ML19319A293

Comments on U.S. Department of Energy West Valley Demonstration Project Final Study Document: Vitrification Facility Air Emissions During Open-Air Demolition, Measured Vs Predicted, WVDP-579

ML19319A293
Person / Time
Site:	West Valley Demonstration Project
Issue date:	12/04/2019
From:	Amy Snyder Division of Decommissioning, Uranium Recovery and Waste Programs
To:	Bower B US Dept of Energy, West Valley Demonstration Project
Snyder A
References
Download: ML19319A293 (7)

Text

UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON, D.C. 20555-0001 December 4, 2019 Bryan C. Bower, Director West Valley Demonstration Project U.S. Department of Energy 10282 Rock Springs Road West Valley, NY 14171-9799

SUBJECT:

COMMENTS ON U.S. DEPARTMENT OF ENERGY WEST VALLEY DEMONSTRATION PROJECT FINAL STUDY DOCUMENT: VITRIFICATION FACILITY AIR EMISSIONS DURING OPEN-AIR DEMOLITION, MEASURED VS PREDICTED, WVDP-579, Rev. 0 (DOCKET NO. 05000201 (POOM-032))

Dear Mr. Bower:

The U.S. Nuclear Regulatory Commission (NRC) staff has reviewed the subject study, dated September 5, 2019, transmitted by letter dated September 18, 2019 (Agencywide Document Access and Management System [ADAMS] Pkg. Accession No.ML19267A228 Pkg). The September 5, 2019, transmittal contains the test plan and alternate method to demonstrate compliance with the National Emissions Standards for Hazardous Air Pollutants. The NRC understands the U.S. Department of Energy plans to use this method for future West Valley Demonstration Project demolition assessments.

Enclosed are the NRC staffs comments regarding the subject study.

In accordance with Title 10 of the Code of Federal Register Part 2.390 of the NRCs Agency Rules of Practice and Procedure, a copy of this letter will be available electronically for public inspection in the NRC Public Document Room or from the Publicly Available Records component of NRCs ADAMS. ADAMS is accessible from the NRC Web site at http://www.nrc.gov/reading-rm/adams.html.

B. Bower 2 If you have any questions or need any additional information regarding our comments, please contact me at 301-415-6822.

Sincerely,

/RA/

Amy M. Snyder, Senior Project Manager Materials Decommissioning Branch Division of Decommissioning, Uranium Recovery and Waste Programs Office of Nuclear Material Safety and Safeguards ENCLOSURE: NRC Staff Comments cc: P. J. Bembia, NYSERDA T. Rice, NYSDEC R. Dansereau, NYSDOH

ML19319A293 *via e-mail OFFICE DUWP DUWP DUWP DUWP DUWP NAME ASnyder CHolston CMcKenney* BWatson* WVonTill*

CRidge for D.Orlando for DATE 11-14-19 11-22-19 12-04-19 11-15-19 12-04-19 OFFICE DUWP NAME ASnyder DATE 12-04-19 NRC STAFF COMMENTS Specific Comments on Vitrification Building Demolition Validation

1. Selection of radionuclides for the validation modeling and sampling approach: The U.S.

Nuclear Regulatory Commission (NRC) staff has technical questions regarding the type of analytical measurements and sampling methods used.

a. What is the basis for the list of radionuclides selected for the AERMOD1 validation study? Please explain why Tc-99 and I-129 were not considered in the validation study (e.g., low risk significance or lack of data on air concentrations for these radionuclides).

b. How are the physical (gas versus particulate) and chemical forms of the constituents considered in the modeling and monitoring? Does using gross measurements and particulate sampling methods limit the usefulness of the sampling data? For example, does the measurement of particulates only leave out significant releases of gaseous radioactive material that should be considered in the model validation?

c. Why is it appropriate to compare a measured weekly average concentration to a source-emissions estimate that occurs in discrete intervals during the week?

2. Validation study methodology: It is unclear to the NRC staff how the study methodology and rationale support significant changes to model parameters.

a. Only two discrete air sampling locations were used in the validation study (see Figure 2). Comparisons of modeled to monitored concentrations could be misleading if the comparisons are focused on a small area that does not provide complete information on the overall goodness of fit of the model to observed concentrations2. Although the validation study has many assumptions and limitations, significant changes were made to the model and model parameters based on the results of the study.

b. The Vitrification Facility and Main Plant Process Building Decommissioning and Demolition Plans discuss the use of low volume air samplers and real time air monitors at the demolition site boundary during demolition, as well as a breathing zone air sampler in the cab of the excavator(s). Were these air samplers and monitors oriented in multiple compass directions surrounding the vitrification facility? Was air sampling data from other sampling locations (other than ANVDEMO1 and ANVDEMO2) evaluated as an additional check on the goodness of fit of the modeled air concentrations to actual measurements and to assess uncertainty in the model given limitations of the validation study?

c. Would it be worthwhile to compare worker exposures based on bioassay monitoring to the exposures based on modeling estimates?

1 The AERMOD air dispersion model was developed by the American Meteorological Society/Environmental Protection Agency Regulatory Model Improvement Committee (AERMIC) Model.

2 Air concentrations could be quite disparate in two discrete locations but be generally consistent considering the overall distribution of activity. Air concentrations could also be very similar (e.g., close to background) in two discrete locations but otherwise be quite different considering higher concentrations in the centerline of the plume.

Enclosure

d. While the Main Plant Process Building is located to the SSE of the Vitrification Facility, Figure 4 shows that the wind is blowing roughly from the direction of air sampler ANVDEMO1 towards the SSE quite frequently. Additionally, low wind conditions near building/structures could also lead to relatively high concentrations close to the source. NRC staff believe it is reasonable that air sampling stations were located to the north of the Vitrification Facility given the wind patterns shown in Figure 4; however, a discussion of modeled versus actual concentrations using additional (more complete) air sampling data around the Vitrification Facility would provide more confidence in the AERMOD modeling results and its ability to estimate plume distributions and assess dose during demolition activities. Plots of the differential between modeled plume distributions and actual data, if available for these other air sampling stations, would be very informative in assessing the predictive performance of the AERMOD air dispersion code.

3. Selection of background concentrations: Table 4 and Figure 5 show that during the baseline period, the ANVDEMO1 and ANVDEMO2 demolition sampler results were consistently less than the 16 sampler results for every compass location for gross alpha (and consistently higher for gross beta). How does the Department of Energy (DOE) ensure that these systematic differences do not negatively affect the model validation study? For example, the AF03 background concentrations were added to AERMOD modeled values for comparison against demolition sampler data which could make it appear that the modeled values were too high for alpha and too low for beta.

4. Uncertainty in air concentrations from hydraulic hammering:

a. Air concentrations at two sampling locations in the validation study were significantly overestimated by AERMOD in weeks that hydraulic hammering demolition activities occurred. The fact that the modeled versus actual concentrations were similar for some types of activities3 and orders of magnitude different for other activities (e.g., hydraulic hammering), suggests there is significant uncertainty in the modeling/validation. Based on the results of this validation study, DOE contractors recommended a two order of magnitude reduction in the Air Release Fraction (ARF) for hydraulic hammering, which would lead to a similar reduction in the source term and dose. To help justify the reduction in source term for hydraulic hammering using the ARF, DOE should comment on the basis for the hydraulic hammering ARF values in DOE-HDBK-3010-94 (DOE, 1994) and explain the differences between the West Valley demolition project data and other studies upon which the ARF values are based.

b. Are there any plans to update the DOE handbook and other documents supporting modeled parameters based upon the West Valley study?

c. What other factors may have played into the large differences between modeled and actual results (e.g., model/parameter uncertainties including inventory, lack of consideration of deposition4, debris pile source term contributions, treatment of low wind conditions, building/structure wake effects)?

Reference:

Airborne Release Fractions/Rates and Respirable Fractions for Nonreactor Nuclear Facilities, DOE-HDBK-30 10-94, US Department of Energy, 1994, Reaffirmed 2013.

3 For shearing, the modeled and measured values were close to background and any added radioactivity from shearing activities appeared to be within the noise. Thus, the modeled and measured concentrations were similar, but it is unclear if the sampling data are useful for validating the AERMOD modeling results.

4 Deposition could lead to higher concentrations closer to the source.

2

5. Uncertainty in air concentrations from hot cutting: Air concentrations for beta emitters at both sampling locations in the validation study were underestimated by AERMOD for hot cutting activities. While DOE attempted to adjust modeling parameters (i.e., physical state factor) to better align the modeled and actual air concentrations, limited data is available to ensure that modeling is reasonably conservative for hot cutting activities. Given the limited data set (two weeks of data at each of the two samplers) and uncertainty associated with the measurements (e.g., only gross measurements of particulate activity on filters collected weekly at two discrete sampling locations is measured), it is unclear that sufficient information is available to adjust model parameters to ensure decommissioning and demolition constraints will be sufficiently protective. Most of the underestimate of gross beta is thought to be associated with relatively high volatility radionuclides such as Cs-137.

a. Was any thought given to comparing Cs-137 (or other radionuclide) concentrations to model results (although the sampling strategy may not be conducive to validating air concentration results for volatile radionuclides in gaseous form)?

b. Are other parameters besides the physical state factor available to adjust the modeling (e.g., Appendix F notes that 1 percent of the inventory is assumed to be vaporized)?

c. What other information (e.g., from other sites and projects) is available to select parameters and simulate radionuclide release from these types of hot cutting activities?

6. Assessment of statistically significant demolition sampling data: Please clarify how the probability density functions (pdfs) for each sampling week during demotion were constructed. For example, how many measurements were used to determine the population variability component of variance when constructing the demolition sampling pdfs found in Appendix D?

7. Rubble Pile vs. Load Out Emissions:

a. How does DOE differentiate between debris pile and load out emissions and other demolition activities occurring at the same time or does DOE assume all emissions come from just one activity?

b. Could DOE comment on the relative risk of various demolition activities?

c. In the case of hydraulic hammering, DOE notes the weeks when only debris pile emissions contributed to the measured concentrations (see Table 24) and independently modified the moisture content parameter to better match data during those weeks.

However, hydraulic hammering, did not occur in isolation and adjustments to hydraulic hammering ARFs to match air concentration data were confounded by the contributions from multiple sources.

8. Estimated Doses:

a. While DOE provided preliminary estimates of dose based on modeling, could DOE provide final estimated worker and member of the public doses from the Vitrification Facility demolition based on measured values?

b. Could DOE comment on the types of activities and radionuclides that are expected to lead to the highest doses?

c. Could DOE provide information on the expected relative risk from demolition of the Vitrification Facility versus the Main Plant Process Building demolition?

3

General Comments

1. Lessons Learned from Hanford Demolition: What lessons learned did DOE incorporate into its demolition plans and procedures, if any, based on the Hanford Plutonium (Pu) finishing plant contamination event? The Hanford Pu contamination event pointed to the need for a good understanding of demolition practices reflected in the modeling and clear documentation of assumptions used in the modeling, as well as consideration of lower likelihood events and pathways. For example, for the Hanford demolition project, uniform demolition rates were assumed in the modeling although actual demolition rates were variable with higher rates of demolition occurring near the end of the project. The Hanford demolition project modeling may have assumed that the demolished building rubble was packaged and disposed in a timely manner while the rubble may have remained on the ground for a few days representing a significant source term. Low frequency meteorological events may lead to under- or over-estimates in public and worker doses (e.g., high winds leading to greater dispersion of particles away from the demolition site and into uncontrolled areas).

It appears that building demolition plans and procedures at West Valley considered building rubble emissions, had constraints on rates of demolition, and considered uncertainty in meteorological conditions through use of 95th percentile concentrations from modeling simulations. Were any changes made to project plans and procedures following the Hanford event to ensure that the modeling is consistent with actual practices and considers potential conditions at the time of the demolition to ensure workers and members of the public are protected?

2. Model Uncertainty: How is uncertainty in modeling parameters, including the material at risk, considered in the modeling to ensure that projected doses are acceptable? For example, the 95th percentile concentration from modeling was calculated based on 5 years of meteorological data to ensure that concentration remained below 0.02 times the Derived Air Concentration (DAC). What other modeling uncertainties (e.g., inventory, deposition, treatment of low wind conditions and building wake effects) were considered or how was uncertainty otherwise managed?

3. Ambient/Offsite Sampling: What were the results of and what analyses did DOE perform with respect to off-site air sampling to evaluate whether air concentrations were indistinguishable from background? Confirm that validation to CAP-88 calculations will not be performed and the reasons for the lack of validation of off-site air concentrations (e.g.,

due to negligible impacts at offsite locations).

Editorial Comments

1. Clarification: What is meant by boiling point of radionuclides. Is this the boiling point of the chemical form of radiological contamination?

2. Apparent cut and paste error: Table 13 has the same values for modeled and predicted concentrations.

4