

U

U

Sheppard (1995)

Stanek and Calabrese (1995)

Assumed for a case study of 0-1.5 y: 0.1-10 mg d-I A TCDD risk assessment at 1.5-5 y: 9-50 mg d-I 1 x loM6risk 6-12 y: 5-50 m g d-' 13-30 y 0.1-50 rng d-I 0.1-30 mg d-I outdoors A 0.1-10 mg d-I indoors I Back-calculated from median = 84.5 mg d-I IM measurements of urinary mean = 261 mg d-I arsenic using EPA ingestion S.D. = 625 mg d-' and inhalation models 95th% = 986 mg d-I I/G Recommended values, all pica child: 500 mg d-I LR geometric mean values (GSD = 12); 2.5 y: 50 mg d-I 6 y: 20 mg d-' adult gardener: 20 mg d-I adult (indoors): 0.4 mg d-I I/G Revision of estimates 75 m g d-I (median M presented in Calabrese et al. daily)" (1989)

"Basisfor ingestion-rateestimates: Assumption (A), estimate (E),indirect measurements (IM),direct measurements 0, literature review (LR). bDerivedfiom lead ingestion-rates reported by Wedeen d al. (1978). 'Young children, 470 ? 370 m g d-I ; 10.5 percent of young children > 1 g d-' , 21 percent > 1 g d-I on at least one occasion, intakerates for single time geophagia incidences of 3.7 to 61 g d-'. dEquivalentto a geometric mean of 59 mg d-' and GSD of 2.81. "nge of median soil ingestion of 64 subjects over 365 d: 1 to 103 mg d-'. Range of average daily soil ingestion for 63 subjects over 365 d: 1 to 2,268mg d-l. Median of the daily average soil ingestion for 64 subjects: 75 mg d-'. Range of upper 95 percent soil ingestion estimates: 1 to 5,263rng d-'. Median upper 95 percent soil ingestion estimate of 64 subjects over 365 d: 252 mg d-I.

C

C

U

Lee and Kissel (1995)

Construction site

Residential

2

+

1

sr

m

Z

0

=!

rn

2

4

gC1

g

E

128

/

5. DOSE FROM INGESTED RADIONUCLIDES

TABLE 5.19 -Soil ingestion rates and number of days of exposure (TIused for screening dose calculations. Land-Use Scenario

Adult

Child

T ( d Y-9

Range

Agricultural Heavily vegetated pasture Sparsely vegetated pasture Heavily vegetated rural Sparsely vegetated rural Suburban Construction, etc.

0.1 0.05 0.1 0.05 0.1 0.05

0.1 0.2 0.1

0.1

-

270 270 270 270 270 270 180

180-360 180-360 180-360 180-360 180-360 180-360 90-270

-

-

% GSD of 4.2 was used for all child dose calculations. A GSD of 3.2 was used to calculate adult doses.

Calabrese and Stanek (1994; 1995) discussed the source of these variations and attributed it to high soillfood ratios, i.e., conditions in which the signal-to-noise ratio for the tracer was excessively high. Errors in specifying the fecal sampling period relative to the ingestion period relative to the ingestion period were also implicated in causing calculational errors. Most of the quantitative estimates reported, other than the previous four studies discussed, have little empirical basis. The few that do [e.g., Linsalata et al. (1986) or Wong et al. (198811, may not be easily generalized except within the similar conditions in the country of origin. The reported intake rate estimates in any of the papers cited here apply either to single individuals or to small groups which were observed for short periods of time. These and other factors contribute to the difficulty in generalizing to population. Some studies indicate unusual exposure and intake conditions. For example, the intake rate estimates by Haywood and Smith (1990; 1992) for the Aboriginal population of the Maralinga district in Australia (up to 10 g d-l) point out the risk to which indigenous people may be subjected by virtue of lifestyle attributes and environment. However, not all rural living conditions will result in extraordinarily high intakes. Linsalata et al. (19861, for example, found relatively modest intakes of 200 mg d-l for a farming community in Brazil, an environment considerably different than the Marlinga Desert. Presumably the degree of vegetative ground cover, a function of the regional climate, as well as life-style, are important factors that influence intake rates. Cultural norms for geophagia, even in the United States, may vary significantly.Vermeer and Frate (1979)investigated geophagia

5.3 DIRECT INGESTION OF SOIL

1

129

in rural Mississippi in 50 households containing 229 individuals. In this study, the individual had to admit to consumption of clay to be considered a practitioner of geophagia. Therefore, the data were believed to represent a minimum of the true incidence. The nutrition survey administered did not uncover any geophagia among adult males or adolescents of either sex but identified 57 percent of the women and 16 percent of the children (ages 1 to 4 y) as geophagic. In particular, 50 percent of the pregnant women practiced geophagia. Average daily consumption of clay was reported to be 50 g; the values were determined as a portion or multiple of the amount held in a cupped hand. Despite beliefs of some authors that geophagia in rural areas or among children with normal mental faculties is diminishing, geophagia has recently been noted in some countries not to be a rare event. In reports by Wong (1988) and Wong et al. (1988), children institutionalized in two homes in Jamaica were studied for geophagous behavior over a period of four months to assess the risk of exposure to geohelminth infection. Wong determined from analysis of fecal levels of dietary silica that in the two homes, 33 and 66 percent of children, respectively, practiced geophagia and that some children ingested up to 10 g soil d-'. However, 20 percent of the children accounted for >60 percent of the soil ingested. Several mentally disturbed children consumed, on occasion, up to 60 g soil d-'. In the study by Wong, 21 percent of the children with normal mental capabilities, with an average age of 3.1 y, displayed geophagia occasionally with ingestion rates >1 g d-' and in 10.5 percent of all observations of 24 children, ingestion was > 1 g d-'. Calabrese and Stanek (1993) reviewed the dissertation by Wong (1988) and concluded that the high prevalence of geophagia among these normal children challenges previous assumptions that geophagia is always a rare condition. The literature quoted in Table 5.18 has derived intake rates for specific groups of study by a variety of methods. The weakest estimates are those produced by assumption, e.g., guessing or assuming the number of times children might put candy which has collected surface dirt or dirty hands in their mouth. These studies would include Gelfandetal. (1975),Wedeenet al. (1978),Vermeer and Frate (1979), NASINRC (1980),Kimbrough et al. (1984), and Haywood and Smith (1990; 1992). Some literature has quoted parametric distributions based on an assumption in which the spread of the distribution represents lack of knowledge uncertainly, eg., Martin and Bloom (1976), Rogers (1975), and Healy (1977). Other literature has quoted intake rates based on measurements of fecal and urinary excretion of an environmental contaminant followed by a back calculation to

130

/

5. DOSE FROM INGESTED RADIONUCLIDES

determine the likely rate of soil ingestion. Because these estimates were not produced by planned intake studies, these estimates are referred to as "indirect estimates." These studies would include Lepow et al. (1974), Day et a.!, (1975), Fisher et al. (19811, and de Silva (1994). The most reliable estimates of soil intake are from planned intake studies which have inferred their findings from the quantity of naturally occurring trace elements measured in feces (Binder et at., 1986; Calabrese et al., 1989; Clausing et al., 1987; Davis et al., 1990; Linsalata et al., 1986; van Wijnen et al., 1990) under conditions designed to determine reliable estimates. These studies are termed "direct estimates"in Table 5.18. The main distinction of these studies from "indirect measurements" is in their planning and design. The more careful of these studies subtract the amount of trace elements assumed to have been ingested from non-soil sources, e.g. food, toothpaste, etc. Finally, some reports have simply reviewed previous literature, e.g., Finley et al. (1994) and Sheppard (1995). The uncertainty of even well planned studies is emphasized by the recent report of Stanek and Calabrese (1995) in which they revised their previous estimates of ingestion rate from their 1989 study. Their new estimates were acknowledged to be a striking departure from their previous recommendations as well as EPA recommended default values. In particular, their new estimate of the upper 95 percent of the distribution of intake rates over a year's time was 1,750 mg d-' compared to 200 mg d-' recommended by EPA, a value that is generally interpreted to approximate the upper 95 percent of the distribution for children (Stanek and Calabrese, 1995). The recent analysis of Stanek and Calabrese also determined that 10 percent of the subjects might ingest 1.2 g d-I on average and that 33 percent of the children would ingest >10 g of soil on 1 to 2 d y-' and 16 percent would ingest >1 g of soil on 35 to 40 d y-I. The estimated median of the 64 subjects averaged over 365 d was 75 mg d-I and the range of average daily soil intake rates varied from 1 to 2,268 mg d-I. Most of the publications cited in Table 5.18 refer to populations in the United States; the exceptions reported findings in Brazil (Linsalata et al., 1986), Jamaica (Wong et al., 19881, Australia (Haywood and Smith, 1990; 1992),and the Netherlands (van Wijnen et al., 1990). A few of the papers include nonwhite populations, e.g., Black (see Gelfand et al., 1975; Wedeen et al., 1978; Vermeer and Frate, 1979), Native American (see Fisher et al., 1981), Jamaican (see Wong et al., 1988), and Aboriginal (see Haywood and Smith, 1990; 1992).Those studies noted in Table 5.18 which have empirical

5.3 DIRECT INGESTION OF SOIL

1

131

measurements from children are probably the most reliable. Intakes of soil by adults is much more controversial among reporting authors. Soil ingestion among adults in western societies is likely to be mainly a function of occupation. These studies cited in Table 5.18 do not address the special situation of the pica child or adult (e.g., pregnant women) who might intentionally eat soil. Ingestion rates for specially defined critical interest groups andlor non-United States' populations must be examined carefully and estimated with sufficient knowledge of the living conditions and cultural attitudes of the population of interest.

5.3.3 Recommended Ingestion Rates Some recommendations have been offered in the literature and by United States' regulatory agencies for default soil ingestion rates [for example, ATSDR (1992) Exhibit D.3; Kimbrough et al. (1984), EPA (1989a; 1989b; 1989c) Exhibit 6-14; EPA (1990a), Calabrese et al. (1992), Calabrese and Stanek (19941, Sheppard (1995)l. For example, ATSDR (1992) provides soil ingestion values for CERCLA (Federal Comprehensive Environmental Response, Compensation and Liability Act) and RCRA (Federal Resource Conservation and Recovery Act) mandated assessments. The Agency for Toxic Substance and Disease Registry (ATSDR) Public Health Assessment Guidance Manual (ATSDR, 1992) gives values of 50 to 100 mg d-I for non-pica children, 50 mg d-' for adults (Calabrese et al. 1990) and states that 5 to 10 g d-I are possible intakes of children with pica behavior. Their primary sources of data were Calabrese et al. (19891, Davis et al. (1990) and Calabrese et al. (1990). In a series of risk assessment guidance publications, EPA (1989a; 1989b; 1 9 8 9 ~recommends ) the use of similar values for soil ingestion: 200 mg d-' for children and 100 mg d-' for adults. Calabrese et al. (1992) and Calabrese and Stanek (1994), in later publications, made the following recommendations which were intended to be conservative, upper-bound estimates: (1)assume that two percent of children aged 1 to 6 y exhibit soil pica of 1g d-', (2) assume that 0.2 percent of children ingest 10 g d-' from ages 1 through 6 y. They also provided estimates intended to be more realistic: (1)assume one percent of children exhibit soil pica of 1 g d-I for 4 d week-' and 500 mg d - ' for 3 d week-' for 4 y, (2) assume that 0.2 percent of children ingest 10 g d-' for 3 d week-' and 200 mg d-' for 4 d week-' for 4 y. The recent revisions (Stanek and Calabrese, 1995)of the Calabrese ingestion r a t e s emphasize t h e need to have wide uncertainty

132

/

5. DOSE FROM INGESTED W I O N U C L I D E S

distributions that allow for high values to occur with finite probability. The recommendations here take into account the near universal values of 50 to 100 mg d-I as a n estimate of central tendency. Furthermore, a lognormal distribution has been assigned to represent the range of possible alternative values with a GSD of 4.2 for children and 3.2 for adults. This degree of uncertainty results in an upper 95th percentile equal to 100 mg d-I x (4.2)2 or 1,764 mg d-' for children, a value in agreement with the Stanek and Calabrese (1995) revised estimates. NCRP recommendations for intake rates for the various scenarios are presented in Table 5.19. In light of the sparse empirical data for many scenarios, Table 5.19 is intended to recommend a wide range of possible alternative values. The ranges may be assumed to cover approximately 95 percent ( + 2 S.D.) of the values likely to occur in a typical United States' population. The purpose of these suggested ranges is for use in assessments which fit the scenario description. It should be understood that the values and ranges presented in Table 5.19 represent a subjective determination of the central tendency and spread of numerous literature citations rather than a statistical analysis of available data.

5.3.4

Use of Soil Ingestion-Rate Data in Screening Calculations

The screening criteria in this Report are defined for critical population subgroups that are composed of individuals whose occupations, recreational activities or lifestyles would likely bring them into contact with soil contaminated with a particular radionuclide. These assumptions take into account exposure situations for people with life habits that are likely to be substantially different from many typical residents of the United States. The definition of critical groups, in relation to soil ingestion, requires careful consideration. Children, because of smaller body mass may be the obvious choice; however, in some circumstances, only adults may be exposed (eg., at construction and agricultural sites). In general, the (unidentified) child who practices geophagia is a t the highest risk. However, whether such individuals are numerous enough to consider in setting guidelines for acceptable soil contamination levels is debatable. The degree of realism or conservatism desired, regardless of the subgroup of interest, is another area requiring consideration. To some extent, this problem is handled in the guidance presented in this Report by a detailed uncertainty analysis that attempts to differentiate between true stochastic variability (i.e., variation

5.3 DIRECT INGESTION OF SOIL

1

133

among individuals) and uncertainty (e.g., the relevance of a parameter value to a particular critical interest group). Finally, for population exposure considerations, it is difficult to know what is the appropriate proportion of the population that is a t risk from soil ingestion. However, just recognizing that there is a segment of the population that practices pica.geophagia is the first step to defining a heretofore neglected critical interest group. Similarly, recognizing that soil ingestion occurs inadvertently is a sufficient argument for including soil ingestion in screening calculations of radiation risk to populations living on, or near, contaminated lands.

5.3.5 Calculation of Screening Doses To calculate the annual E(r)(Sv y-I) in terms of a unit concentration of a radionuclide in soil: where:

CWil

I,d T Dfw OF

average concentration in top 5 cm of soil (Bq kg-') average ingestion rate of soil during the exposure period (kg d-'1 = exposure duration (d y-I) = ingestion dose factor (Sv Bq-I) defined earlier = occupational exposure modification factor = =

In the analysis presented in this Report, a number of land-use scenarios have been defined which specify whether the land is rural or urban and whether the vegetation is heavy or sparse. For each of these land-use scenarios, specific values for the parameters IBOil and T in Equation 5.4 have been chosen (Table 5.19). The same ingestion dose factors used earlier for ingestion of foodstuffs are also used for ingestion of soil. (For sites where children are likely to be the most exposed population, the CIA correction factors described earlier were applied.) The concentration of the nuclide in the ingested soil is assumed to be equal to the mean measured over the top 5 cm. Even if the concentration in the soil actually ingested is higher than the average, the bioavailability of many radionuclides (transfer from gut to blood) is likely to be much lower than for the same nuclide incorporated in vegetables, meat or milk, and thus the calculated E from soil ingestion are still likely to be conservative. A site-specific dose assessment where the soil ingestion pathway is significant should consider the actual distribution of the nuclide with depth in

134

/

5. DOSE FROM INGESTED RADIONUCLIDES

the soil and whether dose factors based on lower gut uptake factors than used in the screening model might be more appropriate. For the screening calculations in this Report, the use of median parameter values is based on soil ingestion by normal adults or children since pica children make up < 1 percent of all children and