ML20248F225

Forwards Detailed Comments on Draft Std N13.31, Assessment of Radiation Doses from Plutonium & Americium from Soil

ML20248F225
Person / Time
Issue date:	06/01/1998
From:	Cool D NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:	Johnson N HEALTH PHYSICS SOCIETY
References
NUDOCS 9806040159
Download: ML20248F225 (9)

Text

. _ _ . _ _ _ _ _ _ _ _ - _ _ _ _ - - _ _ --

June 1,1998

~

2 Ms. Nancy Johnson Health Physics Society 1313 Dolley Madison Boulevard l

Suite 402 McLean, VA 22101

Dear Ms. Johnson:

As promised in my letter to you dated May 6,1998, I have enclosed detailed comments on the draft standard N13.31, " Assessment of Radiation Doses from Plutonium and Arnericium from Soil." As you will gather from the number and extent of the subjects addressed in the comments, NRC views this draft as a very important document that could have a significant impact on our licensees' operations. I would also like to point out that the staff members who prepared these comments are considered experts in this field, and I believe would be able to contribute very significantly to the ultimate high quality of the completed standard if one or more of them were to take part in the continuing efforts of this standard's working group, I therefore reiterate the request I made in my last letter to you that the working group consider inviting one or more of my staff to participate in future work on the standard, either as full members or as technical advisors.

Please call me at 301-415-7179 if you wish to discuss this matter further, and I look forward to a favorable reply to my request. Thank you.

Sincerely, (orig. signed by)

Donald A. Cool, Director Division of Industrial and Medical Nuclear Safety Office of Nuclear Material Safety and Safeguards

Enclosure:

As stated DISTRIBUTION:

\

IMNS C/F 5 NRC File Center, NMSS R/F j SSherbini fU Release to the PDR: /~ Yes No DOCUMENT NAME: G:\ ANSI.SS To reedvs a copy of this document, Indicate in the boa: "C" = Copy without attachment / enclosure J[y//ffC&

"E" = Copy with attachment /arelosure *N" = No copy OFFICE IMNS ce l lMNS ,. C IMNSg-l l l NAME SSherbini:c'e" 3 Moore *de,,, DCooV DATE 5/IV98 5L1/98 5/d98 OFFICIAL RECORD COPY 8 /;/d@33 g j a "g :x d 9806040159 980601 -

EPSANSI PDR ORG w<

PDR spw -

.g9)$_3_g gr y tm u

)

Review Comments On HPS N13."1 Working Group Draft Standard On Assessment Of Radiation Doses Fro,n Plutonium And Americium From Soil General Comments:

1. SurfacetSubsurface Source-Terms: The purpose of the draft ANSI standard (hereafter will be referred to the standard); as stated in Sect;on 2.0, page 6; is "to provide an ANSI-approved method for assessing the potential for radiologicalimpact of americium and plutonium from soil, both sudace and subsurface. on surrounding population." The standard did not provide definitions of surface and subsurface soils. In addition, the standard did not use an intemally consistent source-term for the screening analysis. The standard also lacked meaningful guidance on subsurface soil sampling and how the source-term area could affect dose assessment results. For additional and detailed comments see specific comment #1. We suggest the standard be revised to exhibit internal consistency and to provide adequate guidance on surface / subsurface source-terms.

2. Cric. cal Group: The standard did not establish a dose exposure scenario based on a pre-denned entical arouo. However, the stLadwd described performing dose analysis to an individual of conggr} {the member of the public most likely to receive the highest radiation dose) NRC's regulation for license termination (10 CFR 20 Subpart E) requires dose assessment for the averace member of the critical arouo. We would note that, NRC's regulation focuses on a group of individuals reasonably expected to receive the greatest exposure whereas the current standard focuses on a single individual expected to receive the highest radiation dose. As a recommendation, we request that ANSI standard adopts use of the " average member of the critical group" in the proposed Pu/Am dose assessments.

3. Occupancy Factor / Exposure Time: The standard selected an occupancy factor of 1 and an exposure time of 365 d/y. Thus, the ANSI standard assumed an individuallives on the contaminated soil the entire year for an entire lifetime. This appears to be inconsistent with the typical family farm seenario where the critical group mernber is assumed to spend a '* action of time onsite performing miscellaneous activities on the contaminated soil, a fraction of time living indoors (using a shielding exposure factor),

and other activities performed off the contaminated area within the remaining avalisble time fractions. In many instances, the assumptions and occupancy factors adopted in the ANSI standard sppears to be more conservative and inconsistent with National and intemational practices (e.g., NRC's technical basis for translating contamination to annual Total Effective Dose Equivalent, TEDE, (NUREG/CR-5512); NCRP,1996; and ICRP Publ. 43,1985). For example, NCRP (NCRP,1996) assumes that the maximally exposed individual typically occupies the contaminated area for only 25% of the year.

4. Selection of Default Values: The ANSI standard presented a list of default values in Table 13 without providing a rationale for relection of such values. For example, the standard selected dietary parameters assuming that a major portion of human diet is grown on the contaminated site although ti'e contaminated area was not exp8icitly ATTACHlWENT

~. .

2 defined. The fish and poultry food consumption parameters were not provided. Leafy vegetables were not distinguished as an independent food category. The standard did not discuss how conservative these default values are or the basis for their selection.

NRC staff recommends that the standard adopts a realistic approach in selection of default values and provide a rationale for selection of these default screening values.

5. Elimination of Exposure Pathways: The standard adopted an approach to eliminate exposure pathways contributing less than 10% of the total TEDE. For example, the standard stated in Section 5, page 8, that " measurements sha// be performed for all exposure pathways that can contribute more than 10% of the total TEDE." In addition, the standard contemplated that dose evaluation shall be conducted only for credible exposure pathways that could result in an associated mortality risk that exceeds 104/y.

The standard presumed a prior knowledge of the percentage of dose contribution or knowledge of the exact risk of eacn individual pathway. In actuality, contribution of each individual pathway would only be known after completion of the dose impact analysis. In addition, for a multiple pathway scenario, assuming a potential contnoution of 7 to 8%

from each of 3 to 4 insignificant pathways; the collective dose contribution from the so called . significant pathways" may add tc mcie than one third of the tota! TEDE.

Therefore, elimination of pathways should be based on physicallimitations at a given site regardless of the percentage of contribution of the specific pathway to the total dose.

6. Resuspension Factor vs. Mass Loading Factor: The ANSI standard used the resuspension factor rather than the mass loading factor to calculate radionuclides concentration and subsequently estimate dose impact from the dust inhalation pathways. The standard did not present an argument for its preferential use of the resuspension facter. NRC believes that the standard should use the mass loading factor rather than resuspension factor because the mass loading is a time-independent parameter which can be determined in the field for the average year. Resuspension factor however, is variable with time and depends largely on the thickness of the resuspendable layer. Therefore, we recommend that the standard uses the mass foading factor as an alternate option to use of the resuspension factor.

7. Derivation of Radionuclides Concentration for the ingestion Pathway: The ANSI standard used Equation 4 ( page 9) for calculation of doses from the ingestion pathway.

The term Aywas used to denote concentration of radionuclides iin environmental media j. Thus, the standard assumed that the dose assessor would use readily determined values for radionuclides concentrations in food, milk, meat, and crops. While that may be optional, it is not always practical to limit analysis to existing conditions.

Frequently, radionuclides concentrations in ingested materials are derived from data on bioaccumulation and soil / plant transfer factors of concern. The standard did not discuss or reference these factors. For a proper analysis of ingestion pathways, the standard should include methods that do not depend on direct radionuclides measurements. The standard may use an independent term in Equation 4 to express derivation of radionuclides concentration in the ingested food media using partiai pathway bioaccumulation and transfer factors. In addition, the standard should provide a

reference list of these factors to enable users to determine radionuclides concentration in food media based on soil concentration and nartial pathway analysis.

8. Analysis of Partial Pathway Transfer Factors: The standard did not analyze the following partial pathway transfer factors ;PPTF's) for food ingestion: (a) soil-plant-human pathway, (b) soil-stored hay-animal-human pathway, (c) soil-stored grain-animal-human pathway, (d) soil-forage feed-animal-human pathway, (e) transfer factors for irrigation water to plants, plants to human, and plants to animals to human. These PPTF's could have a significant impact on the concentration of radionuclides in the ingested food media.

9. Inconsistencies Within the Standard: The standard contains some internal inconsistencies due to use of multiple sources for derivation of doses associated with various pathways. For example, the standard derived the screening effective dose equivalent factor for the groundwater ingestion pathway using radionuclides soil concentration, q,. multiplied by the groundwater screening factor (SF,,). The standard used NCRP screening dose factors (NCRP, Report No. 1231,19%) to e>. press the value of SF, The NCR ' dose f eters were derived based on assumptions and default values different from assumptions made in the current ANSI standard (see specific comment #2). As will be discussed shortly (specific comment

1. 1), the source-term assumption in the standard is inconsistent with NCRP model assumptions. Water and food parameters assumed in the ANSI standard are also inconsistent with water and food consumptions parameters adopted in the NCRP report without a discussion of these inconsistencies. In brief, the current ANSI standard is J intemally inconsistent because assumptions and defaults used in the standard are different from those assumptions and defaults used in deriving such dose factors.

10. Limitations of the Screening Analysis: The standard did not address the issue of screening analysis limitations due to screening code /model assumptions. The standard should caution users that screening analysis may be inappropriate for complex situations with subsurface soil and/or groundwater contamination. The simplified source term and transport models used in screening, may preclude its meaninpful use for complex situations. For example, for sites with deep soil contamination and with very large contaminated areas the default assumptions of 30 cm soil thickness and a default area of 100 m2 may be inappropriate. Other possible limitations of screening analysis could be associated with situations where the sites are characterized by substantial surface water and/or groundwater contamination. We recommend that the  ;

standard outlines the limitations of the screening analysis and the scope of its application.

11. Incompleteness of the Full Dose Assessment (e.g., Site-Specific Assessment)

Approach: The standard is incomplete regarding approaches and methods to be followed in conducting Pu/Am full dose assessments and pathway analysis. For example, it was stated, on page 12, "use appropriate model parameter values in a l l

suitable dose assessment model to obtain a best estimate prediction of the TEDE."

The standard described briefly; in Section 6.2, page 13; selection of exposure evaluation method. The standard stated that "the actual method of assessing the dosc/ risk from 1

l elevated plutonium and amencium concentrations in soilis not specified." Further. the standard stated that "Other pathways, such as ingestion of fish and other aquatic animals, shall also be considered if present." Thus, the current standard appears not to contain a structured approach for dose / risk assessment methodology.

12. Compliance With the Dose Criteria - MARSSIM Approach: The dose / risk assessment of the ANSI standard will be used for remedial actions. In this context, the standard is essentially based on determination of the "mean Pu/Am concentration in soil." This approach is generally acceptable; however, compliance with the dose criteria is usually determined through the following steps: (a) establishment of a derived concentration guideline level (DCGL), equivalent to the dose criteria (e.g., 25 mrem /y),

and (b) demonstration of compliance with the dose-based regulations in accordance with the " Multi-Agency Radiation Survey and Site Investigation Manual" ( MARSSIM) methodology as defined in Sections 2. 5 of MARSSIM manual. In general, MARSSIM methodolog"';pically requires that the overall site be divided into survey units and a i

statistical test, Wilcoxon R . ank Sum (WRS), is applied to examine if soil concentrations indeed meet the DCGL. Therefore, the standerd should reference MARSSIM and describe aspects involving derivation ot DCGL s, MARSSIM survey methodology, and the statistical test typically used for demonstration of compliance with the DCGL.

13. Uncertainty of the Dose Estimate: The standard defined the uncertainty of,the dose estimate to be the difference between the 95th and the 5th percentiles of the dose distnbution. However, the standard did not specify a certain percentile (e.g.,50th) to demonstrate compliance with the dose criteria or with the data quality objectives (DQO's). In addition, for performing the uncertainty analysis, the standard requested the use of either site specific values or the default values given in Table 1 of the I standard. We would recommend the standard specifies the ranges of defaults or site specific parameters that can be used in performing the uncertainty / sensitivity analysis.

14. Percentile Distribution Level For the Remedial Action: The standard employed the 95th percentile of the distribution of TEDE estimates as the soil remedial action levelc We believe that this percentile action level doesn't reoresent the best dose estimate but s highly conservative dose estimate. This level also appears to be inconsistent with the above approach of calculating the dose uncertainty using the difference between the 95th and the 5th percentile levels.

15. Integration of Field Survey and Laboratory Measurements: The standard did not address the issue of integrating field survey results with laboratory analytical results for derivation of soil concentration. Field measurements are typically characterized by having lower sensitivity, accuracy, and precision than laboratory measurements. In addition, field samples typically represent much larger surface areas or soil volumes than laboratory discrete sample. The standard should discuss approaches to coupling field survey data with laboratory data 'or the purposes of: (a) establishing, for the dose models, a proper source term of soil contamination, (b) deriving an accurate DCGL for the dose impact analysis, and (c) for demonstration of compliance with cleanup or decommissioning criteria.

16. Laboratory Measurements and Quality Assurance / Quality Control (QA/QC): The standard presented, in Appendix C. a bnef guidance on laboratory measurements and QA/QC. The information presented in Appendix C is incomplete and should be supplemented by referencing established guidance documents. In this context, the ANSI standard workgroup should benefit trom referencing MARSSIM and also benefit from mutual discussions with tr " Multi-Agency Radiation Laboratory Protocols" (MARLAP) workgroup regarding the overall issue of radioanalytical methods, measurements, and QA/QC.

Specific Comments:

1. Specific Comments On the Source-Term issue:

(a) For the screening analyses, the standard contained inconsistencies in the source-term for derivation of the dose conversion factors. For example, the depth of contamination (i.e., the source-term thickness) to the upper 30 cm of soil. The standard presented e: ternal radiation dose conversion factors (DCF,,)

in Table 1b and referred u,ronevumy tc " EPA 1988"(i.e., Fedaral Guidance Report No.11) as the source reference for these dose factors. The DCF , were expressed in Sv/yr/Bq/m2 units indicating an assumption of a thin layer source-term. For calculation of the effective dose equivalent from inhalation, the standard assumed a resuspendable soil layer cf 1 cm thick. For the groundwater ingestion pathway (i.e., drinking water ingestion) the standard used NCRP dose conversion factors (NCRP 1996) that were derived based on a contaminated soil thickness of 15 cm. The standard did not explain the rationale for selecting dose factors that were based on different source-terms. In addition, the standard did not discuss possible internal inconsistencies due to variations in the source term used in the screening analysis. The ANSI standard workgroup may benefit from review of NRC's screening dose impact methodology (NUREG\CR-5512) and NCRP and IAEA disposal screening models (NCRP 1996, Report No.1231, and IAF A,1993).

(b) For the full dose assessmer's (e.g., site-specific analyses), the standard did not define the depth and distribution of contamination and did not address the issue of subsurface sampling frequency. Thus, the standard lacked a meaningful guidance on how to establish an appropriate source-term for conducting the site-specific dose impact analysis.

(c) The standard did not use MARSSIM guidance to address the follow;ng source-term issues: (a) sampling design, (b) the number of measurements to estimate the mean of an unstratified/ stratified site, and (c) selection of samples, sample locations and grids for soil sampling and survey.

(d) The standard did not address MARSSIM guidance regarding the dependence of type of sampling on the class of survey area. In addition, the guidance contemplated that the purpose of subsurface sampling is to define the depth of roet zone. The standard should recognize that, in addition to defining the depth

n of root zone, subsurface sampling is also used to define source-term available to ground transport.

(e) The standard used the average soil concentration as the enteria for the dose impacts and subsequently releasing the contaminated area. The standard lacked any guidance on elevated concentrations and did not address the hot spot issue.

(f) Dose modeling is highly dependent on the source distribution at the site. The dose factors used in this analysis are dependent on a fixed area (e.g.,100 m2 ),

The paper did not address the issue of dose variation with the assumed default source-term area.

2. The NCRP's groundwater screening factors (SF.,,) were derived using a model for groundwater contamination that is based on a leach rate factor (e.g., release rate) which is dependent on infiltration rate, thickness of the layer of buried soil, retardation coe'ficient, and on soil porosity. The NORP employed default values of infiltration rate, poros.ty, and waste thickness of 0.13 m/y,0.30, and 0.5 m respectively. The default retardation coefficients used in the NCRP model were those of NRC's NUREG/CR-5512. The NCRP's source-term assumes a buried waste which is spread uniformly over a circular area of 100 m2 , and a uniformly mixed depth of 0.15 m. The NCRP assumed a maximally exposed individual who spends 0.25 of its time (2,000 h/y) at the center of the contaminated circle, and gets all of drinking water 80 Uy and half of its vegetables (100 kg/y) from the site. AH radionuclides released from te waste were assumed to be diluted in a volume of water equivalent to 91 m8 /y. The above NCRP groundwater model assumptions and the associated source-term and critical group exposure values are incompatible with the assumptions and default values selected in the current standard. NRC recommends that the ANSI workgroup review models assumptions and default parameters for all pathways to ensure internal consistency within the standard.

I

3. The standard contemplated that dose evaluation shall be conducted only for credible j exposure pathways that could result in an associated mortality risk that exceeds 104/y i Based on NRC's Statement of Consideration (FR Vol 62, No.139, p 39061, July 21, 1997) for the license termination rule, a dose of 15 to 25 mrem /y is approximately l

equivalent to 2.3 x 10d to 3.8 x 10" lifetime mortality risk. By simple rick / dose  !

conversion, the credible pathway should account for a dose of at least 0.07 mrem /y. In performing dose impact analysis, partial pathways are typically analyzed. It is unclear if l the 0.07 mremly lower bound dose limit is also applicable to partial pathways. Typically, dose assessors derive the partial pathway dose prior to determining the overall specific 3 l pathway dose. Thus, the merit of using the credible pathway concept is unclear and the )

current standard approach may lead to elimination of partial pathways that could result )

d in a significant dose when these partial pathways are considered collectively,

4. The ANSI standard lacked information and guidance on the following important aspects of dose assessment:

(a) The standard did not address dose modeling or other approaches to meet ALARA requirement. }

i j

u _

(b) The standard did not address the issue of dose modeling associated with potential offsite releases. Offsite releases from subsurface soils. dunng the dose evaluation penod (i.e.,1000 y) could be significant for certain radionuclides.

Therefore, the standard should discuss modeling approaches to assess potential radionuclides transport and associated dose unpacts N The standard did not address the issue of background identification and measurements.

(d)

The standard did not address the issue of index of sensitivity in radiological measurements.

(e) Onsite ra dionuclide transport via groundwater or surface water could have significant impact on the site-specific dose analysis. The current standard did not address the issue of radionuclides, transport within the site boundary, for the site-specific analysis. The standard should include approaches and taodels for assessment of radionuclides transports and their implications on the peak doses specifically those associatew with wt .er-dapendent pathways.

(f) The standard did not address the issue of criteria for selection of proper codes /models in the dose impact analysis. The standard should include minimum information on how to select proper codes for site-specific dose analysis. For example, codes verification and validation and compatibility of the models with site-specific conditions are essential criteria for codes /model selection.

5. The standard used the term 'best estimate ~ without describing its exact definition. It is unclear if it refers to the mean or the average value or simply to an unidentified ,

qualitative term. We recommend that the standard provides in the Definition Section 4.0, "the best estimate" definition.

]

6. The standard listed, in Table C.1, typical laboratory parameters used in transport modeling without discussing how these parameters would be used in the Pu/Am dose impact analysis. In addition, the standard indicated that at least 10% of load of samples would be used as blanks, spikes, and duplicates for QA/QC analysis. Typically, the number of QA/QC samples should depend on several factors such as DQO's, total number of analyzed samples, and QA/QC performance record of the laboratory. We recommend that the ANSI standard workgroup (HPS N13.31) look into the information presented in MARSSIM and MARLAP on QA/QC and laboratory measurements. )

7.- The value for soil ingestion was erroneously reported as 36.5 kgly instead of 36.5 gly.

Transfer factors for meat and milk and vegetable /soit uptake ratios were based on the old RESRAD code version 5.0 rather than en the newer versions like 7.0 or higher.

8. The standard used a 90% confidence for estimation of average soil concentration. We rec:mmend that the standard discuss the rationale for selection of this value.

e

< 1

9. The MDA equation (page 18) should use the cackground standW deviation /C, rather than the measurement standard deviation s: In addition we recon, mend that the quantities encompassed by the proportionality constant. K. (e g., de'ection efficiency and probe geometry) be explained.

10. The standard assumed that size fractions <100 to 125 s are the resuspendable fraction.

The paper should discuss the size fraction which is available for inhalation and the dose conversion factors for such size fraction. Typically, for the inhalation pathway analysis, size fractions >20 u are disregarded.

11. The standard indi.:ated, on page 55, that direct measurements of Pu and Am in the local food and water are preferabie to calculating the concentration. Measuring radionuclides concentration in food is a cumbersome process because it is dependent on ' cod type, seasonal variatiors, and achievement of equilibrium. In addition, due to low concentratica af radionuclides in food material, analytical errors could be large and data quality could be uncertain. Subsequently, coitection of reliable data on radionuclides concent ation in food could be too costly. Therefore, the standard should include alternatJ options for indirect determination f radionuclides concentrations in food and water.

12. The standard used in equation D.1 (page 57), a source-to-intake transport factor designated P(x,t). The standard did not explain how this factor will be derived.

13. The standard discussed ICRP Publications No. 30,48,56,67, and 68. The standard also presented a comparison of ICRP ingestion dose coefficients for ICRP 56 and 67.

The standard did not discuss or recommend which ICRP dose factors to use. However, the standard used for inhalation pathways the dose conversion factors of FR No.11, which is based on ICRP 30. and for ground ingestion pathways used NCRP Report No.123 I (NCRP,1996). We recommend that the standard be explicit and consistent in choosing dose coefficients.

14. For the screening analysis, the current ANSI standard relied heavily for assessment radiation doses from Pu/Am on NCRP Report No.123 I (NCRP,1996). The NCrRP Report No.123 I stated on page 3 that " It is emphasized that " doses" estimated by these screening techniques are strictly for comparison with an environmental standard (limiting value) and are not intended to reoresent estimates of actual dose to individuals." We recommend that the standard discuss how the derived TEDE from all pathways is different from the actual dose or risk to individuals.

.