ML20195C112

Process Safety of Proposed Tank Waste Remediation Sys (Twrs)

ML20195C112
Person / Time
Site:	07003091
Issue date:	05/31/1999
From:	Merritt Baker, Chang L, Murray A NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:
Shared Package
ML20195C109	List: ML20195C102 ML20195C112
References
NUDOCS 9906070084
Download: ML20195C112 (104)

Text

. l PROCESS SAFETY OF PROPOSED TANK WASTE REMEDIATION SYSTEMS fTWRS1 B R MAY,1999 Alexander P. Murray l Merritt N. Baker  !

Lydia W. Chang 1 Makuteswara S. Srinivasan

)

S syB REOy,+ ,

< es

..... -Q l

l Tank Waste Remediation System Section, Special Projects Branch, Division of Fuel Cycle Safety and Safeguards, l Nuclear Material Safety and Safeguards,  ;

U.S. Nuclear Regulatory Commission.

fDO

)

r (This page intentionally blank) 4 9

E

1

o j PROCESS SAFETY OF PROPOSED TANK WASTE REMEDIATION SYSTEMS (TWRS)

TABLE OF CONTENTS jJ Abstract 24 introduction i

ZS TWRS Process Description j M Generic Process 3._2 Contractor Proposed. Conceptual TWRS Process Acoroaches 4_A Potential Chemical Safety issues and Areas of Concern

{

5J Analysis and Consecuences of Potential Chemical Safety issues and Concenis at l TWRS Facilities 5_1 Analysis and Conseauences from Chemical Safety Issues and Concerns at the Generic TWRS Facility 1 5.1.1 Radiochemical Inventories and Larae Tanks 5_1,2 Process Efficacy I

EM Oraanic lon Exchanae Resin and Nitrate Interactions in the Presence of Very Hiah Radiation Fields 5.1.4 Hydroaen and Flammable Gases SM CST Drvina 5.1.6 Oraanic Materials 5.1.7 Radioivsis 5.1.8 Hiah Temperature Operationt 5.1.9 Nonradioactive Chemical Effects 5_2 Analysis and Consecuences from Chemical Safety issues and Concerns with Contractor Proposed Conceptual Acoroaches 5.2.1 Radiochemical Inventones and Larae Tanks in Contractor Proposed Desians 5.2.2 Process Efficacy with Contractor Proposed Desians 5 2.3 Oraanic lon Exchanae Resin and Nitrate Interactions in the Presence of Very Hiah Radiation Fields in Contractor Proposed Desians 5_2_4 Hydroaen and Flammable Gases in Contractor Procosed Desians 5_2A CST Drvina at Contractor Proposed Desians' 5 2.6 Oraanic Materials in Contractor Proposed Desians 5.2.7 Radiolvsis Effects at Contractor Proposed _Desians 5.2.8 Hiah Temoerature Ooerations in Contradfr Procosed Desians 5 2.9 Nonradioactive Chemical Effects with Centractor Proposed Desians

j r i 4 1 l

5_ 3 Potential Methods for Prevention and Mitiaation 5.3.1 Prevention and Mitiaation of Events involvina Radiochemical inventories and Larae Tanks .

512 Potential Methods to Address Process Efficacy Concerns 5)) Prevention and Mitiaation of Events involvina Oraanic lon Exchanae Resin and Nitrate Interactions in the Presenge of Very Hiah Radiation Fields 5.3.4 Prevention and Mitiaation of Events involvina Hydrocen and Flammable Gases 5 3.5 Prevention and Mitiaation of Events involvina CST Drvina 516 Potential Methods to Address Oraanic Materials Concerns 51Z Potential Methods to Address Radiolvsis Concerns 518 Potential Methods to Address Events and Concerns with Hiah Temperature Operations 5.3.9 Prevention and Mitiaation of Events involvina Nonradioactive Chemical Effects 6_J Risk Cons'iderations LQ Summary and Conclusions SJ References Aooendix A: Process Descriotion Accendix B: Samole Calculations

l 5

PROCESS SAFETY OF PROPOSED i ANK

! WASTE REMEDIATION SYSTEMS (TWRS) l GLOSSARY OF TERMS AEGL Acute Exposure Guideline Levels ALARA As Low As Reasonably Achievable ARF Atmospheric Release Fraction, the amount of a material released at A vapor and as aerosol into the immediate atmosphere atm atmosphere, a unit of pressure representing standard atmospheric pressure at mean sea level, equivalent to 14.72 psi or 101,325 Pascals.

BNFL inc. British Nuclear Fuels Limited Incorporated l a DOE TWRS-P contractor, sometimes just called BNFL.

~

CCW Complexed Concentrate Waste, a form of Hanford tank waste containing relatively high levels of organic compounds.

CMPO an extractant sometimes used in solvent extraction systems; it has a high affinity I for actinide species. l l

CST Crystalline Silicotitanate, an absorbent for cesium, and, to a lesser extent, strontium and TRU. l 1

DEPA Di-Ethyl Phosphoric Acid, an extractant sometimes used in solvent extraction systems, with a high affinity for uranium and related species.

DI Deionized, as in D1 water DOE The U.S. Department of Energy DR Damage Ratio, that fraction of the SSC(s) damaged in an event and able to release the material it (they) contain.

DST Double Shell Tank (two carbon steel walls, in a concrete vault), for nuclear waste I

DWPF Defense Waste Processing Facility, at the DOE Savannah River Site EPA The U.S. Environmental Protection Agency ERPG Emergency Response Planning Guidelines, for hazardous chemicals (from the American Industrial Hygiene Association)..

o 6

gew gram equivalent weight; the molecular weight divided by the maximum charge (valency) of the material, usually in aqueous solution, grams gpm gallons per minute (flow)

G value a measure of hydrogen production based upon radiolysis, usually expressed as molecules of hydrogen evolved per 100 ev of radiation HEPA High efficiency particulate air filter, capable of removing 99.97% of all particles with mean diameters over 0.03 microns.

HLW High Level Waste, generally used by DOE to describe the solids containing and sludge fractions of tank wastes, and SST contents.

LAW Low Activity Waste, generally used by DOE to describe the supernate (liquids) in tanks and the content of DST's.

LCO Omiting Condition of Ooeration, a regulatory requirement for continued facility operation.

LET Linear Energy Transfer, as by radiation passing through a medium, ev/ micron.

LFL Lower Flammability Limit, the minimum concentration of a flammable gas or vapor required in air (or oxygen) to sustain a burning flame or deflagration. I Usually used in reference to hydrogen, which has an LFL in air of about 4%.

LLW Low Level Waste, as defined in 10 CFR 61 LMAES Lockeed Martin Advanced Environmental Systems, a TWRS-P contractor LNT Linear, No-Threshold theory for radiation effects upon biological systems.

LOCA Loss Of Cooling Accident M Molar, or molarity, moles / liter MAR Material At Risk, the total quantity / inventory of material or radioactivity within an SSC or area affected or potentially affected by an event, measured in kilograms or curies. ,

MTlHM Metric Tonnes of Initial Heavy Metal, referring to nuclear fuel MWD Megawatt Days, a measure of nuclear fuel /SNF burnup NOx Nitrogen oxides NRC The U.S. Nuclear Regulatory Commission

f l 0

~

I .

l l

I 7 OSR -

Operational Safety Requirement (a technical specification), usually related to facility operations.

PTFE Polytetrafluoroethylene, a polymer containing fluorine and frequently used to <

seal valves and fittings. I RF Respirable Fraction, that fraction of a vapor or aerosol that can be ingested into the lungs and result in exposure. For aerosols and particulates, this is usually assumed to include particles smaller than 10 microns in diameter.

SCBA Self Contained Breathing Apparatus SCR Selective Catalytic Reduction, an offgas treatment process that uses ammonia to reduce nitrogen oxides into nitrogen prior to discharge. j I

SNF Spent Nuclear Fuel l SSC Structures, Systems, and Components (of a facility)

SST Single Shell Tank (one carbon steel wall, sometimes with a concrsie vault or dish around it), for nuclear waste Tank waste Waste contained within the underground tanks (SST's and DST's) at the DOE Hanford site. l 1

TPA Tri-Party Agreement, between the DOE, Washington Department of Ecology, and ,

the U.S. EPA I

TSR Technical Safety Requirement, a technical specification for safety, usually related to the design of SSC's.

TWRS Tank Waste Remediation System, used to reference the program, the approach, and the facilities at Hanford for treating HLW. Although HLW can include SNF and, thus, TWRS may include SNF facilities, in this paper, TWRS is assumed to refer to the facility or facilities handling ano treating the tank wastes.

TWRS-P The privatized TWRS facility for treatment and vitrification of Hanford tank wastes.

WVDP West Valley Demonstration Project, a DOE project near Buffalo, NY l

9

o 8

(This page intentionally blank) 4 0

4

I-i I

9 f

l I

PROCESS SAFETY OF PROPOSED (ANK WASTE REMEDIATION SYSTEMS (TWRSj l l l

10 Abstracj l Technical approaches are being formulated for the treatment of DOE tank wastes into vitrified l waste forms suitable for long term storage and disposal. Tank wastes from the Hanford site are l likely to involve significant chemical processing for separating radionuclides from nonradioactive species and concentrating them in the high-level waste (HLW) glass. Chemical and process l safety are significant contributors to the total risk from the facility. Areas of chemical and process safety that result in the potential for worker and public exposures exceeding regulatory l limits have been identified as follows:

. ' radiochemicalinventories

. process efficacy )

e organic ion exchange resin / nitrate interactions

. hydrogen and flammable gases i

. CST drying

. organic materials

. radiolysis

. high temperature operations

. nonradioactive chemical effects ,

This paper analyzes a generic TWRS facility as well as conceptual design approaches proposed by DOE contractors. Analyses estimate both unmitigated and mitigated consequences from potential accident scenarios in these areas, using a conservative bounding approach suitable for safety categorizations and preliminary designs. Several scenarios involving large radiochemical inventories (in tanks), flammable gases, organic ion exchange resin interactions, and cold chemical effects have potential accident consequences to the workers and the public of sufficient severity such that prevention (reduce probability) and mitigation (reduc 3 consequences) become necessary, requiring the identification of items relied upon for safety, ideally, processes and approaches proposed for tank waste processing should incorporate robust designs with redundant features, and examples of several approaches are given. The paper discusses suitable process accident prevention and mitigation methods that are compatible with the regulations and offer the potential for reducing process accident risk to acceptable levels.

r i O

10 (This page intentionally blank) t t

t '

f .

11 L Introduction l The Department of Energy (DOE) established the Tank Waste Remediation System (TWRS)

[ program at the Hanford site to manage, retrieve, treat, encapsulate / immobilize, and disposition l' radioactive waste materials from the underground storage tanks onsite, in a safe, i environmentally sound, and cost effective manner (Reference 1). These tanks primarily contain i

high-level wastes (HLW) and chemical species from processing spent nuclear fuels for more than fifty years at tha site. ,

The tank contents consist of mixtures of materials from some eight major processes. Even though the radiation levels are high (typically exceeding 100 R/hr), the majority of the waste constituents are nonradioactive. The tanks hold approximately 55 million gallons of waste, ar d amount to some 300 million curies of raoioactivity, primarily from cesium and strontium but with smaller contributions from other fission products and transuranic (TRU) isotopes. Physically, the tank contents exist as liquids, sludges, salts, saltcakes, and mixtures'thereof, and some tanks periodically release gas mixtures.

DOE is pursuing a privatization initiative at Hanford for the construction and operation of contractor-owned, contractor-operated facility or facilities for treating these wastes. The plan calls for processing between 6 and 13% of the tank waste as Phase I, and a subsequent, larger program would process the balance of the tank wastes as Phase ll. Phase I consists of two parts. Part A requires contractors to select safety standards 'and requirements, and to generate conceptual designs and initial safety analyses. DOE awarded contracts for Part A to two teams, one led by British Nuclear Fuels Limited inc. (BNFL Inc.), and the other led by Lockheed Martin Advanced Environmental Systems (LMAES). Currently, Part A is essentially completed; the contractors have each submitted a System Requirements Document (on standards), Hazards

/A nalysis Repod, Integrated Safety Management Plan, and initial Safety Analysis Report. The DOE has generated evaluation reports on the contractor submittals (References 2 and 3). From an evaluation of this information, DOE selected one contractor (BNFL) to participate in Part B.

Part B involves a 24 month facility design phase that will result in closure of the previous Part A activities, regulatory permit applications, fixed unit prices for treated wastes, financing, and a firm schedule, followed by facility construction and operation. Part B is expected to require 5-8

. years for design and . construction activities, ano another 5-10 years for the processing of the waste. The total value of the Part B work may approach $7 Billion. Currently, Phase 11 plans to enlarge and utilize the Phase I facilities instead of constructing and using entirely new facilities.

The Phase ll activities extend the waste processing timeframe by another 10-30 years, depending upon the facWty capacities, waste retrieval activities, and difficulties encountered.

DOE is a self-regulating agency on nuclear safety matters, and has established a Regulatory Unit (RU) at the DOE Richland Operations Office adjacent to the Hanford Site (References 2 and 3). DOE is also considering the potential for external regulation, and has entered into a Memorandum of Understanding (MOU) with the Nuclear Regulatory Commission (NRC) for cooperation and mutual support, with the possibility of transitioning to NRC regelation sometime in the future (Reference 4). The MOU has two main purposes:

Point Paper May 27,1999

o 12

1. The DOE to acquire the capability to implement a program of nuclear safety and safeguards regulation consistent with the NRC's regulatory approach.

2. The NRC to acquire sufficient knowledge and understanding of the physical and operational situation at the Hanford waste tanks and the processes, technology, and hazards involved in Phase I activities, to. enable the NRC (a) to assist DOE in performing reviews in a manner consistent with NRC's regulatory approach and (b) to be prepared '

to develop an effective and efficient regulatory program for the licensing of DOE contractor-owned and contractor-operated facilities that will process waste at Hanford during Phase ll.

The DOE contract award for Part B of Phase 1 envisions the design, construction, and operation of facilities that will be reused and expanded for Phase 11, and, thus, the distinction between Phases I and 11 is diminished. This may affect the regulatory tr'ansition between the DOE and the NRC in the future.

This paper presents a summary of chemical safety approaches related to the TWRS facilities, and identifies 'and discusses potentialissues and areas of concern from NRC analyses of TWRS. The main objectives are:

- Outline the pertinent process steps and operations required for the proposed TWRS-P facility, for both generic and specific contractor approaches (Section 3 and Appendix A).

. Identify potential areas of process and chemical safety concerns (Section 4).

- Estimate the potential consequences and ramifications associated with these areas of process and chemical safety (Section 5).

. Assess the potential for mitigating methods to reduce the consequences (Section 6).

• Discuss the implications from the NRC perspective (Section 6).  !

I l

l I

Point Paper May 27,1999

i ,

13 I

3 a TWRS Process Description The Hanford tank wastes consist of approximately 55 million gallons of highly radioactive materials containing some 200 million curies, from reprocessing, recovery, and decontamination operations at the site over the last 45 years. Eight major and numerous minor liquid waste streams have been discharged to the tanks, and 38 tanks are on the "watchlist" because cf flammable gas, organic compound, or high-heat concerns (Reference 1). The tank contents contain both low-level, high level, and TRU wastes; DOE usually applies the term " low-activity wastes" (LAW) to predominantly liquid phases in the tanks, and "high-level waste" (HLW) to solid phases from the tanks. For practical purposes, there is little difference, as both require handling and processing in heavily shielded facilities. The majority of the activity accrues from the cesium and strontium fission products. The majority of the mass consists of nonradioactive species, such as sodium, nitrates, and metal oxides / hydroxides (e.g., iron, aluminum).

The LAW waste is normally termed supernate. It contains the majority of the cesium and technetium activity, and consists primarily of aqueous and aqueous soluble species. Processing can entrain HLW solid particulates from the tanks, and, thus, the LAW waste envelope allows up to a 5% slurry to be transferred and separated during processing. DOE has identified three representative, LAW compositions (Reference 5). As summarized in Table 1, Envelope A represents a baseline set of radionuclides, Envelope B denotes a high cesium example, and Envelope C contains higher concentrations of strontium and TRU. In contrast, the HLW example is designated Envelope D, and consists of insoluble species, precipitates, and oxides (Table 2). It contains most of the strontium and TRU isotopes.

The objectives of TWRS processing are to:

. Retrieve the waste from the tanks and immobilize it by vitrification.

. Process and treat the tank wastes to address chemical compatibility concerns.

. Separate the radioactive from the nonradioactive species, with the majority of the nonradioactive species going to a LAW stream, and the majority of the radioactive species going to the HLW

. Immobilize the LAW and HLW into glass, within metal containers, suitable for long-term storage and disposal.

. The immobilized LAW should be suitable for near surface disposal, per 10 CFR Part 61 requirements, and the HLW should be suitable for future repository disposal, per 10 CFR Part 60 requirements.

There are implied objectives of reduced waste volumes, an overall reduction of the potential hazards (chemical and radioactive) from the wastes, and limited or no use of unusual, rare / exotic, or expensive reagents. This section presents summaries of generic process technologies to achieve these objectives. and overviews of the BNFL and LMAES conceptual process approaches.

Point Paper May 27,1999

i _

14 i

'11 Generic Process Figure 1 depicts the overall block diagram for the generic process. The overall processing scheme pretreats and separates the LAW supemate into a smaller volume, HLW-like fraction and a significantly less radioactive LAW stream containing most of the salts, which is vitrified.

' The HLW slurry is washed to reduce the soluble species (which go to the LAW pretreatment),

combined with the HLW-like streams from LAW, and vitrified in a separate, HLW melter. The principal activities are: .

. Waste preparation and transfer

. Pretreatment consisting of:

- Suspended solids removal

- ' Cesium separation Technetium removal -

Strontium and TRU treatment and removal

- Treatment and removal of other radionuclides or species

. _ Low Activity Waste Immobilization

.: Immobilization of High Level Waste and of radionuclides removed during pretreatment.

. Off-Gas Treatment

. Supporting Plant Process,es Appendix A provides a more detailed summ ary of the generic process approach, with references for anticipated process steps (References 6 - 18). Planned operations at a TWRS

. facility have similarities with current operatic ns at West Valley, New York (WVDP) and the Defense Waste Processing Facility (DWPF) at Savannah River, South Carolina (References 11 )

and 13)._ However, significant differences exist in the pretreatment area due to the multitudo of different elements and isotopes present in the Hanford wastes, which, in the absence of separations, would generate an inordinately large quantity of immobilized HLW. DWPF and WVDP also use cementitious waste forms for their equivalent of treated LAW waste, while the TWRS approach vitrifies the treated LAW. MRS-P designs envision a facility approximately the same or slightly larger than DWPF, with several HLW and LAW melters. TWRS-P throughput is also likely to be similar to the DWPF capacity. An accelerated cleanup schedule

'for Hanford would increase both the size and throughput capacity of MRS-P, primarily from the additional melter systems required. .

Point Paper May 27,'1999  !

s l 5

15 l

Table 1: Typical LAW Radionuclide Composition (Reference 5)

Radio- Maximum Ratio, Bq/ mole Sodium Curies / Liter at 14 Molar Sodium

"" ~

Envelope Envelope l Envelope inn'sklope' CE kiloN PNndlek A~ l :B bc .~: d%NWF :X89

• O C'@

TRU;_ 4.8E5 4.8E5 3.0E6 1.82E-04 1.82E-04 1.14E-03 l

~

Sr 90MU 4.4E7 4.4E7 8.0E8 1.66E-02 1.66E-02 3.03E-01 TM@ - 7.1E6 7.1 E6 7.1 E6 2.69E-03 2.69E-03 2.69E-03 Csh3$ ' 4.3E9 6.0E10 4.3E9 1.63E+00 2.27E+01 1.63E+00 Table 2: Typical HLW Maximum Radionuclide Concentrations (Referenc,e 5)

Isotope Cilliter Isotope Cl/ liter Isotope Cilliter H-3' 2.00E-05 Cd-115m ' 6.55E-10 )E0f152&ffjh 1.50E-Oe C-14

~

2.00E-06 Sn-119m. 1.00E-08 'E4154Mfi 1.60E-02 W Fe-55 1.00E-03 Sn-121m{ x 9.00E-06 ;E5155WG 9.00E-03  !

l Ni-59 % e - 1.40E-05 Sn-126 r/P 4.80E-05 ru234QM$y. 7.70E-07 l Co40?#" 3.00E-03 . Sb-124:"? '? 2.61 E-09 114-M$$$( 3.20E-08 l Ni-83 da 1.60E-03 Sb-1.26b,M 4.83E-06 ($$#@@ 8.20E-08 Se-795 4.20E-07 Sb-.126m f" 3.43E-05 (U-238 &'.!U @ 5.80E-07 Sr-90 . . 3.10E+00 Sb 125 1.00E-02 rNp-237Ns f 2.30E-05  ;

Y-90 W. 3.10E+00 Te-125m 3.00E-03 PS2383 p 1.10E-04  !

Nb.93m : 8.70E-05 l-129 9.00E-08 ' Pu-239f y;@. 9.50E-04 Zr-931 1.40E-04 Cs-134 , 6.80E-03 1Pu-240MT 2.60E-04 Tc-99 . .o 4.50E-03 Cs-1351, .. 3.00E-05 Pu-241M'Mtn 6.90E-03 Ru-106 * ' ' 2.00E-04 Cs-137 ' 3.00E+00 #4242h9@ 7.10E-08 Rh-106; 1 2.00E-04 Ba 137m il 3.00E+00 Am441%E@ 4.30E-02 Pd-107? , 4.00E-06 :Co-144 ? , 1.00E-04 'Am'-242:; 149 3.10E-05 Ag-110m . 1.00E-08  : Pr-144L 1.00E-04 ;Am-242mM 3.20E-05 Cd-03ma, 1.09E-03 .Pr-144m ~ , - 1.00E-07 ' Am-243N O 5.00E-06 Im113m4. 1.88E-06 . Pm-147 . - 1.60E-01 .Cm-242id$$ 3.70E-05 Sn-113 188E-06 Sm-151 '

9 30E-2 Cm-244 % *W 9 30E-04 Point Paper May 27,1999

)

O

'16' M Waste Receipt, Waste ' Feed Evaporation Cs Removal by Teemum 1 4 Entrained and Sr/TRU + lon Exchange - > . Removal try Solids Removal ton b e ange j ,

.. Acid Acid

'"U ' Elution Elution lf lf U  !

Cs Eluate netium Retum to DOE after 60 (Conditioning ,

'Y' * * ' '" P! ant (Store for 9 yr.)

onto CST h .

Y U V Dned CST Retum glass handling and Retum to DOE product to DOE container after 60 days t.AW Option

> ~

A .

2 .

a te Feed Ev po ion Cs Removal by Te@mtium 4 Entrained and Sr/TRU + lon Exchange + Removal by lon Exchange Solids Removal Acid > Acid j Stuny y Elution Elution J U V U 1 i

Cs Concentrate LAW Tc Concentrate l Entrained Solids Retum to (Store for initial 2 Vitnfication q DOE after 60 days storage yr. pnor to HLW (Store for initial 2 yr. prior to HLW Plant J metter available) '

metter available)

V

. V Glass product to DOE after HLW Retum glass 60 days storage Vitnfiction product to DOE Plant after 60 days g Sr/TRU slurry blended into metter over years 3.4 &5. i e., +V stored for 2-yr.

LAW and HLW Option Figure 1: Generic TWRS process Approach Pc:nt Paper .

May 27,1999 i

i

i 17 12 Contractor Proposed. Conceptual TWRS Process Anoroaches BNFL proposed a conceptual approach based upon the following operations (Reference 2):

. Strontium /TRU coprecipitation from LAW Suspended solids / strontium /TRU removal by ultrafiltration from LAW Two columns in series, organic ion exchange recovery of cesium from LAW Two columns in series, organic ion exchange recovery of technet!um from LAW

=

Optionalloading of radiocesium onto crystalline SPicotitanate (CST)

. LAW vitrification in a joule-heated melter HLW washing and concentration by ultrafiltration

. HLW vitrification in a joule-heated melter NOx treatment by Selective Catalytic Reduction (SCR) using anhydrous ammonia Some aspects'of the conceptual design were more detailed than other parts. For example, each ion exchange column would enclose 3.15 cubic meters, with an aspect ratio of about 7. In use, the columns would only contain about one cubic meter of resin each, for an effective aspect ratio of about 1.5-2. The approach included four ion exchange columns for cesium removal, arranged as two trains of two columns each. The technetium arrangement was identical. The design included two HLW and three LAW metters.

LMAES proposed a similar conceptual process approach with several variations (Reference 3):

Suspended solids removal from LAW by centrifugation Three columns in series, organic ion exchange recovery of cesium from LAW, regenerated by nitric acid and caustic Three columns in series, organic ion exchange, polishing step recovery of cesium (from the effluent of the preceding step), with electrical regeneration An inorganic, " guard" bed for cesium removal (on the effluent of the preceding step)

. Optional loading of radiocesium onto CST '

Removal of technetium from LAW by electroplating Strontium and TRU removal using ozone destruction of organics followed by precipitation and centrifugation LAW vitrification in a joule-heated melter, augmented by fired burners (based upon oxygen-propane combustion) during startup and glass pouring HLW washing and concentration by centrifugation i

. HLW vitrification in a cold-crucible, induction-heated melter j -

NOx treatment using SCR and ammonia from aqueous solution Of particular interest for process and hazard analyses, each ion exchange column would be approximately 0.67 cubic meters, containing about 0.6 cubic meters of resin, with a working l Point Paper . May 27,1999

O 18 aspect ratio of approximately 4.' The approach effectively used six columns in series to remove the radioactive cesium. The design inco'rporated one HLW and three LAW melters.

. 4 May 27,1999 Point Paper

r _ q 4

19:

L- Potential Chemical Safety lasues and Areas of Concern The TWRS-P facility represents a radiochemical facility with significan't usage and consumption of chemical materials and with a relatively large radionuclide inventory. Table 3 compares the potential radionuclide inventories at TWRS-P locations (calculated from Tables 1 and 2 - see r- Chapter 5 and Appendix B) with selected radionuclide quantities at a nuclear power plant

' (calculated from the Radiological Characteristics Database - see Reference 19); the TWRS facility is likely to handle comparable quantities of radioactive cesium, strontium, and

. technetium, in significantly more mobile physical and chemical forms (e.g., as nitrates and i aqueous solutions), as compared to ceramic oxides in power reactors. In addition, while a i reactor has more energy for potential energetic scenarios during operations, the TWRS-P facihty has more stored chemical energy for prompt potential events directly involving the radionuclides in their mobile forms. ,

At NRC-licensed facilities, as stated in the 1988 Memorandum of Understanding between the

-NRC and the Occupational Safety and Health Administration (OSHA), the NRC regula.tes

. chemical safety issues related to the following: ,

I

1. hadiation risk produced by radioactive materials.

2. Chemical risk produced by radioactive naterials. 1

3. Plant conditions which affect or may affect the safety of radioactive materials and, thus, present an increased radiation risk to workers, the public, and the environment I

Severa safety and process safety analyses have been conducted on proposed TWRS-P )

facilities (References 1 - 3). While some aspects of safety are process specific, the following, recurrent areas of major concern are generic and require attention: i

. Radiochemical inventories and Larae Tanks: The TWRS-P facility will contain relatively large quantities of materials. Facility tankage is likely to be between 10,000 and 100,000 i gallons. The facility may contain megacurie quantities of activity; the tankage represents the potential for up to 10 megacuries for cesium alone. The facility may also contain additional, separated radioactivity in smaller volumes but more concentrated and mobile forms, because the DOE contract requires separation and removal of the cesium, technetium, strontium, and TRU materials from the LAW. As discussed in Sections 5.1.1, 5.3.1, and 6.2,'this requires careful inventory management and a robust design with high quality components and controls, with an emphasis on prevention by avoidance of inventory and on mitigation by reducing inventory of mobile forms of the radionuclides.

Point Paper ,

May 27,1999

6 _

) Y .

s M R i

e H D 1, rd 6 6 6 r

u k ae E E E c

l T Fsel Nayo o 9 0 3 8 M 6 n / SC5c 1 2 1 1 D

i R

W P

(

W M

0

.V rd r

o 0

0, e r el oyoo ae 7 E

8 0

0 7

E 2

7 t

c 0 2, E a - 6 C1~ c 1 2 1 2 e

R .

t l ,

a .

y n .M

_.r$%

5' D lP d d i

m H 5 6 o ~T Eskrl l ao 6 0 E E 2

N M .N a eo E 4 2 .

e - / SCyc 1 1 7 1 M 9 H

r W .D l 9

9 o M W T 9 1 t

c a 0 M , - M1 7, e 0 0 ,. r d 2 e 2 R 0, 0 0, .er el aeo 6 E

0 0

6 E

7 E

1 c s

n y la 1

- 0 oy o 2 2, 6 3 nr e a ic 3 C1 c 9 1 6 1 iaf e M n e r t e rli oR m c m n ly m o o*

r laf C f m -

a Tu 5 in e -

msa l

d

.So E -

n -CC 3 0 0 0 ob 0

2 a. na kt a P- s S n ,rs aD R y sin e 0 cs -

it l /'

emli Ru t

0 Fc i W

T i

c a

l 0 0,

) ) )

Nt s Si r

~ .s o 0 2 0 0 0 n F CC1 7 ( ( ( dt e e P- nc e aar w S t

e R )s , Mah b

t c 6 H C2.

s Wie r - ul E l

Tl e 1 u- da og 2 ad it rc o( srK 3 )

0

)

0

)

0 Mcn 0 ig a it n

f g

CP1 1 ( ( (

0o1 a n' 1 los u i i Q d; sdle n

n ' . sg  ;,- al ab ie u o ' 0 iaRa neT t

r 6 6 u B 0 0

.WkK C n0 E 0 E 2, oh mo 1 7, 2 ct f 8 pe.

.La0 7 o ..H T 1 1 1 1 1 lygr nf n

o 2 -

~ laid n s iut e

1 s

i r

a

.l

,aD

)

m )n mdla oeu p e ,sg oo nt al c 6 0 0 0 ri m g1 t

W kn0n K 'v. E 0 0 0 f t u el a o Aa0 6 0, 3, 7, 0l 7o r ocu cs C .LT1E ( 8 1 6 8( s cr la c e u

to sl a r ce v e

- e au p

i od 7 .

elaP a RvS i  ;

3 dl -

3 -

P le Rna uc - 1 9

97 0

~9 U ERR

t n

b a i C

s T

c S

r- R' T'

t oWW io T NPT . P

=

C

.C 21 l

. Process Efficacy: The W/RS-P facility contacts multiple streams with the radionuclides.

If the process does not function as planned or attain the desired efficiencies, significant l quantities of highly contaminated chemicals may result, influencing downstream chemical processing and maintenance approaches. As discussed in Sections 5.1.2, j 5.3.2, and 6.2, this may requiro a thorough test program, with repeat tests on the process steps, multiple separation stages, an adequate design margin, and a robust design based upon 'he most conservative (least beneficial) test results.

l

. Oraanic lon Exchanae Resin and Nitrate interactions. in the Presence of Very Hiah Radiation Fields: ion exchange resins are organic materials and can react with oxidizing  ;

materials. Radiation fields exacerbate the reactions. In particular, concentrated nitrate solutions can attach nitrate groups (-NO2) to organic resins, and, in the case of anionic resins, compete for ion exchange sites (frequently quaternary ammonium) producing an ammonium nitrate-like combination. High levels of radiation can directly initiate the nitrate-resin decomposition reactions, or the decay heat from radionuclides can dry the resins and increase the temperature to the autoignition point. As discussed in Sections 5.1.3, 5.3.3, and 6.2, this requires a careful design with inventory management (i.e., on the loaded resin) and a robust design with high quality components and controls, with an emphasis on prevention by design, by measurement of resin degradation, and by detection of resin reaction initiation.

. Hydroaen and Flammable Gases: Radiolysis of water and other waste constituents produces hydrogen and flammable gases. These gases are continuously generated and l may accumulate in vessels and headers, and, over time, may achieve flammable conditions. As discussed in Sections 5.1.4,5.3.4, and 6.2, this requires a robust design with high quality components and controls, with prevention by ventilation and mitigation by HEPA filters.

. CST Drvina: C rystalline silicotitanate (CST) is an inorganic material. Once a CST container is I?aded, it represents a concentrated source of cesium with the potential for contact radiatian fields approaching 100,000 R/hr gamma. This represents a source of radiolytically-generated hydrogen. If the container is drained, residual water witnin the container (from holdup or absorbed on the CST) may still be present and undergo radiolysis, producing hydrogen. As discussed in Sections 5.1.5,5.3.5, and 6.2, this requires careful adherence to procedures for drying (a preventative control).

. Oroanic Materials: The wastes contain organic chemicals, and the processing may add additional organic materials (e.g., solvents and extractants from solvent extraction, resin degradation products). These can influence process effectiveness, change anticipated radionuclide separation, undergo radiolysis, and increase unintended reactions (e.g.,

flammable gas generation). As discussed in Sections 5.1.6,5.3.6, and 6.2, this requires a thorough comprehension of the chemistry effects from a test program, with a preference for prevention of effects by destruction.

Point Paper May 27,1999

~

11

. Radiolvsis: The high radiation fields produce chemical changes in reagents and/or l initiate unintended reactions. This is particularly appropriate for relatively new reagents I and processes with limited, prior use in high radiation environments. Recent examples  !

' include nitrite (from nitrate radiolysis) participation in Hanford tank waste chemistry and benzene / tar formatinn from the sodium tetraphenol-boron reagent proposed for in-Tank Precipitation at Savannah River. Radiolysis may also influence the performance and chemistry of typical polymeric materials used in equipment. For example, high radiation

, fields accelerate the release of halides from fluorochloropolymers, which may increase corrosion of austenitic stainless steels; such stainless steels are a common choice for HLW processing facilities. As discussed in Sections 5.1.7,5.3.7, and 6.2, this requires  !

I avoidance of materials subject to degradation or testing, monitoring, and replacement of racHation sensitive items prior to excessive degradation (both are preventative -

approaches).

. Hioh Temoerature Ooerations: Vitrification uses melters operating at temperatures around 1,100 C. As discussed in Sections 5.1.8,5.3.8, and 6.2, this requires careful inventory control of the cold cap (the most mobile radionuclides) and a robust design with high quality components and controls, witn an emphasis on prevention (i.e., detect melter problems and stop radionuclide feeding prior to the accident event).

.. Nonradioactive Chemical Effects: Several chemicals potentially used.by the facility have the capability for generating Vapor clouds and/or having interactions that raise operability and habitability issues for the facility. These include ammonia, nitric acid, and sodium hydroxide. Common fuels (e.g., diesel oil, propane) may also be a concem. As discussed in Sections 5.1.9, 5.3.9, and 6.2, this requires careful inventory management and a robust design with high quality components and controls (preventative approaches), and mitigating safety systems (water sprays and confinement).

Quality control, process / equipment monitoring, and inspections represent probable prevention methods. Cell confinement and high efficiency particulate air (HEPA) filtered ventilation are likely mitigation methods for many of these concerns.

I 1

l F

' Point Paper May 27,1999

q

. \

t t

23 52 Analysis and Conseauences from Chemical Safety issues and Concerns at TWRS Facilities This section discusses the analysis of potential consequences from chemical and process safety issues and concerns at, initially, a TWRS facility using the generic approach (from Section 3.1 )

and Appendix A), and then at a potential TWRS facility using contractor proposed, conceptual l process approaches. Potential prevention and mitigation methods, and their effectiveness, are also discussed in general terms. Appendix B provides sample calculations. The reference texts do not include many examples and data for HLW processing and vitrification scenarios, thus necessitating the use of analogies. Some inconsistencies are noted (e.g., releases from solidified HLW glass greater than the releases from the molten, liquid glass). Thus, testing for release behavior and fractions directly related to HLW treatment and vitrification is recommended.

5A Analysis and Consecuences from Chemical Safety issues and Concerns at the Generic TWRS Facility For the purposes of this discussion, the potential consequences of process safety accidents have been semi-quantitatively analyzed using the proposed TWRS-P site in the 200-E area at Hanford, with a facility fenceline (distance to the public receptor) of 100 meters. This is based upon the actual layout of the 200-East area, which includes a road traveled by members of the public cpproximately 100 meters from the proposed TWRS-P facility site, and the occurrence of prompt accidents, for which evacuation scenarios and subsequent site access restrictions are ineffective. Several of the chemical safety issues have discrete events and have been analyzed for the gener.ic approach, using standard handbooks for parameters (References 20 and 21).

Other issues are broader and require specific, detailed designs for a quantitative estimate of consequences, and, because detailed designs are not yet available, these are discussed qualitatively.

5.1.1 RadlochemicalInventories and Larne Tanks it is anticipated that waste will be transferred to the facility in daily batches of 10,000-15,000 gallons. At 14 molar sodium, the 10,000 gallon batch corresponds to 860,000 curies of cesium-137 (Envelope B). Using 100,000 gallons as the receipt tankage volume (i.e., no ullage)in the TWRS facility, this corresponds to 8.6 megacuries in feed storage. Failure of the tankage with a low energy, liquid freefall, with resuspension from the resulting pool, with a surface release, results in inhalation doses of 200-400 rem at 100 meters, depending upon the energy of the release (temperature, height, agitation, etc.). The LAW also contains suspended solids, which may be as high as 5% by weight; these suspended solids are essentially HLW. The contract (Reference 5) provides maximum limits for the radionuclide content per liter of HLW slurry, based upon 10% suspended solide (a light slurry). Adjustment of the radionuclide composition to the 5% level results in an additional dose contribution of 2,800-6,000 rem to the receptor at 100 meters, principally due to the americium contribution. The total dose from the failure of Point Paper May 27,1999

m

.c <

W y l tankage containing 100,000 gallons.of LAW (with 5% suspended solids) becomes 3,000-6,300

- rem to the receptor at 100 meters.' Various mitigation methods exist to reduce this' estimated f dose to acceptable levels; as an example,'use of a cell with HEPA filters on the exhaust reduces

~

' these accident doses by a factor of a thousand.

For HLW, the calculations use'the 10% level of suspended solids. This gives dose estimates of.

16,000 to 12,000 rem for 100,000 gallons of HLW in tankage. Again, various mitigating methods exist to reduce this accident dose to acceptable levels.

. After ion exchange, the radiocesium exists as soluble cesium nitrate in solution. This goes to interim storage in a tank, prior to either incorporation in the HLW glass product or loading upon

CST for long term storage as a solid. Experimental results (References 8-10) show elution substantially complete after passing approximately 6 column volumes of nitric acid through the resin. Tests obtain peak cesium concentrations of 15-35 times the originalinfluent, with an average 7-15 times higher than the inlet concentration. As a first approximation for a bounding.

analysis, the calculations use a concentrating factor of 15, which, for Envelope B waste, L becomes 350 curies / liter of cesium-137. This gives 132 megacuries for a 1,000 gallon batch,

~

corresponding to.a heat load of 4.9 KW(th) (about 17,000 BTU /hr). The estimated adiabatic heat rise is 2'F/hr. .Thus, if the tank is cooled to maintain temperatures below about 122 F, and loss of cooling. occurs, the adiabatic analysis implies boiling could occur in approximately two days. For the purposes of.this analysis, this loss of cooling / boiling cesium product accident scenario is assumed to last for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, resulting in evaporation of 5% of the contents in the tank. The estimated accident dose becomes 25,000 rem to the receptor at 100 meters, it should be noted that this is not a prompt event,'and that mitigating measures (e.g., backup cooiing systems, cell with filters, etc.) would be effective.

A related accident scenario involves the potential boiling of the cesium product tank due to a full

~

vacuum pull on th'e vessel (i.e., boiling at the tank's current temperature). Such vacuum-

. induced boiling is usually very vigorous and would result in aerosol formation comparable to

- vigorous normal boiling. This could occur more frequently than the loss of cooling accident due to higher maloperation frequencies with vacuum (ring) pumps, air injection systems, and vacuum breakers. The accident can also be rendered incredible (frequency less than 1E-6/yr) by the suitable choice'of equipment (e.g., a fan that cannot attain vacuum conditions under all operational circumstances). A specific design is required for adequate analysis of both probability and consequence. However, the accident consequences of an inadvertent vacuum

' draw boiling event would be similar to and bounded by the loss of cooling water, boiling tank i event, and , as a first approximation without a specific design, similar probabilities are assumed.

A breach failure of the cesium product tank is an additional, potential' accident. Using the 1,000 l gallon basis as before, and free-fall of the liquid followed by resuspension from the pool, the analyses estimate a dose of 55 rem to the receptor at 100 meters. J l

l l

Point Paper May 27,1999

25 5.1.2 Process Efficacy Effective separations and pretreatment are crucial to safe operation of TWRS facilities; without them, contamination becomes more prevalent, radiological dose increases, and wastes will not meet their required specifications. The required removal efficiencies are as follows, based upon maximum contract values for radionuclide concentrations (Reference 5 and Table 1):

cesium: for Envelope NC wastes: 99.8 %

for Envelope B wastes: 99.99 %

technetium: for Envelope NB/C wastes: 89 %  ;

i strontium: for Envelope NB wastes: 0% (removal not required) for Envelope C wastes: 93 % )

l TRU: for Envelope NB wastes: 99.9 %  !

- for Envelope C wastes: 99.98 %' l The removal of TRU and cesium requires the greatest effectiveness. TRU species are relatively insolubl,e in the tank wastes, although organic complexing agents and their degradation products do significantly increase the solubility. Removal relies upon rendering the TRU species insoluble, follov!ed by filtration. This should be relatively easy to accomplish by the methods already discussed. On the other hand, the achievement of the high removal efficiencies required for cesium will'oe more difficult, and necessitates close attention to design and operation. For example, using 50% of the influent as the control point for the effluent, for isolating and regenerating a resin column (a typical value), literature data (References 8 - 10) indicates the following results for resin columns in series:

Column 1- Lead Column: 79% cesium removal, final effluent concentration of 50% of influent. l

. Column 2 - First Polishing Column: 19% cesium removal (cumulative removal of 97.6%), final effluent concentration of 2.3% of Column 1 influent.

Column 3 - Second Polishing Column: 2.26% cesium removal (cumulative removal of 99.86%), final effluent concentration of 0.09% of Column 1 influent.

Thus, to a first approximation, three columns may be the minimum necessary to achieve adequate cesium removal. The experimental data also implies an equilibrium limit around a concentration level equivalent to 99.5-99.8% of the influent concentration, and a guard column (say, of CST) or a small column undergoing more frequent regeneration may be necessary.

The resin column itself is likely to require a large aspect ratio (> 15, almost like a chromatography column) to avoid back mixing and higher effluent concentrations of cesium.

Point Paper - May 27,1999

=

26 Conseg ntty, the cesium resin column design requires close attention to achieve the required removal efficiencies and effluent concentrations, and to avoid potentially high cesium levels in downstream operations which could result in worker exposures potentially exceeding allowable limits.

For comparison, the WVDP utilized a set of four zeolite columns in series, with the columns being advanced as the lead column was loaded with cesium and refilled with fresh zeolite (References 11 and 22). In an early supernate processing campaign, WVDP experienced valve

. problems that rendered one of the columns incapable of dumping the loaded zeolite, and,

- consequently, the remainder of that promssing campaign was accomplished by three columns l in series (Reference 22).

Technetium exists in multiple valence states. Effective processing requires removal of multiple valent state technetium, in both cationic and anionic forms, or conversion of technetium into a

' common, physico-chemical form. If technetium removalis inadequate, then the limits for the

. receptor at the waste disposal facility may be exceeded.

Process efficiency.and its trends constitute routine operations and anticipated events, and, as such, are subject to relatively low radiation limits (see Section 6).

5.1.3 Oraanic lon EU. hance Resin and Nitrate Interactions in the Presence of Very Hiah Radiation Fields Adverse ion exchange resin reactions with nitric acid have occurred in nuclear applications, including overpressurizations and releases (References 21 and 23). The cesium removal resin represents the' main concem because of its relatively high concentrations arid dose conversion factors.' Experimental data reported in the literature indicates approximately a ninety-fold increase in the concentration of cesium upon the resin column, as compared to the feed solution (References 8,9, and 10). Thus, as an approximation, treatment of Envelope B waste implies a feed solution of 8 curies / liter at an adjusted (partially diluted) feed level of 5 Molar sodium, corresponding to a loaded resin concentration of approximately 720 curies / liter and a heat load of 3 watts /! iter. Using a 100 liter column as a basis (i.e.,6 inches ID by 18 ft high of resin), the  ;

material at risk becomes 72,000 curies.

The first resin interaction scenario assumes the resin interactions occur after the column is loaded but awaiting regeneration, and is initiated by localized decay heating and radiolysis. The potential for involvement of additional resin columns depends upon the specific design and is not included here. The data set for release fractions from burning plastics typically used in ion exchar ;,e resins (e.g., polystyrene and polymethylmethacrylate [PMMA - lucite]) is limited and based upon uranium contamination - cesium values are likely to be higher. The PMMA data set is significantly larger, and bounding values of SE-2 and 1.0 are suggested for the release and respirable fractions, respectively (Reference 21). The source term becomes 3,600 curies using these values. The dose estimate is 1,400 rem to the receptor at 100 meters.

Point Paper May 27, ?999

1 27 j t ,

The 'secorid scenario assumes the column is isolated and the . event is initiated during regeneration when nitric acid interacts with the resin in the high radiolysis fields. The highest solution concentration occurs about midway through the cycle, and i.s equivalent to approximately 30 times the feed concentration or about one-third of the column's loaded concentration. The peak occurs for approximately one bed volume of eluant. Resin columns t I are usually designed with a solution freeboard (i.e., a resin free space) above the resin of about 25-50% of the resin volume, for expansion effects. Interstitial space within the resin usually amounts to 30-50% of the resin volume. Thus, an amount of eluted solution equivalent to a resin volume (i.e.,50% plus 50%)is bounding. This corresponds to a material at risk of 24,000 curies of cesium in solution. The reactions provide energy to vaporize the solution (or superheat it if it is in a confined space, ultimately leading to failure), and, ultimately, a release occurs. The energy required to vaporize all of the solution can be supplied by buming less than 10% of the resin, and, thus, an atmospheric release fraction of 1 is used. The respirable fraction is 0.3 (from Reference 21). The source term is 7,200 curies. The accident dose estimate becomes

- 2,700 rem from the eluant liquid vaporization release. The solid resin also releases cesium, but the source term is reduced 50% because of the regeneration, and this contributes another 700 rem. The total dose estimate from the accident becomes 3,400 rem to the receptor at 100 meters. ,

lon exchange is also used to separate and recover technetium from the tank wastes.

Manufacturer data indicates a maximum loading of 0.8 gew/ liter of anion resin, corresponding to about 80 g/ liter of technetium. For a 100 liter resin column, this becomes 8 kg, or about 140 curies. The source term becomes 7 curies, and the estimated dose from a fire on the loaded resin is 0.7 rem.

These estimates oo not include any contributions from alpha particle radiation, which can accelerate and increase resin degradation reactions. Consequently, if TWRS processing encounters significant quantities of TRU isotopes around the ion exchange systems, then aojustments may be necessary in the release. fractions.

5.1.4 Hydronen and Flammable Gases -

The tank wastes and intermediate streams within the TWRS facility generate hydrogen and flammable gases at various rates, from radiolysis and other reactions. Hydrogen generation rates are strongly influenced by the solution chemistry and the forms of radiation emitted by the radioisotopes. The presence of alpha emitters (e.g., TRU) greatly increases hydrogen generation due to the relatively high energy transfer of the alpha particle. Hydrogen generation estiniates range from 20-150 liters /hr, depending upon the tank the LAW receipt and Cesium Product tanks have the highest generation rates. These esti- ahow that hydrogen can accumulate to potentially flammable concentrations in less t' o  :

• Jay if active ventilation is not maintained. The presence of organic materials in some ne ,astes is expected to increase hydrogen generation while they are undergoing processng.

Point Paper -

May 27,1999 l

I

28 Accumulation of flammable gases followed by a deflagration event potentially increases the amount of material at risk and the release fractions as compared to a tank leak scenario. The

. lower flammability limit fur hydrogen in air is 4%. In contrast,30% hydrogen in air represents the stoichiometric limit, which corresponds to the maximum energy release per unit volume of ullage space. In a 100,000 gallon tank, these produce 1 and 10 kg maximum quantities of L hydrogen for the 4 and 30% hydrogen concentrations, respectively. Thus, the material ratio (essentially waste to explosive mass ratio) is very large (n 10) in this tank, if the tank heel is small (say, less than 500 gallons, with hydrogen sucked into the tank from the common header during a tank pumping / draining operation), the release fracti'o n (ARF x RF; see Appendix B) can

' be relatively high, on the order of 7E-2 (Reference 21). Thus, this " heel" condition represents

' one bounding condition for analysis (for current purposes, potential blast effects and collateral damage to and releases from other equipment are neglected). Using 500 gallons as the basis, the source term becomes approximately 3,000 curies from the feed tank, the receptor at 100 meters experiences a dose contribution of about 1,200 rem from the cesium. The potential contribution from the entrained solids (at a maximu,n of 5%) is about 19,000 rem, for a total of approximately 20,000 rem to the receptor at 100 meters. For a 500 gallon heel in the HLW receipt tank, the potential dose estimates would be approximately double. For a heel in the cesium product storage tank, the analyses use 0.5% of the volume (the tank is likely to be considerably smaller than the receipt tanks). The source term is about 6,600 curies of cesium, and the estimate corresponds to 2,500 rem.

For the scenario where there is a relatively large fraction of liquid in the tank, venting and containment failure best describe the situation as the potential exists for rapid temperature and pressure increases in the tank (potentially as high as 15 atm). The release fractions (ARF x RF) range from 4E-5 (< 3 atm overpressure, lower flammability limit [LFL] situation of 4-7%

hydrogen) to 2E-3 (3-30 atm overpressure, > 7% hydrogen). The LAW source terms are 390 curies and 20,000 curies for the two hydrogen concentration cases. The total dose estimates become about 2,300 and 100,000 rem to the receptor at 100 meters, for the two hydrogen concentration ranges. It should be noted that the likelihoods differ for the two hydrogen concentration cases; the low concentration cases (4-7% hydrogen) have a higher probability of occurrence than the higher concentration cases. The HLW results are approximately double.

5A5 CST Drvina The CST container is specified (Reference 5) not to exceed the general dimensions of 33 cm in .

diameter and 137 cm in height (13 inches by 54 inches), corresponding to a volume of 117 liters (about 4 cubic feet). Using 100 liters as the basis for the internal volume, the CST occupies i approximately 50%, leaving an interstitial void space of 50 liters. Application of a 1% hydrogen limit (25% of the LFL) implies a maximum, residual water weight of 0.4 grams. Use of a 150 psig burst limit on the container implies a maximum of 240 grams of residual water, with a post-radiolysis gas containing 10% of the initial cover gas,60% hydrogen, and 30% oxygen, and an energy release capability _of 3.3 megajoules (3,200 BTU). Specific container designs may provide for higher pressures and limits.

Point Paper May 27,'1999 I

h 9

29 The contract specifications list a maxirnum, radiolytic heat generation rate of 1.5 KW per CST container. Using this and 0.005 W/ curies as a basis, the CST contains 300,000 curies of cesium. The hydrogen generation becomes 3.6E-8 molelsec hydrogen (H2) with a gamma G value of 0.43 H2 molecules /100 ev, an LET (Linear Energy Transfer) of 0.52 Kev / micron, and an assumed water thickness of one micron on the CST particles (the actual water thickness may be larger). Thus, 0.4 grams would radiolyze to hydrogen and oxygen in about one week; radiofysis of 240 grams of water requires about 12 years (without consideration of extemal sources).

Essentially all other assumptions imp y a higher rate of radiolysis. Thus, hydrogen generation is a likely event if residual water remains within the CST. Efficient drying and monitoring methods require development prior to CST use.

The bounding accident scenario involves a hydrogen deflagration of a loaded canister, resulting in its rupture and dispersion of its CST powder, physical dispersion of 1 canister equivalent of feed solution, and collateral damage to two adjacent canisters involving rupture but no additional deflagration (i.e., gas venting through a powder effect). The estimated dose is 48,000 rem at 100 meters. Mitigation via HEPA filtration reduces this to 48 rem. Additional mitigation is likely to be required. CST operations require the development of technical specifications for drying

- and monitoring CST containers, and the items relied upon for safety would probably include the -

HEPA filtration on the cell, the moisture measuring instrumentation, and an overpack for the container.

i I

jd4 Ornanic Materials The tank wastes contain organic materials, which pose the following concems:

1

.- Adverse Effect uoon Reauired Process Efficiencies l 1

Organic compounds and their breakdown products adversely affect removal by depressing separation factors, due to either competition or inactivation of the separation  ;

1 methoo (e.g., organic compounds coating a resin). This impacts the worker safety and

- the public receptor at the waste disposal site location (see Section 5.1.3).

. . Potential Flammable Gas Generation 1

The presence of organic compounds increases the generation rates of flammable gases

, in the tank wastes, including hydrogt /. and methane. This raises the scenario of a deflagration that can result in doses exceeding worker and public regulatory limits (see l Section 5.1.4).

. Radiolvsis Products i

In addition to flammable gases, radiolysis products may include polymeric species (e.g.,

multi-ring compounds), extremely reactive compounds (e.g., nitrated esters), and Point Paper May 27,1999 i

~

w f 9 30 x ,.

~ potentially hazardous chemical species (e g., benzene). These can initiate events that result in doses exceeding regulatory limits.

These concems are manageable to potential doses below regulato levels provided the design is sufficiently robust and sufficient analytical information is available on the specific organic moieties involved. For example, destruction of the organic species (e.g., by oxidation) early in the process effectively precludes these concems from the tank wastes.

51Z Radiolvsis Proposed 'IWRS facilities contain significant quantities of highly radioactive species. Potential

. contact dose rates for pipes, vessels, and containers are likely to exceed .10,000 R/hr in the pretreatment area, and integrated doses in excess of 1E8 Rads are possible on an annual

- basis.1 Obviously, the facility design incorporates shielding (nominally 4 feet minimum) to protect the workers. However, items on the process side of the cell experience these high radiation fields and may suffer degradation. Degradation is likely to result in accelerated maintenance and higher worker exposure rates than expected. Examples include cables', connectors, graphite seal components (e.g., in rotating equipment and pumps), and elastomers (e.g., in the connectors between resin and CST columns). Radiolysis resulting in failure of the latter items in equipment produces energetic, spray type leaks of process fluid, which, if unmitigated, will allow radiological dose limits for the worker and the public to be exceeded. Quantification requires the review and analysis of specific designs which are not yet available. The WVDP ion exchange I dump valve experience demonstrates how degradation of elastomeric sealing materials in

- service can lead to unanticipated equipment failure and maintenance, with potential exposure and safety considerations (Reference 22). In this case, the PTFE sealing material around a valve degraded under the high radiation fields, resulting in the valve binding in a partially open -

position, preventing the use of a fourth zeolite column in one of the supemate processing campaigns. .

i.1.al Hiah Temperature Operations i Typical vitrification metters operate at 1,100-1,200 C and use 1-2 pours to fill a canister; the pour may even be semicontinuous.- The TWRS facility is likely to be designed to accommodate either the 3 meter WVDP-like canister (nominally 625 liters of glass) or a canister 50% longer, containing about 1,000 liters of glass (Reference 5). As a basis, the calculations here use a metter volume of 1,000 gallons (3',800 liters) of glass (based upon the DWPF metter, Reference 24), and a full,4.5 meter long canister of (molten) glass on the pouring stand (about 1,000 liters). Specific designs may have the melter and canister in the same cell (like DWPF and WVDP), or have the canister in a separate, pour cell below the melter cell, only connected i

through a bellows / airlock area (as practiced at some facilities in Europe). For conservatism, this analysis assumes they are in the same cell, for a total material at risk corresponding to 4,800

- (3,800 from the metter pNs 1,000 from the canister) liters of glass and a 0.25 inch thick by 3 foot Point Paper . ,

May 27,1999 l

l l

t 31 l

diameter (about 17 liters) liquid cold cap. The calculations use a 7:1 concentration factor for the cold cap transition to a glass (i.e., the glass is more concentrated). Calculations should be considered as very approximate due to the lack of experimental information on molten glass

. releases.

The first accident scenario considers simultaneous failure of the melter and the canister and

. - complete loss of the cold cap (based upon its rapid heating and volatilization due to the molten l- glass), with a free-fall spill of the molten glass to the floor of the cell, with resuspension of l _ particles and evaporation of volatile species (cesium and technetium) as the glass pool cools (I j hour duration). The unmitigated dose estimates to the receptor at 100 meters are 14,200 rem, 160 rem, and 110 rem for the cold cap, free-fall, and resuspension effects, for a total of 14,500 i-rem.

The second accident scenario considers an explosive release of materials from the melter due to failure of the water cooling jacket and a subsequent steam explosion. This results in greater aerosol formation. As before, the cold cap is completely lost due to the rapid radiant heating and volatilization in an uncontrolled manner. The aerosol formation uses a model based upon

. high pressure venting of a ges (i.e., the steam from the water leak) in a dense, heavy metal

' solution, based upon the melter's glass volume. Resuspension from the spill is assumed to be the same as before. The potential dose estimates become 14,200 rem,11,600 rem, and 110 rem for the cold cap, steam explosion, and resuspension, respectively, for a total of 26,000 rem.

i l The other metter related accident involves the drop of an unsealed, solidified canister, with the

! resulting resuspension of the glass fines within the canister. The release fraction (ARFxRF) is L 2.5E-5 (References 20,21, and 25). DWPF and WVDP used comparable values. It should be i noted that measured and expected thermal loads from WVDP and DWPF are in the 300-700

watt / canister range, but, with the radionuclide separations proposed for TWRS, it is likely that

. the early canisters may contain more of the separated cesium and strontium than the later canisters, and the bounding condition should be used. Table 4 estmates the radionuclide f content, based upon a maximum thermalload of 2.25 KW to a 4.5 rneter canister (volumetrically

! . equivalent to the 1.5 KW limit for the 3 meter canister). The estimated dose becomes 150 rem to the receptor at 100 meters.

I Point Paper ,May 27,- 1999

32 Table 4: Estimated Bounding Radionuclide Content for a 4.5 Meter High, Vitrified HLW Waste Canister (approximate heat load of 2.25 KW)

Radionuclide J Activity in' Curies:

• Strontium-90 2.17E+05 Technetium-99 360 Cesium-137 2.17E+05 ,

1 Plutonium-239 51 I Plutonium-240 34 Plutonium-241 6,600 kmericium-241 43 Curium-244 431 f

5.1.9 Nonradioactive Chemical Effec'ts The EPA provides methods for estimating the impacts from the releases of hazardous chemicals, such as ammonia and nitric acid leaks (Reference 26). Ammonia is a liquid under pressure, and vaporizes upon release to the atmosphere. In an arid region, such as Hanford, little fallout or absorption due to moisture would be anticipated, and dilution would be based upon the entire contents of the tank. Bounding calculations indicate the potential for toxic  !

ammonia effects (>117 ppm) extending out for several miles from the leakage of multi-thousand l l

pound quantities, and, obviously, the vapor cloud engulfs the TWRS-P and neighboring facilities.

Such ammonia levels would force the immediate evacuation of the TWRS-P plant and pose l

operability and control room habitability issues with the radioactive materials. A similar situation exists for nitric acid. The potential toxic effects from multi-thousand gallon spills of nitric acid are also estimated to extend out several miles, presenting evacuation and operability concerns for i TWRS-P facilities, and impacting receptors beyond the facility's fenceline. f I

SJ Analysis and Conseauences from Chemical Safety issues and Concerns with Contractor Proposed Conceptual Approaches The DOE contractors (BNFL and LMAES) proposed conceptual approaches for the TWRS-P j facility (References 2 and 3) similar to the generic concept. The BNFL design effort is )

continuing because of its selection as the Phase 1B contractor. The discussion uses the designs and status of the two contractors as reported at the end of Phase 1 A (References 2 and 3), focusing on the BNFL approach and only briefly discussing the LMAES concept.

Point Paper May 2' 7,1999 l

1 i

33 Both contractors used the Hanford Site boundary for the locatiorNf the public receptor 4 (approximately 12,000 to 15,000 meters from the proposed facility / release), and use the DOE

. concept of collocated worker at a nominal distance of 100 meters from the release. For some of-the postulated svents, the preliminary design approaches show a distance of under 20 meters from the release to the collocated worker. Both focused on short-term, inhalation pathways of 8-24 hour durations. In contrast, NRC regulations recognize members of the public and occupationally exposed, radiation-trained workers, but do not include collocated workers; all ,

individuals outside the contractor controlled area (typically, around 100 meters from the facility) i are considered to be members of the public. The BNFL documentation did not include I consequences for the facility worker, but qualitatively identified items important for worker safety.

Most of the scenarios analyzed by both contractors constituted prompt events with insufficient

)

warning to allow meaningful evacuation. The BNFL approach extensively relied upon elevated release via a stack for mitigation via dilution, while the LMAES approach used HEPA filtration for i mitigation. For design purposes and categorization of structures, systems, and components

- (SSC's), the contractors analyzed events without mitigation as a ground release. HLW/TRU source terms were somewhat uncertain and not well de"ned; BNFL excluded potential TRU l

. contributions to the LAW (from the suspended solids), wnile LMAES included them.

Categorization of SSC's are also less well defined, although LMAES may have followed a more conservative approach.

5J1 Radiochemical inventories and Larne Tanks in Contractor Proposed Desians -

BNFL's proposed facility includes two nominal 60,000 gallon tanks for LAW storage and three nominal 60,000 gallon tanks for HLW storage within the TWRS facility. The tanks are assumed to be 89% full for LAW, and 100% full for HLW, with radiocesium inventories of 2.3 and 0.68 megacuries, respectively, with cesium dominating the LAW impacts and TRU dominating the HLW impacts. Failure of one individual tank with a low energy, liquid freefall, followed by resuspension from a liquid pool, with a surface release, are assumed; for the public receptor 7.4 miles away, the estimated consequences correspond to 0.057 and 1.1 rem for a LAW and an HLW tank, respectively. For the collocated worker 100 meters away, the corresponding values

. are 99 and 1,820 rem for LAW and HLW, respectively. BNFL suggests using the stack as the mitigating means (via an elevated release) for the collocated worker in these accidents; this  ;

reduces the collocated worker dose to 0.036 and 0.69 rem for LAW and HLW tank failure, respectively.

This approach raises questions about bounding a,? anservative analyses. First, the calculations use a 7 molar sodium basis. However, the sodium concentration ranges from 3 to 14 molar, and, thus, the analysis should use values corresponding to the upper bound. This effectively doubles the LAW source term. Second, the LAW source term calculations do not include any contributions from suspended solids; if the suspended solids contain TRU's, the potential radiological consequences could be substantially increased. THrd, common mode

' failures are overlooked; events (e.g., earthquakes, structural collapse) that cause failure of one l tank may cause the failure of several (or all) tanks and the assumed mitigation means (the stack).' Fourth, a low energy event is postulated, using ambient conditions (25*C). It is likely j Point Paper ,

May 27,1999

\

34 that the spillage will oe more energetic and thh temperature will be higher (40 C or more), which willincrease the release. Fifth, the HLW release fractions are based upon a thick sluny, but the slurry is likely to be thin, and the activity occurs in both the solid and liqd phases. Use of a liquid release fraction would increase the release fraction (and the radiological consequences) for those isotopes by a factor of 4 or more. Sixth, the proposed mitigation means relies upon dilution. Filters are specifically excluded. This is inconsistent with normal nuclear industry practice, and the survivability of an 88 meter high stack during an event (e.g., an earthquake) is not readily apparent and a supporting analysis is not provided. Safety categorizations of SSC's are not clear. Potential radiological consequences may be underestimated by an order of magnitude or more.

BNFL also analyzed a situation where loss of cooling water resulted in boiling in the separated cesium (product) tank, which can contain upwards of 5 megacuries of cesium and 24,000 curies of technetium, producing public and collocated worker doses of 1.2 and 1,360 rem, respectively. BNFL identified a significant time period of 5-12 days for this event to occur, and dismissed safety features for the tank by noting the event would be detected and corrected by restoring cooling water before the tank could boil.

Additional analyses indicate this would not be a prompt event, but that it could occur within a shorter timeframe than anticipated due to the likelihood of higher temperatures in the tank contents and in the operating cells, and potential uncertainties in the waste composition. The release fractions and potential consequences may be underestimated because of the soluble nature of the radiochemical species, which is more readily carried into the vapor phase (as droplets and salts) by the boiling liquid. Nuclear industry precedent would imply monitoring

' systems capable of detecting loss of cooling water and temperature increases, and restoration of cooling, within a time period that avoids boiling of the tank's contents. This would probably involve several items relied upon for safety (e.g., temperature monitors) and technical -

requirements or limiting conditions for operations.

BNFL did not separately analyze the breach failure of a full cesium product tank. Insufficient information is available regarding comparisons of the cesium product tank with the LAW and .

HLW receipt tank failure accidents and their consequences. Such analyses might indicate comparable or even greater potential hazards from a breach failure of a full cesium product tank, l

and, in all likelihood, the tank would become an item relied upon for safety.

k The design proposed by LMAES also includes large tanks and radiochemicalinventories. {

Consequence analysis based upon tank failure and spray leaks indicates potential doses exceeding 30,000 rem, and many additional scenarios with potentially over 100 rem to the receptor at 100 meters. Safety categorizations were identified for these tanks and the ventilation systems, including the HEPA filters. Analyses indicate sirnilar conclusions as with the BNFL design. ,

t Point Paper May 27,1999

C i

\

35 512 Process Efficacy with Contractor Proposed Desians

• BNFL indicates the desired removal percentages will be achieved. However, this paper review raises questions about the pretreatment approach with the ion exchange columns and the removal of strontium and TRU isotopes, and the lack of process redundancy -i.e., downstream  !

processing is not capable of correcting a bleedthrough of radionuclides, so recycling (and throughput derating) might become necessary.' These ion exchange columns are relatively ,

large and have aspect ratios considerably below those recommended for the types of resins under consideration. At best, treatment consists of only two columns in series, with a 200%

freeboard in each column. It is possible that significant backmixing and premature radionuclide breakthrough will occur, resulting in nonattainment of contract required removal efficiencies, reworking of the batch, or excessive radionuclide release to the LAW vitrification areas of the

' plant. Technetium removal assumes the presence of the anionic, pertechnate ion. However, j not all of the technetium is present as anions, and cationic forms are unlikely to be removed.

The BNFL design uses. coprecipitation of strontium and TRU isotopes, and does not destroy l organic compounds in a discrete step; instead, they are routed to the melters for ultimate,  ;

l thermal destruction by vitrification. While coprecipitation may achieve the removal requirements l of Envelope C waste, it is not clear that adequate performance can be achieved with some of  ;

the high organic wastes (e.g., CCW) and there are no downstream unit operations capable of removing strontium and TRU. High concentrations of organic compounds may also adversely impact the other separation processes.

L LMAES also indicates the desired removal efficiencies will be achieved, albeit with a different process. There are some questions about the LMAES pretreatment approach, but there are also some positive attributes in the approach. First, cesium removal consists of 6 columns in series, using the RF (resorcinol-formaldehyde) resin in each bed, with implied high aspect ratios.

While the RF resin does experience some batch variability (References 8-10),6 columns in series essentially overcomes this concern, and, as with all resin systems, new types of resins i can always be substituted. Thus, the design incorporates layers of redundancy in the cesium  ;

removal train, and the desired remova; of cesium and minimization of solution re-treatment is .

I likely to be achieved without potentiai regulatory implications. The design also includes a CST based guard bed, which effectively functions as a seventh column in series. Second, the i approach oxidizes the organic compounds with ozone to precipitate strontium and TRU's. While the oxidation conditions need to be verified, and, potentially, another oxidant used (e.g.,

peroxide), the approach effectively eliminates potential compounding effects from the organic i compounds and should convert the technetium to a common form (the anionic pertechnate i species). Third, the LMAES route uses electroreduction for the removal of technetium. This is j l likely to be effective for cationic technetium. However, a good fraction of the technetium exists as anionic species - after oxidation by ozone, all of the technetium may be in the pertechnate form - and the overall removal of technetium will be inadequate for the vitrified LAW to meet the l performance requirements of the disposal site. An anionic removal process may be more l appropriate for the technetium. Fourth, the process uses several electrolytic steps to regenerate l resins. These generate significant quantities of hydrogen requiring mitigative controls.

Point Paper ,

May 27,1999

s 30 5.,11 QthWc lon Exchanne Resin and Nitrate interactions in the Presence of Very Hiah Radiplon Fields in Contractor Proposed Deslans ,

BNFL proposes to use commercially available, organic ion exchange resins for separation of the cesium and technetium from the LAW (Reference 2). Each column consists of approximately one cubic meter of resin placed in a vessel approximately three cubic meters in size, for the accommodation of resin swelling and density changes. There are a total of eight ion exchange colum,ns, arranged as four trains of two columns each, connected in series. All of the columns

- are located in approximately the same area of the pretreatment cell, without intervening walls or other major components. The proposed facility usos two trains for cesium removal and two trains for technetium separation; one train in each pair is online, while the second train is on standby - undergoing maintenance or resin changeout. Each train is organized in a lead / lag

configuration - once the lead column is loaded (based upon unspecified effluent analyses), the

' train would be taken offline and just the lead column regenerated. The train'is then readied or retumed to service with the partially-loaded lag column arranged as the lead column, and the

regenerated column now arranged as the lag column. The columns concentrate radioactivity.

The 'BNFL analysis considers the resin / nitrate reaction as occurring at the beginning of a regeneration cycle (after passage of one column volume of nitric acid) initiated by a loss of cooling water, and only involving one column. The analysis presents a situation consisting of 10% of cesium in solution and 90% of the cesium remaining on the resin, when degradation occurs and ultimately results in an explosion. The release consists of flashing solution spray and hot gases from the burning resia, both containing cesium. BNFL estimates the consequences as 0.62 and 1,120 rem for the public and collocated worker doses, respectively;

. 29% of the dose comes from the cesium in solution. BNFL identifies mitigation via ventilation system routing of the release to the stack, and atmospheric dispersion and dilution. The stack and ventilation system (without any filters) become items relied upon for safety.

The consequence analysis raises several questions. First, the accident is initiated by a loss of cooling water, but no cooling means are identified in the column design, and this requires further - '

explanation.. . Second, the cesium sventory needs additional discussion and elucidation. The large freeboard in the column impnes the potential for a greater fraction of the cesium in solution than 10%, potentially as high as 50%. Third, the situation is identified as a resin exotherm, but l flashing spray, release values equivalent to an endotherm are used; if the resin dries out, as stated _in the scenario, the liquid release fraction (ARF x RF) will be greater than the assumed ]

value of 0.033.1 Fourth, the resin release fractions use an example for uranyl nitrate ,

hexahydrate, and do not account for the expected greater volatility of cesium. Fifth, the l

- analyses do not evaluate the potential effects upon the second column in the train, the adjacent j trains, or other adjacent equipment (i.e., common mode and common cause effects). Sixth, elements of the design do not match the smallion exchange columns typically used in the nuclear industry, and the actual resin volume ir' the design has a low exterior surface area to volume' ratio.- Lastly, the mitigating features rely upon atmospheric dispersion and dilution, and

. specifically exclude filters; this is not normal practice in the nuclear industry. No additional, independent safety features are identified. Further evaluation of the resin systems seems warranted for identifying other potentialitems relied upon for safety.

Point Paper May 27,1999 1

y 37 LMAES proposes the use of the RF resin, and identifies a resin hre scenario with an unmitigated dose estimate of 200 rem for the collocated worker. LMAES also identifies several HLW/ nitrate / organic powder interactions that produce collocated dose estimates in excess of

' 1,000 rem; this is due to the use of a calcination / induction metter combination, with sugar used for redox and particle size control. LMAES uses single stage HEPA filtration for mitigation (a standard nuclear industry approach), which reduces the dose to under 1 rem.

l W Hydronen and Flammable Gases in Contractor Proposed Desians BNFL does not analyze a hydrogen and flammable gas event; hydrogen generation by

l. radiolysis is estimated to be too slow (hours to days) to achieve flammable levels, and the resin l column fire is used to bound the hydrogen bum events. However, this paper has conducted additional analyses which indicate this may not be the situation. Hydrogen generation by

' radiolysis will continually occur throughout the plant, with the highest rates in the pretreatment areas: Analyses based upon gamma radiolysis and chemistry assumptions imply LFL limits are approached in tank ullage spaces fairly rapidly and potentially within hours, assuming a loss of ventilation. BNFL specifically excludes a safety designation for the process vessel vent system.

Given the uncertainties of radiolysis, the waste chemistries, and the potential contributions from TRU isotopes, it seems prudent to consider the potential effects from radiolytic hydrogen generationdnd means for its detection, and prevention and mitigation of potential deflagrations.

This is likely to identify some items relied upon for safety in the design, potentially including the vent system and hydrogen detectors, and some technical specifications (e.g., percentage of LFL allowed, outage time for ventilation system, etc).

The LMAES design considers hydrogen deflagrations in several vessels, with several collocated worker dose estimates exceeding 1,000 rem (principally from tanks that contain HLW suspended solids). Dose estimates from hydrogen evolution from electrical processes used in pretreatment are relatively low (1-3 rem). LMAES cites (but does not use) release fractions some 40 times higher; analyses indicate these higher values appear to be more appropriate for

- bounding-type analyses early in the design process. LMAES does categorize vessels and various associated equipment (e.g., fans, detectors, air injectors) as safety class. LMAES uses ,

single stage, HEPA filtration as the principal means for mitigation. Additional means for mitigation and defense in depth are likely to be necessary for the tanks with larger source terms.

W CST Drvino at Contractor Proposed Desians The BNFL process relies upon draining of the fluid in the CST canister and flow of air through

- the bed until no more moisture is detected. It does not identify any hazards asxciated with the drying of CST canisters, but does analyze a CST canister overpressurization event based upon inadequate drying and radiolysis of water (no deflagration), resulting in rupture of the container.

.BNFL analyses indicate doses of 2.4 and 4,310 rem for the site boundary and 100 meter receptors, respectively. BNFL identifies mitigation via ventilation system routing of the release to the stack, followed by atmospheric dispersion and dilution. The stack and ventilatior) system

- Point Paper May 27,1999 1

W.

o 1

38

'become items relied upon for safety. The potential benefits of filtration.are specifically excluded and not relied upon for safety, This approach needs further elucidation and identification of process parameters. First, parameters for drying of the material require development, including appropriate monitoring means; as presently proposed,5-10% of the weight may accrue from bound water, which would undergo radiolysis and accumulate in a closed container. Second, the analysis should consider common mode failure effects, including the' potential for several canisters to be ineffectively dried and/or affected by the same accident mechanism (e.g., earthquake, canister drop). Third,

' the scenario does not include collateral damage effects to adjacent canisters impacted by the overpressurized canister. Fourth, increases in release fractions need to be considered due to the potential for greater drop heights and hydrogen deflagration. Lastly, the mitigating features rely upon atmospheric dispersion and dilution, and specifically exclude filters, which is not normal practice in the nuclear industry. No additional, independent safety features are identified. Further evaluation of the CST systems seems warranted for identifying other potential items relied upon for safety, l

' LMAES uses electrically heated al i

ry the CST at temperatures up to 200 C. This is likely to remove a considerable fraction of the bound water on the material. However, operating limits and specifications require development. LMAES identifies a hydrogen deflagration, unmitigated accident dose of 4 rem to the collocated worker. Dose consequence increases seem  ;

appropriate for the quantity of material at risk and the release fractions. However, LMAES identifies HEPA filtration as a mitigating means. Further evaluation of the system would be desirable to verify the adequacy of the approach and identify any other potential items relied upon for safety.

_ 5_.24 Oraanic Materials in Contractor Proposed Desians BNFL does not identify any concerns or process effects with organic materials, and assumes that they are destroyed in the vitrification processes. It is not clear if the potentially deleterious effects of organic compounds upon separation efficiencies have been fully evaluated for the

- BNFL design. LMAES concludes that destruction of the organic compounds early in the processing is beneficial and allows for easier separation of the strontium and TRU materials.

l LMAES proposes to use ozorie for this purpose. Analyses indicate both contractors have assumed uniform levels of organic compounds, whereas actual operations are likely to experience batches of wastes with organic compound levels significantly above the average, perhaps by as much as an order of magnitude, and, thus, it is prudent to address the potential effects from organics proactively. The effectiveness of the use of ozone under alkaline conditions requires additional test information due to the potential for catalytic destruction of ozone by the hydroxide ion.

.May 27,1999 Point Paper ,

4 0

39  ;

ill Radiolvsis Effests at Contractor Proposed Desians BNFL acknowledges high radiation fields but does not verify the lack of potentially deleterious effects.' LMAES notes the effects of radiolysis upon hydrogen generation. Both designs include elastomeric components containing chlorides and fluorides (e.g., in the metal hoses / quick disconnects in several of the cells). Additional reviews will have to be performed as the designs mature to verify the lack of identified radiolysis concerns. 1 ill Hinh Temperature Operations in Contractor Proposed Desians BNFL analyzes one HLW melter accident involving failure and spill of th'e molten material onto i the floor of the cell. Cesium and technetium volatilize from the spill, based upon estimates from the Savannah River site. The analysis assumes nonvolatility of strontium and TRU isotopes, and, thus, cesium and technetium are the only contributors to the dose. No energetic effects are included. The dose estimates become 4.6 and 0.008 rem at the 100 meter and site boundary  ;

receptor locations, respectively. The analysis does not quantify exposure to workers from the i scenario, but states it is likely to exceed the 4.6 rem result. The analysis does not identify mitigating design features, but indicates several features ma/b e relied upon for safety. It should be noted that glass spills are messy but generally have low dose consequences, and principally impact the workers. This occurs because of the presence of shield walls and the maintenance of negative pressure in the cell by the ventilation' systems.- Consequ.ently, it is likely that the cell walls,' shielding,' and ventilation systems will be identified as items relied upon for safety. Also, for this scenario, confirmatory calculations should be performed to verify minimal contributions from TRU and the lack of energetic means of resuspension, such as steam formation / explosion from cooling water leaks and the potential dispersion of the cold cap if the radioactive feed is not discontinued before the postulated event.

LMAES has not analyzed a glass spill accident for the HLW melter in their design. Again, such a scenario needs to be evaluated because of the high curie inventories involved and the potential for impacts upon worker dose exposure.

All Nonradioactive Chemical Effects with Contractor Proposed Desians i

BNFL analyzed two toxic chemical releases from outside storage tanks. The first involves the breach of a 25,000 pound liquid ammonia storage tank. Using the value of 200 ppm ammonia in air as the toxic endpoint (the ERPG-2 limit) and EPA methodology, ammonia concentrations

. from the breach will exceed 200 ppm for 5 miles from the breached tank. BNFL concluded this was not a concern for the public (minimum distance to the public of 5.9 miles, based upon the use of the Hanford Site boundary), but the tank would have a safety categorization for the protection of the workers and the collocated workers. BNFL also presented an analysis for the i breach of a 5.000 gallon tank containing 12.2 molar nitric acid. Using EPA methods, calculations indicate the TEEL-2 limit of 15 ppm would be exceeded for up to 320 meters from the tank, and, thus the tank would have a safety categorization commensurate with the ,

Point Paper ,

May 27,1999 ,

~

40 protection of the workers and the collocated workers. BNFL identifies hardened censtruction of the ammonia and nitric acid tanks to survive design basis ecrthquakes and unspecified physica!

damage.

A review of this approach indicates both accidents are prompt events (i.e., no development time for warnings and evacuations) and could affect receptors beyond the fenceline of the facility (100 meters away). The effects from ammonia potentially extend out for miles. Without mitigation, the TWRS-P facility could be rendered uninhabitable for a period of time.

Consequently, the safety categorization should be commensurate with protection of the public.

The BNFL consequence analysis of the nitric acid tank breach appears to be based on standard conditions (80 F) for a tank maintained in a building instead of the actual design of a tank effectively outside, and, consequently, the analysis should be based upon the potentially higher temperatures (120 F or more) of an outside tank under summer conditions, which will increase the distances affected. The approach needs to consider and evaluate structures around the

' tanks for potential"two over one" effects circumventing tank hirdening benefits. Ideally, a more conservative design approach should be used, with evaluations analyzing the potential advantages from tank enclosure in a building, deluge sorays, and means for protecting TWRS-P operators and providing safe areas for employees.

LMAES did not specifically analyze chemical-only accidents. However, the LMAES accident analysis includes a blast wall protecting the process building from breaches of flammable gas tanks. Insufficient detailis provided on the other chemicals. However, the LMAES approach -

includes ozone for the destruction of organic compounds, with ozone lines routed through the building. This presents habitability concerns from leaks and failures. The design also incorporates high pressure cylinders for breathing air supply, but does not consider potentially deleterious effects from cylinder failures, such as from valve shearing during a seismic event (the cylinder racks are not hardened) and the cylinders as missiles. The LMAES design also uses ammonia for SCR of NOx; however, the design stores the ammonia as an aqueous solution, thus effectively precluding the major concerns associated with anhydrous ammonia.

Although not analyzed, the LMAES facility uses comparable quantities of nitric acid.

Consequently, potential accidents from tank failure and the asse .iated safety-rstated features need to be c6nsidered and evaluated. As with the other contractor's approach, the chemical accidents are prompt events (i.e., no development time for warnings and evacuations) and can affect receptors beyond the fenceline of the facility (100 meters away). The use of ammonium hydroxide instead of ammonia significantly reduces the potential consequences and distances.

' However, as before, without mitigation, the TWRS-P facility would be rendered uninhabitable for a period of time. Consequently, the safety categorization should be commensurate with protection of the public. The approach needs to consider and evaluate structures around the tanks for potential"two over one" effects circumventing tank hardening benefits Ideally, a more conservative design approach should be used, with evaluations analyzing the potential advantages from tank enclosure in a building, deluge sprays, and means for protecting TWRS-P operators and providing safe areas for employees.

4 Point Paper May 27,1999 l

i

l- .'

l l 41 l

l !J Methods for Prevention and Mitiaation i 1

1 5.3.1 Prevention and Mitiaation of Events involvina Radiochemicalinventories and Larae Tanks l The tanks consist of welded components, and the larger tanks wiil be field fabricated. Typical i welds of steels have failure rates of 1E-2 to 1E-3/yr in intended service situations. High quality l welding of stainless steel components and weld inspections will be necessary at TWRS, and should allow a 1-2 order of magnitude reduction in the probability of weld failure. This reduces the tank failure rate to the 1E-3 to 1E-5/yr range. For comparison, chemical process industry data shows ucintended rupture rates of 1E-4 to 1E-5/yr for general tanks, pipes, and reactors i

(Reference 27 Given that the tanks themselves will be quality components on a nuclear grade, I

quality assurance program, it seems reasonable to use the 1E-5/yr value at the lower end of the ranges. The number of tanks increases the rate proportionately; a TWRS facility is likely to have at least two tanks in service, so the estimated tank rupture rate for the facility becomes 2E-5/yr.

Prevention requires a reduction in the probability of occurrence of events involving large ,

inventories of radionuct; des. With leak detectors and other monitoring equipment, testing, and )

suitably conservative design (e.g., field annealed, thicker wall / corrosion allowances), an order of magnitude reduction in occurrence should be possible, reducing the probability to around 2E-6/yr.

A loss of cooling / boiling tank scenario is the only accident at a TWRS facility that is not prompt; it takes a finite amount of time to develop and release its source term, and gives the operators time to conduct corrective actions. Therefore, the frequency of occurrence should be low, and the credible / incredible frequency boundary is used (1E-6/yr). Prevention of a boiling tank scenario relies upon multiple cooling loops, backup cooling systems, instrumentation, and administrative controls (e.g., to connect backup or fire water to the cooling loop, transfer / pump contents). These are likely to reduce the frequency significantly, but, in the absence of a specific design for analysis, the credible / incredible frequency boundary (1E-6/yr) is still used. It should be noted that loss of cooling incidents with HLW have occurred, one of which resulted in significant releases to the environment.

Mitigation reduces consequences. The standard nuclearindustry approach uses cell i confinement with negative pressure, and exhaust through HEPA filters. This approach provides l a source term reduction by at least a factor of 1,000; a second HEPA filter system in series usually achieves another factor of 100 reduction, for a total of 100,000. The HEPA filtration system usually exhausts via a small stack on the roof, which reduces doses by dispersion (i.e.,

dilution). Additional equipment may be placed in the exhaust system to improve either its effectiveness or reliability, such as heaters, coolers, cyclones and impactors, electrostatic l precipitators, roughing filters, and sand filters. Tank size and facility inventories can be reduced l to decrease source terms and subsequent dose estimates from potential events.

Point Paper . May 27,1999 l

l

i i

+ '42 512 Potential Methods to Address Process Efficacy Concerns

- Process ' effectiveness requires a conservative design capahle of addressing process variations and upsets with a minimum of rework and recycle. Specific events, their consequences, and prevention and mitigation methods require more detailed information than is currently available.

Process testing is necessary, under both expected and off-normal conditions, with identification '

of key process parameters. Repeatable results are desirable.. A suitably conservative, design approach accommodates the expected range of process conditions using the least favorable .

(but workable) experimental results. For the TWRS facility,'this is likely to result in the following:

. hiah asoect ratio ion exchanae columns (i.e.! tall and thin. substitute oressure droo for a low effluent concentration)

The Oak Ridge experiments (References 8-10 ) achieved better results with higher

' aspect ratios (15), which is identified as a minimum. This implies that the cell area (s) housing the columns should be able to accommodate the height and allow sufficient

' headspace for other components, piping, and installation / replacement. It is not clear if there is sufficient headspace in the pretreatment areas of the current, contractor-specific designs.

a minimum of 3 columns in series The performance and behavior of organic resins in columns of any significant size have

' not been tested and evaluated, and the chemistry and mechanisms are still not completely understood. Technetium removal requires a process capable of removing it in all oxidation states or conversion to the same state prior to removal. Since all potential waste'combinatio'ns cannot be either tested or foreseen, conservative assumptions are necessary to allow for uncertainties, and they also provide flexibility. Use of additional ion exchange columns in series is a relatively simple and straightforward approach for providing consetvatism and flexibility. As an example, the WVDP uses this approach with four columns in series; most of the removal (of cesium) occurs in the lead column, f and the second and third columns function as polishing units (the fourth column was rendered inoperative by valving problems); WVDP attains h!gh cesium decontamination

[ factors with a variety of feed chemistries and concentrations.

a ouard bed / column / unit _ 5e LAW treatment orocess Effective pretreatment is needed to meet the LAW immobilization limits and near surface disposal criteria, and to' avoid a potential radiation dose hazard if some of the downstream facilities are designed with less shielding and more operator involvement.

This implies an additional guard treatment to prevent breakthrough contamination of the LAW. In a three or more column design, the last column accomplishes this function.

a minimum of two column trains

' Flexibility should be built into the ion exchange system. It is advisable to include piping {

and valving that allows a number of combinations to be used. An additional column train

' and valving provide operational flexibility (resin changes, elution, rinsing, etc.),

continuous operation, and higher operating throughputs.' It also allows protection for Point Paper. ' May 27,1999

< o

I .

1 i

l 43 unplanned equipment events; for example, WVDP was able to recover from two valve'

! failure events due to its train / valving design and the flexibility of a four-column system.

- prevention and mitiaation of the potential effects from oroanic comoounds (most orobab!v by oxidation / destruction) l The organic compounds and their degradation products offer the potential for affecting process chemistries and separation efficiencies for some of the wastes. The effects may depend in a non-linear manner with the concentrations and organic species present in the wastes, and, thus, may be difficult to predict for most of the HLW. For example, EDTA may be rendered ineffective at iron and free sodium concentration levels that have little impact upon HEDTA, bidentate, and tridentate species. The best approach appears to be the prevention of effects by removal or destruction of organic compounds early in the pretreatment process.

additional filters Pre- and post-filters should be installed around the ion exchange columns. These protect the resin from filling with particulates (on the inlet) and prevent carryover of resin fines and insoluble degradation products.

- adeauate intraolant surae capacity The benefits of plant surge capacity have to be evaluated against the risks of increased I I

inventory, particularly with items in mobile, solution forms. However, for the numerous rinsewaters and the pretreatment solution, this may be an acceptable approach (i.e., the l benefits outweigh the risks) due to their (lower) activity and the semibatch nature of their l respective treatment processes (evaporation and vitrification). Both WVDP and DWPF have identified adequate surge capacity as desirable, even for the vitrified waste products.

multiole melters i Melter maintenance and replacement is likely to involve significant down time. As discussed in Appendix A (Sections A.3 and A.4), plant flexibility is improved by having l multiple melters.

5.3.3 Prevention and Mitiaation of Events involvina Oraanic lon Exchanae Resin and Nitrate Interactions in the Presence of Very Hiah Radiation Fields i

Prevention relies upon passive and active design features. Passive features include a maximum diameter (if a column) or thickness (if a slab or an annulus) that minimize deviations between surface and centerline temperatures. Finning (internal and external) and fluting constitute other methods to increase cooling area to column volume ratios. Cooling (by a jacket or internal tubes), a monitoring / cold water injection system, or a dump / quench approach represent active design features. The cooling water system may become an item relied upon for safety, with redundancy and diversity. Cumulative resin exposure (chemical or radiological) and l time limits (e.g., for the resin to be exposed to nitric acid or remain loaded) denote administrative Point Paper May 27,1999 I

l

7~

44 limits. A specific approach and design are necessary for estimating the frequency of occurrence. In the absence of such information, historical experience (Reference 23) implies an unlikely frequency (1E-3 to' 1E-5/yr range), and, given the reporting of nitrate-resin interactions in the past 40 years and the potential presence of 10-20 columns in a TWRS facility, the upper part of the range seems appropriate (1E-3/yr) for a " standard" system without specific prevention measures. Note that a plant operating period of 40 years would increase the probability to over 1E-2, which is often considered the boundary for an anticipated event.

Specific prevention measures require a more detailed level of design incorporating some of the aforementioned features,and would reduce the frequency of occurrence. In the absence of specific design features and as an approximation for specific prevention measures,- the analyses assume an order of magnitude reduction in frequency (i.e., to 1E-4/yr).

Potential mitigation methods also include' passive and active design features. Smaller cclumns minimize inventory of the resin (i.e., the fuel) and the principal hazard (the radiocesium). Since the quantity of resin will most likely remain uniform during the processing, simple administrative limits should be definable.. The cell provides a passive barrier protecting the workers and the public. A separate column enclosure (with or without a filter) allows for expansion of the reacting resin / nitrate mixture after column failure but maintains confinement. Active features

include water sprays (on the ruptursd column or inside the cell / enclosure), pool suppression, offgas water scrubbing, rupture disks / blow-out panels, additional / enhanced cooling systems, the ,

j

'ventilation system (fans and stacks),and HEPA filtration on the offgases. Some of these active systems would be activated automatically by temperature, flow, colorimetry, or other instrumentation, and require a specific design for analysis. The standard nuclear industry approach uses cell confinement with negative pressure and exhaust through HEPA filters. This I approach provides a source term reduction by at least a factor of 1,000; if needed, a second

. HEPA filter system in series usually achieves another factor of 100 reduction, for a total reduction of 100,000.. The HEPA filtration system usually exhausts via a small stack on the roof, '

which reduces doses by dispersion (i.e., dilution). Additional equipment may be placed in the exhaust system to improve either its effectiveness or reliability, such as heaters, coolers, cyclones and impactors, electrostatic precipitators, roughing filters, and sand filters.

5.3.4' Prevention and Mitiaation of Events involvina Hydronen and Flammable Gases Hydrogen and flammable gas generation will occur throughout the facility (a probability of unity), )

with relatively high rates associated with higher cesium and TRU concentrations. High organic l wastes (e.g.,~ CCW) will generate hydrogen at higher rates, and will also generate other flammable gases. Most of the concerns occur in the pretreatment area of the plant. There are essentially no passive, preventative features - active ventilation of vessels containing significant quantities of radionuclides is'necessary. Redundant ventilation systems and monitors provide j

the necessary airflows. Instruments (e.g., bubblers) and fluidic devices introduce air into the tanks and these may assume a safety class function if specific design analyses indicate their function is relied upon for safety. This would also require a safety designation for the power and air supplies. Analyses for specific designs might include an inerting system as a backup to failure of the ventilation systems.' Frequencies for standard systems are likely to be in the 1E-Point Paper ,

May 27,1999

r l 45 S/yr (Iow hydrogen concentration above the LFL) to 1E-6/yr (high nydrogen concentration above the LFL) range. Prevention by safety class and redundant systems reduces the frequencies by

at least an order of magnitude; as a basis, the credible / incredible cutoff of 1E-6/yr is used for both low and high hydrogen concentration events. The safety class and redundant systems might include a backup ventilation system, eductors, and fluidic devices - the latter two are l

pneumatic and supply dilution air.

\.

l Potential mitigation methods rely upon the ventilation systems, double HEPA filtration, and rupture / blowout panels. If the event occurs, the overpressure effects are assumed to render one of the HEPA systems ineffective, but the second system is assumed to continue functioning and provide a reduction factor of 1,000 for particulate contamination.

l 5.3.5 Prevention and Mitlaation of Events involvina CST Drvina CST drying entails similar hazards to the preparation of spent nuclear fuel (SNF) for dry storage, such as high radiation fields and the need for the removal of hydrogen sources. Thus, methods used for SNF dry storage seem reasonable for consideration in the CST case.

Prevention relies upon operations and procedures, as there are no passive means to avoid CST drying concerns. Following the SNF model, CST drying is likely to use the following, in order:

• Draining of the CST canister, until free-flowing water is removed.

• Dry air or nitrogen flow through the canister.

• Monitoring systems on the effluent for its moisture content, as a means for following the water remaining with the CST.

. Higher temperatures, to drive off both residual water (at 100-120 C) and associated water (at 150-220 C). This requires heated air or nitrogen. Heated gases also accommodate higher quantities of water vapor.

. Cessation of the gas flow, and vacuum drying.

. Inert gas backfill and sealing of the canister.

The inert gas may be either nitrogen or helium; helium is preferred for long-term storage due to its superior heat transfer properties. The expected frequency is in the unlikely range; as a first approximation, similar to the unanticipated vessel failure rate of around 1E-5/yr (Reference 26).

Application of safety class systems and technical specifications are likely to reduce this about an I

' order of magnitude, to circa 1E-6/yr. Safety class items could include the moisture detector (s),

drying gas heaters, and inerting system, while technical specifications might include the temperature of the drying gas and the moisture content level for terminating drying.

Potential mitigation methods rely upon the ventilation systems, double HEPA filtration, and rupture / blowout panels. If th'e event occurs, the overpressure effects are assumed to render Point Paper May 27,1999

P .

)

1 46 one of the HEPA systems ineffective, but the second system is assumed to continue functioning and provides a reduction factor of 1,000 for radioactive particles.

111 Potential Methods to Address Concerns with Ornanic Mate _rjgta l The Hanford tank wastes contain numerous organic compounds and their degradation products from radiolysis, thermal, and chemical effects. Current estimates for organic compounds range between 1,500 and 1,800 tonnes, and include some 63 organic raw materials and 23 organic -

, chemical reaction products. Few direct analytical measurements have been made. There are certain wastes (e.g., CCW) which contain significantly greater concentrations and .quantities-than the average for all of the tanks. The potential consequences from high organic

s. concentrations would be similar to the loss of process effectiveness or a flammable gas deflagration (Sections 5.3.2 and 5.3.4). There are three approaches which need to be followed to address concerns with organic materials:

• Define the standard and conservative limits for the tank waste envelopes containing f

, organic materials, and verify by experimental testing; these limits effectively become

. operational or technical safety requirements (OSR's and TSR's), and operation outside these limits is not permitted. This may include process variations for some portions of the envelope, and may be based upon limits for avoiding exotherms upon heating and effects upon the melter redox chemistry, in addition to process efficiency effects, a Analyze for the organic materials by both sampling and online instrumentation.

Instrumentation will likely be more sophisticated than TOC (Total Organic Carbon),'and may include chromatography or mass spectrometry methods, converted to use in highly radioactive environments and shielded cells. Analysis verifies compliance with the organic material limits derived from experimentation.

- Design a robust process that is not affected by the presence of organic materials in the waste. This is likely to involve removal or destnJction of the organic materials as one of the first processing steps. Destruction by oxidation is one such method, and has the additional benefit of converting all of the technetium to the anionic form for more efficient removal by the technetium removal resin.

1.1Z Potential Methods to Address Radioivsis Concerns Radiolytic effects are best addressed by avoiding reagents and materials with known sensitivities to high radiation fields and by verification testing. Radiolysis of water is expected, but the exact products and rates in the presence of tank waste species are largely theoretical, and uncertainties exist. Conservative margins are desirable. Polymeric materials, if used (as in valves), should be carefully screened and tested. If radiation sensitive reagents and components are used (e.g., organic ion exchange resins), then their behavior should be well Point Paper May 27,1999

- I l

l 47 investigated via testing and closely monitored during W/RS use via technical specifications l (TSR's and OSR's).

5.3.8 Potential Methods to Address Events and Concerns with Hiah Temperature

, l Operations The'high temperature hazards associated with operating the melters and offgas systems have some inherent risks. At over 1,000'C, the molten glass can ignite many customary materials of construction. Ordinary electricalinsulating materials become a serious fire hazard in this situation. This implies that electric heaters and instruments (e.g., thermocouples) require special sheathes, insulation, and standoffs to avoid contact with the hot areas of the metter and offgas treatment system. Even with appropriate materials selection, the service life of many components is likely to be severely shortened by these high temperature applications.

Differential expansion effects of the different components and materials also have to be considered.

The frequency of melter failure events is not well known due to the limited operational experience with joule heated metters. Failures have occurred, albeit with pilot plant or prototype designs in LLW processing applications. Failures have also occurred with the smaller induction melters used for HLW vitrification in European facilities. The consequences involved slightly higher worker exposure. The metter experience to date implies that most metter failures are detected and the electricity disconnected prior to large releases of the glass and radionuclide dispersion. The melters include a cold cap of concentrated aqueous feed on top of the molten glass to help minimize the volatilization and carryover of cesium and technetium prior to their incorporation into the glass. Although not estimated,' controls would also stop the waste feed to the metter, essentially reducir'g the cold cap and its dose contribution, and replace it with process water (to perform the cold cap function of suppressing cesium and technetium volatilization) until safe shutdown of the melter is achieved. The melters could be considered to be like vessels and have an unmitigated, catastrophic vessel failure rate of 1E-5/yr. However, the experience from the metal and ceramics industries implies higher frequencies of major failures (i.e., involving spills, and major cleanup and repairs), in the 1E-2/yr to 1E-3!yr range.

Given that a melter for a 'lWRS P facility would be designed, constructed, and inspected to meet high quality assurance standards commensurate with the nuclear industry, the lower end of the range is used (1E-3/yr) as the frequency for failure without additional preventative measures. Addition of preventative measures, such as instrumentation loops and shutdown before failure, would reduce the frequency; in the absence of a specific design, an order of magnitude reduction is assumed, giving a mitigated failure rate around 1E-4/yr.

Steam explosions are possible with water-cooled metters and furnaces. The WVDP and DWPF assume that steam explosions are incredible events with rates below 1E-6/yr. However, steam explosions have occurred with water-cooled equipment (including melters) in the steel and ceramics industries, and, thus, the unmitigated frequency of gross water / steam induced failures may be as high as the 1E-2/yr to 1E-3/yr range. Assurning the melters would be acquired under a quality control program commensurate with nuclear industry standards, with rigorous design Point Paper May 27,1999

c ,

E 48

' (e.g., greater corrosion allowances), testing, and inspection, the analyses use the low end of the range (1E-3/yr).: Mitigation might include additionalitems notnormally present on commercial designs, such as infrared and activity monitors, and high speed shutdown or dump systems, in addition to confinement systems? Consequently, a lower frequency of 3E-5/yr (the geometric mean of 1E-3 and 1E-6, the latter being the incredible / credible boundary) is used for mitigated conditions.

Potential mitigation methods rely upon the ventilation systems, double HEPA filtration, and rupture / blowout panels. If the event occurs, the overpressure effects are assumed to render one of the HEPA systems ineffective, but the second system is assumed to continue functioning and to provide a reduction of 1,000 in the release of radioactive particulates.

jJJ Prevention and Mithtion of Evuts'involvina Nonradistive Chemical Effects ,

The main hazards of cold chemicals involve the storage of bulk quantities of ammonia and nitric acid, in tanks, because of their use in potential quantities sufficient to cause a serious chemical accident and potentially initiate radiological events and accidents. The storage tanks consist of

welded components, and the larger tanks will be field fabricated. Typical welds of carbon and stainless steels have failure rates of 1E-2 to 1E-3/yr in intended service situations. High quality welding of stainless steel components and weld inspections will be necessary at TWRS, and i

should allow a 1-2 order of magnitude reduction in the probability of weld failure. This reduces the tank failure rate to the 1E-3 to 1E-5/yr range. For comparison, chemical process industry data indicates unintended vessel rupture rates of 1E-4 to 1E-5/yr. Given that the tanks themselves will be nuclear quality components on the _ quality assurance program, it seems reasonable to use the 1E-5/yr value at the lower end of the ranges. The number of tanks .

increases the rate proportionately; the quantities of ammonia and nitric acid required'at a TWRS 1

facility could require one large tank each, or several smaller ones. Smaller tanks and inventories may require more frequent shipments. Assuming just one large tank each, the estimated tank rupture rate for the ammonia and nitric acid tanks becomes approximately 1E-5/yr per tank.

Prevention requires a reduction in the probability of occurrence of events involving large inventories of chemicals.- Quality tanks will be used for hazardous chemicals like ammonia and

~

nitric acid, constructed and inspected in accordance with the ASME code. With quality tanks incorporating leak detectors and other monitoring equipment, testing, and suitably conservative design (e.g., field annealed, thicker wall / corrosion allowances), an order of magnitude reduction in occurrence should be possible, reducing the probability to around 1E-6/yr.

Mitigatiori reduces consequences. The standard chemical industry. approach uses enclosure (in effect, cell confinement), sometimes with exhaust up a small stack. Since both ammonia and nitric acid are hydrophyllic, water spray systems are also very effective at minimizing airborne releases, and could be contained within the enclosure (again, a typical chemical industry approach, particularly around tank car loading and unloading stations). The small stack functions in a similar manner as its nuclear counterpart, reducing vapor effects by dispersion l Point Paper May 27,1999 ,

I l

I

n i

53 under all credible conditions, then this would reduce the number of items relied upon for

. safety.

l lon Exchance Column Events: different controls can be used to provide adequate l assurances of safety with the ion exchange columns. Preventative controls rely upon l cooling, detection, and dilution / dispersal of the reacting resin mass. Resin performance L improves with lower feed solution temperatures (20-25*C or lower); cooler temperatures l . also reduce the probability of resin degradation reactions. The cooling means, heat exchanger, and temperature indicators are likely to become items relied upon for safety.

The column itself is likely to be a preventative control, because it confines a significant amount of radioactivity (i.e.,- once loaded - to avoid a loss of confinement) and the design may incorporate specific design features to cool or separate the resin bed and prevent l runaway resin reactions. If mitigation relies upon an enclosure around the columns, l venting to the cell; then the enclosure, vents / relief valves, rupture disks, and the

. ventilation systems (both the process / vessel ullage [2 HEPA banks _in series} and cell [1 HEPA bank)) become the primary controls. Analyses should be performed to verify no l cross effects upon other vessels on the same vent header from potential fumes caused I by the degradation reactions; if such effects are discovered, then the resin columns l should have their own ventilation system. In order to meet defense-in,-depth requirements, the area outside the cell (usually an access or operating corridor) and its ventilation system, and a short exhaust stack represent potential secondary mitigating controls. Specific level and spill detection instrumentation and alarms could also be identified as controls.

Tank Deflaaration Events: these probably require the same items relied upon for safety as with the tank leak events. Hydrogen and flammable gas detection systems constitute i additional potential controls. Redundant exhaust fans, supply air fans, or nitrogen I blanket systems may be warranted as controls if more refined calculations or experimental results indicate higher evolution. rates for flammable gases. j l

CST Drvina Events: Prevention would use OSR's on the CST drying. Temperature and humidity instrumentation are likely to become the primary controls,'with temperature and humidity endpoints determined by experimental testing. The CST canister drying enclosure and its filtered vent represent the first potential mitigating control, but, in the absence of a specific design, are not credited with any source amelioration. The cell and its ventilation system (including two HEPA filter banks) likely become the secondary control; the first HEPA filter bank is assumed to be rendered ineffective by the deflagration. In order to meet defense-in-depth requirements, the area outside the cell (usually an access or operating corridor) and its ventilation system, and a short exhaust stack represent probable tertiary controls.

. Melter Events: Prevention relies upon melter instrumentation and controls to stop the radioactive feed to the melter and replace it with process water (to maintain the cold cap), and terminate operations prior to metter failure. In the absence of a specific design, these are not credited with source reduction. Melter electrical and thermal Point Paper ,

May 27,1999

l l

l

\

54 sensors become the probable controls. The melter, its enclosure and its filtered vent represent the first potential mitigating control, but, in the absence of a specific design, are not credited with any source amelioration. The cell and its ventilation system

_(including two HEPA filter banks) would probably become the primary control. In order to  !

I meet defense-in-depth requirements, the area outside the cell (usually an access or operating corridor) and its ventilation system, the melter offgas system, and a short exhaust stack represent likely secondary controls.

• ' Chemical Events: Prevention would use high quality tanks and components, with 100%

weld inspection and leak testing, and a generous (conservative) corrosion allowance.

Prevention also could use an enclosure (with a spill basin) for the reduction of weather l effects (including diurnal thermal cycles) upon both the components and the chemicals of concern (principally ammonia and nitric acid). Leak detection sensors and a water spray / deluge system probably constitute the primary mitigative method. Exhaust fans I j

(on the enclosure) activated by a separate sensor system denote the secondary i

mitigative method. The third mitigative method locates the cold chemical storage tank area away from the facility, and preferably near a cooling pond or its equivalent.

Potential catastrophic releases of chemicals may affect the operability of the TWRS-P facility, rendering the area uninhabitable for a period of time (possibly as long as several hours). Consequently, additional safety requirements may be needed for control room air, breathing air (SCBA), and/or the ability of the TWRS process to operate and shutdown automatically, without human assistance.

I The average worker and cancer risks presented in Table 7 include contributions from all sources, such as industrial accidents, environmental chemical exposures, and other nonradiological contributors. Therefore, acceptable limits for potential contributions from radiological risks associated with process hazards of TWRS-P are likely to be lower, perhaps a few percent of these averages. This preliminary analysis suggests this is indeed the case for a TWRS-P facility design incorporating standard nuclear industry prevention and mitigation techniques, the estimated risk with prevention and mitigation features is 5% of the average occupational risk and 0.1% of the average, public cancer risk. This is consistent with discussions in the literature (Reference 30).

~

In general, designs proposed by the contractors do not consider prevention and controls and only incorporate one mitigating means to overcome failures. The designs do not include important auxiliary effects in the analyses, such as common mode failures, operability, recoverability, and plant habitability for operators, and means for controlling these effects.

Point Paper May 27,1999

i

? ..

e 49 (i.e., dilution). Additional equipment may be placed in the exhaust system to improve either its effectiveness or reliability, such as scrubbers, heaters, coolers, cyclones and impactors, electrostatic precipitators, roughing filters, and sand filters. Tank size and facility inventories can be reduced to decrease source terms and subsequent chemical effects from potential events.

- Alternatively, several smaller tanks can be used in place of one large one. These mitigation methods are likely to reduce chemical effects from ammonia and nitric acid tank failures to levels below which serious injuries are likely to occur (essentially around ERPG-1 limits).

1 I

Y 9

j l

4 Point Paper May 27,1999

50

'o l

(This page intentionally blank) t l

l l

2 I

1 i

i l

Point Paper .

May 27,1999 O

g

'y f 51 l

l3 Q .

Risk Considerations f Table 5 summarizes potential consequences from unmitigated events, and Table 6 lists potential mitigating controls and their beneficial impacts. All of the accidents listed in Table 5 have potential consequences exceeding the thresholds and guidelines in regulations, including the proposed revisions to 10 CFR 70 (Reference 28), referred to as the "New Part 70" in Tables 5 and 6. Many of the events have frequencies in the 1E-2 to 1E-4 range and would be considered to reside in the "unlikely probability bin (References 2 and 3).

I l Using the U.S. national average for workplace fatalities of 4.8E-5/yr (Reference 29) for l comparison, ten process scenarios exceed that national average (at 100 meters). The total estimated, unmitigated risk from the TWRS-P facility at 100 meters due to process incidents involving radionuclides is approximately 2.4E-2/yr, some 500 times larger than the U.S.

l. workplace average risk. For contrast and comparison, Table 7 displays additional risk l comparisons, and shows that the U.S. average background radiation dose dominates individual public radio!ogical risk (at 1.8E-4/yr). Table 7 also lists the average risk due to cancer (2E-3/yr) from Reference 30.- By comparison, the potential unmitigated risk from TWRS-P exceeds the background dose risk by two orders of magnitude and the average cancer risk by a factor of ten.

Four accident scenarios involving two forms of melter failure, and two forms of resin interactions dominate the risk by accounting for 90% of ihe total risk.- A large portion of the risk from the two metter accident scenarios accrues from rapid thermal volatilization and dispersal of the aqueous cold ca,p from a catastrophic release of the high temperature; molten glass. Limi,ted  !

j

! experimental data and experience are available for these melter failure scenarios. If these i metter and resin accidents are effectively prevented and/or mitigated, the TWRS-P risk l l decreases to around 1.4E-3/yr, a level commensurate with the risk associated with occupational exposure limits, but still some 10 times greater than the risk due to average background j ,

exposure to radiation. Several accident scenarios involving tank failures or deflagrations also exhibit the potential for very high doses. In the case of chemical storage tank failure, the f!

I potential ammonia and nitric acid releases would result in irreversible, deterministic health effects around the TWRS-P fac"ity and its environs, and would render the facility uninhabitable for operating and control purposes. Thus, prevention and mitigation are required to minimize the impact of these chemical effects upon radioactive materials.

l ,

it should be noted that the above values represent radiological risks and do not include risks -

posed by the chemicals and the industrial nature of the TWRS-P operations.

Fortunately, relatively simple and effective, prevention and mitigation methods are available, and Table 6 displays this situation. Prevention and mitigation methods reduce the total risk to the receptor at 100 meters from the TWRS-P facility to about 2.5E-6/yr. This result is about 5% of

- the average occupational risk and around 1.4% of the risk due to the average background dose.

Incorporation of prevention and mitigation controls is likely to be acceptable to the Part 70 undergoing revision, although further analysis may be necessary for the melter failure scenarios.

Consequently, the preventative and mitigating design features are likely to become controls and items relied upon for safety, and are discussed as follows:

4 Point Paper May 27,1999

I.)~

+

52-

• LAW and HLW Tank Leak Events: Prevention could be accomplished by high quality tanks and components, with 100% weld inspection and leak testing, and inventory controls. The tanks and the ullage ventilation system (including the first HEPA filter bank) could become the primary mitigative control. The second ullage ventilation system HEPA filter bank and the cell, with its ventilation system and (single-stage) HEPA filter bank and sump, constitute likely secondary controlsJ In order to meet defense in depth requirements, the area outside the cell (usually an access or operating conidor) and its ventilation system, and a short exhaust stack represent probable tertiary controls. .

Specific level and spill de'tection instrumentation and alarms may also be identified as i controls. Ideally, the inventory of liquid LAW and HLW in the TWRS-P facility should be

' minimized,' perhaps by some form of an inventory limit (e.g., such as a potential LCO). A specific design might include several smaller tanks (instead of one large one), of which one or more are maintained empty as a dedicated spare (s) in case of a leak in a full tank. The requirement for a spare tank or spare tank capacity could become a

.  ; requirement (perhaps another potential LCO).

. Cesium Tank Events: Prevention should be' accomplished by high quality tanks and components, with 100% weld inspection and leak testing, and inventory controls. The cesium product tank (s) and the ullage ventilation system (including the first HEPA filter bank) are likely to become the primary control. The second ullage ventilation system HEPA filter knk and the cell, with its ventilation system and (single-stage) HEPA filter bank and f.wp, constitute probable secondary controls. In order to meet defense in

- depth requirements, the a.ua outside the cell (usually an access or operating corridor) and its ventilation system, and a short exhaust stack represent likely tertiary controls. If these are (physically) the same items relied upon for safety by the LAW and HLW tanks, then an analysis should be performed demonstrating no deleterious effects upon the safety functions from the different tank systems and their events. Specific level and spill detection instrumentation and alarms could also be identified as controls, ideally, the inventory of liquid cesium solution in the TWRS-P facility chould be minimized or even eliminated, perhaps with an inventory limit (a potential LCO). It is also possible to add a requirement for conversion of the cesium into a solid, such as loading onto CST. A l specific design might include several smaller tanks (instead of one large one), of which one or more are maintained empty as a dedicated spare (s)in case of a leak in a full

~

. tank. The spare tank or spare capacity could also become a requirement (e.g., another 1 potential LCO).

I The cesium solution in the tank requires cooling to prevent temperature increases and, ,

ultimately, avoid solution boiling. Thus, the cooling means and its backup potentially become items relied upon for safety. This includes the separate cooling coils or jackets (zones), the cooling sources (usually cooling water, with process and firewater backup),

the pumps or means of recirculation, the piping, and redundant flow / temperature instrumentation. Interruption of cooling is likely to become an LCO, and the respective temperatures (of the tank and cooling means) OSR's. If it could be demonstrated by calculation or experimentation that a tank design is adequately cooled by passive means l

Point Paper May 27,1999

y ,2-J2 ; '

) )

.' k d .~

i' 5-4-

r ,

s ,kis t

e , . ' E E 2 5 a ir lar t

atck1> ,6 7- t d 3- od ga 6 1 E E

4o e t t e 4-i pi s-r i

t o o 5 5 Ef oa t

Ef t a E t

t 2 7 ig 7. o g 1 m m R"y J ' 5-5- 1 5

%i t 1

%iit nI m E E 3m 7m U'- 2" - 3 6

(

n (

n u u A, {*;i"Qp[ s A _

rp

f Q:jWy3Q )4}fD t

y g%; le y

le y

le y

le y

le y ly e le y

i4 a k k 9 l ' k k k k k d n i 9

ai]q'=_

' i i i i i i l

ln n ln ln l l 7.l

ln n n 9 b Be U U U U U U U 1 _

r o sg:: ly 7 U (rpsd s

h 2

~- y ig H a M

y  ; ,~

+ .

t )

li 5- 5- 6- 5- 3- 3- 5-d .r ic a Jye .cl y E E E E E E E F ^

. er nlol t

" 2 2 1 2 1 1 1 S . ut n 5

5 R 1qne" e ov ..' .

W EFn

- r cE _

T icr

.(u 44?

e n

ygf3yQ? (

e ,

G ,c e- -

e ,nf h ht 0e- h ig h

ig h

ig h

ig h

ig h

ig ig t 7u  :- H H H H H H H a e^4'i a t q r

s t

g'_

n e

v P sno Q' .'

E _C 1y l' _

d t

e a dc ge. .,y jg^ ,

g t e n ,.,t 0 0

0 0 5 0 0 0 a.W*

it i a et 0 0, 0 5 0 0 0 _

m m 3, 0, 4, 4, 0, n

i qa a 2 3 0 guc%t t

iep 6- 1 5 1 U mspr r 0-

- 2 2 f

o y

r n ni Uot C _s 0U 0 ;L 1

i 0

0, 3

0 0

0, 6

a r i '

yQi$

[pn pge$

m  ;

y; , 4 *

';e g.w mep

) ul m tpN yn u ,. .

r r .f.f Qla

~

Mw.y&ea s s S

u '

',l i,s ac

=,1 i c -0 gc a;

.N 6g n gf g t; 0f.w:. e$ .,,i ep$ m?

pt,JtM$

g i.

$w&jm _

.&f$N:i n t $ p' ~ F s.e ,,Ft e -

.ki0 c.

l no0 .k n

.a o 0, .. a .0,$l/n 0&e i

, l (~

ai

,t t a pf . r v h( ~kn v . . k :-

. n- TC( 1

.Tw1l ri

9. m t

,1 r

gw a

ep

Ee$e.p[Y y ~a. _ Ty a .

.i f.

n og t

" i r

i

.l l

o P a

y. .u 5 -

e n s .

~>

'W T;;r

7$?M 7c j isi .-

5t@

t s su le w.,W.g; -@fr.. sie

) n b ,h 1Q 4

Au L A

H l

e oo CLB n a '.'M .' .

el Rf .(

t ,.%DM io _

a T Il I i

P _

e r )

k

, is d",'

e t

4- 5 5-4-

iar t

4 tode at ,

E 5-gac k'1 E E E Ef t t pi s-r E 4 o a 2' 5 i

9 2 1 6 1

a i

m mR y 1 1 1 1

2%i i t

ni '

1

( m U~ n u

e f; 1 '

2 )

1 y y y ly ly ly t

y.

le e e e le le k

9 k k k 9

..M: ii J t?

l k i

k i

ln i

ln i

ln i

l n

li n 9 7.bn ai ' ln U U U U 1

- b B~ ' U U y y 7

)

d 9 or '

lh lh 2 y

e u

Lp

(

ig ig a n H H M it n y '

o c

(

y t )

5- 5- 5- 6- 6- 5-li ic yed E E E E E E a cl rV 1 1 1 1 1 F Meut nlo/

er t 1

S iqne r n R

s re ov 6 t 5

W cE -

n T

c EFu (

i r E 'c e

s c

e G e c

e ny h h h h h h h 0er g ig ig ig ig ig t

t 7u tr q o g" H H H H H a iH s a sa ete t

n e PnC o v C E

d e

t  ;,-

a c a t

sr it g 4^e~

. et,re n 0 0

0 0

0 0

0 0

0 0

0 0

gucotem 0, 5, 3, 0 0, 0, im n Gqatpmr iep e e 8 3

2 2 5 1

6 1

8 4

U msmc0' 1 2 f

o n nie0 y

r Uo' . C' .. R 1' a ;2 , .,t m .

.i #

f' f, rd m  ;

2,  :

s f at. 'y. c9 rFS2 H.%MY y[r [yaN E't.wkp2 2 i .

u }' c$r, S

k2 :d i -

j/9 eI4E k

. t a a nH ._ ng H c a(tc 0

, ni g$ r S e' n .w 'g Le .EgaT / l /a n N +Tow Th) ~T. y a; i  ;  ;

ep ae .f m' Wt

. i v 3rlf a

. ,Ef .j eH e ut .W / =W 4Ag l

l L.s. D eM P l

f

.Aig H- .T D f 5 Ll a L f

t e ;15Q;;')4 '~'2 DW .D ss .

i i

/f 2"

/ I in H ;H.kee CH .'H D.HC l HG . .CH . S '2F 2E  ; 2 o -

lb P -

a i T l I1llIf l\lf if

m .

~ -

. 'y7 ~

)

)

,k ,k ) ) _

,' _ e e dl e ,x ; lais t r lai t

s r lb lb _

t at 3od 2od a a _

gack1

- t e - t e ic i c

i i pi r s- Ef t Ef oag t

lp lp t

mmRy 3.

7%i o ag 3.% i t 1 t p

a p

a ni .

0i 3 m 4i 5m t

o t

o U~ i '

( n ( n n n u u ( (

~  ;-

jQ T .

1)jfe -

y y y y 2- l e le le le

i t s/~y; .~979 l

il~ k k k k 9

. 7.

i l li l i i l 9 ai ~  ; n n n n 9

)

d

- bob B,4e r

n1!MFU U U. U 1 7

e 'U p'; 3 2 _

u ( y n -

a it n . M o

c

(

y ,

it )

5-dye .d l 3- 3- 5-ic E E E E a e cl r y F t aer nlol t 1 1 1 1 S mut n 1

5 R

W i

t s qne-e ov r cE T EFn yt ,( u-icr e yy 6 s "'

n e i G ce g. :- # -

h e ny h

0er h h h t

7uq o .' ig g g ig t

a t g .

H iH iH H s

r a sa ete PnC t >

n -

e v o .

E C d

e e , -

t a dc2 t, , .

e a sr a n et3, r e -

g 0 0 3- 3- _

it t r _

i 0 0 y _

m i gucotem qat 5, 0, G P

G P

/

2-n isp pne t 4 6 U eJ r 1 2 R R E f

o menn n ke0 c 0' - E E 4 2

y U o"C>R1 r = _

a k s -

m ,

. - ' -.. pB/ Qpg d 4~ i 5>

7 i r

M{3[$

N$ h)2M m ."

[gf+

u y; F- -. d y[c 3s p7 e _

h;,S7Tw.nMg M$ 3kd g.-

1. k S w.?7$

3 t

Af e i a yr t a

g f neJ;e k$

C o

,',. ' iM NTd i 2:?

.e nY it i

r e

ii v 5 .

nB r

a m p g

e T T e n a gttiit.eW EJr,ge 5 g . w% u P t,Q r

i k u e

f.icMp r h'~ i n

l k ," 4 l i

l a

t n

lb b?' a 2 h 1 .

t o io a F ,~

F T Il T P

a 9'

l; 1_ o a

^

,1 t 9 t o8 - 8- 0 7o t

_ k r-s 9 - 9E E 1 8-Ef t 8-T ia

^

Rey E6 E

E -

2 3 E E

7 7 o E 1

1 3 61 1 1 Z -

7

(

r o?

p0 yl e7 t l

ebr s s s s s s s kaa e e e e e e e Lp P Y Y Y t

i Y Y Y Y ew ce AcN

)

d dy-lt te oi oi' 6- 6- 6- 6-ah b glia E 4- 4- 6- 9 E E E E E E 9 i

t eb 2 2 1 2 1 1 1 9 y

iko-Mir 1 lit LP ( 7 ic 2 a o y

F e e e t a S dny c t t e t e M e r ia i a t a

R t aug eo d e

d e ih a

w i

d W gqe dg ei o w

o w o e T i et m m m ic t

i sa r e

r e mH L L L r e

r e MnC o t

n t

n t e

r t

n n C I I In I e

8

'5 G h e d ht  :

~m e t

e s )e r a 7; 'o r r l ir e s  ;

.  : t s au d) pr f t ne lo gete 6 Ai s r

^ .

. 2 ev ri t ce 3 5Po 1 4 4 Aec t

1 0 n

o e 2Em 0 2 c C+ Rm 3

^ 1 6 Hot Pf C F d 0 t Ee f

l i

t a eo01 t t a

se ir u Hn t

i f

n g t ( d s t

e i t

i a ir f

(

o .

.M P

f o A A A t A

t s Pp Pp P np P t c c a m n'y Em Hus Em Hus Eg n

e vm A P

A P

E H

e stepe o i p i t / Hil l/u n l

/ l s E E i im

/-./Uf I

s a cren le/k le/k l

c en 2o o

/. c e/kd cenl i e/H e/ H 2

/. 2 n n/N r

e .l dS .ut a .ut a t n .

r ao rt un r t un t t h "

e 2 s 2s.ut s e v eo 2s r t

c. trol r eg se se f

l o klo e r klo e r vr klo res o/ v lov /v s o

y PnoF oR nca anp nca anp

/

l lem e nca anps a l cl ne l

cll ne

/

s/n aee l

l r

a C' .

TES TES ce TESC EC EC GCS m 1 . 34. 1 . 34 1 2. 1 345 1 . 2 1 2 1 23.

m 7 ( - ,

i  ; ,

N: t5 -B u 4- I

l ~

S f.*'t j9j~; N'hn C

A '0

,a" gt f r9hf' N7 e,

u t

hL3 kM 2i

~

N N k~ 2 l

. ne ,"

4, O0~ i~

ka' ff Mi a~ .

r

v r i

a il L 0,'

1 F

kn a" l

g' n 1' N,N' i

r n a'

e p

.E n 6

le pQ' k

[

T' 9 [N [ ;g.y T' a'

2 T0 a0j Ni d i

s h~ P a

n' K i s0,gJ,. ks f@W t b

S n T

a  ;

ao I~ '

Y io P

f _

d_

^

,1 8-9- 9-8- 7o)

_ sk r-E%a l

E E E

- ia E 3 2 1 3. to E R. ey

2 1 1 6

1 4(

t r _

e p?'

0 -

yl et'7

s s s s s aa e e e e e R.

Lp tbPr Y Y Y Y Y ew ce

.A

.cN U '

) '

d g dy )T e

u n d oi te oi t'

l '

6- 6- 6- 6- 6- 9 it n ah glia' b" E E E E E 9 o i eb 1 1 1 9 c

(

y ikoS t

Mir 1 1 1

it l

.LP' .(

~

7 2

ic y a e a F e M S 1nyc r t

ia R

A ^

_ug eo d e w o

w h ig h

ig V  ; qe m L L o

H H

l

"- sa et r

'c i

ii e i

r inC t n

i o

e I n

e C .

9 G e 5

e c d d d n d e

d e e e e .

ht

u .e .m

.o r e

r e

r e

r e

r e

t qr r a d) d) d) d) d) iet so ne ne ne ne ne lo s

Wspr n n et e ev ri t ev ri ev ri 5

ev rt c i

6 ev rt i

Aefc t t r c c c t

n oce 3 PfAe 8 5.A 2 Pfe 3.A s 2 Pi f 1

Aef 1 nCem 2Pf o 1 Pf C o R Ee f

Ee f

Ee Ee Ee l

P deo00 Hni Hni Hni Hni Hni a t t a t t t t s

t s

s s it 1 s n g ;t r r r i r ir t

e i t

i a if

(

i f

(

if

(

f

(

f

(

o .M P

f ,

o '

t c

t c

t c

t c

t c

t c .s' e e e e ie a mnoy j n

j n

j n

j n n atope t p l 2

i 2

i 2

i 2

i 2

i 2 1 2 l 2 /. 2 /. 2

- m i t/Uf t /.

t N t N t N t N t t N I

e ns a n t n/ r n t n/ r n t n/ n t n/ r n e n/ r el dS e e e e v o r e e v o e v o v son e

h t

t otlrl r oe v sv s o v /

v s /

v s /

v / s o /

s /

n

/

s /

l s l

/

n s /

l n s l n

f o PnoF .oR' aAee ll aAee l

aAee GPCS l

aAee GPCS l

aAee GPCS l

y r .C GPCS GPCS a .E . . .E . . .E .. . E'. . .E . .

A .

2 1 H23 1H23 1 H23 1 H23 1 H23 m

.j)Q yg!3

- m f;O 3,d 2 i: ~ . .

u - .

ese .@Cj;;1; ? s,G[.g+t%

ii%i

}s*%b S f

,Li1.zW N

C@ l e

.i i J.tN1 n*f .?

s i

e 1. sani

.asD@l. Mj*^

,/gHi.

t7 n /,

O dd T " Tf y r e

. e ;-

Wl2$jc~ r. ~ -. p i

.in aa~5  :

,,vELR mR'T ,s 3S e a 9 m.W$f$

l a kn k-@ A~ H a Q ~LA^ Hy f

P

_ 6 .

O ,L a t le w n

_ b a

RT1- w.O#)?

lf g$Wi ..,H J 2sDCT@+.M .J y2"

,H 6.$ _2 /

H M$

w.

p io 1

P T l

e d f f e ,1 8- 7o) 6o) t -

la E%t la E%t

) )

a ks r- E 0 0 gi a ire t

y 4 5.

73T 0 o 3. 1 o 15T

( (

i 2 ( (

M r

e7 is is p0 s s lyyr lyyr yl .e7 a nsa a

nsa s s ibtw r

a s

e As e As e e e k ;P Y r Y Y i

L r r

ec ec ew ce h

t r N e h e tr N cN u u A F F

)  ;

d t

)

e u

n d_oi= dty l

^ _oi 6- 4- 4- 6- 6- 9 it n  ;-

h b i a E E E E E 9 o l 9

(

c ~ eb ko- 1 1 1 1 1 1 ir y ~ LP( 7 it l

2 y

ic a a e o F e t M

dnyc e t

S r ia t a

R ^2eugo h d ih w w ig e d g o o W  ;- qe H m ei L L _

T c

^

et sa r e mH r

i r " nC t n e e o I t

n n C i e

0 G e ) t 6 e c t ds n d d a e

r e h e- m er ela r

t t

uLp

.qro, r e eh e b a d) d o d o ) 1 1 s

l a et s ne nt nt eea t - -

io ispr e ev rt i ee r u r ue G G tn n et P P t

r n

o t eCem oce 8 4 Pf Ace 51Pe Ad 6 Adh 2

Ped E R R E

C o R Ee f

Ev Evn _

P deo00 Hn Ht c i

i Htc a < <

la i

t t t tsfe t e it ag 1 s sf n t r irfe irfe t

e .i i

t a i f

(

f

( n f

( n i i

o M P

f o A A A t s P Pn Pn o s s c E Eo Ei t ap l mn e oty H Ht a i

Ha y

a v

a r

at pe r m i t!

! 2 e 2t

/ n 2 t

n ep eo ns Uf a r

/ r i t nu r tn e /

t e uS uS r e

wlo dS es em nm e s/ t s/

oc t y rl e r /vr u

/

h vo vu klo c kl 6-o e nct e t / r

/

lc nct

+tlnoF ll l t t f

o len esn l

le s ane ane E y oR CE CI ci n TED TED 5 r c' -

2 a p . .

1 2 1 2 1 23 1 23

=

m 1 2

v .

A 0 k m

S u

e, rQ Ji, )d1 j Mkf $}O ie

/r t a li

$h?$ ' >

6y)fW l

n7 s.7'(

. nf 99 ir d

s s s n rh- s a ,, - e t'

1

/

gi

. i nr a T'

• ir r t r n~ mg~ epe en a ..d ulw a g e

.ev ya rl f 'C s i t

S o~ i

?y i~

e:

ii

.caf Ap p i

it p

a -

r E ; '- .

/re D " r m P 6

e

~ ,YJ/v De@

T D kMC ea p9 t e

l%~T;i e

%luN eihe Ni~a; r ci r

ii ut t

l a

t n

lb T

a ,

IIIIII S2@ CH ME AF' N TU t

T o io P _

E ,. - ,

2 i

61

. Table 7: Different Sources of Risk Limits l

Risk Source / Basis Dose Equivalent, Frequency, Risk, Re'm yr-1 yr-1

., a '

, w,:.

. s 3.,

c,mug yx v O Worker Limitsg.;, gig 3 g.:,., . j 4gg;j.

9,. 'g a ,; ]

a. , ;p .

i 4,- , y Part 20, Worker - 5 1 2E-3 Limit Part 20, Typical 0.31 1 1.2E-4 ALARA Value U.S, Worker (-) (-)- 4.8E-5 Average, AllCauses

a; . .

_ m p,gyg - - ,

4. g y,,

Part 20, Public Limit 0.1 1 SE-5 Part 20, D&D and 0.025 1 1.3E-5 Part 61, Public Limits -

v -

,.# Typical Public Values s

, ; .5 p. Mi % '

U.S. Average 0.350 1 1.8E-4 Background  !

Background 0.500 1 2.5E-4 Difference between Denver and U.S.

Average Average U.S. Public (NA) (NA) 2E-3 Cancer Fatality Rate  ;

Average Public Dose <0.001 1 < SE-7 ,

from Commercial i Nuclear Plant

- Note: Radiological comparisons, assume Linear No Threshold (LNT) theory, with risk factors of  ;

2,500 rem / fatality for workers and 2,000 rem / fatality for members of the public. These rates are kept constant, and not reduced for higher acute doses (e.g.,1,000 rem / fatality for individual,  ;

acute doses over 10 rem).

.. Point Paper ' May 27,1999

I 62 (This page intentionally blank) 1 I

l Point Paper May 27,1999

.. )

63-

-L- . Summary and Conclusions i

Technical approach 6s are being formulated for the treatment of DOE tank wastes into vitrified waste forms suitable for long teim storage and disposal. Tank wastes from the Hanford Site are likely to involve significant chemical processing for separating radionuclides from nonradioactive species, with the intent of concentrating the radioactive species in the vitrified HLW and concentration of the nonradioactive species (including most of the sodium) into vitrified LAW. q The current plan envisions treatment in a private contractor owned facility on the Hanford site, with subsequent storage of the vitrified HLW at a DOE facility onsite. Disposal of vitrified LAW occurs in a near-surface, DGE disposal unit at Hanford.

This paper presents analyses for chemical and process safety at potential TWRS-P facilities using generic and conceptual process approaches proposed by DOE contractors. These analyses identify the following areas of concern from the perspective of chemical and process safety:

. Radiochemicalinventories

.- Process efficacy

. Organic ion exchange resin / nitrate interactions

.- CST drying

. Organic materials

. Radiolysis

. High temperature operations

. Nonradioactive chemical effects upon radiochemical processing Several of these areas of concern have events that can be analyzed at this early stage of design using a conservative, bounding approach suitable for SSC categorization. The remaining areas of concern are discussed qualitatively, but require a more detailed process and facility design for quantification. Analyses estimate both unmitigated and mitigated consequences and potential  !

risks from these events. Several event scenarios involving resin interactions, CST, large 1

radiochemicalinventories, metter coldcap, and cold chemical rele ases have potential accident i consequences to the workers and the public of sufficient severity such that the corresponding risk may not be acceptable, based upon comparisons to current limits. Consequently, prevention (reduce frequency and probability) and mitigation (reduce consequence severity) become necessary, requiring the identification of items relied upon for safety. Analyses estimate the combined, unmitigated risk to the receptor at 100 meters as approximately 2.4E-2/yr, about an order of magnitude above the equivalent risk of the 10 CFR Part 20 radiation )

worker annual dose limit of 5 rem. Melter and organic resin scenarios dominate the potential,  !

unmitigated risk at 100 meters, accounting for about S4% of the risk total. However, the remaining risk from the identified potential events still is an order of magnitude greater than the average occupational risk for U.S. workers of 4.8E-5/yr.

Point Paper ,

' May 27,1999

o 64

. Fortunately, relatively simple and effective, prevention and mitigation methods are available to reduce the potenti? ! risk from the TWRS-P facility. Passive prevention methods rely upon high quality, inspected, ~and tested components, with conservative design and corrosion margins.

This is a standard approach for nuclear facilities. Active prevention uses controls to avoid 1 operating sequences outside the design envelope that are precursors to events with i consequences; the controls can be based upon instrumentation or administrative procedures.  !

Mitigation reduces consequences primarily by confinement (e.g., cells with HEPA filters), with adequate levels of defense-in-depth. Again, this is a standard approach in the nuclear industry.

Prevention and mitigation offer the potential to reduce the risk from TWRS-P operations to ,

around 2E-6/yr. This result is less than 5% of the average occupational risk, around 1% of the l

.< risk from the average background dose, and around 0.1% of the average cancer fatality rate, l and is likely to be acceptable.

Obviously, DOE and its contractors will include experimental testing as part of the program leading to the design, construction, and operation of the TWRS-P facility. Few appropriate safety related parameters, such as failure rates, modes, and re' ease fractions; are available for l HLW'processing and vitrification facilities. It would be beneficialif the measurement of such  !

safety parameters could be included in the DOE program.

A potential rupture of cold chemical storage tanks containing ammonia and nitric acid would have onsite and offsite effects exceeding ERPG-3 levels and require evacuation of the facility.

Consequently, the facility design should include provisions to address such an event, by either dedicated breathing air to control and operator areas of the facility during such an event, a remote control facility, automat id operation / shutdown, effective mitigation of the chemical releases, or other means. In addition, some portions of the facility should be designed as

. shelter areas from potential events, including chemical releases.

l l

l l

t Point Paper May 27,1999

g 65 '

L. Bafa_geseg r

1. R.E. Gephart and R.E. Lundgren, "Hanford Tank Cleanup: A Guide to Understanding the Technical issues," PNL-10773, Third Printing, February,1997.

2. U.S. DOE, " DOE Regulatory Unit: initial Safety Evaluation Report of the BNFL Inc. Initial Safety Assessment," RUREG-98-09, Revision 0, March,1998.

\'

3. U.S. DOE, " DOE Regulatory Unit: Initial Safety Evaluation Report of the LMAES ISA

. Package,". RUREG-98-10, Revision 0, March,1998.

4. U.S. DOE and NRC, " Memorandum of Understanding Between the Nuclear Regulatory Commission and the Department of Energy; Cooperation and Support for Demonstration Phase (Phase 1) of DOE Hanford Tank Waste Remediation System Privatization Activities," January 29,1997.

5. U.S. DOE Contract Number DE-AC06-RL13308, "TWRS Privatization," September 9, 1996. -

l

6. Fourth International Conference on Nuclear Fuel Reprocessing and Waste Management,  !

April 24-28,1994, hosted by the British Nuclear Industry Forum (Recod'94).

7. - British Nuclear Fuels plc, " Low Active Effluent Management Group,"_1997.

8. D.D. Lee, J. F. Walker, P.A. Taylor, and D.W. Hendrickson, " Cesium Removal Flow

' Studies Using lon Exchange," AIChE 1997 National Spring Meeting, Houston, Texas, March 9-13,1997.

9. D.D. Lee, J.R. Travis, and M.R. Gibson, " Hot Demonstration of Proposed Commercial Oesium Removal Technology," ORNL/TM-13169, December,1997.

10. D.D. Lee, J.R. Travis, and M.R. Gibson, " Hot Demonstration of Proposed Commercial Cesium Removal Technology: Progress Report," ORNL/TM-13363, December,1997,

11. W.F. Hamel, Jr., M.J. Sheridan, and P.J. Valenti, " Vitrification at the West Valley Demonstration Project," Radwa.ste Magazine, March,1998.

12. U.S. Department of Energy, Status of ANL programs on Strontium recovery by solvent extraction, Efficient Separations and Processing Integrated Program.

13. U.S. Department of Energy, Efficient Separations and Processing Integrated Program.

14. S.A. Colby, " Ozone Destruction of Hanford Site Tank Waste Organics," WHC-SP-1004,

'.1993.

1 Point Paper May 27,1999 i

O i

i

!- 66

15. olational Research Council, " Glass as a Waste Form and Vitrification Technology:

Summary of an International Workshop," National Academy Press,1996,

16. V. Jain et al, " Review of Process Safety issues Relevant to Vitrification of Radioactive Wastes," Waste Management'98, March 2-5, Tucson, AZ.

- 17. . D. Martineau, " Safety in the Design and Operation of the Vitrification Facilities at La Hague," NRC/DSIN/IPSN Meeting, November 3,1997.

18. See "BNFL Vitrification Status" in Reference 6.

19. 1 " Radiological Characteristics Database," U.S. Department of Energy,1995.'

20.' U.S. NRC, " Nuclear Fuel Cycle Facility Accident Analysis Handbook," NUREG/CR-6410,

. , March,1998.

~ 21. DOE Handbook," Airborne Release Fractions / Rates And Respirable Fractions For Nonreactor Nuclear Facilities," Volumes 1 and 2, DOE HDBK-3010-94, December, 1994.

22. M.N. Baker and R,.J. Fussner, " Integrated Radwaste Treatment System: Lessons

. Learned from 2 % Years of Operation," DOE /NE/44139-68, May,1997.

23. C. Calmon, " Explosion Hazards of Using Nitric Acid in lon Exchange Equipment,"

Chemical Engineering, November 17,1980.

24. - WSRC, "S-Area Defense Waste Processing Facility (S-Area DWPF) - Final Safety Analysis Report," Volumes 1 and 6, WSRC-SA-6, Revision 13, November,1995.

25. DOE, " Tank Waste Remediation System, Hanford Site, Richland, Washington, Final Environmental Impact Statement." Volumes 1 and 2, DOE /EIS-0189, August,1996.

26. Center for Chemical Process Safety, " Guidelines for Process Equipment Reliability' i

Data," American Institute of Chemical Engineers,1989.

27. EPA RMP (Risk Management Program) Guidance,1997.

28. U.S. NRC, "Public Meeting on Part 70 Activities," May 28,1998.

29. Washington Times, Monday, October 5,1998, page D20

30. U.S. Department of Energy, " Regulatory Unit Position on Radiological Safety for Hanford Co-Located Workers," RL/Rrg.98-18, September 16,1998.

Point Paper May 27,1999

V e

\~7 4 67-Anoendix A l Generic Process Description l'

Figure A.1 depicts the overall block diagram for the generic process, and Tables A.1 and A.2 l provide information on the LAW and HLW tank waste envelopes. The overall processing scheme pretreats and separates the LAW supernate into a smaller volume, HLW-like fraction and a significantly less radioactive LAW stream containing most of the salts, which is vitrified. The HLW slurry is washed to reduce the soluble species (which go to the LAW pretreatment),

combined with the HLW-like stream from LAW, and vitrified in a separate, HLW melter. The principal activities are:

. Waste preparation and transfer q

. Pretreatment consisting of: 1

- Suspended solids removal I

- Cesium separation l

- Technetium removal j

- Strontium and TRU treatment

- Other radionuclides or species

. Low Activity Waste Immobilization

. High Level Waste Immobilization

. Off-Gas Treatment

. Supporting Plant Processes i

Planned operations at a TWRS facility have similarities with current operations at the WVDP and  ;

the DWPF. However, significant differences exist in the pretreatment area due to the multitude of different elements and isotopes present in the Hanford wastes, which, in the absence of separations, would generate an inordinately large quantity of immobilized HLW. DWPF and WVDP also use cementitious waste forms for their equivalent of treated LAW waste, while the TWRS approach vitrifies the treated LAW. TWRS-P designs envision a facility approximately the same or slightly larger than DWPF, with several HLW and LAW melters. MRS-P throughput is also likely to be similar to the DWPF capacity. An accelerated cleanup schedule for Hanford  ;

would increase both the size and throughput capacity of MRS-P, primarily from the additional l melter systems required. .

A.1 Waste Preparation and Transfer LAW is to be transferred into a DST (probably 241-AP-106) within the AP tank farm for bulk mixing and sampling. Envelopes A, B, and C are received as separate batches, prior to

[ .

Point Paper May 27,1999 l ,

1

• i 68 Table A.1: Typical LAW Radionuclide Composition Radio- Maximum Ratio, Bqlmole Sodium Curles/ Liter at 14 Molar Sodium nuclide Envelope Envelope Envelope - Envelope Envelope Envelope A B C A B C 4 TRU 4.8E5 ' 4.8E5 3.0E6 1.82E-04 1.82E-04 1.14E-03.

Sr-90 4.4E7 4.4E7 8.0E8 1.66E-02 1.66E-02 3.03E-01 Tc-99 ' 7.1 E6 7.1E6 7.1E6 2.69E-03 2.69E-03 2.69E-03 )

Cs-137 4.3E9 6 OE10 4 3E9 1.63E+00 2.27E+01 1.63E+00 Table A.2: Typical HLW Maximum Radionuclide Concentrations i

Isotope Cl/ liter Isotope Cilliter Isotope Cl/ liter

~

H-3 2.00E-05 Cd-115m ' 6.55E-10 Eu-152 > J% 1.50E-04 q C-14 2.00E-06 Sn-119m 1.00E-08 Eu-154 Esp 1.60E-02 l Fe-55s' 1.00E-03 Sn-121m 9.00E-06 Eu-155%'e ' 9.00E-03 Ni-592 1.40E-05 Sn-126 4.80E-05 !U-234@h 7.70E-07 I

Co60 **

3.00E-03 Sb-124 2.61E-09 ' U;235M'*? 3.20E-08 Ni-634G4 1.60E-03 1 Sb-126 , . 4.83E-06 ,U-2362M 8.20E-08 Sar 79 n:S- 4.20E-07 Sb-126m ' -

3.43E-05 U-238W#4 5.80E-07 Sr-90.1 .

3.10E+00 Sb 125 , 1.00E-02 'Np 237Mbgr 2.30E-05 e Y-90c:..n 3.10E+00 Te-125m . 3.00E-03 Pu-2384nn 1.10E-04 Nb-93m4 +

~

8.70E-05 l-129 9.00E-08 Pu-239 F 'm:L 9.50E-04

's 2.60E-04

~

Zr-93J i; . 1.40E-04 Cs-134 6.80E-03 Pu-240 G .

Tc-99 % . 4.50E-03 Cs-135 . 3.00E-05 Pu-241rl /G 6.90E-03 Ru-106 : 2.00E-04 Cs-137 3.00E+00 Pu-242fM , 7.10E-08 Rh-106L e 2.00E-04 .Ba-1?!m 3.00E+00 Am-241.d5 4.30E-02 l Pd-107 " 4.00E-06 Ce.144 1.00E-04 Am-242M ' 3.10E-05 l Ag-110m'.a 1.00E-08 P .-144- i.00E-04 Am-242m , c:q 3.20E-05 Cd-113mi. 1.09E-03 ?r.144m1 1.00E-07 Am2243%U 5.00E-06 in;113ml 1.88E-06 Pm-147 1.60E-01 Cm-242 i ' . 3.70E-05 Sn-113 1.88E-06 Sm-151 9.30E-2 Cm-244 9 30E-04 Point Paper May 27,1999

1 e

1 l

69 d

. Waste Receipt, Feed Evaporation Cs Removal by The ' '(

l 4 Entrained and Sr/TRU + lon Exchange + Removal by Solds Removal lon Exchange .

Acid Acid Stuny qy Elution Elution U U lI j Cs Eluate J Retum to DOE after 60 (Conditioning a V nfication

'O * "" '" Plant (Store for 9 yr.)

ont CST)

Ik '

V V U Dned CST Return glass j handling and Retum to DOE product to DOE container after 60 days .

LAW Option l

Y JL i

Liquid Waste Receipt, Waste Feed Evaporation h ' Cs Removalby Technetium

• 4 Entrained and Sr/TRU + lon Excnange

• by f,'n'"Ex ange Solids Removal Acid Acid Slurry y Elution Elution U -U lI Cs Cone.ntrate Tc Concentrate LAW ,

Entrained Solids Retum to (Store for initial 2 -

(Store for initial 2 Vitnfication DOE after 60 days storage yr. pnor to HLW yr. pnor to HLW Plant rnelter available) melter available)

U U

Glass product to DOE after HLW Retum glass 60 days storage C Vitnfiction product to DOE Plant after 60 days Sr/TRU slurry blended into metter over years 3.4 &5. I e., +U stored for 2.yr.

LAW and HLW Option Figure A.1; Generic TWRS Process Approach Point Paper ,

ay 27, M i

I70 transfer to the TWRS processing facility. Transfer uses in-tank pumps to transport the LAW to

)0 the facility via two, double-contained pipes, configured in a recirculation arrangement to facilitate l uniformity and suspension of any entrained solids. Each pipe run amounts to approximately .

1,000-1,500 linear feet, with a useful inner diameter of 3-4 inches. Tankage at the TWRS facility accepts the LAW and any rinse solutions used to clean the pipes and tank farm equipment.

Typical daily batch sizes are 10,000-15,000 gallons, at flow rates of 10-20 gpm. Recirculation flow rates are approximately ten times higher. Even though the DST and receiving tank are not pressure vessels and are unpressurized, the transfer pumps require 100-150 psi to overcome frictional losses in the transfer piping, and temperatures may increase to the 30-60*C range.

HLW is transferred in a similar manner, using either a separate DST or pump pit (probably around 241-AP-104), and a separate pair of double-contained, transfer pipes.

A.2. LAW and HLW Protreatment The LAW may vary. in composition, with the predominant species (sodium) varying over the 3-14 M (molar, moles / liter) range. Thus, it requires treatment by either evaporation or dilution to a

- uniform sodium concentra':on, nominally assumed to be around 7 M in sodium.

A.2.1.' Suspended Solids Removal

. Coprecipitation, filtering, hydrocycloning, centrifuging, and evaporation are the typical unit

. operations for separating solids from liquids. For TWRS, a combination of several unit

. operations may be necessary. The LAW may contain up to 5% solids as a thin slurry Metal

' filters, as a cartridge or a membrane, allow for rinses that remove soluble salts from the (concentrating) suspended solids, and can be backfiushed to recover the solids as a thicker slurry.: Ultrafilters are crocsfilters with a high fluid velocity for reducing flitercake and high pressure differentials. In combination with iron coprecipitation, these have been effective for 1

- removing fine particles in alkaline waste solutions (References 6 and 7). For some contract options (e.g., the LAW-only attemative), the solids would be returned to DOE as a thick slurry. In the complete flow sheet (LAW and HLW), the solids are stored for subsequent vitrificaticn with the HLW. i A.2.2 Cesium Separation ,

Cesium is present as a soluble species in LAW and HLW, and can be substantially removed by treatment of the supernatent and washings from the sludges and saltcakes. Competition from other alkali metal species in the tank wastes (primarily sodium and potassium) decreases the selectivity and removal efficiency; i.e., the nonradioactive species are also partially removed with the cesium. The main methods considered for cesium removal from Hanford tank wastes use

~

either ion exchange or absorption'(References 8,9, and 10), although solvent extraction has also l

- been proposed. All three methods have been tested in the laboratory. I Point Paper May 27; 1999

~, -

_t 71 lon exchange and absorption methods use similar equipment. The materials are received as solid particulates, typically 0.5-5 mm in diameter. A mixing tank combines these particulates with water or other solutions to form a slurry. A pump transfers the slurry to a cylindrical column with internal supports for the bed formed by the particulates. The column and bed are usually designed to provide at least a 1:1 help,ht: diameter (aspect) ratio, with distributors at either end to minimize backmixing and end effects, Experimental tests with organic resins proposed for cesium removal from HLW have required aspect ratios in excess of 15 in order to achieve fixed bed behavior (essentially plug flow), and avoid back mixing and premature breakthrough. In i operation, a filter clarifies the feed LAW /HLW wash solution and protects the bed. The solution normally flows down through the particulate bed, and, since one pass may not be sufficient, two or more columns are arranged in series to increase the cesium removal and decontamination factor; for example, the West Valley plant uses a minimum of three columns in series. There may also be a separate train of columns for additional redundancy and flexibility. The design typically includes valving and other means to isolate the individual columns for backwashing of the calumn, resin / absorbent removal (by elutriation or. sluicing), and. regeneration. For resins, the i design incorporates tanks and piping for regeneration. A post-filter removes any fine particulates from the column' effluent prict to further downstream processing.

Cesium removal by ion exchange uses cationic organic resins' Organic resins function primarily by ionic charge effects, although added specificity is achieved by also incorporating " molecular recognition" groups into the material. Manufacturers normally ship cation resins in the hydrogen i form, as damp solids. At the TWRS facility, the resins require wetting with DI water (resulting in swelling), rinsing with a dilute acid (to remove manufacturing residues), rinsing with DI water, and, because the tank wastes are alkaline, conversion to the sodium form by contact with dilute .

sodium hydroxide. After depletion of the resin effectiveness (as detected by a rise in the effluent l concentration of cesium), the column is isolated and rinsed with Di water Regeneration uses an 1

- acid stream, usually consisting of nitric acid at moderate concentrations. This produces a soluble, cesium nitrate solution for storage and subsequent processing (usually vitrification).

Rinsewater romoves residual nitric acid, and a' sodium hydroxide solution reconverts the resin to I the 'sodium form ready for reuse. Spent rinses require treatment for reuse or disposition. ' Over time, the use/ regeneration cycles attrite the resin and decrease its functionality, and, hence, the

- effectiveness, of the resin column, and, ultimately, the resin has to be removed and replaced.

Cesium removal by absorption utilizes inorganic materials, such as zeolites and crystalline silicotitanates (CST), that function by physical inclusion (cavity size) effects instead of ionic forces. Consequently, these sorbents tend to have slower kinetics and lower capacities as compared to ion exchange resins, and they cannot be regenerated. However, they are essentially unaffected by radiation, and raquire less preconditioning of the feed and the absorbent (typically, only water rinsing). Zeolites consist of aluminum silicates with various other metal compounds present (e.g., potassium, calcium, zirconium) in the matrix. Zeolites have

- been successfully used at West Valley for cesium removal from an alkaline supernate, followed by incorporation into the vitrified HLW product (Reference 11). Oak Ridge has used CST for treatment of solutions similar to HLW supernate, with the loaded CST column dried in place (i.e.,

in the column), ready for near surface disposal (Reference 8). The used and loaded sorbent is transferred to a storage vessel for subsequent processing (e.g., vitrification, as practiced at West Point Paper May 27,1999

' )

1 72 Valley), or the column itself is removed from the system and, after dewatering, is stored as the inner canister within an overpack.

The LAW-only option for Hanford tank wastes implies loading of the cest.um onto a sorbent and the return of the loaded sorbent to DOE. In laboratory testing, high concentrations of CST have been achieved in vitrified waste simu ants, and, consequently, cesium Joaded CST may also be used as an interim TWRS-P produd prior to its incorporation into HLW glass.

A.2.3 Technetium Removal Technetium removal and its effectiveness depend upon the oxidation state of the technetium-containing moiety. A considerable fraction (waste analyses suggest 60-80%) is normally present in solution as pertechnate, TcOi, an oxidized, anionic form. The remainder exists in cationic forms, such'as technetium (ll).. This is likely to be the case with the Hanford tank wastes.

Technetium removal has primarily focused on ion exchange using anionic organic resins with pyridine-type functionalities.

The mechanical arrangement is very'similar to the cesium ion exchange, with two or more columns in series. In use, the resin is received and wetted with DI water. Pumps transfer the resin as a slurry to the column. A dilute caustic rinse displaces the water from the interstitial spaces within the column, in order to avoid aluminum precipitation when the LAW is introduced.

After the column is loaded, it is isolated, and a water rinse removes traces of the LAW solution remaining in the free and interstitial spaces within the column. A dilute acid solution (again, usually nitric acid) elutes the resin, and recovers the technetium as pertechnate in a nitrate / nitric acid solution. For the LAW-only option, the technetium is returned to DOE in solution. For the complete process, the technetium solution is stored, for subsequent incorporation into the HLW vitrified product. The resin is rinsed with water, and the water displaced with dilute caustic to complete the regeneration cycle. Operation degrades the resin's functionality over time, and, after performance becomes inadequate (as measured by the column's technetium capacity and ,

effluent concentrations), the resin is either sluiced or slurried out for waste treatment (usually vitrification). Fresh resin is then conditioned and introduced into the vessel to form another usable column.

i A.2.4 Strontium and Transuranic Treatment Strontium and TRU removal from the LAW can be accomplished by either coprecipitation / filtering or solvent extraction (References 7,12, and 13). Some of the LAW may contain organic-based complexing agents and their fragments, and, thus, for favorable equilibria, destruction of these materials (e.g., by oxidation, such as with ozone [ Reference 14)) may be necessary.

Coprecipitation can use iron or manganese, followed by filtration. Solvent extraction requires the strontium and TRU isotopes to be in a soluble form. Solvent extraction contacts the clarified solution with an organic solvent containing an extractant (e.g., DEPA, CMPO), which causes the ,

transfer of the strontium and TRU into the organic phase as complexes. The process chemically Point Paper . May 27,1999 4

p 1 i

73 adjusts the loaded organic phase by contact with other solutions, as necessary to improve the separation, and then strips the strontium and TRU from the organic solution back into the aqueous phase with a dilute nitric acid solution, in the LAW-only option, the strontium and TRU are returned to DOE either as a slurry or a solution. The complete, LAW /HLW process temporarily stores the separated strontium and TRU, for subsequent vitrification with the HLW.

A.2.5 Other Radioisotopes or Species Analytical and historical uncertainties as to the actual tank contents may result in other species requiring treatment and removal (References 2 and 3). The tank wastes contain radioactive cobalt and europium. Uncertainties in the sample analyses overlap the concentration ranges at which cobalt and europium removal from the LAW would be necessary to meet radiation limits on l the container for the vitrified LAW. The tank wastes also contain sulfur, added as sulfamates and i

other species during chemical processing. The sulfur may be present at levels that interfere with vitrification operations and chemistry, such as by secondary phase formation (e.g., a molten salt phase on top of the glass) and altered chemistries (e.g., separate glass phases and/or inclusions), that couid have detrimental effects upon the performance of the radioactive waste form. Its removal may be necessary from the LAW stream prior to vitrification. ,

l Several methods exist for removal of cobalt, europium, and sulfur, lon exchange is the most probable approach because these elements are likely to be preser,t as multivalent species that l can more readily be differentiated from the abundant univalent ions (e.g., sodium and nitrate).

A.3- Low Activity Waste Immobilization l i'

The Tri-Party Agreement (TPA) requires immobilization of LAW via vitrification. This corresponds to glass production rates of 5-30 te/ day. For these rates, vitrification applications with radioactive materials usually utilize joule melting (i.e., melting achieved by the ohmic resistance heating between electrodes placed within the melter), with smaller radiant heaters (usually electrical) for j startup and for the pouring area (s) of the metter. Reliability and maintainability concems imply several smaller melters instead of one largo metter; the largest radwaste melters operated to date are in the 2-5 te/ day capacity range (Reference 15). Operations and maintenance may be either remote or manual, depending upon the radiation fields from the (treated) LAW. LAW from pretreatment normally requires concentration (sometimes to a slurry state) and mixing with glass forming chemicals (e.g., silica and boric acid) prior to actual feeding to the melter; this is accomplished by a separate evaporation step prior to the melter. Redox control chemicals, such

' as sucrose and formates, are added. These also reduce the level of nitrates and subsequent l

NOx evolution. Once started, the melter is usually kept at operating temperatures around 1,000 C with a molten pool of glass between the electrodes, even when the feed is stopped, to  !

minimize thermal stresses. Melters usually have cooling water jackets on the outside. Operation balances the thermal energy input (electricity) and feeo rates in order to maintain a small liquid / slurry phase (termed a cold cap) on top of the molten glass for reoucing volatile species

- carryover into the offgases.- Melters may function with either continuous or batch filling of the Point Paper , May 27,1999

.c ,

.1 74 waste containers, and the glass product may be either continuous or cut (usually while molten),

into small, discrete pieces about 3 cm in diameter (" gems"). Pouring uses spouts on side drains (or arms) of the melter, and suction or bubblers lift the molten glass up the discharge spout, from whence it pours by gravity into a waste container below the melter. Bottom drains are not usually incorporated into the designs. Melters have'a finite operating life, based upon the corrosion of the refractory linings and the electrodes, and the accumulation of salt'or metal phases not removed by the main glass pourings. ' Instrumentation monitors temperature and electrical parameters, and the melter is removed from service when indications of excessive corrosion are detected. The removed melter is either refurbished or dismantled for waste disposal. Melter longevity, maintenance, and dismantling / disposal are key considerations for LAW processing; to date, no large-scale LAW metter has operated for longer than 12 months without major refurbishment. It is anticipated that future developments will improve metter life but it will still be limited (probably to 'a few ye,ars), and, thus, melter replacement becomes a requirement at

, vitrification facilities and must be factored into the design. ,

A.4 High Level Waste Immobilization HLW immobilization also utilizes vitrification at temperatures around 1.100 C. The HLW melter (s) are smaller than their LAW counterparts, and correspond to total glass production rates of 1-3 te/ day. The melter design may be either joule heated, such as at the DWPF and West Valley facilities (References 15 and 16 ),' or use induction heating (i.e., extemal coils induce a current, and the electrical resistance in the molten glass (and its metallic liner) generates the heat), such as at the European HLW facilities (References 17 and 18). Induction metters are smaller because of the limiting distances for induction effects. induction methods usually operate on dried HLW material, and, thus, operate in tandem with a rotary calciner. Typicalinduction metters last several months, and their design incorporates rapid melter maintenance and replacement. ,

Multiple year experience is available with both types of melters. Joule-heated, HLW melters at WVDP and DWPF correspond to capacities of 1 and 2 te/ day, respectively. The European induction melter designs are smaller, with a capacity of about 0.7 te/ day for each unit, but are normally configured with 2-3 units per vitrification facility, thus giving a facility capacity comparable to that of the DWPF.

A.5 Off-Gas Treatment Off-gases from the melter require treatment for the removal of particulates, volatile species, and nitrogen oxides (from nitrate decompotition); The sequence consists of cooling, scrubbing, NOx

' treatment, temperature adjustment, filtering, sampling, and exhaust removal, and multiple stages may be used in each step. The initial treatment uses mixing with cooler air to reduce the overall off-gas temperature and solidify any molten glass entrainment from the metter. A film cooler ameliorates glass accumulation at the beginning of the off-gas system by introducing a film of cooler air adjacent to the pipe wall. The warm gases enter a water scrubber or impinger (various Point Paper May 27,1999 l

l 75 geometrical configurations are possible - spray, packed column, submerged etc.) for removal of particulates and (partially) volatilized radionuclides such as cesium, technetium, and iodine (if present). The spent scrubbing solution is ultimately returned to the melter via its feed / adjustment tanks. A subsequent scrubber system removes NOx with a dilute caustic stream. If the residual  !

I NOx level is still elevated, it can be further reduced by additional caustic scrubbing or selective catalytic reduction (SCR) with ammonia. SCR ultimately produces nitrogen by the ammonia-NOx reaction, but requires temperature adjustments to the air stream. Heaters raise the gas )

temperature above its dewpoint, and the ga.ses then flow through roughing and HEPA filters.

~

Additional heaters may be used to improve the buoyancy of the gases prior to discharge via an elevated stack.

A.6 Supporting Plant Processes Other plant process systems of concern are water (management), clecontaminatior, gas I systems, and cold chemical management. .

A TWRS facility is likely to use copious quantities of water as part of the process, ano thi;; in turn generates aqueous waste streams of various levels of radioactivity. Potential process water additions may exceed the size of the LAW /HLW batch undergoing processing. Unless these process water additions are effectively managed, they adversely affect plant operations and generate additional secondary LLW.' The process offers the potential for recycling significant quantities of contaminated water within the facility. For example, used rinsewaters from ion  !

exchange regeneration can be used for the preparation of the subsequent regeneration solutions (acidic or alkaline, as the case may be). Some of the contaminated process water is likely to require evaporator or ion excnange treatment, and some may be discharged directly to the Hanford Site's Liquid Effluent Treatment Facility (LETF).

Decontamination of waste containers (after filling) and plant SSC's is a necessary part of operations and helps to avoid contamination spreading within the facility over time. Existing HLW i vitrification facilities have limited in-cell surge capacity for filled containers awaiting decontamination or repeat decontamination, and an inadequate canister decontamination approach can effectively shut the process down. Effective canister decontamination also depends upon the canister filling approach and the adequacy of the seal between the melter and the canister. Mechanical decontamination with glass frit, with subsequent recycle of the spent (contaminated) glass frit to the melter, has been successfully applied at DWPF (with only limited repeated decontamination necessary), while the French facilities have obtained success (with some acceptable repeated decontamination rate) with high pressure water decontamination provided parameters are closely controlled and the process maintains an effective seal between the metter and the canister. The British facilities have also successfully used high pressure water decontamination, but with a higher repeat decontamination rate. West Valley uses a chemical decontamination and polishing system based upon cerium (IV) nitrate solutions without any problems; the spent solution is recycled to the melter feed tank. Any one of these approaches is feasible, although equipment and performance prob! ems have to be evaluated for the mechanical approaches.

Point Paper May 27,1999

E-i 76 Facility SSC decontamination may be accomplished by water and dilute chemical (e.g.,

surfactant and acid) sprays, particularly within process cells. If additional decontamination is

- necessary, then the SSC would be removed and transferred to a dedicated decontamination cell where more complete and/or aggressive methods are applied.

Cold (nonradioactive) chemical management includes storage and preparation / mixing of the reagents at the facility. Offgas treatment will likely use reductants, such as sugar, formic acid, or

' ammonia, in relatively large quantities.' For example, at a 10,000 gal LAW / day flowrate, with 7 molar nitrate, and 80% removal or destruction by other processes (s.g., Redox control in the metter, caustic scrubbing), selective catalytic reduction of NOx requires approximately 2,100 lbs (about 500 gallons) of ammonia daily. This requires a relatively large storage tank. Storage of concentrated nitric acid and caustic will also be necessary, with likely tankage of 5,000-10,000 gallons each.

l i

Point Paper. May 27,1999

r-77 1

Appendix B l

Sample Calculations

• l 31 Basic Source Term Formula:

ST = (MAR x DR) x (ARF x RF) x LPF l

l where:

ST = source term, a quantity; curies are usually used for radioactive materials. The equation assumes that an airbome inhalation pathway will dominate the prompt, overall dose to non-l worker receptors from most accident scenarios. For workers, additional pathways (e.g., direct i shine) may have to be considered. A few accident scenarios may also involve significant l

. contributions from non-inhalation pathways. More detailed design information may be needed for l reasonable estimation of non-inhalation dose effects.

l MAR = material at risk, representing the maximum quantity of radioactive material that can be present. This is estimated from the physical attributes (e.g., tank and container size) and l chemical properties (e.g., concentration) of the SSC under analysis. The accident scenario has to be considered in the determination of MAR - for example, a tank failure due to a weld defect would use only that single tank's inventory for the MAR, but a tank failure due to a hydrogen explosion or an earthquake might impact adjacent tanks and areas containing handling radionuclides, and .have more MAR.

1 l DR = damage ratio, the fraction of material at risk actually impacted by the accident scenario.

There is some interdependence between the MAR and DR values. The DR is estimated based upon an analysis (qualitative or quantitative) of the SSC's response to the accident scenario, l such as strength / stress, or fraction in mobile form (e.g:, present as a liquid).

1 ARF = airbome release fraction, the fraction of material suspended in air and available for transport and inhalation from an accident scenario. This is highly dependent upon the conditions assumed in the accident scenario. For example, temperature has a significant impact for liquids and solutions because of vapor pressure effects, phase changes (boiling), and increased reactivity (at higher temperatures).

I RF = respirable fraction, that fraction of the aerosols and particles in the air that can be inhaled l into the human respiratory system. This is commonly assumed to be particles with a diameter of l 10 microns or less. There is some interdependence between the ARF and RF values, and sometimes the product (ARF x RF) is used.

LPF = leakpath factor, the fraction of the radionuclides in the aerosol transported through the SSC/ confinement / building / filtration mechanism to the release point. The LPF may be estimated based upon established models for agglomeration and deposition, and filtration efficiencies, and l specific SSC and building designs. There may be several LPF's involved in a scenario - from the tank to the enclosure, from the enclosure to the cell, in the ductwork, etc. - and a simplified 1

Point Paper . May 27,1999 l

L

i i 78 L

summary value map be used. LPF's are difficult to ascertain'on preliminary designs. Bounding analyses for determination of unmitigated consequences, safety categorizations, and an initial set of items relied upon for safety usually use an LPF of unity. LPF's of mitigating equipment (e.g.,

filters) use a factor speicific for that type of equipment, without inclusion of physical processes

, (e.g;, deposition in duct work) that might occur prior to the equipment.

R,2 , Standard Dose Conversion Calculation:

Dose = [ Sum {each isotope x isotopic dose conversion factor}] x { chi /q} x { breathing rate}

where:

- { chi /q) represents the dispersion effects at a specific location; for Hanford conditions, chi /q ranges from about SE-3 to 4E-2 for the receptor at 100 meters, depending upon the conditions assumed,' averaging of conciitions, and the inclusion of building wake and plume meander effects; these estimates use approximately 3.41E-2 s/m3 at 100. meters from a point source. (Chi /q) value's drop to around 1E-4 for a receptor at 1,600 meters, and to the 1E-5 level for the receptor at 10,000 meters.

. The breathing rate is approximately 3.47E-4 m3/s, a value corresponding to light activity. This

= may not be conservative for a 12 hr shift during an accident condition. Standard dose conversion factors are:

• Table B.1: . Standard Dose Conve'rsion Factors ~

Radionuclide - Inhalation Dose Conversion Inhalation Dose Conversion Factor, Sv/Bq Factor, Sv/ curie Cobalt-60 ' 5.9E-8 2,183 Strontium-90 3.5E-7 12,930 l Technetium-99 2.25E-9 83.3 Cesium-134 1.25E-8 463 1 Cesium-137 8.63E 9 319 Europium-154 7.73E-8 2,860 Europium-155 1.12E-8 414 ,

Plutonium-239 1.16E-4 4.29E6 Plutonium-240 1.16E-4 4.29E6 Plutonium-241 2.23E-6 82,500 Americium-241 1.2E-4 4.44E6 ' '

Point Paper May 27,1999 l

]

j 79 )

Curium-244 6.7E-5 2.48E6 3,.2 Radiochemical Inventories and Laroe Tanks This provides supporting information for Section 5.1.1. Dose estimate bases have been )

calculated for the feed materia!s as follows:

J Table B.2: Bounding LAW Feed Tank Radionuclide Inventory J Radionuclide Low Activity Waste (LAW),100 Kgal Basis, Envelope B Clear 5% Suspended Total [ Curies x DCF],

Supernate, Solids, Curies Sv Curies Curies Cobalt-60 (0) 570 570 1.24E6 (<1%)

Strontium-90 6,300 587,000 593,000 7.67E9 (16%)

Technetium-99 1,000 850 1,850 154,000 (<1%)

Cesium-134 (0) 2,570 2,570 1.19E6 (<1%) -

Cesium-137 8.6E6 570,000 9.17E6 2.93E9 (6%)

Europium-154 (0) 3,030 3,030 8.67E6 (<1%)

Europium-155 (0) 1,703 1,700 7.04E5 (<1%)

Plutonium-239 0 180 180 7.72E8 (2%)

Plutonium-240 0 49 49 2.1E8 (<1%)

Plutonium-241 0 1,306 1,306 1.08E8 (<1%)

Americium-241 70 (assumed) 8,140 8,210 3.65E10 (75%)

Curium-244 0 176 176 4.36E8 (<1%)

Totals 8.61E6 '1.18E6 9.7E6 4.86E10 .

Point Paper May 27,1999

E e

. j 80-HLW estimates are based upon a 10% suspended solid loading. -.

Table B.3: . Bounding' HLW Feed Tank Radionuclide Inventory .

Radionuclide High Level Waste (HLW),100 Kgal Basis j

.10% Suspendsd Total (Curies x DCF),- l Solids, - Curies Sv ]

1 Curies Cobalt-60 1,140 1,140 2.48E6 (<1%)

Strontium-90 1.2E6 ~.2E6 1 1.55E10 (17%) 4 Technetium-99' 1,700 1,7001 142,000 (<1%) l Cesium-134 5,140 5,140 2.38E6 (<1%)

Cesium-137 1.14E6 - 1.14E6 3.64E8 (<1%)  !

1 Europium-154 6,060- 6,060 1.73E7 (<1%)

Europium-155 3,400 3,400 1.41 E6 (<1%) i Plutonium-239 360 360 1.54E9 (2%)

Plutonium-240 - 100 100 4.2E8 (<1%)

Plutonium-241 2,610 2,610 2.16E8 (<1%) .

Americium-241 16,280 10,280 7.3E10 (79%)

Curium-244 352 352 8.72E8 (1%)

Totals 2.36E6 2.36E6 9.19E10 Tank failure / spill: Use bounding values from Reference 21 for aqueous solution, undergoing a liquid free-fall spill (these values are identical to containment failure below the liquid level, in a tank, for aqueous solutions):

ARF = 2E-4 '

RF = 0.5 (ARF x RF) = 1E-4 Some of the tanks also contain solids, in dilute slurries. Reference 21 presents the following bounding values for a free-fall of a slurry:

ARF = 5E-5 RF = 0.8 (ARF x RF) = 4E-5 Point Paper May 27,1999

m l

\ *

, 81 l

The results for LAW using the higher, (ARF x RF) values are:

Table B.4: LAW Feed Tank Failure - Free-Fall Spill Effects i Radionuclide Low Activity Waste (LAW),100 Kgal Basis,5% Solids I l

Material At Source Term, (Curies x DCF), Dose to 100 m  ;

Risk (MAR), Curies , Sv Receptor, Sv Cdries Cobalt-60 570 5.7E-2 124.431 0.001472355 Strontium-90 593.000 59.3 766749 9.0727108923 Technetium-99 1,850 0.185 15.4105 0.00018235 j Cesium-134 2,570 0.257 118.991 0.001407985 Cesium-137 9.17E6 917 292523 3.4613369021 Europium-154 3,030 0.303 866.58 0.0102539812 Europium-155 1,700 0.17 70.38 0.00083279 Plutonium-239 180 1.8E-2 77220 0.913721094 Plutonium-240 49- 4.9E-3 21021 0.2487351867 Plutonium-241 1,306 0.131 10807.5 0.12788190525 Americium-241 8,210 0.821 3645240 43.133031348 Curium-244 176 1,76E-2 43648 0.5164736896 Totals 9.7E6 970 4858404.2925 57.4880404719 (5,750 rem)

Point Paper May 27,1999 l

z

82 Resuspension effects: assume 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, and use value from Refererte 21 for indoors, low air speeds:

ARF = 4E-7/hr x 24 hr = 1E-5 RF=1 Table B.5: LAW Feed Tank Failure - Resuspension from Spill Effects Radionuclide Low Activity Waste (LAW),100 Kgal Basis,5% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies Cobalt-60 570 0.0057 12.4431 0.00014724 Strontium-90 593,000 5.93 76674.9 0.90727108923 Technetium-99 1,850 0.0185 1.54105 0.0000182 Cesium-134 2,570 0.0257 11.8991 0.0001408 Cesium-137 9.17E6 91.7 29252.3 0.34613369021 Europium-154 3,030 0.0303 86.658 0.001025398 Europium-155 1,700 0.017 7.038 0.0000833 Plutonium-239 180 0.0018 7722 0.0913721094 Plutonium-240 49 0.00049 2102.1 0.0248735187 Plutonium-241'. 1,306 0.01306 1077.45 0.0127491426 Americium-241 8,210 0.0821 364524 4.3133031348 Curium-244 176 0.00176 4364.8 0.051647369 Totals 9.7E6 97.82641' 485837.12925 5.74876499928 (575 rem)

Total estimated, unmitigated consequence is 5,750 + 575 = 6,300 rem. Use of the slurry .

parameters would give around 3,000 rem.

Point Paper May 27,1999

_-m.___ -_

83 Table B.6: HLW Feed Tank Failure - Free-Fall Spill Effects Radionuclide High Level Waste (HLW),100 Kgal Basis,10% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies Cobalt-60 1,140 0.114 248.862 0.002944709 Strontium-90 1.2E6 120 1551600 18.35961732 Technetium 99 1,700 0.17 14.161 0.00016756 Cesium-134 5,140 0.514 237.982 0.00281597 Cesium-137 1.14E6 114 36366 0.4303079682 Europium-154 6,060 0.606 1733.16 0.0205079623 Europium-155 3,400 0.34 140.76 0.001665571 Plutonium-239 360 0.036 154440 1.827442188 Plutonium-240 100 0.01 42900 0.50762283 Plutonium-241 2,610 0.261 21532.5 0.25478761275 Americium-241 16,280 1.628 7228320 85.530542064 Curium-244 - 352 0.0352 87296 1.0329473792 Totals 2.36E6 237.7142 9124829.425 107.971369137 Point Paper May 27,1999

84 Resuspension effects: assume 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, and use value from Reference 21 for indoors, low air speeds:

ARF = 4E-7/hr x 24 hr = 1E-5 ,

RF=1 Table B.7: HLW Feed Tank Failure - Resuspension from Spill Effects Radionuclide High Level Waste (HLW),100 Kgal Basis,10% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies Cobalt-60 1,140 0.0114 24.8862 0.00029447 Strontium-90 1.2E6 12 155160 1.835961732 Technetium-99 1,700 0.017 1.4161 0.0000168 Cesium-134 5,140 0.0514 23.7982 0.0002816 Cesium-137 1.14E6 11,4 3636.6 0.0430307968 Europium-154 6,060 0 0606 173.316 0.002050796 Europium-155 3,400 0.034 14.076 0.00016656 Plutonium-239 360 0.0036 15444 0.1827442188 Plutonium-240 100 0.001 4290 0.050762283 Plutonium-241 2,G10 0.0261 2153.25 0.0254787613 Americium-241 16,280 0.1628 722832 8.5530542064 Curium-244 352 0.00352 8729.6 0.10329473792 Totals 2.36E6 23.77142 912482.9425 10.7971369137 Total estimated, unmitigated consequence is 10,800 + 1,080 = 12,000 rem. Use of the slurry ,

parameters would give approximately 6,000 rem.

Similar calculations for the cesium product tank (1,000 gallon basis) produce a dose of 55 rem for the receptor at 100 meters.

For the cesium product tank boiling scenario (5% boils away):

DOSE = (1.32E6 Ci) x (319 Sv/Ci) x'(0.05 x 1) x (3.41E-2 s/m3) x (3.47E-4 m3/s)

= 250 Sv (25,000 rem)

Point Paper May 27,1999 l I

)

~

pi 85 9.4 - Oraanic lon Exchance Rosin and Nitrate Interactions in the Presence of Very Hinh Radiation Fields This provides' supporting information for Section 5.1.3.

l l The experimental data indicates approximately a ninety-fold increase in the concentration of l . cesium upon the resin column,'as compared to the feed soluton (References 8,9, and 10).

Thus, treatment of Envelope B waste implies a feed solution of 8 curies / liter at an adjusted feed level of 5 Molar sodium, corresponding to a loaded resin concentration of approximately'720 curies / liter and a heat load of 3 watts / liter. Using a 100 liter column as a basis (i.e.,6 inches ID by 18 ft high of resin), the material at risk becomes 72,000 curies.

The first resin interaction scenario assumes the recin interactions occur after the column is loaded but awaiting regeneration, and is initiated by localized decay heating and radiolysis. The potential for involvement of additional resin columns depends upon the specific design and is not included here. The data set for release fractions from burning plastics typically used in ion i exchange resins (e.g., polystyrene and polymethylmethacrylate [PMMA - lucite]) is limited and l based upon uranium contamination - cesium values are likely to be higher. The PMMA data set

! is significantly larger, and bounding values of SE-2 and 1.0 are suggested for the release and respirable fractions, respectively (Reference 21). The source term becomes 3,600 curies using the'se values. The dose estirnate is:

DOSE =_ (3,600 Ci) x (319 Sv/Ci) x (3.41E-2 s/m3) x (3.47E-4 m3/s) = 13.6 Sv (1,360 rem) to the receptor at 100 meters.

1 I The second scenario assumes the column is isolated and' the event is initiated during li regeneration when nitric acid interacts with the resin in the high radiolysis fields. The highest solution concentration occurs about midway through the cycle, and is equivalent to approximately l 30 times the feed concentration or about one-third of the column's loaded concentration. The peak occurs for approximately one bed volume of eluant. Resin columns are usually designed ,

with a solution freeboard above the resin of about 25-50%, for expansion effects. Interstitial space within the resin usually amounts to 30-50% of the resin volume. Thus, an amount of eluted solution equivalent to a resin volume (i.e.,50% plus 50%) is bounding. This corresponds to a material at risk of 24,000 curies of cesium in solution. The reactions provide energy to l vaporize the solution (or superheat it if it is in a confined space, ultimately leading to failure), and, l . ultimately, a release occurs. The energy required to vaporize all of the solution can be supplied by burning less than 10% of the resin, and, thus, an atmospheric release fraction of 1 is used.

' The respirable fraction is 0.3 (from Reference 21). The source term is 7,200 curies. The accident dose estimate becomes:

DOSE = (7,200 curies) x (319 Sv/Ci) x (3.41E-2 s/m3) x (3.47E-4 m3/s) = 27 Sv (2,700 rem) frnm the eluant liquid vaporization release. The solid resin also releases cesium, but the source >

term is reduced 50% because of the regeneration, and this contributes another 700 rem. The total dose estimate from the accident becomes 3,400 rem to the receptor at 100 meters.

l Point Paper - May 27,1999

i 86 lon excnange is also used to separate and recover technetium from the tank wastes.

Manufacturer data indicates a maximum loading of 0.8 gew/ liter of anion resin, corresponding to  !

about 80 g/ liter of technetium. . or a 100 liter resin column, loaded and awaiting regeneration,

. this becomes 8 kg, or about 140 curies. The source term becomes 7 curies, and the estimated dose from a fire on the loaded resin is:

DOSE = (7 Ci) x (83.3 Sv/Ci) 2. (3.41E-2 s/m3) x ( 3.44E-4 m3/s) = 7E-3 Sv (0.7 rem).

There is insufficient information in the published literature to calculate a regeneration accident consequence, but, by analogy to the cesium case, it is likely to be around 2 rem.

RJ Hydrocen and Flammable Gases '

This provides supporting information for Section 5.1.4.

Feed Tank: 100,000 gallon basis Assume G = 0.43 (molecules H2) per 100 ev beta-gamma,'and G = 2 (molecules H2) per 100 ev alpha. Note that this value is highly dependent upon the solution chemistry - more acidic conditions (e.g., nitric acid) tend to decrease the value.

LLW Feed Tank: 9.69E6 curies beta-gamma (assume all 0.7 MEV),9,920 curies alpha (assume all.5.5 MEV), for a 100,000 gallen basis:

(H2 - beta-gamma) = (9.69E6 Ci) x (3.7E10 dps/Ci) x (0.7E6 ev/dps) x (0.43 molecule H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 143 liters /hr at STP (H2 - alpha) = (9,921 Ci) x (3.7E10 dps/ci) x (5.5E6 ev/dps) x (2 molecules H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 5.4 liters /hr at STP j

Total = 148.4 liters /hr at STP HLW Feed Tank: 2.36E6 curies bet ~a-gamma (assume all 0.7 MEV),19,700 curies alpha (assume all 5.5 MEV), for a 100,000 gallon basis:

> l Point Paper ,

May 27,1999 I

E

, 87 (H2 - beta-gamma) = (2.36E6 Ci) x (3.7E10 dps/Ci) x (0.7E6 ev/dps) x (0.43 molecule H2/100 l

ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole) i

= 35 liters /hr at STP

! (H2 - alpha) = (19,700 Ci) x (3.7E10 dps/Ci) x (5.5E6 ev/dps) x (2 molecules H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 10.7 liters /hr at STP Total = 45.7 liters /hr at STP l Cesium Product Tank: 1.32E6 curies beta-gamma (assume all 0.7 M'iV), for a 1,000 gallon basis j (note - neglect nitric acid effects upon the "G" value):

l (H2 - beta-gamma) = (1.32E6 Ci) x (3.7E10 dps/Ci) x (0.7E6 ev/dps) x (0.43 molecule H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 20 liters /hr at STP For the large tanks (100,000 gallons), assume additional ullage of 10% (10,000 gallons):

LFL = 4% = 1,520 liters Time to reach LFL = 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br /> for' LAW tank 33 hours3.819444e-4 days <br />0.00917 hours <br />5.456349e-5 weeks <br />1.25565e-5 months <br /> for HLW tank 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> for Cesium Product tank l For the loaded CST container:

300,000 curies of Cs-137.

(H2 - beta-gamma) = (3E5 Ci) x (3.7E10 dps/Ci) x (0.7E6 ev/dps) x (0.43 molecule H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

Point Paper May 27,1999 L )

, 1 88

= 4.5 liters /hr at STP (complete absorption / poor dreining, maximum generation condition) 1 (H2 - beta-gamma) = (3E5 Ci) x (3.7E10 dps/ci) x (SE2 ev/ micron /dps) x (1 micron water film) f

(

x (0.43 molecule H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 3.2 cc/hr at STP (thin film / poor absorption / good draining / low temp. drying, minimum generation condition) l The heel cases are based upon flammable gases (hydrogen) being sucked into the tank through f a common vent header during a draining operation, without adequate removal by ventilatien. A tank heel of 500 gallons is assumed. This gives the maximum energy and dispersion of radionuclides.

,/ Maximum energy in air occurs at 30% hydrogen.

In 100,000 gallons of gas at atmospheric,30% hydrogen = 1.13E5 liters = 4,600 moles.  !

Energy release = 56 Kcal/ mole x 4,600 moles = 258,000 Kcal = 1.02E6 BTU Adiabatic temperature rise = TO + (Q/m/C) = 50 + 2.58E8 cal /(15,400 moles

• 5 cal /c/ mole) =

3,400 C Pressure rise = 1 atm x (3,400 + 273)/(298) * (1/1.5) = 8.22 atm = 121 psia Maximum energy with co-evolved oxygen occurs at 67% hydrogen.

In 100,000 gallons of gas at atmospheric,67% hydrogen = 2.52E5 liters = 1.029E4 moles.

Energy release = 56 Kcal/ mote x 10,290 moles = 576,000 Kcal = 2.29E6 BTU Adiabatic temperature rise = TO + (Q/m/C) = 50 + 5.76E8 cal /(15,400 moles

• 5 cal /c/ mole) =

7,500 C Pressure rise = 1 atm x (7,500 + 273)/(298) * (1/1.5) = 17.4 atm = 256 psia (ARF x RF) = 7E-2 Sourceterm based upon 500 gallon neel = 0.5% of contents Point Paper May 27,1999

i 89 1 l

' Table B.8 LAW Feed Tank Failure - Hydrogen Detonation with Heel Radionuclide Low Activity Waste (LAW), 500 gal Heel Basis, 5% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies ]

Cobalt-60 2.85 0.1995 435.5085 0.005153241 Strontium-90 2,970 207.9 2688147 31.8080370069 Technetium-99 9.3 0.651 54.2283 0.00064167 Cesium-134 12.9 0.903 418.089 0.004947122 Cesium-137 45,900 3213 1024947 12.1278903669 Europium-154 15.2 1.064 3043.04 0.0360073794 Europium-155 8.5 0.595 246.33 0.002914749 Plutonium-2'39 0.9 .

0.063 270270- 3.198023829 Plutonium-240 0.25 0.0175 75075 0.8883399525 Plutonium-241 6.53 0.4571 37710.75 0.44621999153 l

Americium-241 41.1 2.877 12773880 151.149489876 Curium-244 0.88 0.0616 152768 1.8076579136 Totals 48968.41 3427.7887 17026994.9458 201.475323095  ;

Total estimated, unmitigated consequence is 20,000 rem to the receptor at 100 meters.

i e

., Point Paper May 27,1999

90 Table 8.9: HLW Feed Tank Failure - Hydrogen Detonation with Heel Radionuclide High Level Waste (HLW),500 gal Heel Basis,10% Solids Material At Source Term. (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies Cobalt-60 5.7 0.399 871.017 0.0103064829 Strontium-90 6,000 420 5430600 64.25866062 Technetium-99 8.5 0.595 49.5635 0.00058647 Cesium-134 25.7 1.799 832.937 0.009855894 Cesium-137 5,700 399 127281 1.5060778887 Europium-154 - 30.3 2.121 6066.06 0.0717778682 Europium-155 17- 1.19 '492.66 0.005829498 Plutonium-239 1.8 0.126 540540' 6.396047658 Plutonium-240 0.5 0.035 150150 1.776679905 Plutonium-241 13.1 0.917 75652.5 0.89517333675 Americium-241 81.4 5.698 25299120 299.356897224 Curium-244 1.8 0.126 312480 3.697482096 Totals 11885.8 832.006 31944135.7375 377.985374941 Total estimated, unmitigated consequence is 37,800 rem to the receptor at 100 meters.

For the cesium tank, the heel is smaller, so use the same percentage (0.5% = 5 gal):

Sourceterm = (1.32E6) * (0.005) = 6,600 curies DOSE = (6,600 Ci) x (319 Sv/Ci) x (3.41E-2 s/m3) x (3.47E-4 m3/s) = 24.9 Sv (2,500 rem).

For the scenario where there is'a relatively large fraction of liquid in the tank, containment failure followed by pressurized venting of the liquid approximates the situation. The low pressure scenario (< 3 atm) has a release fraction of 4E-5, while a high pressure venting scenario (3-30 atm overpressure, from higher concentrations of hydrogen and flammable gases) increases the release fractions to around 2E-3.

Point Paper May 27,1999

l

~.

91 ,

. Table B.10

• LAW Feed Tank Failure - Low Pressure Hydrogen Deflagration with Liquids Radionuclide Low Activity Waste (LAW),100 Kgal Basis,5% Solids ,

Material At Sourc'e Term, '(Curies x DCF), Dose to 100 m Risk (MAR), Curies ' Sv Receptor, Sv _

Curies Cobalt-60 570 '

O.0228 49.7724 0.00058894 ,

Strontium-90 593,000 23.72 306699.6 ' 3.62908435692 Technetium-99 1,850 0.074- 6.1642 0.0000729 Cesium-134 - 2,570 0.1028- - 47.5964 0.0005G319 Cesium-137 9.17E6 366.8 117009.2 1.38453476084 Europium-154 3,030 0.1212 346.632 0.004101592 Europium-155 1,700 0.068- 28.152 0.00033311 Plutonium-239 180 0.0072 30888 0.3654884376 Plutonium-240 49 0.00196 8408.4 0.0994940747 Plutonium-241 1,306- 0.05224 4309.8 0.0509965705 Americium-241 8,210 0.3284 1458096 17.2532125392 Curium-244 176- 0.00704 17459.2 0.20658947584 Totals 9782641 391.30564- 1943348.517 22.9950599971 Total estimated, unmitigated consequence is 2,300 rem to the receptor at 100 meters.

j 9

Point Paper _ -

May 27,1999

. . i r

i:

92

. Table B.11 LAW Feed Tank Failure - High Pressure Hydrogen Deflagration with Liquids Radionuclide - Low Activity Waste (LAW),100 Kgal Basis,5% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv. Receptor, Sv

) Curies j Cobalt-60 570 1.14- 2488.62 0.0294470939 l

Strontium-90 593,000 1186 15334980 181.454217846 Technetium-99 1,850- 3.7 308.21 0.003646956 Cesium-134 2,570 5.14 2379.82 - 0.0281596961 Cesium-137 9.17E6 18340 5850460 69.226738042 Europium-154 3,030 6.06 17331.6 0.205079'62332 Europium-155 1,700 3.4 1407.6 0.0166557085 Plutonium-239 180 0.36 1544400 18.27442188 Plutonium-240 49 -

0.098 420420 4.974703734 Plutonium-241 1,306 2.612 215490 2.549828523 Americium-241 8,210 16.42 72904800 862.66062696 Curium-244 176 0.352 872960 10.329473792 Totals 9782641 19565.282 97167425.85 1149.75299986

-- Total estimated, unmitigated consequence is 115,000 rem to the receptor at 100 meters.

Note that if the LAW can be considered as similar to a " concentrated heavy metal solution," then a lower release fraction (ARF x RF) of 4E-4 can be used, which reduces the estimated doses by a factor of 5.

Point Paper May 27,1999 I

i

___ _ _N

93 Table B.12: HLW Feed Tank Failure - Low Pressure Deflagration with Liquids Radionuclide High Level Waste (HLW),100 Kgal Basis,10% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies Cobalt-60 1,140 0.0456 99.5448 0.001177884 Strontium-90 1.2E6 48 620640 7.343846928 Technetium-99 1,700 0.068 5.6644 0.000067 Cesium-134 5,140 0.2056 95.1928 0.001126388 Cesium-137 1.14E6 45.6 14546.4 0.17212318728' Europium-154 6,060 0.2424 693.264 0.008203185 Europium-155 3,400 0.136 56.304 0.00066623 Plutonium-239 360 0.0144 61776 0.7309768752 Plutonium-240 100 0.004 17160 0.203049132 Plutonium-241 2,610 0.1044 8613 0.1019150451 Americium-241 16,280 0.6512 2891328 34.2122168256 Curium-244 352 0.01408 34918.4 0.41317895168 Totals 2377142 95.08568 3649931.77 43.1885476549 Total estimated, unmitigated consequence is 4,300 rem to the receptor at 100 meters.

Point Paper- May 27,1999 W

l l

94 Table B.13 HLW Feed Tank Failure - High Pressure Deflagration with Liquids Radionuclide High Level Waste (HLW),100 Kgal Basis,10% Solids Material At Source Term, (Curies x DCF), Dose to 100 m Risk (MAR), Curies Sv Receptor, Sv Curies 1,140 2.28 4977.24 0.0588941877 Cobalt-60_

Strontium-90 1.2E6 2400 31032000 367.1923464 Technetium 1,700 3.4 283.22 0.003351257 Cesium-134~ 5,140 10.28 4759.64 0.0563193922 Cesium-137 1.14E6 - 2280 727320 8.606159364 Europium-154 6,060 12.12 34663.2 0.41015924664 Europium-155 3,400 6.8 2815.2 0.033311417 Plutonium-239 360 0.72 3088800 36.54884376 Plutonium-240 100 0.2 858000 10.1524566 Plutonium-241 2,610 5.22 430650 5.095752255 Americium 241- 16,280 32.56 144566400 1710.61084128 Curium-244 352 0.704 1745920 20.658947584 Totals 2377142 4754.284 182496588.5 2159.42738274 Total estimated, unmitigated consequence is 216,000 rem to the receptor at 100 meters.

Note that if the HLW can be considered as similar to a " concentrated heavy metal solution," then a lower release fraction (ARF x RF) of 4E 1 can be used, which reduces the estimated doses by a factor of 5. .

BJ CST Drvina ,

1 This provides supporting information for Section 5.1.5.

For the loaded CST container:

300,000 curies of Cs-137.

Point Paper May 27,1999

1

. 1

.. )

95 i

1 (H2 - beta-gamma) = (3E5 Ci) x (3.7E10 dps/Ci) x (0.7E6 ev/dps) x (0.43 molecule H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 4.5 liters /hr at STP (complete absorption / poor draining, maximum l generation condition)

(H2 - beta-gamma) = (3E5 Ci) x (3.7E10 dps/Ci) x (SE2 ev/ micron /dps) x (1 micron water film) x (0.43 molecule H2/100 ev) x (1 mole /6.02252E23 molecule H2) x (3,600 sec/hr) x (22.414 liters / mole)

= 3.2 cc/hr at STP (thin film / poor absorption / good draining / low temp. drying, minimum generation condition)

Pressure accumulation:

assume an inert cover gas at 1 atm (abs) - SNF storage usually uses a pressurized cover gas.

Assume a design rating of 150 psig (a typical standard rating for components) and an internal void volume of 50 liters, based upon'the canister design. J Gas from radiolysis (moles) = (P x V) / (R x T)

= (11 atm x 50 liters) / (8.205E-21-atm/(mole - K) x 298 K)

= 22.5 moles.

l Assume both hydrogen and oxygen evolved, implies 2 moles cover gas,13.7 moles hydrogen, I and 6.9 moles oxygen. Minimum pressure rise due to water (no radiolysis)is 100 psig.  ;

Quantity of residual water is 247 grams (0.25 liters, or 0.5% of the void volume). j The first scenario assumes a canister filled with feed solution (i.e., the organic ion exchange eluate) and the CST loaded with cesium. A hydrogen-oxygen bubble develops and detonates.

Collateral damage occurs to two adjacent, loaded (but already drained) canisters. There are three contributing sources to the release:

The wet. loaded CST:

. MAR = 300,000 curies DR=1 There is relatively little data on the effects of detonation upon brittle ceramics like CST. Values for gas deflagrations go up to an ARF of 1, while an ARF of 0.1 and an RF of 0.7 are recommended as bounding for pressurized gas venting through a powder-like bed of solids.

ARF = 0.5 RF = 0.7 Point Paper May 27,1999

96 DOSE = (3E5 Ci) x (0.5 x 0.7) x (319 Sv/Ci) x (3.41E-2 s/m3) x (3.47E-4 m3/s)

= 396 Sv (39,600 rem)

Use of a reduced ARF of 0.1 produces a lower result of approximately 8,000 rem.

The solution: .

Assume 100 liters of solution (eluant) affected - 50 liters within the column and 50 liters in the immediate piping / manifolds. The ion exchange data implies'a 10 fold increase in the cesium concentration of the eluant as compared to the original feed solution - assume this also holds true for the Envelope B waste. Thus, the concentration is increased from 22.7 Ci/ liter in the feed to 227 Ci/ liter in the etuant.

MAR = 1.00 liters x 227 Ci/ liter = 23,000 curies.

DR=1 Vaporization of 100 liters of fluid requires 220,000 BTU, and the hydrogen burn releases about 3,200 BTU. .Thus, at most,1.4% of the solution vaporizes. The liquid venting can also be considered a high-pressure venting (i.e., no thermal or evaporation effects), which has a bounding ARF of 2E-3 and a bounding RF of 1. These calculations use the 1.4% value based upon the maximum amount that can vaporize.

ARF = 1.4E-2 RF = 1 DOSE = (23,000 Ci) x (1.4E-2) x (319 Sv/Ci) x (3.41E-2 s/m3) x (3.47E-4 m3/s)

= 1.22 Sv (122 rem)

The drained CST in the two adiacent containers:

MAR = 300,000 x 2 = 600,000 curies DR = 0.5 The release fractions depend upon the gas pressure and any metallic remnants that disperse the material more (e.g., in the tests, parts of the rupture disks; in the scenario, shrapnel from the failed containers) (see Reference 21, page 4-71). The bounding values are:

ARF = 0.1 RF = 0,7 DOSE = (6E5 Ci) x (0,.5) x (0.1 x 0.7) x (310 Sv/Ci) x (3.41E-2 s/n-3) x (3.47E-4 m3/s)

= 79.3 Sv (7,900 rem).

Point Paper May 27,1999

r..

g 97 Total estimated dose from all three contributors becomes-Total = 39,600 + 122 + 7,900 = 47,600 rem to the receptor at 100 meters.

I Sl Canister Glass Radionuclide Content Calculations:

This provides supporting information for Section 5.1.8. l Assume a 4.5 meter canister is produced, and is subject to the same volumetric thermal density limitations as the 3 meter high canister.

3 meter canister = 1.5 KW(th) maximum, approximate glass volume of 625 liters.

4.5 meter canister = (4.5/3) * (1.5) = 2.25 KW(th) maximum, approximate glass volume of 1,000 liters.

Assume cesium and strontium quantities are aoproximately the same, based upon standard l fission yields and the DWPF SAR (Reference 24).  !

2250 watts = { strontium curies}

• 0.00669 W/Ci Sr-90 + { cesium curies}

• 0.00369 W/Ci Cs-137 subject to { strontium curies} = { cesium curies}

Solution gives 217,000 curies for strontium and 217,000 curies for cesium in each 4.5 meter high j canister.

Ratio approximate values for TRU isotopes from DWPF SAR (Reference 24):

Plutonium 239: 51 curies Plutonium-240: 34 curies Plutonium-241: 6,600 curies Americium-241: 43 curies Curium-241: 431 curies Total: 7,200 curies of TRU.

Note that the values for curium and americium appear to be juxtaposed in the original reference, but the effect has little calculational value because of the similarity between the two isotopes' dose conversion factors.

Ratio te.chnetium content using Envelope A/C ratio to cesium:

{ technetium activity per 4.5 meter canister} = (7.1E6/4.3E9)

• 217,000

= (0.0017)

• 21.7,000 = 360 Ci Tc.

Point Paper -

May 27,1999

98 Glass Oanister Droc Scenario Calculations:

This assumes the canister is filled with glass and that the glass has cooled to room temperature, and that only the temporary plug is installed. During transfer' operations to the welder, the canister is dropped, the temporary plug falls out, and glass fines and whiskers are released as fine particulates. The scenario assumes only one canister is involved. The review of a specific design might indicate the consideration of two or more canisters being affected.

Use (ARF x RF) = 2.5E-5, DR=1, LPF=1. -

The results become:

Table B.14: Glass Canister Drop Scenario lsotope MAR Source Term, (Curies x DCF), Dose to 100 m Curies Sv Receptor, Sv Strontium-90 217,000 5.43 70,000 0.828289 Technetium-99 360 9E-3 0.75 0.000009 Cesium-137 217,000 5.43 1,700 0.02011559 Plutonium-239 51 1.3E-3 5,580 0.066026466 3

Plutonium-240 34 8.7E-4 3,700 n.04378099 Plutonium- 241 6,600 0.17 14,000 0.1656578 Americium-241 43 8.4E-4 3,700 0.04378099 Curium-244 431 1.1 E-2 27,000 0.3194829 Total 442,000 11.1 126,000 1.48714261053 I

. Cold Cao Dispersion:

Cold Cap Volume: about 17 liters.

Cold Cap Concentration: about 1/7 of glass product.

Cold Cap radionuclide quantities = (17/1000) * (1/7) = 2.4E-3 of 4.5 meter high canister.

Assume instantaneous vaporization of cold cap by the molten glass spill: (ARF x RF) = 1.

Point Paper May 27,1999

- l

l l 99

. l The results become:

Cold Cap Dispersion l Table B.15:

isotope MAR Source Term, (Curies x DCF), Dose to 100 m Curies Sv Receptor, Sv -

Strontium-90 521 521 6.74E6, 79.752398 Technetium-99 0.86 0.86 72 0.00085195 Cesium-137 521- 521 166,000 1.9642282 Plutonium 239 0.12 0.12 515,000 6.0938405 Plutonium-240 0.082 0.082 352,000 4.1651104 )

Plutonium- 241 16 16 1.32E6 15.619164 Americium-241 0.1 0.1 444,000 5.2537188 l

Curium-244 1 1 2.48E6 29.345096 l Total 1,060 1,060 1.2E7 142.194407854 Catastrophic Molten Glass Soill From Melter / Full Canister Accident:

Molten glass volume of melter = 3,800 liters l Molten glass volume of canister = 1,000 liters l Total volume of molten 91 ass = 4,800 liters (MAR) l Model as a free-fall liquid spill, with a viscosity > 8 centipoise. l Bounding ARF = 7E I Bounding RF = 0.8 (ARF x RF) = 5.6E-6 l l

The accident handbooks (References 20 and 21) do not have any data on molten glass spills.

l The liquid value (above) may not underestimate the (ARF x RF) product, as it is a factor of 5 l lower than the corresponding value for solid HLW glass fines.

l i

Point Paper May 27,1999 L-

C 100 The results for the spill become:

. Table B.16:

Catastrophic Glass Spill from Melter and Full Canister.

Isotope MAR Source Term, (Curies x DCF), Dose to 100 m -

Curies Sv Receptor, Sv Strontium 1.04E6 5.82 75,600 0.89455212 Technetium-99 1,730 9.69E-3 0.81 0.00001 Cesium-137 1.04E6 5.82 1,840 0.021772168 Plutonium-239 245 1.37E-3 6,030 0.071351181 Plutonium-240 163 9.13E-4 4,000 0.0473308 Plutonium- 241 31,680 0.18 15,100 0.17867377 Americium-241 206 1.15E-3 5,110 0.060465097 Curium-244 2,070 1.16E-2 29,200 0.34551484 Total 2.12E6 11.9 136,000 1.61966956049 Assume the glass remains sufficiently liquid for one hour, and include one hour's worth of resuspension effects:

.(ARF x RF) = (4E-6/hr x 1 hr x 1) = 4E-6 I

Point Paper ,

May 27,1999 l

l 101 Table B.17: Catastrophic Glass Spill from Melter and Full Canister - Resuspension Effects Isotope MAR Source Term, (Curies x DCF), Dose to 100 m . I Curies Sv Receptor, Sv Strontium-90 ' 1.04E6 4.13 53,700 0.63541599 Technetium-99 1,730- 6.88E-3 0.58 0 Cesium-137 - 1.04E6 4.13 1,310 0.015500837 ,

Plutonium-239 245 9.73E-4 4,280 0.050643956 Plutonium-240 163 6.48E-4 2,840 0.033604868 Plutonium-241 31,680 0.13 10,700 0.12660989 .

Americium-241 206 8.17E-4 3,630 0.042952701 Curium-244 2,070 8.24E-3 20,700 0.24493689 Total 2.12E6 8.45 96,600 1.149665132 The total estimate for a major failure of the metter is:

DOSE = (Cold Cap Dispersion) + (Catastrophic Glass Spill) + (Resuspension) '

= 14,200 + 162 + 115 = 14,500 rem Steam Explosion in Melter - Contribution from Molten Glass in Melter and Canister:

1 For a steam explosion in a melter, include the filling (full) canister inventory, and use the values for high pressure venting from a vessel containing a heavy solution:

ARF = 1E-3, RF=0.4, (ARF x RF) = 4E-4.

l Point Paper May 27,1999

[ .

O 102 Table B.18: Steam Explosion in Melter 5 isotooe MAR Source Term, I (Curies x DCF), Dose to 100 rri Curies Sv Receptor, Sv Strontium-90 1.04E6 416 5.4E6 63.89658 Technetirn-99 1,730 '.692 0 57.8 0.00068393 Ceskum-137 1.04E6 416 131,000 1.5500837 Plutonium-239 245 0.098 431,000 5.0998937 Plutonium-240 163 0.065 286,000 3.3841522 Plutonium- 241 31,680 12.9 1.08E6 12.779316 365,000 4.3189355 Americium-241 206 0.082 Curium-244 2,070 0.828 2.08E6 24.612016 Total 2.12E6 850 9.71E6 115.64166103 ,

The total estimate is:

u DOSE = (Cold Cap Dispersion) + (Steam Explosion) + (Resuspension)

= 14,200 + 11,600 + 115 = 26,000 rem.

BJ Nonradioactive Chemical Effects Both ammonia arid nitric acid are likely to be present in sufficient quantities (5,000 - 10,000 gallons or more, each) to present occupancy concerns for operators in the TWRS-P facility. The EPA's Risk Management Program (RMP) provides a computer code (RMP* Comp) for estimating the affected environment from chemical releases, as a distance from a release for the chemical concentrations to fall oelow ERPG-2 levels (also referred to as the toxic endpoint). This code has been exercised for potential releases of ammonia, nitric acid, and propane, a typical fuel gas for industriai facilities.

l Ammonia can be stored and used in several different forms. Anhydrous ammonia is a pressurized liquid under pressure, and represents the worst case condition from the perspective of potential releases. The code assumes the ammonia is liquified, under pressure, in a rural area (flat /few obstacles and turbulence generators), without an enclosure,1.5 m/s (3.4 mph) windspeed, F Class stability, and 77 F temperature.

i .:;5 Point Paper Ma'y 27,1999

c: .

)..' 3 103.

- Table B.19: Toxic Endpoints (Distances) for Anhydrous Ammonia Releases ,

I Quantity of Release, Distance to Toxic Endpoint, miles (km) 77"F Temperature 1,000 0.6 (1) '

5,000 1.3 (2.1) 1 10,000 1.8 (2.9) I 20,000 2.6 (4.2) 25,000 2.9 (4.7) 50,000 4.0 (6.4) 100,000 5.6 (9.0)  !

- Toxic Endpoint = 0.14 mg/ liter (ERPG-2)  !

Table B.20: Toxic Endpoints (Distances) for Aqueous Ammonia (30%) Releases Quantity of Release,. Distance to Toxic Endpoint, miles (km) pounds 77'F Temperature 1,000 0.3 (0.5) 5,000 0.7 (1.1) 10,000 0.9 (1.4) l l

20,000 1.3 (2.1)  :

25,000 1.4 (2.3) 50,000 2.0 (3.2) l 100,000 4.3 (6.9) i The nitric acid calculations assume 1.5 m/s (3.4 mph) windspeed, Class F stability, and an air

. temperature of 77'F.

Point Paper . May 27,1999 4

1 i

104 Tab.e 8.21: Toxic Endpoints (Distances) for Anhydrous Nitric Acid Releases I

l Quantity of Release, Distance to Toxic Endpoint, miles (km) )

pounds 1 80 F Temp,erature 140 F Temperature '

1,000 0.9 (1.4) 3.8 (6.1) 5,000 1.4 (2.3) 5.5 (8.9) 10,000 1.4 (2.3) [ dike} 5.5 (8.9) [ dike] j 3.8 (6.1)[ loss of dike) 12 (19) [ loss of dike) 20,000 1.4 (2.1) 5.5 (8.9) 25,000 1.4(2.3) 5.5 (8.9) 50,000 1.4 (2.3) (dike] 5.5 (8.9) [ dike) 8.7 (14) [ loss of dike) >25 (>40)[ loss of dike) 100,000 8.7 (14) >25 (>40)

- assumes 25ft by 25 ft by 1 ft high dike around the tank, unless noted.

- air temperature of 77 F, liquid temperature as noted.

- toxic endpoint is 0.026 mg/l (IDLH) )

i Table B22: Toxic Endpoints (Distances) for Aqueous Nitric Acid (80%) Releases Quantity of Release, Distance to Toxic Endpoint, miles (km) i 80 F Temperature 140 F Temperature 1,000 0.7 (1.1) 2.2 (3.5) 5,000 1 (1.6) 2.9 (4.7) 10,000 1 (1.6) 2.9 (4.7) [ dike) 6.2 (10)[ Loss of dike) 20,000 1 (1.6) 2.9 (4.7) 25,000 1 (1.6) 2.9 (4.7) ,

l 50,000 1 (1.6) 2.9 (4.7) [ dike]

11 (18)[ Loss of dike]

100.000 4.5 (7.2) 11 (18)

- assumes 25 ft by 25 ft by 1 ft high dike around tank, unless noted. ,

- 80% is the lowest concentration allowed by the computer code. I i

Point Paper May 27,1999

105 Table B.23: Toxic Endpoints (Distances) for LPG / Propane Releases Quantity of Release, Distance to Toxic Endpoint, miles (km) pounds 77 F Temperature 1,000 0.08(0.13) 5,000 0.1 (0.2) l 10,000 i 0.2 (0.3) l I'

20,000 0.2 (0.4) 25,000 0.2 (0.4) 50,000 0.3 (0.5) 100,000 0.4 (0.6)

- Toxic Endpoint for propane is equivalent to the distance where the overpressure from an ,

explosion falls below 1 psi. I RJ Inclusion of Freauencies and Risk Frequencies are estimated from published documents and experience. Risk is estimated as the product of frequency and consequence, divided by dose / cancer conversion factor, i.e.

Risk = Consequence x Frequency / conversion factor (yr-1]

The estimates use 2,500 rem per fatal cancer for worker populations, and 2,000 rem per fatal cancer for the general population. Some reference sources recommend a lower population dose / cancer relationship for higher acute exposures (e.g ,1,000 rem population dose per fatal cancer for acute individual doses above 10 rem). Such potential effects are not included here because they do not affect the conclusions regarding items relied upon for safety and prevention / mitigation requirements. Thus, the aforementioned 2,500 and 2,000 rem values are used as constants, regardless of the acute dose.

Point Paper ,

May 27,1999 u