ML23104A056

STJ-02MSRE-D992 Msre Engr Eval 2016

ML23104A056
Person / Time
Site:	Abilene Christian University
Issue date:	11/03/2016
From:	CH2M Hill, Oak Ridge
To:	Office of Nuclear Material Safety and Safeguards, US Dept of Energy, Office of Environmental Management
References
DE-SC-0004645 STJ-02MSRE-D992
Download: ML23104A056 (1)

Text

STJ-02MSRE-D992 SAFELY DELIVERING THE DEPARTMENT OF ENERGYS VISION FOR THE EAST TENNESSEE TECHNOLOGY PARK MISSION Molten Salt Reactor Experiment Engineering Evaluation and Extended Life Study, URS l CH2M Oak Ridge LLC, Oak Ridge, Tennessee This document is approved for public release per review by:

Guy Wilson 11/3/2016 ORNL Classification Office Date

Global Energy Services LLC, Restoration Services, Inc., and Value Added Solutions, Inc. contributed to the preparation of this document and should not be considered eligible contractors for its review.

STJ-02MSRE-D992 Molten Salt Reactor Experiment Engineering Evaluation and Extended Life Study, URS l CH2M Oak Ridge LLC, Oak Ridge, Tennessee Date IssuedNovember 2016 Prepared for the U.S. Department of Energy Office of Environmental Management URS l CH2M Oak Ridge LLC Safely Delivering the Department of Energys Vision for the East Tennessee Technology Park Mission under contract DE-SC-0004645

This page intentionally left blank.

APPROVALS Molten Salt Reactor Experiment ST J -02MSRE-D992 Engineering Evaluation and Extended Life Study, USQDReview Determ inatio11 Exemptio11 Criteria USQD Preparer:

Exllibit L Mandatory Co11tractor Docume11t PCCB Reviewer:

Prepared by:

URS I CH2M Oak Ridge LLC, November 2016 Oak Ridge, Tennessee OUSQD ouco 0CATX D Exempt (Select Criteria 1-3 below.)

USQDIUCD/CAT X No.:

Non-Intent Change DOE-Approved Safety Basis Document DOl om 0(3J Chief Accounting Officer. Internal Audit. Labor Rt!lations. General Counsel. Outreach &

Public Affairs. or Project Controls Services 181 No (No PCCB Reviewer Signature Required) 0 Yes

<Requires review by the Proforma Change Control Board.)

Name Tommy Morgan, Nuclear ngmeer Nuclear and High Hazard Operations URS I CH2M Oak Ridge LLC b./5>~~

Brent Rankin. Senior Technical Advisor Reactor Operations/Engineering Global Energy Services LLC Ill Date Date Date Date 1o/h/zott, Dak 1

Concurred/

Reviewed by:

Approved by:

Operations and Engineering Global Energy Services LLC Craig Eutz, MS E Faci y Man ger URS I CH2M Oa

• 1 ge LLC J.d.~

Nuclear and High Hazard Operations URS I CH2M Oak Ridge LLC

~eld, Technical Support Manager Nuclear and High Hazar Operations URS I CH2M Oak Rid e L Harold Conner, Manager Nuclear Services & Engineering URS I CH2M Oak Ridge LLC and Reporting Jeffrey Paris, Engineering Mana r URS I CH2M Oak Ridge LLC 111A(J~

Malcolm Carman, Manager Nuclear Facility Safety and Criticality URS I CH2M Oak Ridge LLC iv loizbtJ/6 Date Date~ 1 112/;tP/i~

Date I

10-20 *-zoM Date

/0-,U) -llo Date

!O(~o;/b Datl I

/6 -)-o- ('

Date Date

/4-ZJ/-16 Date

REVISION LOG Revision Number Description of Changes Pages Affected 0

Initial issue of document.

All v

This page intentionally left blank.

vi

CONTENTS FIGURES..................................................................................................................................................... xi TABLES..................................................................................................................................................... xii TERMS AND ACRONYMS..................................................................................................................... xiii EXECUTIVE

SUMMARY

........................................................................................................................ xv

1. PURPOSE........................................................................................................................................... 1-1

2. BACKGROUND................................................................................................................................. 2-1

2.1 REFERENCES

........................................................................................................................... 2-3

3. SCOPE................................................................................................................................................. 3-1 3.1 SYSTEM DESCRIPTION......................................................................................................... 3-1 3.2 CURRENT CONDITION EVALUATION............................................................................... 3-1 3.3 NEAR-TERM UPGRADE RECOMMENDATIONS............................................................... 3-2 3.4 LIFE-EXTENSION RECOMMENDATIONS.......................................................................... 3-2 3.5 WASTE MANAGEMENT PROGRAM.................................................................................... 3-2 3.6 SALT REMOVAL OPTIONS EVALUATION........................................................................ 3-2

3.7 REFERENCES

........................................................................................................................... 3-2

4. DRAIN TANKS, FLUSH TANK, AND DRAIN TANK CELL........................................................ 4-1 4.1 SYSTEM DESCRIPTION......................................................................................................... 4-1 4.2 CURRENT CONDITION........................................................................................................ 4-10 4.2.1 Tank Integrity and Salt Status.................................................................................... 4-10 4.2.2 Fluorine Gas Generation and Management in Drain Tanks....................................... 4-12 4.2.3 Drain Tank Cell and Tank Heaters............................................................................. 4-15 4.3 NEAR-TERM RECOMMENDATIONS................................................................................. 4-15 4.4 LIFE-EXTENSION RECOMMENDATIONS........................................................................ 4-16

4.5 REFERENCES

......................................................................................................................... 4-16

5. REACTIVE GAS REMOVAL SYSTEM........................................................................................... 5-1 5.1 RGRS SYSTEM DESCRIPTION.............................................................................................. 5-1 5.1.1 Piping/Tubing and Valves............................................................................................ 5-1 5.1.2 RGRS Glovebox........................................................................................................... 5-1 5.1.3 Chemical Traps............................................................................................................. 5-2 5.1.4 Vacuum pumps............................................................................................................. 5-3 5.1.5 FTIR Data Acquisition System..................................................................................... 5-4 5.2 CURRENT CONDITION EVALUATION............................................................................... 5-6 5.2.1 Operational Observations............................................................................................. 5-6 5.2.2 Corrosion Issues with the RGRS................................................................................ 5-12 5.3 NEAR-TERM UPGRADE RECOMMENDATIONS............................................................. 5-14 5.4 LIFE-EXTENSION UPGRADE RECOMMENDATIONS.................................................... 5-15 5.4.1 Option 1: Continuous Purge of FDTs and FFT Headspace Gas Using CVS with No Traps............................................................................................................. 5-16 5.4.2 Option 2: Continuous Purge of FDTs and FFT Headspace Gas Using CVS with a NaF trap and Booster Pump............................................................................. 5-17 vii

5.4.3 Option 3: Continuous Purge of FDTs and FFT Headspace Gas Using Both CVS Fans with a NaF Trap......................................................................................... 5-17 5.4.4 Option 4: Routine Frequent Pumpdown of FDTs and FFT to Minimize Fluorine Inventory...................................................................................................... 5-17 5.4.5 Option 5: Design and Build a New Fluorine Gas Removal System to Replace RGRS.......................................................................................................................... 5-17 5.4.6 Recommendations...................................................................................................... 5-17

5.5 REFERENCES

......................................................................................................................... 5-18

6. SUPPORT AND ANCILLARY SYSTEMS....................................................................................... 6-1 6.1 ELECTRICAL SYSTEM DESCRIPTION................................................................................ 6-1 6.1.1 Current Condition Evaluation....................................................................................... 6-1 6.1.2 Near-Term Upgrade Recommendations....................................................................... 6-4 6.1.3 Life-Extension Upgrade Recommendations................................................................. 6-4 6.1.4 References.................................................................................................................... 6-5 6.2 CONTAINMENT VENTILATION SYSTEM DESCRIPTION............................................... 6-5 6.2.1 Current Condition Evaluation....................................................................................... 6-5 6.2.2 Near-Term Upgrade Recommendations....................................................................... 6-6 6.2.3 Life-Extension Upgrade Recommendations................................................................. 6-7 6.2.4 References.................................................................................................................... 6-8 6.3 FIRE PROTECTION AND FACILITY LIFE SAFETY SYSTEMS........................................ 6-9 6.3.1 Current Condition Evaluation..................................................................................... 6-10 6.3.2 Near-Term Recommendations.................................................................................... 6-10 6.3.3 Life-Extension Recommendations.............................................................................. 6-11 6.3.4 References.................................................................................................................. 6-11 6.4 STRUCTURAL SYSTEM....................................................................................................... 6-11 6.4.1 Current Condition Evaluation..................................................................................... 6-13 6.4.2 Near-Term Upgrade Recommendations..................................................................... 6-21 6.4.3 Life-extension upgrade recommendations.................................................................. 6-23 6.4.4 References.................................................................................................................. 6-23 6.5 CRANES.................................................................................................................................. 6-24 6.5.1 Current Condition Evaluation..................................................................................... 6-24 6.5.2 Near-Term Recommendations.................................................................................... 6-25 6.5.3 Life-Extension Recommendations.............................................................................. 6-25 6.6 AIR COMPRESSORS............................................................................................................. 6-25 6.6.1 Current Condition Evaluation..................................................................................... 6-26 6.6.2 Near-Term Recommendations.................................................................................... 6-26 6.6.3 Life-Extension Recommendations.............................................................................. 6-26 6.7 CHARCOAL BEDS................................................................................................................. 6-27 6.7.1 Main Charcoal Bed, Auxiliary Charcoal Bed, and Charcoal Canister....................... 6-27 6.7.2 References.................................................................................................................. 6-30 6.8 STEAM.................................................................................................................................... 6-31 6.8.1 Current Condition Evaluation..................................................................................... 6-31 6.8.2 Near-Term Recommendations.................................................................................... 6-31 6.8.3 Life-Extension Recommendations.............................................................................. 6-32

7. WASTE MANAGEMENT PROGRAM............................................................................................. 7-1 7.1 PROGRAM DESCRIPTION..................................................................................................... 7-1 7.2 CURRENT CONDITION.......................................................................................................... 7-1 7.3 NEAR-TERM RECOMMENDATIONS................................................................................... 7-2

7.4 REFERENCES

........................................................................................................................... 7-2 viii

8. SALT REMOVAL OPTIONS EVALUATION.................................................................................. 8-1 8.1 SALT REMOVAL

SUMMARY

............................................................................................... 8-1 8.2 SALT REMOVAL OPTIONS ANALYSES............................................................................. 8-1 8.2.1 Tank Removal/Sectioning............................................................................................ 8-1 8.2.2 Thermal Salt Transfer................................................................................................... 8-2 8.2.3 Chemical Dissolution................................................................................................... 8-4 8.2.4 Mechanical Removal with Needle Scaler..................................................................... 8-5 8.2.5 Entombment................................................................................................................. 8-6 8.3 SALT REMOVAL OPTIONS EVALUATION AND CONCLUSIONS.................................. 8-6

8.4 REFERENCES

........................................................................................................................... 8-7

9. RECOMMENDATIONS.................................................................................................................... 9-1 9.1

SUMMARY

............................................................................................................................... 9-1 9.2 NEAR-TERM RECOMMENDATIONS................................................................................... 9-1 9.3 LIFE-EXTENSION RECOMMENDATIONS.......................................................................... 9-3

9.4 REFERENCES

........................................................................................................................... 9-4 APPENDIX A. GENERAL CORROSION INFORMATION................................................................. A-1 A.1 PUBLISHED CORROSION RATE INFORMATION............................................................. A-4 A.1.1 Materials of Construction Guideline for Anhydrous Hydrogen Fluoride.................... A-5 A.1.2 Corrosion Resistance Tables....................................................................................... A-5 A.1.3 Corrosion Resistance of Nickel-Containing Alloys to Fluorine Compounds.............. A-6 A.1.4 Corrosion Characteristics of Stainless Steel Alloys Exposed to F2............................. A-6 A.1.5 Corrosion Rates for Selected Alloys in Hydrofluoric Acid......................................... A-7 A.2 ADDITIONAL REFERENCES................................................................................................ A-8 APPENDIX B. ULTRASONIC THICKNESS AND X-RAY FLUORESCENCE MEASUREMENTS OF RGRS TUBING........................................................................................................................... B-1 B.1 REFERENCES.......................................................................................................................... B-5 APPENDIX C. MCL METALLURGICAL AND ANALYTICAL RESULTS ON RGRS PUMP TUBING/DEPOSITS......................................................................................................................... C-1 C.1 REFERENCES........................................................................................................................ C-12 APPENDIX D. CONCENTRATION OF HF RELEASED UNDER CONTINUOUS RGRS SYSTEM D-1 D.1 EXECUTIVE

SUMMARY

....................................................................................................... D-3 D.2 PURPOSE................................................................................................................................. D-3 D.3 CALCULATIONS.................................................................................................................... D-3 D.3.1 Assumptions................................................................................................................ D-3 D.3.2 Diagram....................................................................................................................... D-4 D.3.3 Variables...................................................................................................................... D-4 D.3.4 Equations..................................................................................................................... D-4 D.3.5 Conversions................................................................................................................. D-5 D.3.6 Methodology................................................................................................................ D-5 D.4 RESULTS.................................................................................................................................. D-6 D.5 DISCUSSION........................................................................................................................... D-7 D.6 REFERENCES.......................................................................................................................... D-7 APPENDIX E. PREVIOUS MSRE SALT REMOVAL STUDIES......................................................... E-1 ix

This page intentionally left blank.

x

FIGURES Figure 2.1. MSRE facility showing the reactor and drain tanks................................................................ 2-2 Figure 4.1. Fuel Drain Tank....................................................................................................................... 4-2 Figure 4.2. FDT thimble assembly (prior to installation).......................................................................... 4-3 Figure 4.3. Steam dome-bayonet tube assembly installation into FDT..................................................... 4-4 Figure 4.4. Drain Tank Cell....................................................................................................................... 4-5 Figure 4.5. Drain Tank Cell (during construction).................................................................................... 4-6 Figure 4.6. Penetrations in Drain Tank Cell.............................................................................................. 4-7 Figure 4.7. FDT heating element............................................................................................................... 4-8 Figure 4.8. FDT heating element, nichrome wires..................................................................................... 4-9 Figure 4.9. Simplified gas-flow diagram for FDTs, FFTs, and related process lines.............................. 4-14 Figure 5.1. Temperature readings for the alumina trap during the April 2016 pumpdown....................... 5-7 Figure 5.2. Plugged outlet of vacuum pump from April 2016 pumpdown................................................ 5-8 Figure 5.3. Line from pump outlet............................................................................................................. 5-9 Figure 5.4. Internal heavy deposits on scroll pump................................................................................... 5-9 Figure 5.5. Deposits on exhaust and ballast valve assembly................................................................... 5-10 Figure 5.6. Deposits on the inlet screen of the vacuum pump................................................................. 5-10 Figure 5.7. Molecular sieve inlet September 2016.................................................................................. 5-11 Figure 5.8. Molecular sieve outlet September 2016................................................................................ 5-11 Figure 6.1. Simplified electrical distribution diagram for MSRE complex............................................... 6-2 Figure 6.2. Cutaway illustration of MSRE looking east (from DAC-EA-020794-A018)....................... 6-12 Figure 6.3. Floor plan of MSRE High Bay.............................................................................................. 6-15 Figure 6.4. Elevation of building frames on CL 1 to CL 6 looking north (from Drawing 22-S2 Issue B, Framing Details).................................................................................................. 6-16 Figure 6.5. Elevation of building frame on CL 7 to CL 9 looking north (from Drawing D-KS-19404 G2, Rev. 0 and Drawing D-KS-19404 C3, Rev. 0)...................... 6-17 Figure 6.6. Construction of reinforced concrete piers supporting building columns............................... 6-18 Figure 6.7. Crane/building column C-7 moment connection east face.................................................... 6-20 Figure 6.8. Crane/building column C-7 moment connection west face trimmed.................................... 6-20 Figure B.1. RGRS tubing failure, discharge of vacuum pumpSpring 2016.......................................... B-4 Figure B.2. UT and XRF measurements of stainless steel RGRS tubing downstream of Al2O3 and molecular sieve traps......................................................................................... B-6 Figure B.3. UT and XRF measurements of Monel RGRS tubing upstream of Al2O3 and molecular sieve traps.......................................................................................................... B-7 Figure B.4. RGRS discharge tubing failure, downstream of vacuum pumpSeptember 2016............... B-8 Figure B.5. Calibration data and NDE technician certification................................................................ B-9 Figure C.1. Section of type 304L stainless steel tube with the fracture surface on the upper right......... C-5 Figure C.2. Section of type 304L stainless steel tube cut from the left end of the tube sample in Figure C.1.................................................................................................. C-5 Figure C.3. Secondary electron micrograph showing a portion of the fracture surface of the tube sample in Figure C.1........................................................................................ C-5 Figure C.4. EDS spectra from an area of the fracture surface where charging occurred.......................... C-6 Figure C.5. EDS spectra from an area of the fracture surface where no charging occurred..................... C-6 Figure C.6. Secondary electron image of the fracture surface and the internal surface (top left) of the tube section in Figure C.2.......................................................................................................... C-7 Figure C.7. EDS spectra from the fracture surface shown in Figure C.6.................................................. C-8 Figure C.8. EDS spectra from the inner surface of the sample shown in Figure C.6............................... C-8 Figure C.9. EDS spectra collected from a cut surface of the tube section in Figure C.6.......................... C-9 xi

Figure C.10. Secondary electron micrograph of the deposit removed from the internal surface of the short section of tubing shown in Figure C.2Area 1............................... C-10 Figure C.11. Secondary electron micrograph of the deposit removed from the internal surface of the short section of tubing shown in Figure C.2Area 2............................... C-10 Figure C.12. EDS spectra of the scale shown in Figure C.10................................................................. C-10 Figure C.13. EDS spectra of the cubic shaped material in the scale shown in Figure C.11................... C-11 Figure D.1. Drawing of the tank headspace and streams.......................................................................... D-4 Table D.1. CVS measurements used in this calculation........................................................................... D-6 Table D.2. Constants used in this calculation........................................................................................... D-7 TABLES Table A.1. Typical physical properties of Type 13X molecular sieve media............................................ 5-4 Table A.1. Overview of currently installed software versions................................................................... 5-5 Table A.1. RGRS thermocouple assembly locations................................................................................. 5-5 Table A.1. Water solubility of salt components and other species............................................................ 8-5 Table A.1. Evaluation of salt removal options........................................................................................... 8-8 Table A.2. Salt removal options ranking results...................................................................................... 8-10 Table A.1. Nominal composition of alloys used in the MSRE FDTs and RGRS..................................... A-4 Table A.2. Environments to which the components of the drain tanks and RGRS have been exposed... A-4 Table A.3. Corrosion rates for selected alloys in 3 hydrofluoric acid concentrations at 24°C and 50°C............................................................................................................................. A-7 Table C.1. Semiquantitative results of EDS examination of selected areas on the sample in Figure C.3........................................................................................................................... C-7 Table C.2. Semiquantitative results of EDS examination of selected areas on the sample in Figure C.6........................................................................................................................... C-8 Table C.3. Semiquantitative results of EDS examination of freshly cut tube surface.............................. C-9 Table C.4. Semiquantitative results of the EDS examination of the deposit shown in Figure C.10 and Figure C.11................................................................................................................ C-11 Table C.5. Identification of molybdenum isotopes in deposits collected from failed stainless steel tube. The isotopic composition of naturally occurring molybdenum is included for reference........................................................................................................................... C-12 Table D.3. Results of HF concentration calculations................................................................................ D-8 Table D.4. Spreadsheet formulas.............................................................................................................. D-9 xii

TERMS AND ACRONYMS ACB Auxiliary Charcoal Bed AOV air-operated valve ASME American Society of Mechanical Engineers CBC Charcoal Bed Cell CFR Code of Federal Regulations CL column line CVS Containment Ventilation System DOE U.S. Department of Energy DOE-ORO DOE Oak Ridge Office DSA documented safety analysis EM Office of Environmental Management FDT Fuel Drain Tank FFT Fuel Flush Tank FHA fire hazards analysis FTIR Fourier Transform Infrared Spectroscopy FTIR-DAQ FTIR data acquisition system HEPA high-efficiency particulate air JCO Justification for Continued Operation LANL Los Alamos National Laboratory LOV loss of ventilation LSS life safety system MAR material at risk MCB Main Charcoal Bed MCC motor control center MSRE Molten Salt Reactor Experiment NDA nondestructive assay NESA North Electrical Service Area NHHO Nuclear and High Hazards Operations NRTL Nationally Recognized Testing Laboratory ORNL Oak Ridge National Laboratory PISA Potentially Inadequate Safety Analysis PT pressure transducer RGRS Reactive Gas Removal System ROD record of decision S&M surveillance and maintenance SME subject matter expert thimble FDT thermal well TSR technical safety requirement TRU transuranic VCR vacuum coupling radiation WAC waste acceptance criteria WHP waste handling plan WIPP Waste Isolation Pilot Plant UCOR URS l CH2M Oak Ridge LLC UT-B UT-Battelle, LLC XRD X-ray diffraction XRF X-ray fluorescence Y-12 Y-12 National Security Complex xiii

This page intentionally left blank.

xiv

EXECUTIVE

SUMMARY

The Nuclear and High Hazard Operations (NHHO) organization of URS l CH2M Oak Ridge LLC (UCOR) performs surveillance and maintenance (S&M) activities at the Molten Salt Reactor Experiment (MSRE) facility at the Oak Ridge National Laboratory (ORNL). In 2015 and 2016, three Potential Inadequacies in the Safety Analysis (PISAs) were identified at the MSRE. The issues involve potential release of fluorine (F2) and hydrogen fluoride (HF) from the salt tanks due to corrosion of the tank thimbles, potential release of these gases from spent sodium fluoride (NaF) and alumina (Al2O3) traps, and corrosion issues in the Reactive Gas Recovery System (RGRS) pump and in the piping immediately past the pump. As a result of these issues and others, the MSRE Engineering Evaluation team was chartered to assess the current facility conditions and to develop near-term and life-extension recommendations to both reduce risks and to maintain the facility and facility operations in a safe state until decommissioning and demolition (D&D). D&D is currently scheduled for 2060. The team was also chartered to evaluate salt removal options as a part of accelerating D&D.

The team evaluated the current condition of ten systems and one program and developed near-term and life-extension recommendations as applicable for each system or program. The current condition evaluation included a review of system descriptions, historical studies, walk downs, subject matter expert (SME) interviews and review of operations and maintenance issues.

As a result of the PISAs, the facility has been in a state of transition from managing the key risks of 233U with fluorine/hydrofluoric acid also present to a state where the dominant risk is from the fluorine/hydrofluoric acid. There are two key areas for risk reduction that the team identified:

Improve the reliability and increase the redundancy of the Central Ventilation System (CVS).

Repairing Fan #2, providing backup power capability for the CVS fans and ensuring an adequate supply of spare parts are all actions requiring increased efforts. Reliance on the CVS has become more critical as a result of the PISAs. The current Justification for Continued Operation (JCO) requirements include evacuation of Bldg. 7503 if CVS shuts down.

Fluorine buildup in the headspace of the Fuel Drain Tanks (FDTs) dominates the documented safety analysis (DSA) accident scenario. Managing the fluorine source term in the headspace of the Drain Tanks to minimize the source term available for accidental release and reduce the corrosive impacts on the Reactive Gas Removal System (RGRS) components will improve the safety margin of the facility and facility operations. Key recommendations include beginning more frequent pumpdowns of the drain tank gas headspace and development of a continuous purge system that does not allow buildup of the fluorine source term in the headspace of the drain tanks. This continuous purge system would ultimately vent to the CVS and would likely not require traps, since continuous operation minimizes the source term available in the tank headspace. This passive system would reduce risk and provide significant cost and waste savings. Additional recommendations for this area include sampling and analysis of the headspace gases in the FDTs, and integrity measurements of the FDTs and associated thimbles.

A number of recommendations were made to reduce the risk associated with the operation of the RGRS.

Increased rigor on procedural controls for the RGRS flow rates and Fourier Transform Infrared Spectroscopy (FTIR) operation is needed. The recommendations also include more increased documentation of system performance during each pumpdown, beginning with the October 2016 pumpdown, to capture critical data and allow trending to determine comprehensive corrective actions.

Follow-up actions after the October pumpdown are needed to determine the cause of the embrittlement of xv

the tubing downstream of the vacuum pumps. Metallurgical examination of a section of the Monel1 400 tubing in the RGRS to determine if embrittlement is occurring is also needed. A number of other recommendations to improve the overall operation and reliability of RGRS are detailed further in the report.

The electrical system components at MSRE in many instances are original to the facility (circa 1960).

Several near-term recommendations were made based on immediate issues identified by the team. Also, because of team concerns and issues identified by facility personnel in the September planned power outage, it is recommended that dedicated, independent additional resources be applied to evaluate the electrical system in more detail and develop specific near-term corrective actions.

Inspections of the MSRE structure were performed; however, only about 1% of key structural components could be evaluated. Based on the result of this limited evaluation, a more intrusive structural inspection, prioritized towards prevention of further water intrusion is recommended.

In an effort to significantly reduce the material at risk (MAR), the charcoal canister containing approximately 4 kg of 233U should be relocated to another facility that has more capability to process this material for final disposition. Previous studies suggested removing the canister may allow downgrading the facility hazard category by analysis. Whether this is possible or not, reduction in Security and S&M and operational controls make this a prudent course of action. Also, sampling of the headspace gases in the charcoal beds and the canister would provide information to better manage the potential long term risks from these components.

The team developed a total of 40 near-term recommendations and 11 life-extension recommendations for the 10 systems and one program. The near-term recommendations were all prioritized based on technical feasibility, overall cost, impact on facility safety, and integration of the recommendations with other systems. The life-extension recommendations were prioritized based on the technical feasibility, safety impact, initial capital cost, and potential for operating and maintenance cost savings over the facility life.

These recommendations, in order of priority, are discussed in Sect. 9.

The team also reviewed the numerous MSRE salt removal studies, and evaluated the following options:

1. Tank removal/sectioning

2. Thermal salt transfer

3. Chemical dissolution

4. Mechanical removal with needle scaler

5. Entombment Tank removal with sectioning of the contents into Waste Isolation Pilot Plant (WIPP)-acceptable packaging led the list of options, based on low technical and safety risk (proven, remote technology), and among the lowest regulatory risk (no heel left in facility).

The majority of the salt removal studies assumed (as this team did also) that the salt waste would be disposed of at WIPP. Technical evaluations performed by Los Alamos National Laboratory (LANL) concluded that the salt waste met the WIPP technical criteria. The team recommended that the transuranic 1 Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise, does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof or its contractors or subcontractors.

xvi

(TRU) waste determination be formally approved in a timely manner so that actions taken toward salt removal will be based on a final waste disposal decision.

The team understands that NHHO plans to assign a dedicated Project Manager to implement the recommendations through their integrated schedule. The team fully endorses this approach.

xvii

This page intentionally left blank.

xviii

1. PURPOSE The Nuclear and High Hazard Operations organization of URS l CH2M Oak Ridge LLC (UCOR) performs surveillance and maintenance (S&M) activities at the Molten Salt Reactor Experiment (MSRE) facility at the Oak Ridge National Laboratory (ORNL). In 2015 and 2016, there have been three Potential Inadequacies in the Safety Analysis (PISAs) at the MSRE. The issues involve potential release of fluorine (F2) and hydrogen fluoride (HF) from the salt tanks due to corrosion of the tank thimbles (commonly called thermal wells in contemporary literature), potential release of these gases from spent sodium fluoride (NaF) and alumina (Al2O3) traps, and corrosion issues in the Reactive Gas Recovery System (RGRS) pump and the piping immediately past the pump. The current plan is to maintain the S&M activities at the MSRE facility through 2060, which will require continued repairs and upgrades to facility equipment and processes to ensure safe operations until final decommissioning and demolition (D&D).

The current remediation strategy includes removal of the salts from Fuel Drain Tanks (FDTs) 1 and 2 and the Fuel Flush Tank (FFT), and ultimate disposal of the salt waste at the Waste Isolation Pilot Plant (WIPP) in Carlsbad, NM.

This evaluation assessed the facilitys current state, and recommended short-term and life-extension actions to maintain MSRE in a safe and compliant state until D&D. The current state evaluation includes a description of the operating systems, a current condition evaluation based on walkdown results, interviews with subject matter experts (SMEs), recent operating and maintenance history, and testing results as applicable. Recommendations for upgrades were developed for each system for the near term and for the life-extension case of operation until 2060 upgrades. The systems/areas to be evaluated included the Drain Tank Cell, FFT, and FDTs; electrical; RGRS; ventilation; cranes; air compressors; fire protection; structural; and waste management. Results from these evaluations were used to identify and prioritize recommendations for near-term and life-extension upgrades, with the focus on maintaining safe and compliant operations within the MSRE documented safety analysis (DSA) and technical safety requirements (TSRs). Alternatives for the current remediation strategy for salt removal were also evaluated.

1-1

2. BACKGROUND This section summarizes information about MSRE with a focus on the past key processing activities in the facility relevant to this evaluation. Details on individual systems or equipment are included in the subsequent system discussions in Sect. 4 through Sect. 6.

The Molten Salt Reactor Experiment (MSRE) is located in Building 7503 at ORNL, a U.S. Department of Energy (DOE) facility in Oak Ridge, Tennessee. The MSRE was developed as part of the ORNL-led Molten Salt Fuel Feasibility studies, the first of which was the Aircraft Reactor Experiment (operational 1954-1956). Construction of the MSRE began in 1962 and ended in 1964. The MSRE was a graphite-moderated, liquid-fueled reactor operated from June 1965 through December 1969 as a demonstration of the technology needed to develop a large-scale molten salt breeder reactor.

The liquid fuel for the reactor was a fuel salt composed of uranium tetrafluoride (UF4) dissolved in a carrier salt of lithium fluoride (LiF), beryllium fluoride (BeF2), and zirconium fluoride (ZrF4). It was initially fueled with 235U, but in 1968, the fuel mix was changed to 233U. Plutonium from the 235U operations was already in the fuel salt when, in September 1969, about 200 g of plutonium trifluoride (PuF3) was added. The fuel salt was circulated by a fuel salt pump, at temperatures above 600°C, through the reactor vessel and a primary heat exchanger, and drained into two fuel salt drain tanks (FDTs) located within the stainless steel-lined, concrete-shielded Drain Tank Cell adjacent to the Reactor Cell (Figure 2.1). A flush salt similar in composition to the fuel salt (without the uranium fuel) was used twice to flush the reactor system to remove residual pockets of fuel salt. The flush salt, which became contaminated with a small amount of residual fuel salt after each flush, was drained to the FFT, also located in the Drain Tank Cell.

The Charcoal Bed Cell (CBC) was designed to house the filtration systems for the reactor off gases during reactor operation. The CBC contains the Auxiliary Charcoal Bed (ACB) and the Main Charcoal Bed (MCB). When the reactor was operational, the ACB filtered the offgas from the FDTs, and the MCB filtered the offgas from the reactor fuel loop and sampling systems.

From 1971 until 1989, following reactor shutdown, a salt annealing process was part of the shutdown procedures (DOE/OR/02-1671&D2). Fluorine generation from radiolysis of the salt was expected based on knowledge of chemical stability of the salt. The annealing process heated the fuel salt to below melting temperatures to force fluorine in the salt matrix to recombine before it would migrate from the salt.

However, in 1994, significant migration of uranium from the stored fuel and flush salts was discovered.

High concentrations of fluorine (F2) and uranium hexafluoride (UF6) gases were found to be present in the offgas system piping, and a significant deposit of uranium was found in the auxiliary charcoal bed as a result of leak through a valve. After significant effort and facility modifications, approximately 80% of the uranium was removed from the auxiliary charcoal bed and placed in a canister for safe storage.

As one of the modifications from this event, the Reactive Gas Removal System (RGRS) was designed and operated, beginning in 1996, to remove F2 and UF6 deposits from the Off-Gas System piping network and the ACB. RGRS was subsequently modified to support defueling operations, which were performed from 2004-2008. The fuel salt underwent hydrofluorination and fluorination as part of the defueling process. The defueling process removed uranium from the salt by heating and melting the salt (~450ºC),

fluorinating it, then circulating the gas through cold traps to collect the uranium. An alumina trap was used for fluorine removal. Defueling was determined to be completed in 2008. RGRS has been used since defueling was completed to treat and remove the fluorine from the drain tanks 1-2 times per year before the drain tank pressures reach atmospheric.

2-1

Figure 2.1. MSRE facility showing the reactor and drain tanks.

Defueling of the fuel salt in the FDTs and flush salt in the FFT was determined to be complete in 2008, based on reaching the goal of 50 ppm of uranium. The 50 ppm goal was determined via extrapolation of measured UF6 gas removal rates leaving the drain tanks (i.e., not direct measurement of remaining uranium in the fuel and flush salts). In 2011, UCOR performed nondestructive assay (NDA) measurements of the fuel and flush salts to address uranium concentration uncertainties associated with the 2008 measurements and historical uranium inventory records (DOE/OR/01-2513&D1). The gamma spectroscopy NDA measurements indicated that 280 ppm of uranium remained in FDT-1. Although a greater concentration than the 2008 indirect measurements, 280 ppm of uranium continues to support the conclusion that the fuel salt has been defueled. In this report, the defueled fuel salt will be referred to as fuel salt, and the salt in the FFT will be referred to as flush salt.

In 1999, nondestructive testing, including ultrasonic thickness measurements and eddy current testing, was performed to assess the integrity of the FDTs and the FFT. The tanks were fabricated from INOR-8 (now commercially known as Hastelloy N), as part of the MSRE project. A summary of relevant background information on corrosion is contained in Appendix A. The testing concluded that any corrosion caused by previous operations was not significant enough to affect tank integrity for a planned pressurized salt transfer at that time. However, the planned salt transfer did not occur.

In July 2008, DOE issued an Engineering Evaluation Work Plan (DOE/OR/01-2386&D1) that described a thermal method for removing salts. In September 2009, an Engineering Evaluation report (BJC-OR-3301) recommended not proceeding with salt removal until the structural integrity (for both corrosion and stress-cracking concerns) of the FDTs and FFT are confirmed, the continued generation of fluorine inside the tanks considered, and the possible existence of multiple phases (stratification) of material in the salt be determined. In 2010, an engineering evaluation of options for removal of the salt from the FDTs and the FFT was performed. This evaluation examined multiple options for salt removal, 2-2

including thermal, entombment and two mechanical options. However, the maintain as-is option for 50 years was recommended as the overall preferred near-term option at that time.

A Potentially Inadequate Safety Analysis (PISA) for the MSRE facility was declared on August 4, 2015.

New information was identified that brought into question an increased consequence from a F2/HF release from the MSRE FDTs due to the potential for a thimble breach. A release of F2/HF from the RGRS was analyzed in the MSRE DSA, but a release of F2/HF from the FDTs of the magnitude now postulated was not analyzed. A second PISA was then declared for lack of analysis relative to release of F2/HF from NaF and alumina traps and spent media. One of the corrective actions for the August 2015 PISA was to pump down the headspaces of the FDTs. While conducting this activity on April 21, 2016, the vacuum pump discharge line plugged, resulting in a work stoppage. During replacement of the vacuum pump in the RGRS, green, white, and gold colored deposits were observed at the pumps discharge connection fitting and discharge line. While disconnecting the discharge line from the pump, the line sheared due to corrosion. The evaluation of this condition determined that corrosion issues were present at the pump and in the piping downstream of the pump. A third PISA was declared for this condition.

A revised DSA incorporating lessons learned from these PISAs was submitted to DOE for approval in September 2016. The October 2016 pumpdown of the FDTs/FFT gas headspace was performed under an approved Justification for Continued Operations (JCO).

2.1 REFERENCES

BJC-OR-3301. Engineering Evaluation of the Proposed Alternative Salt Transfer Method for the Molten Salt Reactor Experiment Oak Ridge National Laboratory Oak Ridge, Tennessee, October 2009, Bechtel Jacobs Company LLC, Oak Ridge, TN.

DOE/OR/02-1671&D2. Record of Decision for Interim Action to Remove Fuel and Flush Salts from the Molten Salt Reactor Experiment Facility at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1998, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR/01-2386&D1. Engineering Evaluation Work Plan for Molten Salt Reactor Experiment Residual Salt Removal, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2008, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR01-2496&D1. Engineering Evaluation of Options for Molten Salt Reactor Experiment Defueled Coolant Salts, Oak Ridge, Tennessee, 2010, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR/01-2513&D1. Nondestructive Assay Measurements of Defueled Salts at the Molten Salt Reactor Experiment, Oak Ridge, Tennessee, 2011, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR/01-2560&D2. Remediation Strategy for the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2013, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

JCO-OR-MSRE-0117/R6. Justification for Continued Operation for the Molten Salt Reactor Experiment Facility Under the Documented Safety Analysis, 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

2-3

UCOR-4424. Potential Cost Savings During Salt Disposition from the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory Oak Ridge, Tennessee, 2013, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

2-4

3. SCOPE The scope of this current-condition and life-extension study includes the MSRE facility systems or programs consistent with the current remediation strategy and plan for D&D in 2060. The requirements or impacts of each system relative to the safety envelope were evaluated based on review of the current DSA for MSRE and the draft DSA under development during this evaluation. The MSRE facility systems that are included in this evaluation include the following:

Salt Drain Tanks, Flush Tank, and Drain Tank Cell Reactive Gas Removal System (RGRS)

Electrical system Central Ventilation System (CVS)

Fire protection system Structural Charcoal Beds and Charcoal Canister Cranes Compressed air system Steam The Waste Management program is also included in this evaluation.

The options for salt removal that were evaluated include the following:

Tank removal/sectioning Melting Chemical dissolution Mechanical removal by needle scaler Entombment 3.1 SYSTEM DESCRIPTION The system description includes a summary level description of each system identified above, including major components (e.g. tanks, pumps, piping, traps), system function, and importance to maintaining operations within the DSA and TSRs.

3.2 CURRENT CONDITION EVALUATION The current condition evaluation is based on system walkdown; interviews with SMEs for the system; review of recent operating and maintenance issues; and, for the RGRS, nondestructive examination (NDE) testing of portions of the tubing upstream and downstream of the traps, evaluation of tubing failures, and characterization of solids removed from the vacuum pump and tubing.

3-1

3.3 NEAR-TERM UPGRADE RECOMMENDATIONS The near-term upgrade recommendations are based on the evaluation of the current condition of each system, and reflect the repairs or modifications needed either immediately or within the next 1-5 years to ensure continued operations within the safety envelope. Recommendations are presented in order of priority in Sect. 9 as determined by the evaluation team judgement of technical feasibility, relative cost, safety impact and dependence on integration with other systems.

3.4 LIFE-EXTENSION RECOMMENDATIONS The life-extension recommendations are based on the system evaluation and reflect the repairs or modifications needed in the future, in conjunction with the near-term recommendations, to maintain continued operations within the safety envelope for the next approximately 40 years.

3.5 WASTE MANAGEMENT PROGRAM The Waste Management Program evaluation included the identified inventory of various waste volumes and the current waste removal schedule negotiated with the various regulatory agencies. The remediation strategy plan for disposition of the salt waste will be addressed. Recommendations for improvement of the Waste Management Program and to accelerate final disposition of the salt waste will be identified.

3.6 SALT REMOVAL OPTIONS EVALUATION Each option for salt removal identified in the introduction to this section was evaluated. The evaluation sought to advance future options for salt removal versus previous studies that concluded with a leave as-is option, since that option does not support D&D of MSRE. Salt removal options were prioritized by the evaluation team and UCOR D&D SMEs on the basis of technical risk, cost, safety and regulatory constraints.

3.7 REFERENCES

DOE/OR/01-2560&D2. Remediation Strategy Plan for the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, January 2013, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

DOE/OR/01-2496&D1. Engineering Evaluation of Options for the Molten Salt Reactor Experiment Defueled Coolant Salts, Oak Ridge, Tennessee, December 2010, Bechtel Jacobs Company LLC, Oak Ridge, TN.

DSA-OR-7503-0007/R25. Documented Safety Analysis for the Molten Salt Reactor Experiment Facility, Oak Ridge National Laboratory, Oak Ridge, Tennessee, July 2014, URS CH2M Oak Ridge LLC, Oak Ridge, TN.

3-2

4. DRAIN TANKS, FLUSH TANK, AND DRAIN TANK CELL 4.1 SYSTEM DESCRIPTION The two MSRE fuel drain tanks, FDT-1 and FDT-2, are used to store the fuel salt drained from the reactor. Either of the two drain tanks can store the entire salt content of the primary circulating system.

The FFT is used to store the salt that was used to clear the primary system of oxides and other contaminants. The geometry of the FDTs is such that the concentration of fuel in the fuel salt could not produce a critical mass under any conditions. The FDTs are 50 in. in diameter and ~86 in. high, and hold

~80.2 ft3 of salt. The tare weight of each drain tank was 7,000 lb when originally installed. The maximum weight is about 17,000 lb. See Figure 4.1. When the reactor was shut down and drained in 1969, the fuel salt was divided in approximately equal amounts between the two FDTs. The fuel salt consisted primarily of a eutectic mixture of lithium, beryllium, zirconium, and uranium fluorides.

During reactor operation, offgas was generated by the following:

1. The continuous discharge of helium containing highly radioactive fission product gases swept from the fuel salt circulating pump bowl

2. Intermittent, relatively large flows of helium containing, at times, significant amounts of radioactive gases and particulates, such as that discharged during salt transfer operations

3. Flows of up to 100 cubic feet per minute (CFM) of very low activity cell atmosphere gas (5% O2, 95% N2), which was ejected either intermittently or continuously to maintain the Reactor and Drain Tank Cells at sub-atmospheric pressure.

The offgas System disposed of these gases.

The offgas vented from the FDTs and FFT flowed through lines 573, 575 and 577, which then form a common header and enter line 561. Line 561 was connected to the charcoal beds to scrub the gases prior to release via the CVS. Each of these three lines contains a bellows-sealed, air-operated, air-to-close-type control valve, likely constructed of 347 or 304 stainless steel located in the Drain Tank Cell.

Molten salt was transferred between the reactor and drain tanks by pressurizing the system with helium gas, which was fed to the FDTs and FFT via pressurization lines 572, 574, and 576. A common 0.5-in.

nozzle on the top of the FDTs and FFT served pressurization and offgas venting purposes. The control valves in each of the three lines, 572, 574, and 576, are followed by two check valves and a hand valve in series. (The check valves and the hand valves are contained in a pot in the north electric service area, and the 0.25-in. line from each hand valve is run inside a 0.5-in. pipe to the Drain Tank Cell wall penetration).

Pressure transmitters for measuring the drain tank pressures, PT-572, PT-574, and PT-576, are located in each of the three lines downstream of the check valves. The control valves are air operated, 0.5-in., fail-closed valves. The RGRS utilizes the former pressurization lines 572 and 574 to vent gas from FDT-1 and FDT-2, respectively, tapping off of the lines such that the PTs can be utilized to monitor tank pressure. The nozzle and short length of connected piping common to the former pressurization and offgas vent line and the disconnect flange are made of Hastelloy N. Beyond this point the lines are believed to be fabricated of stainless steel.

4-1

Figure 4.1. Fuel Drain Tank.

Each tank is equipped with 32 cooling thimbles (commonly called thermal wells in contemporary literature) that extend into the fuel salt. The cooling thimbles were designed to provide for decay heat removal when the hot fuel salt was drained to the tanks during reactor operation. Bayonet tubes inside the thimbles circulated water, which formed steam that then condensed in a steam dome atop the tanks. The steam domes and bayonet tubes have been removed such that the thimbles are now empty and open-ended at the top of the tanks. The thimbles are 1.5-in., schedule 40 pipe and are constructed of Hastelloy N. See Figure 4.2 and Figure 4.3. The tanks also have a transfer line (1-in. schedule 40 pipe) and reactor drain line (0.5-in., schedule 40 pipe) that run from the bottom to the top of the inside of the tanks, both of which were permanently closed off prior to defueling.

4-2

Figure 4.2. FDT thimble assembly (prior to installation).

The FDTs and the FFT are located in the Drain Tank Cell, which is just north of the Reactor Cell and connected to it by a short 36-in.-diameter tunnel. The Drain Tank Cell is 17 ft 7 in. x 21 ft 2.5 in., with the corners beveled at 45-degree angles for 2.5 ft. The flat floor is at the 814-ft elevation (below the water table), and the stainless steel membrane between the two layers of top blocks is at the 838.5-ft elevation.

The open pit extends to the 852-ft elevation. The cell was designed for 40 psig and, when completed in 1962, it was hydrostatically tested at 48 psig (measured at the elevation of the membrane at 838.5 ft).

Currently, the Drain Tank Cell is inspected on an as-required basis (e.g., after an earthquake, process fluid excursion, or dropped load), and is not pressure rated. See Figure 4.4 and Figure 4.5.

Numerous lines penetrate the walls of the Drain Tank Cell to provide for process and service piping, electrical and instrument leads, and for other accesses. These enter through 0.75-in. to 6-in. pipe or pipe sleeve penetrations that are welded to the stainless steel liner and cast into the concrete walls. Lines are installed in these individually or grouped in plugs that are filled with concrete. See Figure 4.6.

4-3

Figure 4.3. Steam dome-bayonet tube assembly installation into FDT.

4-4

Figure 4.4. Drain Tank Cell.

4-5

Figure 4.5. Drain Tank Cell (during construction).

The Drain Tank Cell floor elevation is 814 ft at the highest point along the west wall and slopes 0.125 in./ft to a trench along the east wall. The trench slopes 0.125 in./ft toward the south and terminates in a sump located in the southeast corner. The sump consists of a 10-in.-long section of 4-in., schedule 40 pipe, and a 4-in. butt-welded cap (347 stainless steel). Jet air ejectors are designed to remove water from the Drain Tank Cell sump. The ejectors are actuated by station service air. The 100-psig supply air is passed through a pressure regulator, PCV-332; and distributed to the Drain Tank Cell through line 342, and to the Reactor Cell through line 332. Both of these lines are 0.75-in., schedule 40 stainless steel and have hand-operated valves and check valves to prevent back flow from the cells. The Reactor Cell jet discharges into line 333 and the Drain Tank Cell jet into line 343, both of which are 0.75-in., schedule 40 stainless steel pipe leading to the liquid waste storage tank. Before entering the storage tank, each line is provided with two flow-control valves in series, FCV-333-A/FCV-333-B and FCV-343-A/FCV-343-B.

These normally-closed, air-operated valves (AOVs) are interlocked to close on a sudden rise in Reactor Cell pressure. Lines 334 and 344 permit the application of a test pressure between the two flow-control valves to check for leak tightness. The Reactor and Drain Tank Cells were designed to be depressurized for leak testing through the jet discharge lines.

4-6

Figure 4.6. Penetrations in Drain Tank Cell.

The two FDTs and FFT are heated within the Drain Tank Cell by cylindrical "furnaces" surrounding the tank walls. The tank tops, bottoms, and heater units are enclosed in canned insulation units. The tank wall heaters are arranged in removable vertical sections, or panels, each about 2.375 in. thick. The heaters are arranged into two groups, an upper and a lower, each group being supplied with electric power from a separate induction regulator. Each of the large panels contains four heating elements consisting of curved ceramic plates in which nichrome heating wire is embedded. The elements are mounted in stainless steel frames and enclosed in 16-gauge, 347 stainless steel to complete the panel assembly. See Figure 4.7 and Figure 4.8.

4-7

Figure 4.7. FDT heating element.

4-8

Figure 4.8. FDT heating element, nichrome wires.

4-9

4.2 CURRENT CONDITION 4.2.1 Tank Integrity and Salt Status FDT-1, FDT-2, and the FFT are approximately 44%, 40%, and 85% filled with salt, respectively.

Defueling of the salt in the FDTs and FFT occurred between 2004 and 2008. The defueling goal of 50 ppm of uranium was declared to have been achieved in 2008 based on extrapolation of FTIR data that measured uranium removal rate in gas leaving the salt. In 2011, NDA measurements were taken on FDT-1, resulting in an estimated uranium inventory of 280 ppm.

In 1999, prior to uranium removal and anticipated salt removal, the FDTs and FFT were examined by means of nondestructive testing, including visual examination, ultrasonic thickness measurements, eddy-current-wall-loss measurements, thermal analysis and finite element stress analysis (BJC/OR-445, Final Report of the Molten Salt Reactor Experiment Drain Tank Qualification, Oak Ridge National Laboratory, Oak Ridge, Tennessee). The conclusion of the testing was that the mechanical integrity of the tanks was adequate to withstand pressures up to 80 psia at 1100°F. Since then, all three tanks have been exposed to both hydrogen fluoride (HF) and fluorine (F2) gas during the uranium removal steps while the salt was in a molten state (~850°F).

In order to remove the remaining fuel salt from the FDTs and FFT, the mechanical and physical condition of the vessel shell as well as all the associated piping, particularly the 32 thimbles (and a 33rd pipe used for temperature measurements), must be determined. The vessel and all internal components, including the thimbles, are reported to have been constructed entirely of Hastelloy N. Once outside the vessel, at the first available transition point, the piping not used to handle molten salt is reported to have switched to stainless steel (ORNL/TM-728).

The thimbles and the thermowell terminate a short distance above the dome of the tank, while several other pipes and tubes used for flow of salt and gases make connections outside the tank. The presence of a significant amount of molybdenum in deposits found in RGRS traps, filters and plugs has led some authors to make the assumption that this molybdenum is a product of corrosion of the Hastelloy N vessel and associated piping. If the molybdenum is totally a product of Hastelloy N corrosion, then it is not unreasonable to propose that the 1.5-in. schedule 40 thimbles (nominally 0.145-in. wall thickness) have undergone considerable corrosion and may present a potential failure point for release of headspace gas into the Drain Tank Cell. In addition, this could be a major barrier to some modes of salt removal being considered. If serious corrosion of the Hastelloy N has occurred, it is a reasonable possibility that corrosion would have been more severe at the salt-vapor interface. It is always a possibility that corrosion could have been even more localized and is more severe at welds or sites of local flaws. However, a specific molybdenum isotope, 100Mo, is a fission product of 233U, while 99Mo is a fission product of 235U.

The molybdenum observed in the RGRS downstream sites could be decay and/or fission products and not due to corrosion. In evaluating the integrity of the FDTs and FFT, the source of the molybdenum is a significant issue; results on isotopic distribution reported in a 2014 publication (UCOR-4610/R0) indicated 100Mo was the only molybdenum isotope identified while an even earlier analysis of samples collected after operation with 235U identified 99Mo as the molybdenum isotope present (ORNL/TM-1853 and ORNL/TM-3884). However, significantly different results on the isotopic distribution will be presented in the report on analysis of deposits from the pump outlet that is included in the discussion on corrosion in the RGRS section of this report.

Wall thickness measurements will help address the issue as to whether the molybdenum deposits are a result of extensive corrosion of the Hastelloy N, but the measurements would not necessarily determine whether any pitting had occurred or if the integrity of the welds might have been compromised. To provide the most information, thickness measurements of the shell and thimbles need to be made in close 4-10

proximity, preferably in a continuous manner, especially in areas where the molten salt surface contacted the metal. In addition, areas on either side of welds, both vertical and circumferential, need to be given close attention because of the susceptibility of the heat affected zone to corrosion. Overall, thickness measurements would help provide an assessment of the extent of general corrosion and the integrity of the Hastelloy N components in the FDTs and FFT.

In order to identify discontinuities in the walls of the drain tanks and the thimbles, eddy current testing is required. If pitting is present or if localized corrosion has occurred at the salt-gas interface, the eddy current technique offers the most promising method of detecting this degradation. As discussed above, in 1999, ultrasonic and eddy current measurements were used to document the condition of the drain tank walls and the thimbles. Any new measurements need to examine no less than the same sites previously examined using techniques at least as good as those previously used. Since the thimbles have a thinner wall (0.140 in.) than the drain tank, it is more urgent to determine if the previously identified pits have increased in depth and if any new pits have formed. As will be proposed, a more thorough examination would be very much preferred in light of the uncertainty on the integrity of the drain tank vessel walls as well as the thimbles and thermowells that have been exposed to molten salt as well as the fluorine and hydrogen fluoride that were introduced into the tanks during defueling. There would also be some benefit to eddy current inspection of selected areas of the tank wall. Areas of the tank wall contacted by the surface of the molten salt have an increased likelihood for pitting or localized corrosion. Eddy current examination offers the best means for finding such degradation.

Another tank integrity concern is the welds that attach the thimbles to the tank dome. Reports reveal that at least two connection methods were considered (UCOR-4807). One method showed the thimble pipes were continuous lengths of pipe penetrating the dome. A second method showed separate sections of pipe welded on either side of the dome. Photographs of the thimbles do not make it clear which of these methods was used to attach the thimbles to the dome. Use of expandable plugs is one of the approaches being considered to seal the thimbles if thinning and/or pitting is shown to have occurred in these pipes.

However, this may only be an effective means to seal the thimbles if it can be established that the inside surface of each thimble is not irregular because of weld penetration if the second attachment method were used. Consequently, in addition to ultrasonic thickness and eddy current measurements, borescope examination of the inner surface of the thimbles would be a valuable component of thimble inspection.

Additional concerns associated with the version of INOR-8/Hastelloy N that was used to fabricate the components of the MSRE is radiation embrittlement and grain boundary penetration by tellurium. Each of these phenomena has the potential to cause significant degradation to the drain tanks and associated components, but the limited time at temperature as well as time exposed to the degrading conditions/environments has to be assessed.

First, tellurium is produced by the fissioning of uranium. This would be a reaction occurring primarily in the reactor; therefore, it is thought to have occurred to an extremely limited level in the drain tanks.

According to separate documents (ORNL/TM-3884, ORNL/TM-1853, ORNL/TM-2578) tellurium did not form stable fluorides in the MSRE but was present in the metallic state. Consequently, it was speculated that the tellurium adhered to the surface of the reactor vessel and the graphite moderator and very little moved with the salt into the FDTs and FFT. Furthermore, any tellurium that made it into the drain tanks was very likely removed as TeF6 during the fluorination process. As a result, it is speculated by some authors (BJC/OR-3301) that tellurium embrittlement of the drain tanks and the internal components would be very limited if it occurred at all. However, since tellurium was produced in the salt, there would seem to be a possibility that tellurium could have been transported into the drain tanks during the many transfers of salt between the reactor vessel and the drain tanks. If that were the case, then tellurium embrittlement cannot be ruled out as a possible contributor to degradation of the Hastelloy N that comprised the drain tanks, thimbles and thermowell.

4-11

The most recognized modes of radiation damage to Hastelloy N are attributed to thermal neutrons which were only found in the reactor vessel where the graphite moderator was present. However, researchers Cunningham (ASTM STP 457) and Angeliu (MDO-723-0043) have reported there are other mechanisms through which fast neutrons can produce radiation damage in nickel-base alloys like Hastelloy N.

Consequently, the possibility of radiation damage in Hastelloy N cannot be discounted, but there is not a practical method to confirm whether or not such damage has occurred without obtaining a sample for destructive analysis.

There are a significant number of reports which provide information on the extent of corrosion of Hastelloy N in molten salts and which speculate, based on corrosion results of somewhat similar nickel-based alloys, on the corrosion of Hastelloy N that might have occurred in the drain tanks during treatment of the molten salt with fluorine and hydrogen fluoride. These estimates of corrosion rates are based on studies made in other salt environments and in conditions that were not necessarily representative of the MSRE drain tank environments. It has been shown that the purity of the salt and the rate that contaminants can reach the salt can significantly affect measured corrosion rates. Nevertheless, corrosion rates reported by these other researchers were so high that the rates were reported in terms of mils/h rather than the usual units of material loss over a year. Starting with the thimble wall thickness measurements reported from the 1999 examination and assuming the estimated corrosion rates are somewhat realistic, then it is likely that significant thinning has occurred in both the thimbles and the drain tank walls. Based on the projected remaining wall thickness in pitted and uniformly thinned regions that were summarized previously, the drain tanks were projected still to be able to sustain sufficient pressurization to move the molten salt out of the tanks, as discussed above. However, that assumes tellurium and radiation embrittlement have not compromised a portion of the remaining wall thickness, and the melting point of the salt is not so far above the eutectic temperature that the drain tanks and the salt have to be heated to a considerably higher temperature in order to move the salt out of the tanks.

4.2.2 Fluorine Gas Generation and Management in Drain Tanks Fluorine gas is generated in the head space of the FDTs and FFT due to radiolysis of the solid fuel salt.

Subsequent to reactor shutdown in 1969, it was thought that fluorine gas recombination into the fuel salt could be accomplished through annual heating to > 200ºC, using the original tank heaters. Although fluorine gas pressure buildup in the headspace of the tanks was somewhat managed by periodic heating, fluorine gas continued to be generated, resulting in creation of fluorine compounds with constituents of the salt, most principally UF6. In 1994, significant migration of uranium from the stored FDTs and FFT to the Auxiliary Charcoal Bed (ACB) was discovered. The charcoal was denatured, and most was removed from the ACB in 2001 and placed in a shielded canister, which is currently located on top of the Reactor Cell in the High Bay. This uranium migration event resulted in the design and installation of the Temporary Vent and Trap System to remove F2 and UF6 from the FDTs and FFT. This system was later renamed the RGRS. Subsequently, the fuel salt underwent hydro fluorination and fluorination as part of the defueling effort. In the current MSRE DSA, DSA-OR-7503-0007/R25, the prohibition of drain tank heater operation is a Specific Administrative Control to prevent fuel melting and to minimize potential formation and release of UF6.

The FDTs and FFT are described in the Rev. 25 of the DSA (July 2014) as confinement boundaries. The hazards related to headspace gas are mitigated by controls in the Hazardous Material Protection Program (e.g., HF monitoring, CVS, FDT/FFT pressure limits, heater use prohibited, crane operation restrictions).

The FDTs and FFT are not described as Design Features or Safety Significant in the DSA.

The fluorine gas generation rate in the FDTs and FFT was likely much greater when fuel remained in the fuel salt. Currently, the gas generation rates are calculated to be 2.87E-4 psia/hour and 2.17E-4 psia/hour for FDT-1 and FDT-2, respectively. The current JCO limit for gas pressure in the FDTs and FFT is 4-12

14.0 psia to reduce the chances of a confinement breach (note: the tanks are backfilled with argon gas between pumpdowns to 12.5-13.0 psia). Between pumpdowns, which occur approximately twice per year, the headspace gas in the FDTs and FFT is confined within the following boundaries:

The headspace volume of each tank Lines 573, 575, and 577 up to the first Off-Gas System AOVs located in the Drain Tank Cell:

HCV-573, -575, and -575 for FDT-1, FDT-2, and FFT, respectively (air-to-close)

HCV-544, -545, and -546 for FDT-1, FDT-2, and FFT, respectively Lines 572, 574, and 576 up to RGRS valves PTF1B (actuator ITS1B) and PTF2B (actuator ITS3B):

Valves located in the North Electric Service Area (NESA)

Sometimes referred to as key-operated valves, and/or boundary valves See Figure 4.9 for a simplified drawing of the gas process lines and valves that carry headspace gas.

Based on interviews with facility personnel and a review of available drawings, it is understood that a complete loss of air would result in HCV-544, -545, -546, -573, -575, and -577 failing to the OPEN position, which would extend the headspace gas confinement boundary to valve V-561 and process lines not permanently isolated from the Off-Gas System, including lines in the High Bay. Air supply to these valves is backed up by nitrogen bottles stored in the Transmitter Room. The backup nitrogen bottles supply air to a common header, and are expected to last approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, if called upon. In addition, previous reports have noted that valves HCV-545, -546, -575, and -577 are known to leak.

Leakage past these valves has not prevented RGRS operations to date, and pressure indicator readings do not indicate measurable leak-by; therefore, no efforts have been made to repair or replace these original valves.

The -544 through -546 and -573 through -577 valves are primarily utilized to ensure the three (3) tanks are isolated from each other, and to provide cross-connection between the FFT and one of the FDTs for pumpdown and backfill of the FFT. The FFT must be cross-connected with an FDT for pumpdown and backfill because there is no dedicated RGRS process line on the 576 line, such as the 572 line for FDT-1 and 574 line for FDT-2. With regard to tank isolation, should these air-to-close valves deteriorate such that they cannot reliably provide the tank isolation and gas confinement functions, a design change could be implemented to do the following:

Permanently close valves HCV-544, -545, -546, -573, -575, and/or -577, if possible, or employ another means to isolate the lines.

Install a dedicated RGRS line, identical to the RGRS lines for FDT-1 and FDT-2, from line 576 to the common RGRS header in the NESA.

4-13

4-14 Figure 4.9. Simplified gas-flow diagram for FDTs, FFTs, and related process lines.

The new RGRS line tapped off line 576 would be used for headspace gas pumpdown and backfill just as lines 572 and 574 are used for FDT-1 and FDT-2, respectively. This would also provide more substantial tank isolation than the current configuration, and the effective headspace gas confinement boundary valve(s) would be more easily accessible in the NESA as opposed to the Drain Tank Cell. Cross-connecting tanks would also still be available. The 2016 MSRE Engineering Evaluation Team assessed this option and deemed it unnecessary at this time, given the controls currently in place for the loss-of-confinement accidents analyzed in the DSA, however, it should be considered should the aforementioned original air-to-close valves cease to effectively provide their functions.

The consequences of a breach in an FDT/FFT or process gas line were analyzed in the July 2014 DSA.

The loss-of-confinement accident is controlled in the DSA and TSR by the following:

Drain Tank Cell and Portable Maintenance Shield (credited Safety Significant SSC)

Prohibition of drain tank heating, in addition to periodic FDT headspace pumpdowns and restriction of crane loads when the Portable Maintenance Shield is open (credited Specific Administrative Controls)

Hazardous Material Protection Program: HF detectors for intrusive work, area evacuation if CVS is not operating (credited Programmatic Administrative Control)

Pressure transducer (PT)-576, which is the primary means of monitoring pressure in the FFT, has been out of service since ~1997. In addition, facility personnel suspect a leak may exist near a valve in the enclosure containing PT-576. UCOR staff is planning to credit PT-576 in the next revision to the DSA and TSR in a Specific Administrative Control for monitoring FFT pressure. During the 2016 MSRE Engineering Evaluation, the repair and calibration of PT-576 was in planning.

4.2.3 Drain Tank Cell and Tank Heaters The Drain Tank Cell and associated maintenance shield are described in the July 2014 DSA as Safety Significant Design Features. The Drain Tank Cell sump level indicator is non-functional, and the functionality of the sump air ejector is not clear. Records of the last date the Drain Tank Cell sump level indicator and air ejector functioned could not be located. Facility personnel believe the Drain Tank Cell was visually inspected via remote camera in the 2008-2011 timeframe, with no signs of standing water.

Based on document reviews and interviews with facility personnel, the status of the electrical leads to the drain tank heaters within the Drain Tank Cell could not be verified. As a part of evaluating the feasibility of the thermal fuel salt disposition option, the condition of the drain tank heaters needs to be assessed.

4.3 NEAR-TERM RECOMMENDATIONS

1. Obtain gas samples directly from the tank headspace (FDT-1 and FDT-2).

This will allow evaluation of chemicals that cannot be detected using the FTIR alone.

2. Perform ultrasonic testing (UT), eddy current testing, and borescope examination on all 32 thimbles and the thermowells for FDT-1 and FDT-2.

UT thimbles and FDT tank for wall thickness and evidence of pitting.

Eddy current testing of full circumference of thimbles for entire length (if possible, ECT FDTs).

4-15

Borescope thimbles, focusing on tank head-thimble interface at welds.

Visual inspection via remote camera of salt, internal tank welds, tank supports, and external tank welds (if possible).

3. Repair PT-576 for pressure indication of the FFT.

4. Address power loss issue for original design air-to-close valves associated FDTs and FFT offgas.

5. Check functionality of the FDT and FFT heaters (note: does not include energizing heaters).

4.4 LIFE-EXTENSION RECOMMENDATIONS

1. Restore capability to monitor for water in Drain Tank Cell.

2. Permanently isolate FDTs and FFT from those process lines not related to RGRS.

Failure of air-to-close AOVs HCV-573, -575, -577, -544, -454, and -456 extend the headspace gas confinement boundary to valve V-561 and process lines not permanently isolated from the Off-Gas system, and RGRS operations could be adversely affected Station air load will decrease, and reliance on an active component to maintain confinement boundary would be eliminated Length of process gas lines would be decreased, reducing probability of leakage and overall risk of releasing headspace gas

4.5 REFERENCES

ASTM STP 457. Development of Fuel Cladding for Fast Reactors, 1969, American Society for Testing and Materials, Philadelphia, PA.

BJC/OR-445. Final Report of the Molten Salt Reactor Experiment Drain Tank Qualification, Oak Ridge National Laboratory, Oak Ridge, Tennessee, February 2000, Bechtel Jacobs, LLC, Oak Ridge, TN.

BJC/OR-2794. Engineering Evaluation of the Leakage Across IXH2 and IXH5 and Extent of Condition Review at the Molten Salt Reactor Experiment Facility at Oak Ridge National Laboratory, Oak Ridge, Tennessee, April 2007, Bechtel Jacobs Company, LLC, Oak Ridge, TN.

BJC/OR-3301. Engineering Evaluation of the Proposed Alternative Salt Transfer Method for the Molten Salt Reactor Experiment, Oak Ridge National Laboratory, Oak Ridge, Tennessee, September 2009, Bechtel Jacobs Company, LLC, Oak Ridge, TN.

CAJ-027503-D699. Pressure Forecast for Fuel Drank Tanks at MSRE from December 3, 2015 Through Projected Atmospheric Pressure, December 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

CONF-96061216-7. Characterization of the Molten Salt Reactor Experiment Fuel and Flush Salts, Prepared for Presentation at the 1996 American Nuclear Society Meeting in Reno, Nevada, June 1996, Lockheed Martin Energy Systems, Inc., Oak Ridge, TN.

4-16

DOE/OR/01-2513&D1. Nondestructive Assay Measurements of Defueled Salts at the Molten Salt Reactor Experiment, Oak Ridge, Tennessee, August 2011, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DSA-OR-7503-0007/R25. Documented Safety Analysis for the Molten Salt Reactor Experiment Facility, Oak Ridge National Laboratory, Oak Ridge, Tennessee, July 2014, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

MDO-723-0043. Assessing the Effects of Radiation Damage on Ni-Based Alloys for the Prometheus Space Reactor System, 2006, Knolls Atomic Power Laboratory, Schenectady, NY.

NCSE-OR-MSRE-1507, Rev. 3, Nuclear Criticality Safety Evaluation, MSRE Fuel Drain Tanks, Fuel Flush Tank and Offgas System, December 2006, Bechtel Jacobs Company, LLC, Oak Ridge, TN ORNL/CP-98146. Removal of Uranium and Salt from the Molten Salt Reactor Experiment, 1998, Bechtel Jacobs Company, LLC, Oak Ridge, TN.

ORNL/RAP-17. Decommissioning of the Molten Salt Reactor ExperimentA Technical Evaluation, January 1988, Martin Marietta Energy Systems, Inc., Oak Ridge, TN.

ORNL/TM-728. MSRE Design and Operations Report, Part I, Description of Reactor Design, January 1965, Union Carbide Corporation, Oak Ridge, TN.

UCOR-4610/R0. Isotopic Distribution in MSRE Chemical Traps, July 2014, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

ORNL/TM-1853. Chemical Research and Development for Molten-Salt Breeder Reactors, 1967, Union Carbide Corporation, Oak Ridge, TN.

ORNL/TM-2578. Processing of the MSRE Fuel and Flush Tanks, 1969, Union Carbide Corporation, Oak Ridge, TN.

ORNL/TM-3884. The Migration of a Class of Fission Products (Noble Metals) in the Molten-Salt Reactor Program, December 1972, Union Carbide Corporation, Oak Ridge, TN.

PROC-MSRE-581/R3. Periodic Inspection of Design Features, May 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

PROC-MSRE-506/R2. Routine Checks at the Molten Salt Reactor Experiment Facility, February 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4807. Engineering Evaluation of Plugging Thermal Wells and Drying Drain Tank Cell Air in the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, December 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

4-17

5. REACTIVE GAS REMOVAL SYSTEM 5.1 RGRS SYSTEM DESCRIPTION The RGRS is a complex gas-handling system, containing numerous valves, temperature indicators, PTs, flowmeters, chemical traps, and analytical instrumentation, that has been used to treat process gases at MSRE since the mid-1990s. The RGRS was initially designed to remove fluorine (F2) and uranium hexafluoride (UF6) deposits from the offgas system piping system where they had migrated from the FDTs and FFT following reactor shutdown in 1969.

The original system included a NaF trap to capture UF6 and an alumina trap to capture F2. The process gases were monitored using Fourier Transform Infrared Spectroscopy (FTIR) and an NDA system was used as a level indicator to monitor NaF trap filling. The system was modified prior to defueling the salts in the FDTs and FFT. A second NaF trap equipped with an NDA system was installed and a molecular sieve was added to the RGRS.

Following defueling operations at MSRE, the RGRS has been used to remove gases evolved from the FDTs and FFT. The primary system components in the current configuration are two NaF traps, one alumina trap, and one molecular sieve arranged in series. The NDA system was removed from the RGRS following defueling. The two FTIRs are still used to monitor gases as they move through the system. One provides chemical composition data on the incoming gas and the second monitors the gas as it leaves the RGRS. The system is equipped with multiple PTs, temperature indicators, and flowmeters to allow operators to monitor and control the RGRS during headspace pumpdown.

See drawing J3E020794A066 for a detailed diagram of the MSRE RGRS.

5.1.1 Piping/Tubing and Valves Three types of tubing are used with the RGRS system. Tubing upstream of the chemical traps is Monel 400 alloy. Downstream tubing appears to be a mixture of stainless steel 304 and 316 alloys. A combination of 0.25-in. and 0.50-in. diameter tubing is used in the RGRS. Some of the tubing has a 0.065-in. wall thickness; however, a majority of the tubing is 0.035-in. wall thickness.

The valves in the RGRS system are primarily bellows valves. There are needle valves associated with the auxiliary cabinet. The valves on the RGRS system are used to open flow paths for the process gases, isolate sections of the RGRS, and control the flow rate through the RGRS. The maximum flow rate allowed during tank headspace pumpdown is 5 L/min. per PROC-MSRE-517, Reactive Gas Removal System (RGRS) Operations.

Flex lines with VCR fittings connect the chemical traps to the hard tubing of the RGRS. The flex lines facilitate changing of the traps.

The RGRS is heat-traced from the six-valve manifold to the molecular sieve. The chemical traps and the flex lines running to and from the traps are not heated. Heat trace is maintained at 42-60°C.

5.1.2 RGRS Glovebox The RGRS glovebox serves as part of the secondary containment for the NaF traps used to process gases at MSRE. It contains numerous valves that must be adjusted to control gas flow. The 10-cm and 20-cm FTIR cells are located in the glovebox. Access to the top of the NaF traps is through the glovebox when 5-1

NaF traps are attached to and detached from the system. The NaF traps are lowered into the carriers from the glovebox. The carriers comprise the rest of the secondary containment for the NaF traps in service at any time. Any tools or equipment that must be used inside the glovebox are bagged in through the glove ports.

5.1.3 Chemical Traps Three types of chemical traps are employed in the RGRS: NaF traps, alumina traps, and molecular sieve (NCSE-OR-MSRE-1507). All traps are constructed of 3-in., schedule 10 Monel pipe with welded Monel top and bottom heads. The effective trap height is ~16.5 in. with a nominal inner diameter of 3.26 in.

(wall thickness is 0.12 in.). Trap diameters were selected to ensure a critically safe geometry in accordance with an ANSI/ANS-8.1 subcritical cylinder diameter of 4.13 in. for 233UO2F2.

Trap inlet and outlet lines are 0.25-in. Monel tubing equipped with manual valves and VCR fittings. The overall height, including inlet and outlet tubes with valves and upper and lower skirt, is approximately 21 in. The traps have upper and lower baffle plates approximately 0.6 in. from the top and bottom of trap, respectively. The trapping media in each trap is located between upper and lower baffle plates.

All traps have thermal wells for insertion of thermocouples at the top of the trap. Five zone thermocouples are used in the two NaF traps and the alumina trap to allow operators to monitor the reaction front in the trap. The molecular sieve has a single point thermocouple to monitor trap exit gas temperature.

5.1.3.1 NaF traps The NaF traps primary function is to trap UF6. The traps also capture HF, MoF6, and other species in the process gas stream. HF is easily displaced by MoF6 and UF6. NaF traps are changed on a regular basis when their capacity has been reached. Temperature increases on the thermocouples associated with the trap are used to monitor the reaction front and determine when the trap has reached capacity. At present, two NaF traps in series are used on the system. The dual-trap configuration was employed during defueling to ensure any potential UF6 carryover from the first trap would be adsorbed on the second trap.

This configuration has been maintained for post defueling operations.

The primary adsorption reactions on the NaF traps are as follows:

UF6 + 2NaF UF6

• 2NaF MoF6 + NaF MoF6

• NaF HF + NaF HF

• NaF The NaF traps are contained in overpacks constructed of 4-in. stainless steel tubing with a welded steel bottom and a threaded top adapter. A stainless steel plug that threads into the top adaptor is used to seal the overpack. The overpack height is approximately 23 in., with an internal diameter of 4.01 in. and a wall thickness of 0.12 in. The trap/overpack assembly is placed in a shielded carrier prior to attachment to the RGRS. If the trap is loaded with uranium, the trap/overpack assembly remains in the carrier until it is received at Bldg. 3019A for storage.

The stainless steel carrier is a cylindrical vessel with an outside diameter of 10.75 in. and a 4.75-in.

diameter hole bored in the center to within 2.5 in. of the bottom. The top and bottom of the carrier have welded 14-in. diameter flanges. The bottom flange is a sealed blank flange. A stainless steel top is bolted to the carrier that results in an overall height of ~28.3 in. with a bore depth of ~23.1 in.

5-2

Overpacks and carriers are not used for the alumina traps and molecular sieves since there is no significant uranium loading on these traps 5.1.3.2 Auxiliary cabinet The auxiliary cabinet serves as secondary containment for the alumina trap and the molecular sieve. It is vented to the CVS.

5.1.3.3 Alumina trap The primary function of the alumina trap is to capture fluorine from the process gas lines. Fluorine reacts with the alumina and drives off water in the process. The progress of the fluorine absorption reaction with the alumina can be monitored with the thermocouples, which shows the reaction front as it progresses through the trap. Alumina traps are changed out when they are exhausted. The alumina trap is contained in the auxiliary cabinet, which also contains the molecular sieve trap.

Several reactions of interest can occur on the alumina trap, which are detailed below.

6F2 + 2Al2O3 4AlF3 + 3O2 The F2 can also react with the water of hydration present in the alumina to form HF and O2. Much of the HF formed (as well as any that might pass through the NaF traps) reacts with the alumina and is retained in the trap by the following reaction:

6HF + Al2O3 2AlF3 + 3H2O Any UF6 escaping the NaF traps can react with alumina hydrate by the following reaction:

UF6 + 2Al2O3

• H2O UO2F2 + 2Al2O3 + 4HF 5.1.3.4 Molecular sieve trap The molecular sieve, which is located in the auxiliary cabinet with the alumina trap, removes moisture from the system by adsorbing it onto an alkali metal aluminosilicate media (Type 13X molecular sieve, Na86[(AlO2)86(SiO2)106] *x H2O; see Table A.1). It can also capture other species present in the process gas lines. The molecular sieve is changed out when it is exhausted. A reaction of concern with the molecular sieve involves HF that may get past the alumina trap.

4HF + SiO2 = SiF4 + 2H2O SiF4 can undergo further reaction with HF to form fluorosilicic acid:

6HF + SiO2 H2SiF6 + 2 H2O 5.1.4 Vacuum pumps Two vacuum pumps configured in parallel are used to pull gases through the RGRS. Historically, only one pump is in use at any given time. Danielson Barodyn' pumps, designed for use in corrosive environments containing HF and F2, were used during defueling operations. These pumps are no longer available. Scroll pumps are currently installed in the RGRS. Switching from one pump to the other is controlled in accordance with PROC-MSRE-0517, Operation of the Reactive Gas Removal System; 5-3

PROC-MSRE-601, Vacuum Leak, FT-1A, and SF6 Test; and PROC-MSRE-596, Alumina and/or Molecular Sieve Trap Changeout.

Table A.1. Typical physical properties of Type 13X molecular sieve media 1/16-in.

Pellets 1/8-in.

Pellets Nominal Pore Diameter (angstroms) 8 8

Particle Diameter (mm)*

1.6 3.2 Bulk Density (lb/ft³)

40 40 Heat of Adsorption (Btu/lb H2O) 1800 1800 Crush Strength (lb) 7 15 Equilibrium H20 Capacity (wt%)**

26 26 Moisture Content (wt%)

1 1

• The length of the pellets is approximately twice the diameter

• • Measured at 17.5 mmHg and 25°C Source: http://www.asge-online.com/AdsbMSievespg176.pdf, last accessed 10/7/2016 5.1.5 FTIR Data Acquisition System The FTIR data acquisition system (FTIR-DAQ) is a distributed network of instruments, data processors, computers, and software that collects and processes signals from two FTIR spectrometers; and multiple PTs, temperature probes, and flowmeters. Signals are processed using a Yokogawa data collection system that combines the signals from all of the sensors including the FTIRs. The Yokogawa system consists of a DX2040 recorder and three MW100 I/O modules for data collection and processing. The MW100 modules collect data from the RGRS, field sensors and transmitters in the Vent House; and field sensors and transmitters in the lower-level electrical area of Bldg. 7503. The system is user-configurable and has the capacity to accept additional input beyond the current design needs. System details are provided in UCOR-4847, System Description for the MSRE FTIR-DAQ.

5.1.5.1 FTIRs Two Bomem Model MB-104 FTIRs collect data associated with the RGRS process gases. FTIR-1 has a 10-cm cell before the first NaF trap and a 20-cm cell between the two NaF traps. A switch moves the optics to allow a single FTIR to monitor both cells; however, simultaneous monitoring is not possible.

FTIR-2 has a 4-m cell to monitor the gases that have passed through all of the traps. It is located after the molecular sieve and before the vacuum pumps on the interior wall of the South Truck Bay. The 4-m cell indicates the composition of the gases being exhausted to the CVS and ultimately to the stack.

5.1.5.2 Software/hardware The FTIR-DAQ employs multiple software releases on three different computers to acquire, process, and manage the data associated the RGRS operations; see Table A.1. Bomem FTSW100 and Horizon MB are used to operate the FTIRs. DAQStandard Configurator and MW100 DAQMaster are used for the Yokogawa systems. The LabView and Kepware Open Platform Communications software packages are used to manage and present data from the RGRS to the system user. LabView' allows the operator to visually see reactive gas concentrations, current temperatures in the chemical traps, and current pressures throughout the RGRS system. LabView is also used to calculate the final concentration of the reactive gases.

5-4

Table A.1. Overview of currently installed software versions Configured Software Version Kepware/

LabVIEW OPC Server Computer Version on FTIR-1 Computer Version on FTIR-2 Computer FTIR-DAQ Application 1.0 None None LabVIEW 2013 with Professional Developer Suite 13.0 (32 bit)

None None DAQStandard Configurator 9.1.1 None None MW100 DAQMaster R3.04.01 None None FTSW100 Suite None 2.80.00.00016 2.80.00.00016 Windows 7 7 Pro Service Pack 1 7 Pro Service Pack 1 7 Pro Service Pack 1 Internet Explorer 10.0.9200.16686 10.0.9200.16686 10.0.9200.16686 KEPServer Ex5 5.13.191.0 None None HorizonMB Mathematics Module None 3.3.0.0 None Source: UCOR-4847, System Description for the MSRE FTIR-DAQ System, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN 5.1.5.3 Pressure transducers (PTs)

Numerous PTs are used in the RGRS system during leak tests and processing of reactive gases. PTs are distributed throughout the RGRS gas flow path. During leak tests, the RGRS is isolated into segments by closing specified valves and the associated PTs are used to evaluate the integrity of each section. Once the system has passed the leak test it is ready for operation. PT-8 is used in conjunction with the FTIR-1 data and flowmeter FT-1A to calculate NaF trap loading during headspace pumpdown.

5.1.5.4 Temperature elements (TEs)

Temperature elements, commonly called temperature probes, are used with all four of the original chemical trap positions in the RGRS (see Table A.1). The probes allow operators to monitor the temperature and establish the reaction front associated with each trap. For reactive gases like fluorine the temperature probe in the alumina trap is the only way to determine that the alumina trap is trapping fluorine. Monitoring the reaction front on NaF traps and the alumina trap provide the operator critical information about the remaining capacity of the traps.

Table A.1. RGRS thermocouple assembly locations Identification Location TE-1 (bottom) through TE-5 (top)

NaF #1 TE-6 (bottom) through TE-10 (top)

NaF #2 TE-11 (bottom) through TE-15 (top)

Alumina Trap #1 None installed Alumina Trap #2 TE-16 Molecular Sieve 5-5

5.1.5.5 Flowmeters The RGRS system has two flowmeters that are used to monitor total gas flow through the system. The flow rates combined with the FTIR data and the data from the PTs to determine the composition of the process gas at any given time. The resulting data is used to calculate the final mass of each reactive gas removed from the headspace of the FDTs and FFT. The performance of FT-1A is evaluated in accordance with PROC-MSRE-601, Vacuum Leak, FT-1A, and SF6 Test.

5.2 CURRENT CONDITION EVALUATION 5.2.1 Operational Observations The function of the RGRS has shifted from trapping UF6 to managing the F2 inventory in the tank headspace generated by the gamma radiolysis of the BeF2-LiF-ZrF2 salts remaining in the drain tanks.

The estimated UF6 concentration in the FDTs is 280 mg/kg based on an NDA evaluation performed by Los Alamos National Laboratory (LA-UR-13-22367). No UF6 has been detected by the FTIR in the last several pumpdowns; however, low levels of MoF6 have been detected on the 10-cm FTIR cell during most pumpdowns since 2008.

Fluorine cannot be seen by the FTIR. Indirect evidence of F2 in the gas stream is based on the behavior of the alumina trap during pumpdown. The temperature increase on the alumina trap is an indication that F2 is being trapped. The reaction front is monitored with the 5-zone thermocouple assembly (TE-11 through TE-15). The trap is considered full when the reaction front reaches the final thermocouple (TE-15).

Figure 5.1 shows a plot made from the readings of the five thermocouples on the alumina trap (TE-11 through TE-15) during the pumpdown in April 2016. As the reaction front moves throughout the trap, the temperature will rise in the area where the tank headspace gas is interacting with the alumina. The regions of highest temperature represent the largest reaction density. Temperatures will drop when either the alumina is spent or the gas flow is stopped. Using the thermocouple data in conjunction with the flow rates provides an accurate indication of when the alumina is spent.

Trace quantities of HF are observed on the 10-cm FTIR cell at the start of each pumpdown. The HF concentration then drops to undetectable levels after a few minutes. Carbon tetrafluoride (CF4) and several oxyhalides (chloride and fluoride compounds containing oxygen) are observed on the 10-cm FTIR cell during pumpdown. The presence of HF and oxyhalides indicates water is reacting with the F2 in the system, suggesting that air in-leakage is occurring.

The location of the in-leakage is unclear. The drain tanks take several months for the pressure to increase from ~8 psia to ~14 psia. If in-leakage were occurring in the drain tanks the resulting HF would be detectable in the 10-cm FTIR cell throughout most of the pumpdown. Considering the low vapor pressure of HF compared to F2 there is no plausible mechanism that would cause HF in the drain tanks to be removed preferentially before the F2. Alumina trap temperatures continue to increase long after the HF is no longer present on the 10-cm FTIR cell. These factors suggest the in-leakage is somewhere in the piping between the drain tanks and the 10-cm FTIR cell. If the in-leakage is in the stainless steel line running from the drain tanks to the dual isolation valves at the front of the RGRS it is possible HF would lay on the piping in the liquid phase until vacuum is placed on the system. The in-leakage may be downstream of the isolation valves in the RGRS. The most likely source of in-leakage is at a compression fitting. It is also possible that the integrity of the piping has been compromised at some point in the system. The HF that is trapped on the alumina trap is not detected on the 4-m FTIR cell.

5-6

Figure 5.1. Temperature readings for the alumina trap during the April 2016 pumpdown.

Silicon tetrafluoride (SiF4) is generated in the molecular sieve as a byproduct of the reaction of HF with the silicate trap media. SiF4, or a similar compound containing Si-F bonds, is observed on the 4-m FTIR cell throughout headspace pumpdown as it was during defueling operations. SiF4 does not pose a significant problem under anhydrous conditions; however, it reacts with water to produce HF. A second reaction leads to fluorosilicic acid. Fluorosilicic acid is highly corrosive.

There is no evidence of general corrosion of stainless steel tubing downstream of the chemical traps or of the Monel 400 tubing upstream of the RGRS traps. There have been two stainless steel tubing failures downstream of the vacuum pumps in the last six months. One of the breaks occurred in April 2016 and the second occurred in September 2016. Both failures were associated with compression fitting locations and were attached to tubing being disconnected during maintenance activities. These events indicate a more thorough evaluation of tubing integrity in the RGRS is necessary to ensure safe operations can continue. It was noted that straight runs of the tubing remain ductile despite the failures at compression fittings. Details of corrosion evaluation are discussed in Sect. 5.2.2 and Appendix B.

The vacuum pump experienced severe plugging during the April 2016 pumpdown. The outlet was completely blocked by deposits of corrosion products (Figure 5.2) and the downstream tubing contained substantial deposits (Figure 5.3). The interior of scroll pumps used with the RGRS typically exhibit deposits in the scroll. The heaviest internal deposits are located near the port to the exhaust and ballast valve assembly (Figure 5.4) and on the valve assembly itself (Figure 5.5). The associated spring failed due to the corrosion. The most significant deposits were on the outlet (downstream) side of the pump; however, deposits were also observed on the inlet (upstream) side of the pump captured by the inlet 0

50 100 150 200 250 300 350 400 450 500 4/18/16 0:00 4/19/16 0:00 4/20/16 0:00 4/21/16 0:00 4/22/16 0:00 4/23/16 0:00 Temperature (°C)

Time Alumina Trap Temperatures During April 2016 Pumpdown TE-11 TE-12 TE-13 TE-14 TE-15 5-7

screen (Figure 5.6).

Figure 5.2. Plugged outlet of vacuum pump from April 2016 pumpdown.

There are significant solid and amorphous deposits throughout the system. A thin residue is observed each time the system is opened. Significant deposits were noted in 2012 during window replacement on the 4-m FTIR cell. Pale green dusty residue (fixed) and dark green amorphous deposits were present on the cell walls, windows, and mirrors. The windows were replaced and the mirrors were cleaned. The remaining residue did not interfere with the light path or the operation of the FTIR. The 10-cm and 4-m FTIR cells continue to degrade. The 10-cm cell windows have not been changed since 2003. As noted, the 4-m cell windows and mirrors were serviced in 2012; however, the single beam spectrum indicates substantial degradation and possible damage from fluorosilicic acid. The light transmission on the 10-cm cell is poor, but the system is still functional at this time. The 10-cm cell windows will require replacement in the near term. Service of the 10-cm cell will require removal of the back of the RGRS glovebox. Extensive control of radiological contamination will be required.

The alumina trap and the molecular sieve in the system at the beginning of this evaluation had been used for multiple headspace pumpdowns. The molecular sieve had not been changed out in several years.

Facility personnel executed work package WP-2016-MSRE-7503-TY1-7154 to replace the alumina trap with two alumina traps in series, replace the molecular sieve, and develop a detailed inspection plan following the October 2016 pumpdown. These actions were initiated based on the vacuum pump plugging event that occurred during the April 2016 pumpdown and were completed during September 2016 (RPJ-027503-D842). The evaluation team reviewed the modifications planned by the facility and concurred with the approach. The inlet of the molecular sieve showed some surface deposits on the tubing surface (Figure 5.7). The outlet of the molecular sieve showed significant deposition of material (Figure 5.8). It is not clear if these are corrosion products or whether they result from reactions occurring in the RGRS. Further evaluation of the deposits may be warranted.

5-8

Figure 5.3. Line from pump outlet.

Figure 5.4. Internal heavy deposits on scroll pump.

5-9

Figure 5.5. Deposits on exhaust and ballast valve assembly.

Figure 5.6. Deposits on the inlet screen of the vacuum pump.

5-10

Figure 5.7. Molecular sieve inlet September 2016.

Figure 5.8. Molecular sieve outlet September 2016.

Pass-through valve leakage is evident from the leak tests executed prior to each pumpdown of the FDTs and FFT. The lack of substantial water and carbon dioxide peaks on the FTIRs indicate leakage is primarily within the RGRS and not from the atmosphere. Leakage across valves in the RGRS have been observed and documented (BJC/OR-2794, Engineering Evaluation of the Leakage Across IXH2 and IXH5 and Extent of Condition Review at the Molten Salt Reactor Experiment at Oak Ridge National Laboratory, Oak Ridge, Tennessee, April 2007).

The 3-in. neoprene flex hose that serves as secondary containment material is brittle and corrugated bands are protruding from assembly in numerous locations. This degradation is evident from where process 5-11

lines enter the NESA and extend into the High Bay to the six-valve manifold. Process lines from the six-valve manifold on are heat traced and cannot be visually inspected without removal of the heat trace.

Preventative maintenance is not current on many valves, vacuum pumps, PTs, temperature probes, flowmeters, or the 10-cm and 20-cm FTIR cells. Maintenance consists of repair of failed devices. PROC-MSRE-601, Vacuum Leak, FT-1A, and SF6 Test, is used to verify leakage levels, performance of the primary flowmeter used in trap loading calculation, and to execute the FTIR sensitivity check. System components are not routinely exercised. The RGRS is dormant between tank pumpdowns unless maintenance is performed on the system, which generally requires execution of PROC-MSRE-517, Reactive Gas Removal System (RGRS) Operations, prior to opening the system. The current pumpdown schedule is 1-2 times annually.

Some capabilities of the data acquisition system installed in October 2015 have not been utilized to date.

For example, several PTs have not been integrated into the system, and additional regions of interest on both the 10-cm and 4-m cells should be monitored. Operation of the FTIR has become less rigorous over time. The startup checklist has been removed from the operational procedures and periodic performance checks are not run on the systems. Performance checks will assist facility personnel in taking proactive steps to maintain the FTIRs.

The team, in conjunction with UCOR Industrial Hygiene SMEs, reviewed the HF monitoring and controls for personnel exposure as documented in UCOR-4799, UCOR Molten Salt Reactor Experiment (MSRE)

Hydrogen Fluoride (HF) Protection, Monitoring and Response Strategy. No issues with the use of the breathing zone monitors for personnel monitoring and protection in place of design and installation of new HF room monitors were identified.

The planned life expectancy of the RGRS was 3 years; the system has now been in operation for almost two decades. Any decisions regarding future S&M or remediation activities at MSRE will need to take into account the original design life of the system.

5.2.2 Corrosion Issues with the RGRS A significant concern is the integrity of the RGRS, and more specifically, the Monel 400 tubing, the stainless steel fittings, and the stainless steel tubing used in the RGRS after the most corrosive gases were presumed to have been removed from the system. In order to confirm the composition of the tubing in the RGRS system, an X-ray fluorescence (XRF) analyzer was used to determine the composition at many locations in the tubing runs presumed to be Monel 400 and a 300-series stainless steel. These measurements showed the sections of the tubing leading up to the glovebox were composed of Monel 400 at all 12 locations where measurements were made. XRF examination of the tubing after the molecular sieve trap showed the tubing was type 304/304L stainless steel except for a recently replaced section that was identified as type 316/316L stainless steel.

An instrument which uses ultrasonic waves to measure the thickness of metallic materials was used to measure the wall thickness of both the Monel 400 and the stainless steel tubing. The measurements of the stainless steel tubing showed an apparent mixture of wall thicknesses, both 0.035 and 0.065 in., and assuming there actually is a mixture of these sizes, the measured thicknesses of all the tubes were within the nominal thickness values, so no clear evidence could be found for thinning of any of the stainless steel. However, even though wall thickness measurements did not indicate any meaningful thinning of the stainless steel tubing, the presence of stainless steel corrosion products provides evidence that some corrosion of the stainless steel tubing might have occurred. Details on sample locations, composition and wall thickness are presented in Appendix B.

5-12

Measurements of the wall thickness of the Monel 400 tubing ranged from 0.065 in. to 0.070 in. The thinnest wall measurements were seen in the transmitter room, which was as close to the fuel salt drain tanks as measurements could conveniently be made. The thinnest wall measurement was 0.065 in., which is generally a nominal thickness for tubing. However, six of the 12 measurements were 0.069 in. or 0.070 in. They were all made at points farthest from the drain tanks, and could well have been normal variations in response to different temperatures. Alternatively, there could be some significance to the thinner tube wall being in the sections closest to the fluorine source. However, no coppera major component of Monel 400was found in the analysis of deposits collected from the outlet of the pump used to move gas through the RGRS.

With regard to the Monel 400 tubing, previous measurements have detected radioactive hotspots at locations assumed to be fittings that connect one length of tubing to the next. With the recognized susceptibility of Monel 400 to stress corrosion cracking in moist HF and hydrofluoric acid environments, the possibility of tube cracking cannot be dismissed. However, given the secondary containment used around the Monel 400 tubing, extensive examination for evidence of cracking would be problematic. The negative pressure maintained for months in the drain tanks and tubing leading up to the isolation valves indicate there are no leaks in those components. No reports have been found that assess the capability to maintain negative pressure in the Monel tubing that makes up the RGRS between the isolation valves and the NaF traps. Since it is speculated that localized deposition of UO2F2 has occurred in the Monel tubing, this has to be viewed as an indicator that there likely are leaking joints or even cracks in the Monel tubing in this portion of the RGRS, and that this has permitted access of moist air and the subsequent formation of reaction product(s) with that air.

There are several concerns regarding the integrity of the RGRS, particularly the previously mentioned possibility of stress corrosion cracking of Monel 400 in HF that was cited in a number of the references used to collect corrosion information. Also, the discovery, after the April 2016 pumpdown of the drain tanks, of deposits in the pump outlet line and the reported brittle failure of the 304/304L stainless steel tube at the outlet of the pump used to help move gas through the RGRS tubing has to be a cause for concern (RPJ-027503-D842). Subsequent discovery of other failed stainless steel tubes in the exhaust system downstream of the pump brings greater emphasis to the need to conduct more extensive examinations of all tubing in the RGRS.

One of the most obvious ways to determine if a material is corroding in a particular environment is to analyze the corrosion product that develops after exposure of the material in question. Because of the radioactive contamination of the vessels of interest and much of the associated piping and tubing, easy access is not available to any corrosion products. However, when plugging was observed in the pump that removes gas through the RGRS, some of that material was collected and analyzed.

An analysis, by scanning electron microscopy with energy-dispersive X-ray spectroscopy (SEM/EDS), of deposits collected from the pump outlet line in April 2016 showed the presence of chromium, molybdenum, iron, and silicon. Crystalline phases identified using X-ray dispersive analysis (XRD) included CrF3*H2O, NiSiF6*6H2O and FeF3*3H2O. The source of the metallic elements could be the Hastelloy N although the iron and chromium are only present in very limited concentrations in that alloy.

The 304L stainless steel contains everything but molybdenum, and that element could be present as a fission product rather than a corrosion product. Consequently, analysis of the deposits from the pump outlet showed corrosion products reached the pump and collected in the outlet line, but the source of the metallic elements in the corrosion products could not be positively determined. A subsequent analysis of corrosion product was conducted on deposits collected during examination of a section of stainless steel tubing that was reported to have suffered a brittle failure. Appendix C contains details on the analytical work performed on the deposits removed from the vacuum pump outlet and the stainless steel tubing.

5-13

A scanning electron microscope was used to examine the two pieces of fracture surface. Deposits from inside the larger piece of tubing and deposits from the internal diameter of the second piece of tubing were also examined. The report of the analysis, included in Appendix C, describes the results of the examination. The EDS spectra of a cut surface of the tubing, selected since it would presumably be free of any deposits or reaction products, confirmed the composition was definitely 304 or 304L stainless steel.

The fracture surface examination did not show any evidence of necking or the cup and cone fracture surface typical of a ductile fracture. The edges of the fracture surface were too damaged for determination of the origin of the fracture, but no evidence was found that was inconsistent with a brittle fracture.

Composition measurements of the inner diameter surface of the second tube section, as well as the presumably brittle fracture surface of the first tube section, showed presence of small amounts of molybdenum along with significant amounts of iron, chromium and nickel, which would be expected given the stainless steel base material. Examination of the deposits removed from the two tube sections showed significant amounts of oxygen and fluorine as well as iron, chromium and nickel. Lesser amounts of molybdenum and silicon were present as well as even smaller amounts of elements like aluminum and manganese.

Of particular significance is the presence of molybdenum on nearly all surfaces even though no molybdenum was present in the stainless steel tubing alloy or the Monel 400 tubing. This led back to the question of the source of the molybdenum corrosion product from the Hastelloy N or fission or radioactive decay product. Appendix C provides a detailed evaluation of the analytical data. It is possible that determination of the molybdenum isotopes in filters and traps as well as the deposits on the surface of the tube samples could help identify the source of the molybdenum. Samples containing molybdenum isotopes were collected shortly after the U-235 and U-233 reactor operations campaigns. Analysis of these samples indicated that all molybdenum was due to fission products, and that there was no evidence of corrosion of system components. The distribution of isotopes found in one scale sample collected in August 2016 clearly show that the molybdenum is a corrosion product and not a fission product. This result is surprisingly different from previously compiled data on molybdenum isotopes which were all fission products. Further isotopic molybdenum analysis may clarify what fraction of MoF6 collected during defueling and subsequent headspace pumpdowns resulted from corrosion and how much was actually residual fission product. Appendix C contains a more detailed discussion of the isotopic distribution of molybdenum and what it may indicate regarding the extent of general corrosion of components at the MSRE.

5.3 NEAR-TERM UPGRADE RECOMMENDATIONS Based on the evaluation teams assessment of the current condition of the RGRS and its role in reducing the risk associated with the fluorine inventory in the FDTs and FFT for the next 1-5 years, the following recommendations should be implemented to support facility operations:

1. Implement routine, more frequent pumpdowns to minimize safety and corrosion risk from head space fluorine buildup and to improve system reliability.

2. Obtain gas samples directly from the tank headspace (FDT-1 and FDT-2). This will allow evaluation of chemicals that cannot be detected using the FTIR alone. The data obtained from these analyses can also be used to verify fluorine generation rate calculations used in the DSA.

3. Initiate actions to evaluate feasibility of continuous purge life-extension option for RGRS using CVS with no traps.

5-14

4. Implement improved procedural controls for flow rates for the October 2016 pumpdown to minimize the risk of channeling through traps.

5. Document system performance after each pumpdown to use in trending and analysis of system issues.

Begin this process with the headspace pumpdown in the fall of 2016. Compare the condition to that found following the April 2016 pumpdown.

a. Obtain intact section of stainless steel tubing with compression fitting for evaluation of embrittlement.

b. Collect samples of solid residues if system plugging is found.

6. Determine the integrity of the Monel tubing in the RGRS by removing a section in an area upstream of the NaF traps where air in leakage is suspected and examine the tubing for evidence of stress corrosion cracking. A similar evaluation should be performed on a section of stainless steel tubing between the molecular sieve and the vacuum pump. The presence of significant residue or solids in either section of tubing should be analyzed for structural details and elemental composition.

7. Calibrate the 10-cm and 4-m FTIR for the primary gases of interest (HF, UF6, MoF6, SiF4, CF4, ClF3, and SF6).

8. Proceed with the current operations team plan for installing sacrificial wool traps to consume corrosive gases that pass through the entire RGRS. These traps serve two purposes: (1) protection of the vacuum pumps, and (2) reduction of possible emission of reactive gases via the CVS. Installation of wool before and after the 4-m FTIR cell would provide additional protection for the windows and mirrors in the cell.

9. Implement increased rigor in operation of the FTIR-DAQ system through development and use of a startup checklist and performance testing.

10. Configure the FTIR systems to identify other components in the headspaces gases and collect the data. Additional regions of interest can be set up for several oxyhalides believed to be present in the gas mixture. The operations team may identify additional gases of interest.

11. Proceed with the current operations team plan for installing a second vacuum pump to provide redundancy.

12. Update the Nuclear Criticality Safety Evaluation to reflect the defueled state and potentially reduce some controls currently in place for RGRS operations.

13. Install 5-zone thermocouple assembly in the molecular sieve to allow for monitoring the reaction front during pumpdowns. Also, install 5-zone thermocouple assembly in second alumina trap to better ensure that monitoring capability of reaction front during pumpdowns is available.

14. Replace HV-59 with a valve better suited for regulating inert gas flow rates through the RGRS.

5.4 LIFE-EXTENSION UPGRADE RECOMMENDATIONS The RGRS was originally intended to be used for removal of UF6 from the offgas piping from 1996-1998. The lifetime was extended to allow RGRS to be used in defueling operations from 2004-2008. The system has now been in use for two decades, greatly exceeding the original planned lifespan. Several questions remain on final disposition of defueled salt in the FDTs and FFT. The final path for salt disposition will greatly impact the role the RGRS plays in the process. The RGRS may also be used in the purging of ancillary and obsolete systems at MSRE that must occur before or during initial phases of facility D&D.

5-15

The RGRS, in its current configuration, cannot be adequately maintained to continue removing reactive gases from the FDT and FFT headspace until 2060. The process lines have been subjected to aggressive corrosive gases through the hydrofluorination and fluorination processes used on all three tanks, as well as ClF3 treatment on FDT-2. Pass-through leakage is already evident on several valves in the system. It is likely that this type of leakage will render the system incapable of passing leak checks in the future. To date, the concentration of water and carbon dioxide observed on the FTIRs remains low and does not change significantly during gas flow indicating leakage is minimal or internal. Over time, the likelihood of bellows seal failure, resulting in leaking into or out of the RGRS, will increase. At a minimum, operators will lose the ability to control flow rates and isolate sections of the RGRS to conduct maintenance activities. It is not possible to predict the time to failure on any given valve, only that, as time passes, one or more will cease to function acceptably.

There have been multiple reports developed over the last several years identifying various significant improvements to the RGRS. These reports and recommendations were reviewed and several options were developed by the team. These options included the following:

1. Continuous purge of FDTs and FFT headspace gas using CVS with no traps

2. Continuous purge of FDTs and FFT headspace gas using CVS with a NaF trap and booster pump

3. Continuous purge of FDTs and FFT headspace gas using both CVS fans with a NaF trap

4. Routine frequent pumpdown of FDTs and FFT to minimize fluorine inventory

5. Design and build a new Fluorine Gas Removal System to replace RGRS Each option is discussed in more detail below.

5.4.1 Option 1: Continuous Purge of FDTs and FFT Headspace Gas Using CVS with No Traps Continuous purge of FDTs and FFT headspace gas using CVS with no traps provides a passive system that significantly reduces the safety risk by minimizing the fluorine source term in the tank headspaces.

The current practice allows the fluorine concentration to build up to hazardous concentrations over the extended time between pumpdowns. The concept would utilize a nitrogen or argon purge to sweep the tank headspace, either in series flow from one tank to the next, or in parallel. This option will provide significant cost and waste savings from the elimination of the various components of the RGRS system and elimination of the waste from the traps. A scoping calculation (Appendix D) was performed that shows that with a minimal purge flow rate through the tank headspace the HF concentration in the stack effluent would be orders of magnitude below the air permit release limits as well as below the odor threshold for HF. However, this option does require a paradigm shift in that the headspace gas is being released with no treatment. New tank piping fabricated from Hastelloy C-276 would be required for the purge gas inlet and exhaust and could be installed in the 3-in. port at the top of each tank. The headspace gas samples and analyses performed to provide better information for the DSA would also be used as design input information for required purge flow rates. The effluent from the system could be monitored if required with either FTIR or UV detection system or sampled routinely for laboratory analysis. The use of a mass spectrometer MS for monitoring fluorine levels in the effluent was evaluated; however, the relatively high capital cost and ongoing maintenance costs for a mass spectrometer render this less favorable than the recommended options.

5-16

5.4.2 Option 2: Continuous Purge of FDTs and FFT Headspace Gas Using CVS with a NaF trap and Booster Pump The continuous purge of the FDTs and FFT headspace gas using CVS with a NaF trap and a booster pump has the same safety risk reduction by minimizing the fluorine source term build-up in the tank headspace. The booster pump will be required to provide the required vacuum on the system with the NaF trap in the flow path. Cost savings would be slightly impacted from the waste generated by the NaF trap and the addition of a pump to the system, but would still be significant. The NaF trap would provide some removal capability for both uranium and fluorine.

5.4.3 Option 3: Continuous Purge of FDTs and FFT Headspace Gas Using Both CVS Fans with a NaF Trap The continuous purge of the FDTs and FFT headspace gas using both CVS fans with a NaF trap is another alternative to Option 2, but using both CVS fans to provide the motive force versus a booster pump. Again, it provides the same safety risk reduction as Options 1 and 2 along with the gas treatment capability in Option 2. However, the capability of using both CVS fans to provide sufficient vacuum is uncertain at best. Also, the loss of redundancy for the CVS and the lack of additional benefits make this option less beneficial than either Options 1 or 2.

5.4.4 Option 4: Routine Frequent Pumpdown of FDTs and FFT to Minimize Fluorine Inventory Routine frequent pumpdown of FDTs and FFT to minimize fluorine inventory provides almost the same level of safety risk reduction as Options 1 and 2, depending on the frequency of pumpdowns (weekly to biweekly). A key benefit of this option is that it could be implemented with the existing RGRS in the immediate future. This option would require additional costs for trap disposal and additional staffing to support the more frequent pumpdowns. However, system reliability should increase from the more frequent operation of the system components and the proficiency of the staff will be improved. After additional evaluation, given the key benefit of being able to implement this option in the near future, this option was made a near-term recommendation.

5.4.5 Option 5: Design and Build a New Fluorine Gas Removal System to Replace RGRS Design and build a new Fluorine Gas Removal System as a replacement for RGRS would incorporate many of the improvements discussed in the near-term recommendations. However, a new system would be designed for fluorine gas removal as opposed to the existing system which was designed primarily for uranium. This system could be operated on a routine basis as Option 4, which would provide the same safety risk reduction benefit. The system would use the headspace gas sample analyses as design input.

The glovebox for the NaF traps should be able to be eliminated based on the low levels of uranium remaining in the system, and an enclosure or hood could be used if necessary. Hastelloy C-276 tubing should be used throughout the system to provide improved corrosion resistance. Tubing runs should be more direct, with minimal changes in tubing size and the number of fittings reduced significantly. Larger traps with improved trap media, specific to fluorine, should be utilized, assuming the current controls from the nuclear criticality safety evaluation are eliminated. Design and installation of this system will require significant funding and time, but it will reduce the risk from current operations and would provide a system for use until salt removal is completed.

5.4.6 Recommendations Based on the teams assessment of the RGRS, the following recommendations should be implemented to support facility operation through 2060:

5-17

1. Establish a project to implement the continuous purge option for RGRS using CVS with no traps.

a. Alternatively, implement the purge option with a NaF trap, fluorine trap, and booster pump with interlocks if the continuous purge with no traps is not an acceptable option.

2. Establish a project to design and build a fluorine gas removal system replacement for RGRS if Recommendation 1 is not an acceptable option.

a. New design would use Hastelloy C-276 for tubing and trap housings.

b. Use shorter tubing runs while minimizing use of fittings.

c. Eliminate use of glovebox.

d. Use larger trap sizes and improved trap media specific for fluorine and other gases identified from the Drain Tank headspace gas sample analyses.

3. Install an ultraviolet (UV) detection system to measure F2 levels in the effluent prior to reaching the CVS.

5.5 REFERENCES

ANSI/ANS-8.1. Nuclear Criticality Safety in Operations with Fissionable Materials Outside Reactors, 1998, American National Standards Institute, Washington, D.C.

BJC/OR-2794. Engineering Evaluation of the Leakage Across IXH2 and IXH5 and Extent of Condition Review at the Molten Salt Reactor Experiment at Oak Ridge National Laboratory, Oak Ridge, Tennessee, April 2007, Bechtel Jacobs Company LLC, Oak Ridge, TN.

J3E020794A066. Reactive Gas Recovery System Process Gas Valve and Pump Status Board, 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

LA-UR-13-22367. Evaluation of Molten Salt Reactor Experiment (MSRE) Remnant Salt Properties for Disposal Eligibility at the WIPP, March 2013, Los Alamos National Laboratory, Los Alamos, NM.

NCSE-OR-MSRE-1507, Rev. 2. Drain TankFuel Flush TankOff-Gas System, 2006, Bechtel Jacobs Company LLC, Oak Ridge, TN.

ORNL-4658. Chemical Aspects of MSRE Operations, Thoma, Roy E., December, 1971, Union Carbide Corporation, Oak Ridge, TN.

ORNL/TM-13579/R1. Technical Competencies for the Safe Interim Storage and Management of 233U at U.S. Department of Energy Facilities, March 1998, Lockheed Martin Energy Research Corporation, Oak Ridge, TN.

ORNL/TM-13770. Evaluation of Fluorine-Trapping Agents for Use During Storage of the MSRE Fuel Salt, May 1999, Lockheed Martin Energy Research Corporation, Oak Ridge, TN.

P3E020794A108/R2.

Reactive Gas Removal System Chem Trap

Notes, August 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

P3E020794A109/R2. Reactive Gas Removal System Chem Trap Sections and Details Sheet 1, August 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

5-18

P3E020794A110/R2. Reactive Gas Removal System Chem Trap Sections and Details Sheet 2, August 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

PROC-MSRE-517.

Reactive Gas Removal System (RGRS)

Operations, latest revision, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

PROC-MSRE-601. Vacuum Leak, FT-1A, and SF6 Test, latest revision, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

PROC-MSRE-596.

Alumina and/or Molecular Sieve Trap Changeout, latest revision, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

RPJ-027503-D842. Engineering Evaluation of the Plugging Event of April 20, 2016 at MSRE, May 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4596. Isotopic Distribution at the Molten Salt Reactor Experiment, Oak Ridge National Laboratory, Oak Ridge Tennessee, June 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4610. Isotopic Distribution in MSRE Chemical Traps Oak Ridge National Laboratory, Oak Ridge, Tennessee, July 2014, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4668. White Paper: MSRE Piping Disposition, Oak Ridge, Tennessee, January 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4699. Long Range Plan for the Reactive Gas Removal System at the Molten Salt Reactor Experiment Oak Ridge, Tennessee, April 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4765. 7503 Molten Salt Reactor Experiment HF Release at Oak Ridge National Laboratory, Oak Ridge, Tennessee, September 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4779. URS l CH2M Oak Ridge LLC (UCOR) Molten Salt Reactor Experiment (MSRE)

Hydrogen Fluoride (HF) Protection, Monitoring and Response Strategy, March 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4804. Fluorine Inventory Control in the MSRE Defueled Salt Drain Tanks at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, November 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4847. System Description for the MSRE FTIR-DAQ, latest revision, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4857. User Guide for the MSRE FTIR-DAQ System, Oak Ridge National Laboratory, Oak Ridge, Tennessee, latest revision, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4877. Technical Basis for Reactive Gas Management at the MSRE Facility, Oak Ridge National Laboratory, Oak Ridge, Tennessee, May 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

WP-2016-MSRE-7503-TY1-7154. Installation of Additional Alumina Trap, September 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

5-19

This page intentionally left blank.

5-20

6. SUPPORT AND ANCILLARY SYSTEMS 6.1 ELECTRICAL SYSTEM DESCRIPTION Electric power is supplied at 13.8 kV to the MSRE site from the ORNL electrical distribution system. The MSRE facilities are supplied with electric power at 480 V from three pole mounted 167 kVA transformers (station 294/F23) located at the west side of the reactor building, in front of the motor generator house. All MSRE loads are supplied from transformer 294-23. The main breaker, which is located in the motor generator house, is operated and maintained by UT-Battelle, LLC (UT-B). The 480 V main breaker feeds the 480 V switchgear (also known as the TVA Bus) with the following load sets:

480 V Bus # 3 Switchgear 480 V Bus # 5 Switchgear MCC T-1 MCC T-2 Non-process loads (via Main Breaker #2)

Facility loads are carried primarily from 480 V Bus #3 Switchgear, 480 V Bus #5 Switchgear, and via Main Breaker #2, however, several active loads are carried by other feeds from the TVA Bus. (Note: the naming nomenclature for loads fed from the 480 V main breaker does not logically follow. The MSRE formerly contained three 250 kVA, 13.8 kV-to-480 V transformers. These transformers have been abandoned and the loads transferred to the 13.8 kV-to-480 V pole mounted transformers 294-23)

See Figure 6.1 for a simplified one-line diagram of the MSRE electrical system.

6.1.1 Current Condition Evaluation The electrical system at MSRE is currently functional and supports routine activities as well as periodic evolutions such as pumping down the gas in the headspace of the FDTs with the RGRS. The team interviewed facility personnel and performed a detailed walkdown with a UCOR electrical engineer familiar with the facility to obtain an understanding of the current condition of the electrical systems at MSRE. The breakers, switchgears, and motor control centers (MCCs) appeared to be original to the facility. Several inactive breaker and switchgear locations were blanked, however, the majority of breaker and switchgear locations associated with inactive systems had overcurrent protection devices racked-in. Facility personnel expressed concern that some breakers might not reclose after routine maintenance is performed.

Issues identified through interviews and walkdown include the following:

Cut and exposed electrical wires in cable trays.

One (1) confirmed case of live 120 V wires PM for main breaker last performed in 2005.

PMs for several 480-V feeder breakers not performed since 2004, and the balance of 480 V breakers have no record of PMs (10 known).

No record of PMs performed for any MCCs.

6-1

6-2 Figure 6.1. Simplified electrical distribution diagram for MSRE complex.

Inconsistent and in some cases inadequate labeling of electrical components.

Unclear, in some cases, which components were energized and which were abandoned in place.

Many system panels, breaker locations, switchgear locations, and MCC locations associated with inactive equipment/systems are located adjacent to active electrical components.

Unknown state of some supply cables (e.g., CVS motor electrical supply cables that run underground).

Parts obsolescence, including inability to obtain Nationally Recognized Testing Laboratory (NRTL) program-approved components for required systems.

Arc flash labeling needs to be updated, and in some cases initially completed, for equipment currently in service.

Overcurrent protection devices in Arc Flash Model not reflective of installed equipment in many cases, therefore current arc flash labels may not be accurate During the walkdown, several locations were identified in the 7503 basement where 120-V or 480-V wires were cut and loose in cable trays. These were identified as immediate safety concerns, relayed to UCOR management and promptly addressed. In addition, facility personnel pointed the team to a known condition in which an active 480-V distribution panel in the 7503 basement was in poor condition, requiring the area to be roped off.

UCOR is in the process of creating a focused team to evaluate electrical overcurrent protection devices at facilities it operates. Included in the scope of the UCOR overcurrent protection team are the status of associated PMs and the identification of equipment that can be air-gapped, and arc flash calculations and labelling. It was identified to the team that for many electrical components at MSRE, an engineering evaluation is required to operate the components due to the state of arc flash labelling.

Maintenance activities and PMs on several electrical components, including PM of the main breaker and repair work for the degraded 480-V distribution panel described above, require a facility outage to be worked. During the course of this evaluation, the MSRE facility outage was completed to perform PM on the 480-V main breaker, inspect the 480-V feeder breakers in the main switchgear for last PM dates, inspect several sections of the CVS power supply raceway, and perform electrical air gapping in the distribution panel for the drain tanks heaters and flush tank. The results of the facility outage were as follows:

The MSRE main breaker was replaced with a spare that had a current PM.

The MSRE main breaker and one (1) of the 480-V feeder breakers in the main switchgear were found to be in a degraded condition. The metal fingers that plug into the rear stabs fell off during breaker removal; these metal fingers make the electrical connections between the breaker and the stabs in each compartment. Four spare and unneeded feeder breakers were removed PMs for the 480-V feeder breakers were found to range from 1966 to 2009. Several breakers did not have PM labels.

Several junction boxes that contain CVS power cables were found to be in poor condition.

Protective coating and sealant was applied after inspection was complete, however, the junction boxes had corroded through in several locations.

Air gapping the conductors to each breaker in the heater distribution panel was successful.

The state of the heater distribution panel now matches the description in the DSA.

6-3

Air gapping of the energized wiring was unsuccessful due to as-found modifications to the internal switchgear wiring 6.1.2 Near-Term Upgrade Recommendations Based on the teams assessment of the electrical system and the extent to which it is currently relied upon in the DSA, the following recommendations should be implemented to support facility operation:

1. Isolate, protect, and/or de-energize (if possible) exposed wires in cable trays.

2. Provide capability to tie in backup power for facility in a timely manner.

Reliable, timely backup power source for CVS, fire protection dry pipe air via station air compressors, and balance of facility loads.

3. Execute MSRE facility outage as soon as possible.

4. Assign a dedicated, independent electrical team to evaluate facility and develop corrective actions for safety and critical system needs.

5. Obtain spares for overcurrent protection devices and motor starters required for facility operation.

This will potentially require DOE approval for non-NRTL-approved electrical components.

6. Ensure electrical PM program is in compliance with regulatory and contractual requirements.

PMs for electrical components required for facility operation.

Arc flash calculations and labelling.

7. Label electrical components appropriately in accordance with UCOR procedure PPD-FO-1036, Conduct of Operations Program Description.

Example: #1 and #2 CVS fans start/stop buttons on their respective MCCs.

6.1.3 Life-Extension Upgrade Recommendations The MSRE electrical system is required to support facility operations described in the DSA and for life safety systems (LSS) in MSRE Complex buildings. The current electrical system at MSRE is very unlikely to operate reliably through 2060. Parts obsolescence issues will continue to be a concern as the facility ages, and increased failure rates should be expected. Costs associated with maintaining aging and obsolete electrical equipment will increase over time.

Based on the teams assessment of the electrical system and the extent to which it is currently relied upon in the DSA, the following recommendations should be implemented to support facility operation through 2060:

Design and install a dedicated electrical distribution system to support future salt disposition options and facility continued operation to 2060. Systems of concern include:

CVS RGRS Support systems (cranes, air compressors, etc.)

Fire protection and LSS Air gap electrical feeds associated with unneeded loads.

6-4

6.1.4 References PPD-FO-1036/R2. Conduct of Operations Program Description, 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

6.2 CONTAINMENT VENTILATION SYSTEM DESCRIPTION The purpose of the CVS is as follows:

1. Ensure contaminants in containment cells and confinement enclosures do not migrate into the operating areas, offices, or the environment

2. Provide fresh air for the operating areas

3. Provide a safe means for releasing process gases to the atmosphere to protect workers.

The CVS includes the 100-ft. stack, stack fans, high-efficiency particulate air (HEPA) filter array, and vented enclosures of the MSRE complex. The CVS vents the containment cells, such as the Reactor Cell, CBC, and FDT cell; and the vented enclosures, such as the RGRS glovebox and the six-valve manifold enclosure. The CVS also provides exhaust pathways for the treated RGRS gas stream and the charcoal canister. Before release into the CVS, potentially hazardous gases from the charcoal canister offgas are diluted with helium.

The CVS uses an array of differential pressure indicators to monitor the direction of flow. Manually operated dampers are used to adjust the pressure between ventilated areas. Exhaust from these ventilated areas is taken through ducts and is split through three arrays of seven roughing filters and six Nomex HEPA filters before being sent out the stack to the atmosphere. The CVS has two exhaust fans that operate at approximately 16,000 ft3/min of airflow at 6 in. water vacuum. Only one fan operates at a time; the other fan is on standby should the primary fan fail. If the operating fan fails, the standby fan must be started manually.

Differential pressure gauges monitor the pressure drop through the HEPA filters. If a drop in pressure is detected, sensors signal the system when the filter vacuum reaches 1 in. water vacuum that there is a possible loss of ventilation. The loss-of-ventilation (LOV) alarm sounds over the MSRE intercom system and alerts the operator stationed in the Waste Operations Control Center (WOCC), since the MSRE facility is not staffed on a 24/7 basis.

The majority of the air flowing through the CVS is ambient air from the supply air filter house. The supply air filter house contains steam coils that heat the air during the winter before drawing it into the High Bay.

The CVS has installed radiation monitors and gas sampling equipment. Alpha, beta and gamma levels are transmitted to a local read-out in MSRE, and UT-B analyzes gas samples for radon three times a week.

Additionally, radiological particulate and tritium analyses are performed every other week. These analyses are compiled with other analyses from the ORNL site to calculate the offsite dose to comply with the air permit.

6.2.1 Current Condition Evaluation Based on the walkdowns of the CVS and discussions with the systems engineer and operations personnel, the following issues were identified:

6-5

Fan #1 was not operational.

Fan #2 has suspected bearing issues.

CVS is operating without a backup fan.

The fans do not have a backup power source.

CVS pulls air from the South Truck Bay into the 7503 Observation Room, which could be an issue if there was a release in the South Truck Bay.

During the report period, the following actions were taken:

Fan #1 had a new base installed.

The most pressing concern during the early weeks of the report period was the condition of both fans. Fan

1. 1 was non-operational because the motor stand was not adjustable, and Fan #2 was experiencing bearing issues at the time of writing this report. If Fan #2 had failed while Fan #1 was out of service, MSRE would have lost the CVS, which, under current JCO requirements (JCO-OR-MSRE-0117/R5), would require an evacuation from Bldg. 7503. The facility did replace the motor base of Fan #1, and the investigation of the bearing issues for Fan #2 is in progress.

In 2011-2012, there was an expansion of the Bldg. 7511 HEPA filter structure that allows the HEPA filters to be changed inside a ventilated containment area. According to UCOR-4162, the DOE recommends changing the HEPA filters every 10 years or when filters do not meet the efficiency standards. The last filter change was performed in 2012. While CVS is operating, two of the three filter arrays are kept active while the other array is on reserve.

According to BJC/OR-2671, a smoke test was performed in the South Truck Bay in 2007 to test the air flows in that area. The results of those tests showed smoke leakage into the RGRS observation room and the High Bay area. Based on this test, an administrative control to evacuate the entire 7503 building in the case of an HF alarm was implemented. The 2007 reports recommendation was to not seal all of the space between the South Truck Bay walls and adjacent areas because the work to both find the leaks and seal them would require extensive effort and might not be effective.

Surveillance of the CVS performance is done through regular checks of the pressure readings, as specified in PROC-MSRE-506, Routine Daily Checks at the Molten Salt Reactor Experiment Facility. Under the current practice, the MSRE operator performs daily checks (Monday through Friday) on the pressure indicators, temperature readings, alpha monitor readings, and other instrumentation. These readings are used to verify the operability of the CVS. In addition to daily readings, the stack fans are switched quarterly and the CVS alarm is tested every quarter. The LOV alarm is tested every quarter for audibility as well as once a year for functionality.

6.2.2 Near-Term Upgrade Recommendations Based on the teams assessment of the current condition and facility plans, the following recommendations should be implemented to support facility operations and maintain the safety envelope:

1. Provide a timely source of backup power for the CVS fans. This could be done with either a backup diesel generator for the facility or through a connection for a portable generator at the fans.

6-6

2. Immediately return both fans to operable conditions.

Fix the motor base of Fan #1. Completed 8/2016.

Investigate the bearing issue and repair Fan #2.

3. Purchase spare parts for the fan system.

4. Label the fan start/stop buttons in the motor control system.

5. Protect the vents and structural supports outside the Bldg. 7511 HEPA house.

6. Check the underground electrical connections and address deficiencies.

A near-term concern with the CVS is the lack of a backup power source. Under current operating procedures, if power to MSRE is lost, the CVS system shuts down, which then necessitates the evacuation of 7503. A plan should be developed and documented with all necessary information and facility modifications needed to support installation of a backup generator that could be readily rented from UT-B or another source when needed. This would allow backup power to be provided to the CVS system without having to maintain an installed generator at the MSRE site. The tie in for the generator could either be in the MCC or an immediate connection to the fans.

At the beginning of the MSRE Evaluation, both CVS fans were not in optimal operating condition. Later in the study, the MSRE facility team repaired the base of Fan Motor #1 and switched Fan #1 to be the primary fan, but work is still needed on Fan #2. Having both fans in working order is crucial to the continued reliability of the CVS. Under current safety basis requirements, a CVS outage requires the evacuation of the MSRE building, stopping all work in that area. Maintaining a supply of fan parts and motors for the system would also reduce the risk of a longer term outage.

Another near-term upgrade is to label the start and stop buttons for each fan in the MCC. The buttons are located directly beneath the breakers for the fans; however, there is no label in close proximity to the buttons. A label directly next to the buttons should eliminate possible confusion during CVS operation.

The duct supports outside of 7503 are in close proximity to a driveway for vehicles. To keep the supports and ducts from being accidentally hit by a vehicle, physical barriers, such as Jersey barriers, should be placed around the ducts and supports, or this portion of the roadway should be closed.

The electrical lines powering the CVS are in an unknown condition. Testing the condition of the wire insulation by a megger test would test the integrity of the underground electrical feeds to the CVS fan motors. If the wires are significantly degraded, new wires would have to be run to the fans.

6.2.3 Life-Extension Upgrade Recommendations Based on the teams assessment of the current condition and facility plans, the following recommendations should be implemented to support facility operations and maintain the safety envelope until the final D&D date:

1. Perform a current smoke test and address any deficiencies.

2. Evaluate elimination of the HEPA filter system.

To ensure CVS functionality during salt removal and final D&D of the building, a CVS functionality test should be performed. This analysis would include, but not be limited to, air flow testing, leak testing, and filter performance. These tests should evaluate the capability of the system to effectively remove airborne particulates in the High Bay and South Truck Bay that could be released during salt removal and D&D of 6-7

the facility. A smoke test reported in BJC/OR-2671, Molten Salt Reactor Experiment Engineering Evaluation of Passivation Cabinet, HF Cabinet, South Truck Bay and High Bay Ventilation Air Flow Patterns, expressed some concerns with ventilation of the South Truck Bay. Under current operating conditions at the time of the MSRE 2016 Evaluation, the South Truck Bay did not contain materials that needed to be ventilated, so ventilation issues were not a concern from this area. However, as stated earlier, CVS would have an important role in salt removal and D&D, so ensuring system integrity and functionality for future work is crucial.

Since defueling of the salt in 2008, the emphasis on RGRS gas treatment has shifted from capturing UF6 to treating the generated F2 gas and fluorinated byproducts generated in the salt. HEPA filters are designed to remove particulates from the air, but these filters would not capture any fluorinated gases that pass through them. Although the Nomex HEPA filters provide improved acid and HF resistance, the filters would not offer any hazard reductions to the system and could potentially be removed from the CVS. If the HEPA filters are determined to be unnecessary because the system does not produce any radioactive or hazardous particulates from any of the cells, then it would be cost effective to discontinue their use. This would eliminate the cost of planning the change, procurement of the filters, and changing the filters as well as maintenance of the HEPA filter shelter. Although the HEPA filters may not be required for RGRS operations, they may become necessary during salt removal. Careful consideration should be made about both the method of salt transfer and the new design or renovation to the RGRS system before removing the HEPA system from service.

The main action needed to prolong the service life of the CVS is to keep the current PM schedule. This system will most likely be running until final D&D, so keeping the system in good condition through regular checks and maintenance is essential.

6.2.4 References BJC/OR-1494/R4. Facility 7503 Containment Ventilation System Description, July 2007, Bechtel Jacobs Company LLC, Oak Ridge, TN.

BJC/OR-2671/R1. Molten Salt Reactor Experiment Engineering Evaluation of Passivation Cabinet, HF Cabinet, South Truck Bay and High Bay Ventilation Air Flow Patterns, January 2007, Bechtel Jacobs Company LLC, Oak Ridge, TN.

BJC/OR-2898. Active Confinement System Evaluation Summary Report for DNFSB 2004-2 Molten Salt Reactor Experiment Facility, Oak Ridge, Tennessee, September 2007, Bechtel Jacobs Company LLC, Oak Ridge, TN.

JCO-OR-MSRE-0117/R6. Justification for Continued Operation for the Molten Salt Reactor Experiment Facility Under the Documented Safety, 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

PROC-MSRE-506. Routine Daily Checks at the Molten Salt Reactor Experiment Facility, latest revision, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4162. Final Report for the Molten Salt Reactor Experiment Containment Ventilation System Filter Change in 2012 at the Oak Ridge National Laboratory Oak Ridge, Tennessee, September 2012, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4777. High Risk Review for Building 7503 Molten Salt Reactor Experiment (MSRE) at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, February 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

6-8

6.3 FIRE PROTECTION AND FACILITY LIFE SAFETY SYSTEMS The fire protection system is functional and relied upon as part of the defense-in-depth fire protection approach at MSRE. The MSRE complex does not contain any rated fire barriers. Fire suppression water is supplied from the potable ORNL water system, which is gravity fed from a water treatment plant north of the Y-12 National Security Complex (Y-12). Sprinkler systems are for life safety, ordinary industrial hazards, and property loss minimization, and are not credited in the fire hazards analysis (FHA). The Fire Protection Program is an Administrative Control described in Sects. 5.5 and 5.6.3 of the DSA. The sprinkler system is a combination of wet pipe and dry pipe. Plant air from the air compressors in Bldg. 7555 maintain pressure in the dry pipe system. In 2007, an Equivalency Request was submitted to and approved by DOE to address two (2) areas not covered by automatic sprinklers due to obstructions.

Fire detection and alarm systems at MSRE include the following:

Master fire alarm panel Auxiliary fire alarm boxes Evacuation horns Heat detectors Sprinkler system pressure switches Sprinkler system water flow switch Sprinkler system control valve supervisory switches Dry pipe low-air supervisory pressure switch Heat detectors annunciate on the master fire alarm panel and at the ORNL Fire Department. The fire suppression system has temperature-activated sprinkler heads.

According to Rev. 25 of the DSA, the MAR associated with MSRE are uranium, fluorine, HF, and fuel and flush salt. In general, the MAR is contained in robust structures such as the FDTs, NaF traps, and charcoal canister in the High Bay. MAR also exists in small quantities in waste containers and on process equipment (e.g., probes).

In the MSRE FHA, credible fires are postulated assuming that the sprinkler system is not operating.

Based on the analyzed combustible loading, current quantities and locations of MAR, and MSRE structures, postulated fires could not cause a breach of the NaF traps and charcoal canister. Chapter 3 of the DSA describes the consequences of radioactive events resulting from fires as being Low to Moderate.

Combustible control measures, such as transient combustible limits and operational controls, are in place to protect the MAR in the High Bay and charcoal bed cell.

MSRE is provided emergency response support from the ORNL Fire Department. Fire hydrant 8-11 is located immediately adjacent to the north side of Bldg. 7503, and secondary hydrants are located within 1,000 ft of the MSRE complex. The MSRE pre-fire plan is maintained by the ORNL Fire Department, and is updated as needed to address the hazardous conditions and tactical considerations for a fire at MSRE.

The fire suppression system undergoes routine inspections and testing including periodic supply system flow tests, internal sprinkler system checks, and destructive testing of sprinkler heads.

6-9

The FHA describes life safety measures and compliance with NFPA 101 Life Safety Code for the MSRE complex. Emergency lighting, interior building materials, egress pathways, and the occupant notification system are provided as appropriate for the MSRE buildings.

6.3.1 Current Condition Evaluation The team interviewed facility personnel and performed a detailed walkdown with the UCOR Fire Protection Lead and Fire Protection Engineer to obtain an understanding of the current condition of the fire protection and LSS at MSRE. Overall, fire protection and life safety equipment appeared to be in good condition with no significant deficiencies noted during the walkdown. The Fire Protection Lead provided an overview of system operation, described maintenance and testing practices, and identified key components for buildings with fire protection features in the MSRE complex. The walkdown included: 7503 general areas, 7503 High Bay, 7503 basement and service tunnel, 7503 remote maintenance control room, 7555 MCC building, and fire protection features on the exterior of the MSRE complex buildings.

During the walkdown, the team noted that combustible materials such as wood, paper, plastics, and an oil drip pan were in the High Bay and in the 840 ft. elevation of building 7503, where many breakers and electrical components are located. In addition, combustible control zone markings were not readily evident around the RGRS glovebox. The transient combustibles did not appear to exceed limits described in the DSA. However, the level of rigor in implementing the combustible control program requires improvement in operating areas of the MSRE complex.

The Fire Protection Lead noted that the master fire alarm panel for MSRE is nearly obsolete and in need of replacement in the near future.

Fire hydrant 8-11, which is located immediately adjacent to the north side of Bldg. 7503, is the primary means for the ORNL Fire Department to access the ORNL water supply system in the event of a fire at MSRE. Should this fire hydrant be used, pressure and flow at the five (5) hose stations and sprinkler locations throughout MSRE will be reduced below minimum requirements. According to the Fire Protection Lead, the ORNL Fire Department pre-fire plans take into consideration this limitation. During the course of this evaluation, hydrant 8-11 was taken out of service due to a leak across an underground valve. UCOR and UT-B are in the process of installing a new hydrant directly off the main Melton Valley header on the north west side of MSRE. The new hydrant will eliminate the above concern and enhance ORNL Fire Department response to MSRE due to the new proposed fire hydrant location on the Melton Valley header. This work is anticipated to be completed in October 2016, with the permanent isolation of hydrant 8-11 to follow.

6.3.2 Near-Term Recommendations Based on the teams assessment of the Fire Protection and Life Safety Systems and the extent to which they are currently relied upon in the DSA and FHA, the following recommendations should be implemented to support facility operation:

1. Address fire hydrant 8-11 design issuePromptly resolved by UCOR personnel

2. Provide capability to tie in backup power for facility in a timely manner Reliable, timely backup power source for CVS, fire protection dry pipe air via station air compressors, and balance of facility loads 6-10

3. Strengthen combustible control markings and level of rigor

4. Replace the master fire alarm panel 6.3.3 Life-Extension Recommendations The team has no additional recommendations to support the extended life of MSRE.

6.3.4 References DSA-OR-7503-0007/R25. Documented Safety Analysis for the Molten Salt Reactor Experiment Facility, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2014, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

FHA-OR-7503-0010/R1. Fire Hazards Analysis for Molten Salt Reactor Experiment Facility, Building 7503 Complex, Oak Ridge National Laboratory, Oak Ridge, Tennessee, 2013, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

NFPA 101. Life Safety Code Handbook, 2006 ed., National Fire Protection Association, Quincy, MA.

6.4 STRUCTURAL SYSTEM This part of the MSRE engineering evaluation is to determine the current state of the MSRE structure, and what short-term and long-term tasks would be necessary to ensure that the structure continues to meet structural design requirements throughout the extended period. The MSRE facility is classified as Performance Category PC-2. Per DOE Standard DOE-STD-1020-2002, Natural Phenomena Hazards Design and Evaluation Criteria for Department of Energy Facilities, PC-2 structures, systems and components are to be analyzed and designed in accordance with the International Building Code (IBC).

The original building and addition were designed in the 1950s. In 2002-2003 the MSRE steel superstructure was re-analyzed and determined to meet the then-current IBC seismic and wind loading requirements. The analysis is documented in calculation DAC-EA-020794-A018, NPH Evaluation of the MSRE Facility, Building 7503. The analysis model was based on available drawings; most of the structure was not accessible for inspection. The two cranes were included in that analysis, without loads and in their parked positions.

The MSRE structure is a safety-significant, credited design feature in the July 2014 version of the Documented Safety Analysis for the Molten Salt Reactor Experiment Facility, DSA-OR-7503-0007/R25.

The MSRE structure is credited for protecting materials stored on the operating floor of the MSRE high-bay structure. Those materials have been removed from the operating floor, and the most recent release of the DSA, Rev. 26, no longer credits the MSRE structure as a design feature. This is yet to be determined; however, whether or not this change to the DSA occurs, sometime in the future it will likely be necessary to credit the MSRE structure as a design feature to support salt removal.

The MSRE structure consists of a steel frame superstructure and a reinforced concrete substructure. The west half of the building is a high crane bay while the east half is a one-story office area. A cross-section of the high-bay is illustrated in Figure 6.2. Most of the structure is not currently accessible for inspection; therefore, this description is based primarily on available drawings.

6-11

Figure 6.2. Cutaway illustration of MSRE looking east (from DAC-EA-020794-A018).

The building measures approximately 159 ft north-south in eight bays (column lines [CL] CL-1 to CL-9),

and 83 ft east-west in three bays, CL-A to CL-E. The superstructure includes a High Bay industrial area from CL-A to CL-C, with an attached low bay office area from CL-C to CL-E. There are no internal columns on CL-B; only the wind columns at the ends of the building. The original building included CL-1 through CL-6. The later addition of CL-7 through CL-9 on the south end of the building includes the reactor and FDT cells. A floor plan of the High Bay area is shown in Figure 6.3.

The High Bay structure consists of nine east-west moment-resisting frames (CL-1 through CL-9).

Intermittent cross-bracing between columns provides stability in the north-south direction. Each frame consists of columns that support bridge crane rail girders as well as the building walls and roof. There are two bridge cranes on common rails; a 30-ton crane and a 6-ton crane. Built-up columns extend up approximately 24 ft from the floor level to support the crane rail girders. From that elevation, smaller single rolled columns extend to the roof. The roof parapet is approximately 43 ft above the floor level.

The roof structure consists of nine six-panel trusses built with L-sections welded to the columns via gusset plates, forming moment connections.

The built-up columns supporting the crane rail girders from CL-7 to CL-9 are heavier than the columns in the original high-bay structure. Additionally, the columns from CL-7 through CL-9 have moment connections into the concrete foundations at their bases; the columns from CL-1 to CL-6 do not have moment connections at their bases. Figure 6.4 shows the building frames from CL-1 to CL-6, and Figure 6.5 shows the heavier building frames from CL-7 to CL-9. It is evident that the high-capacity bridge crane was upgraded to 30-ton capacity when the south addition was built; and that the stronger structure at the south end of the building was designed to support the higher capacity crane.

1 2

3 4

5 6

7 8

9 6-12

The high-bay structural frames are intermittently cross-braced in the north-south direction, and the roof trusses are cross-braced horizontally. The outer walls are corrugated asbestos-cement panels attached to C-section girts. The roof deck consists of wide gypsum planks, 2 in. thick by 1 ft 3 in. supported on steel wide-flange purlins at 10-ft intervals. The roof slopes down to the east, toward the low-bay portion of the building.

There is a South Truck Bay between CL-8 and CL-9, and a North Truck Bay between CL-1 and CL-2.

From CL-2 to CL-9 the High Bay is enclosed by a 16-ga. steel liner on the walls and ceiling. There is a hatch in the ceiling of the south truck bay for crane access from above. The steel liner is attached to the interior sides of the building columns and the bottom chords of the trusses. The sheeting is supported between the main building frames by lightweight steel framing studs and channels.

There is a heavy, reinforced-concrete, shielded Maintenance Control Room elevated above the operating floor at the south part of the west wall, from CL-6 to ~CL-7.4. The room is supported by the concrete substructure, independent of the steel structure. The 3-ft-6-in. thick concrete walls contain shielded windows.

The low bay structure is a typical one-story braced frame consisting of wide-flange columns and beams.

The roof slopes down to the west, toward the high-bay structure. There are roof drains at the interface of the high and low bays. The parapet height of the low bay structure is approximately 14 ft.

There is a steel radiator stack bolted to a concrete foundation outside the south end of the facility. The stack is inactive. The radiator stack is braced to the south end of the High Bay structure. There is an active, steel-lined ventilation stack (the CVS stack) further south that is not structurally attached to the MSRE building, although it is connected through ducting.

The MSRE substructure consists of reinforced concrete foundations, grade beams, walls, cell vaults, and tunnels. The west wall of the substructure is partially exposed; all other parts are below grade.

The building columns are supported by concrete piers on spread footings. The concrete piers range in depth up to ~37 ft. See examples in Figure 6.6.

There is a reinforced concrete spectrometer room and tunnel below the office area floor at the southeast corner of the building. The spectrometer tunnel extends out beyond the south wall of the building.

The basement substructure of the building extends west of CL-A and south of CL-9 at the southwest corner of the building. This part of the substructure includes a ventilation tunnel and the foundation for the ventilation stack.

6.4.1 Current Condition Evaluation A scoping inspection of the high-bay structure was performed during the first week of August 2016.

As mentioned in Sect. 6.4, there is a sheet steel liner attached to the inside of the structural steel building frames. With the exception of the far north and part of the far south frames (CL-1 and CL-9), most of the high-bay steel structure is concealed between the inner steel liner and the outer building envelope. There is an access opening to the attic roof trusses, but there is not a safe way to physically inspect the trusses without installing walkways. It would require substantial resources and time to inspect a statistically-meaningful sample of the steel structural members and connections. The limited portions of the structure that are readily accessible were visually inspected as discussed below. Proposed means of conducting detailed inspections are discussed in the later sections of this report.

6-13

The key to inhibiting further degradation of the structure is preventing water intrusion. There were roof leaks in the past; however, the roof was repaired approximately 5 years ago by application of a polyurethane foam and elastomer sealant. The effects of past leaks were seen in the limited portions of the structure that could be inspected. No active roof leaks or other points of water intrusion were identified by the facility.

6.4.1.1 Steel protective coating deterioration Peeling and alligatoring protective coatings are common in the accessible areas of the High Bay structure.

This alone is not structurally significant; however, it does indicate a general lack of maintenance. If the structural steel is kept dry, deteriorated protective coating would not significantly impact the life of the structure. However, in the presence of moisture, deteriorated protective coating would accelerate oxidation of the structural steel.

6.4.1.2 Loose anchor rods A loose baseplate anchor rod nut was observed on Column C-6. There is an approximate 0.5-in. gap between the nut and the base plate. The nut on the other side is snug.

While not part of the superstructure per se, another loose nut was observed on the reactor shield anchor rods.

6-14

6-15 Figure 6.3. Floor plan of MSRE High Bay.

NORTH TRUCK BAY SOUTH TRUCK BAY

6-16 Figure 6.4. Elevation of building frames on CL 1 to CL 6 looking north (from Drawing 22-S2 Issue B, Framing Details).

14WF68 + 8WF24

6-17 Figure 6.5. Elevation of building frame on CL 7 to CL 9 looking north (from Drawing D-KS-19404 G2, Rev. 0 and Drawing D-KS-19404 C3, Rev. 0).

24WF76 +15C33.9 MOMENT CONNECTION

Figure 6.6. Construction of reinforced concrete piers supporting building columns.

Steel building column Concrete foundation pier on footing 6-18

6.4.1.3 Modification Figure 6.7 shows one side of a moment connection for crane/building column C-7. The opposite side of that column moment connection, shown in Figure 6.8, has been trimmed ~0.375 in., evidently by flame cut. It is not known if the modification was authorized; it is not shown on available drawings. This modification was evaluated by UCOR engineering after discovery and determined not to affect the integrity of the structure. UCOR Engineering Report RPS-027503-D943 was issued to document the engineering evaluation.

6.4.1.4 Corrosion Steel oxidizes from exposure to water and air. If the building envelope was maintained for the life of the facility, there should not be any significant rusting of the structural steel. However, it is evident from the small portion of the structure that was inspected that water has breached the 7503 building envelope in the past. Significant rusting was observed on several members and connections. The inspection was only visual. To determine the extent of deterioration of these accessible members, scraping, wire brushing, probing and/or measuring would be required. Several examples of corroded steel members were observed.

A corroded column baseplate and anchor rod were observed on a wind column of exterior wall of south truck bay.

Near the north east column of the North Truck Bay (C-1), it is evident that water penetrated the roof and caused rusting on the column, bracing and connections. Closer physical inspection would be required to determine the depth and extent of damage.

The moment-resisting column base connection located below an access hatch in the floor of the maintenance control room floor (Column A-7 or A-8) exhibited rusting when viewed from above.

The steel radiator stack outside the south end of the MSRE High Bay is connected to its foundation by a baseplate ring and anchor rods. The baseplate and anchor rods are rusted. Closer physical inspection would be required to determine the extent of damage. There is also some corrosion on the base connection of the active CVS ventilation stack. Also, the large carbon steel duct from the Reactor Cell and Drain Tank Cell has been buried underground in wet clay soil since 1958 without cathodic protection, so there is the potential for corrosion of this duct work.

6.4.1.5 Concrete substructure Most of the concrete substructure was inaccessible at the time of the initial scoping inspection. However, a large part of the exterior face of the west foundation wall is exposed above grade and appears to be in good condition.

There is an access ramp under the fan building into the MSRE substructure. Access was limited due to posted chemical contamination, but one corner of a reinforced substructure column and wall was visible.

The concrete in that area is severely spalled and the reinforcing corroded. Further physical inspection would be necessary to determine the depth of concrete degradation and steel section loss. This represents only a very small fraction of the substructure. Further inspection is planned for the near future.

6-19

Figure 6.7. Crane/building column C-7 moment connection east face.

Figure 6.8. Crane/building column C-7 moment connection west face trimmed.

6.4.1.6 Summary of initial inspection results A very small sample of the superstructure elements was accessible for inspection (estimated ~1%). An even smaller portion of the substructure was accessible.

Of the superstructure elements that were accessible, a significant portion exhibited issues that require closer physical inspection before an engineering opinion can be rendered regarding these elements. In 6-20

order to render an engineering opinion regarding the state of the overall structure, provisions would need to be made to examine structural members and connections concealed within the walls and above the ceiling.

6.4.2 Near-Term Upgrade Recommendations Given that maintenance of the structure and building envelope was neglected for a number of years, it is likely that some concealed parts of the structure have degraded, as was seen in some of the accessible members.

A baseline inspection of the structural components should be performed to determine the extent of degradation of the structure. Because of the nature of the facility, such an inspection would require significant resources, planning and preparation. Recommendations can then be made if repairs are deemed necessary to extend the life of the structure. Once the structure is deemed to be in good repair, further structural inspections should not be necessary, if the integrity of the building envelope is maintained. Periodic inspections should be performed on the roof and walls to detect any water intrusion and then repairs to minimize water intrusion should be performed immediately to minimize potential for further degradation of the structure.

Near-term actions are recommendations for performing comprehensive inspections of the structure and building envelope, and restoring components that have degraded. The items are listed in the recommended order of execution, mostly for practical concerns. With the exception of Item 1, below, there are no concerns of an immediate safety issue.

6.4.2.1 Exterior stairways The stairway going up to the Remote Maintenance Control Room exhibits significant corrosion, including on the handrails, making it an immediate safety concern. The facility responded by tagging out access to this stairway.

Other exterior steel stairways also exhibit corrosion, particularly at the connections of the stair treads to the stringers. For personnel safety concerns, all exterior stairs should be inspected immediately and repaired as necessary. At a minimum, rust should be removed from corroded connections and fresh protective coatings applied. If connections are structurally compromised they should be repaired.

6.4.2.2 Accessible roof trusses Inspect the roof trusses as much as possible from the access platform near Col. C-2 in the North Truck Bay. If rust is seen, wire brushing, scraping, and/or chipping will be performed to determine the extent of damage, if the locations are accessible for physical examination.

6.4.2.3 Accessible roof deck While inspecting the trusses, use lights to search for evidence of past roof leaks. Note such locations for future inspections. If evidence of active leaks is observed, then repair leaks immediately.

6.4.2.4 Accessible structural components Perform a detailed physical inspection of the rusted columns and anchorages, and connected braces and beams that were identified during the initial scoping inspection. Inspections will require wire brushing, scraping, and/or chipping. If section loss is evident, measurements will be taken. An aerial lift will be 6-21

needed to inspect elevated sections. Accessible steel is primarily exposed in the North and South Truck Bays, and below the concrete maintenance control room.

6.4.2.5 Stack inspections Inspect the base anchorages and bracing of the CVS stack and radiator stack. An aerial lift will be required. Inspect the underground Reactor Cell ventilation duct work with a camera to confirm no significant corrosion of the duct work has occurred.

6.4.2.6 Viewable west wall structure View parts of the steel structure through the windows of the west exterior wall, using an aerial lift.

6.4.2.7 Roof trusses, bracing, purlins, and deck The ceiling of the Containment Enclosure does not provide an adequate walk surface from which to safely perform inspections. Therefore, walking planks would be required to access the lengths of the trusses, and along the east wall (CL-C) and west wall (CL-A). Safety considerations might require guard rails along the walking planks.

Once access is established, the trusses, purlins, roof deck and bracing will be inspected for signs of water exposure and corrosion. If possible, the connections for the stack braces at the south end of the building will be inspected from within the attic space. Corroded areas, if found, would be physically inspected to determine the extent of damage, and recommendations for repairs will be made.

6.4.2.8 Building columns The top sections of the building columns can be inspected during the truss inspections. The trusses are welded to the columns via gusset plates. It may be possible to look down at least the upper portions of the columns from above.

A video camera may be lowered down the cavities between the columns for closer visual inspection. Any suspect areas, particularly where water intrusion is evident, will be noted for further inspection when access becomes available. If there is evidence of active water intrusion, then repair the leak immediately.

If there is no evidence of water intrusion or corrosion, it might not be necessary to physically inspect inside the wall cavities at those locations; i.e., it may not be necessary to perform items 9 and 10, below, if the camera inspections do not reveal any significant damage.

6.4.2.9 Building/crane columns and bracing, west wall Although the building/crane columns are concealed within the wall cavities, there are windows in the west wall of the High Bay. There is an estimated 18-in. space between the framing of the inner liner and the outer wall.

The windows are elevated well above grade on the west wall of the High Bay. A lift would be required for access. If the windows can be opened or removed, the inspector(s) can enter the space and perform inspections. If steel corrosion is found, a ladder may be necessary to perform physical inspections of elevated sections. This would be a confined space operation.

6-22

6.4.2.10 Building/crane columns & bracing, east wall There are no windows in the east wall of the High Bay. The roofs of both the High Bay and low bay slope toward the east wall (CL-C); therefore, water leakage is more likely on this side of the High Bay.

Columns to be inspected would be selected based on the inspection from above at the roof truss level.

Where no evidence of past or present roof leaks exists, a visual camera inspection should be sufficient unless corrosion is identified.

Lower portions of the columns could be exposed by removing interior gypsum board from the low-bay offices. Portions of several columns are currently exposed in the offices. Access to the mid portions of the columns, i.e., between the low bay roof and the High Bay roof trusses could be accessed by removing selected exterior wall panels. The panels contain asbestos and appropriate work controls would be necessary. If exterior wall panels are removed, the lower portions of the columns can be accessed by ladder.

6.4.2.11 Radiator stack bracing The steel radiator stack outside the south end of the High Bay is braced to the building structure. The bracing and connections should be inspected. Parts of the connections may be accessible from inside during the roof truss inspections.

At a minimum, a lift will be necessary to reach the braces. It may be necessary to loosen or remove wall panels to access the connections.

6.4.2.12 Miscellaneous supports The only miscellaneous supports that have been identified as important to the ongoing functionality of the MSRE are those supports associated with the CVS stack and ductwork. Some of those supports are mounted to the building structure and may require loosening or removal of exterior wall panels to gain access for inspection.

6.4.3 Life-extension upgrade recommendations As mentioned previously, once the integrity of the MSRE structure is established, ongoing intrusive structural inspections should not be required, provided that the steel structure is kept dry. Therefore, ongoing inspections should be performed to detect and correct leaks in the roof and walls of the building.

It is recommended the external and internal inspections of the building envelope be performed annually, and that leaks be remedied promptly.

6.4.4 References DAC-EA-020794-A018. Calculation: NPH Evaluation of the MSRE Facility, Building 7503, January 14, 2003, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

DOE/OR/01-2496&D1. Engineering Evaluation of Options for Molten Salt Reactor Experiment Defueled Coolant Salts, December 2010, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE-STD-1020-2002. Natural Phenomena Hazards Design and Evaluation Criteria for Department of Energy Facilities, 2002, U.S. Department of Energy, Washington, D.C.

6-23

Drawing 22-S2 Issue B, Framing Details, April 6, 1952, The Austin Company, Atlanta, GA.

Drawing D-KS-19404 C3, Rev. 0, Bldg. 7503 Alterations, Structural Cross Section & Details, 1955, Carbide and Carbon Chemicals Corporation, N.Y.

Drawing D-KS-19404 G1 Rev. 0, Bldg. 7503 Alterations, North Wall Framing, 1955, Carbide and Carbon Chemicals Corporation, N.Y.

Drawing D-KS-19404 G2, Rev. 0, Bldg. 7503 Alterations, Crane Bay Plan & Sections, 1955, Carbide and Carbon Chemicals Corporation, NY.

Drawing S-20794-EB-001-D-2, Rev. 3, Containment Encl. & Change Room Plan, 1981, Union Carbide Nuclear Company, Tuxedo, NY.

Drawing S-20794-EB-004-D, Rev. 0, Containment Encl.North Wall Framing Plan, Elev. & Details, 1964, Union Carbide Nuclear Company, Tuxedo, NY.

Drawing S-20794-EB-005-D, Rev. 0, Containment Encl.North Wall, Wall Liner, Elev. Sects. & Details, 1964, Union Carbide Nuclear Company, Tuxedo, NY.

Drawing S-20794-EB-110-D1, Rev. 1, Containment Encl.Framing Elev. East & West Walls, 1961, Union Carbide Nuclear Company, Tuxedo, NY.

Drawing S-20794-EB-111-D-3, Rev. 3, Containment Encl.Framing Elev. South Wall, Plan & Sect.,

1961, Union Carbide Nuclear Company, Tuxedo, NY.

Drawing S-20794-EB-112-D-1, Rev. 1, Containment Encl.Framing Sect. A-A Col. No. 2 through No. 6, 1961, Union Carbide Nuclear Company, Tuxedo, NY.

Drawing S-20794-EB-113-D-2, Rev. 2, Containment Encl.Framing Sect. B-B Col. No. 6 through No. 9, 1961, Union Carbide Nuclear Company, Tuxedo, NY.

DSA-OR-7503-0007/R25. Documented Safety Analysis for the Molten Salt Reactor Experiment Facility, July 2014, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

RPS-027503-D943. Engineering Evaluation of Structural Column C-7 Modification, August 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

6.5 CRANES The High Bay area has two bridge cranes, a 6-ton crane and a 30-ton crane. These cranes were used in past work packages in the High Bay for lifting shields and other heavy equipment. There is also a 3-ton auxiliary hook on the 6-ton crane bridge, but it is not currently operational.

6.5.1 Current Condition Evaluation The cranes undergo an annual functionality test along with monthly wire rope inspections. The 6-ton crane is currently in use for MSRE High Bay work, but the 30-ton crane is currently not operational. To return the 30-ton crane to service, a wire rope replacement is required prior to use. The 30-ton crane would also require a new bus bar since the bus bar was removed from the 30-ton crane and installed on the 6-ton crane. Also, either limit switches or bumpers must be installed to comply with American 6-24

Society of Mechanical Engineers (ASME) B30.2-2005 2-1.8.2. Under ASME B30.2-2005 2-1.8.2, cranes are required to have a device that would control damage from cranes contacting each other.

The 3-ton auxiliary hook on the 6-ton is out of service. At this time, there are no plans to bring the 3-ton into service.

6.5.2 Near-Term Recommendations After discussions with the rigging engineer and other personnel, the teams recommendations for the cranes are as follows:

1. Identify and purchase critical spare parts for the crane systems.

2. Replace either the limit switches or install new bumpers on the cranes.

3. Make the 30-ton crane operational.

Install a new bus bar on the 30-ton crane and purchase a backup for the 6-ton.

Wire rope replacement.

Two options for controlling crane contact during use are either installing new limit switches or new bumpers on the cranes. The decision on which path to take will be made after consulting with the crane vendor. The vendor will inspect the cranes and provide a recommendation for either installing the switches or the bumpers.

The bus bar on the 30-ton crane was removed and installed on the 6-ton crane to keep it operable. A new bus bar will have to be purchased to install in the 30-ton crane to return it to service. To install the new bus bar, additional electrical components will also have to be replaced.

There is a wire rope purchased for the 30-ton crane. It will be changed when the work package is scheduled and performed.

Maintaining the proper inventory of spare parts for the cranes is critical for continual crane functionality.

Having spare parts for the cranes on hand would limit down time significantly in the event of a component failure.

6.5.3 Life-Extension Recommendations Although the 30-ton crane is not operational, the current inspection and maintenance schedule for the cranes has kept the 6-ton crane operational and has also maintained a list of required repairs to return the 30-ton crane to operating condition. The inspection and maintenance programs should be continued to ensure reliable and operable crane performance.

6.6 AIR COMPRESSORS There are three MSRE air compressors housed in Bldg. 7555 on the west side of Bldg. 7503. The two primary compressors are recently installed Ingersoll Rand compressors. The third, manufactured by Joy, serves as the backup in case of the primary compressor failure. The compressors supply air through the two common air receivers that feed into the plant air system. All of the compressors are powered by the MSRE main power system. In the event of a power outage, compressed nitrogen gas bottles are used to maintain the air pressure in the system for up to 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.

6-25

The compressors supply air to pneumatic valves, such as those used in the process lines to the charcoal bed, RGRS, and drain/flush tank isolation piping systems. Most of the valves in the system are air to open (air-to-open) valves, except for six drain tank air-to-close isolation valves in the Drain Tank Cell; these can only be accessed remotely. In addition to providing air to these valves, the air compressor system also provides air to the dry pipe fire suppression system.

6.6.1 Current Condition Evaluation The operator round sheet includes daily air compressor checks that monitor pressure, run time, and temperature of the air outlet.

The two primary Ingersoll Rand compressors were recently purchased, and are running without issues.

The Joy compressor has been maintained in accordance to the PM program. To minimize wear, the dual-compressor system automatically cycles one unit as the running compressor while the other is on standby.

However, the air compressor system is currently operating ~90% of the time; this indicates system air leakage. The increased strain on the air compressors will lead to a shortened life and increase the chance of compressor failure.

All three of the compressors are currently operating on the main MSRE power. There is no backup power option at this time. In the event of a power outage, the air system relies first on the pressure in the compressor receivers, then the nitrogen gas bottles. During the planned power outage on September 16, 2016, the air system experienced a pressure drop of approximately 170 psi/hr. At that usage rate, the current gas bottle inventory can provide air for approximately 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. If power is not restored or more gas bottles obtained and installed after the current supply ran out, the pressure will drop in the lines and cause the FDT and FFT boundary valves to fail open. When these valves open, the fluorinated headspace gas flows into the piping downstream of the boundary valves. However, these lines do have pressure-monitoring capabilities, so this event can be monitored and controlled. Further discussion of a loss-of-air event is found in Sect. 4.2.2.

6.6.2 Near-Term Recommendations The following recommendation would better ensure the compressed air system availability and reliability:

1. To mitigate extended plant air loss, develop procedures and equipment to maintain continuity of the air supply after the current 24-h backup air supply is exhausted, either by restoring power to the compressors or obtaining sufficient gas bottles.

As stated earlier, losing power at MSRE would shut down both the primary and secondary air compressors. The current inventory of gas bottles will provide plant air for 24 h, but there is currently no provision for a more extended outagefor instance, in the case of a broken air line or catastrophic compressor failure. The evaluation team recommends that a plan should be in place to either restore power to the air compressor system with a backup generator; or obtain additional nitrogen bottles until plant air is restored. In addition, continuing the current PM program will be essential to keep the compressors in a good condition while they are needed by the facility.

6.6.3 Life-Extension Recommendations After discussions with the system SMEs, the following action would improve the functionality of the compressed air valve system:

6-26

1. Install manual valves downstream of the air-operated drain tank valves. The best way to increase the lifespan of the air compressor system is to decrease the overall load on the system. The best and most effective way to accomplish this is to replace the air-to-close FDT boundary valves with manual valves. The beneficial outcome of this is twofold. First, the air-to-close boundary valves may be consuming a large proportion of the system air generated by the compressors, in which case, eliminating these valves would reduce the amount of air required by the compressors. Secondly, the manual valves are more reliable and would not be dependent on the plant air system. The manual valves would also not be dependent upon the air reserve bottle supply to maintain the headspace gas boundary. Additional discussion of the FDT boundary valves can be found in Sect. 4.2.2.

6.7 CHARCOAL BEDS 6.7.1 Main Charcoal Bed, Auxiliary Charcoal Bed, and Charcoal Canister The CBC is a 10-ft-diameter, 23-ft-deep, circular concrete cell positioned below grade on the exterior southeast corner of the High Bay. The CBC was designed to house the filtration systems for the reactor offgases, the ACB and the MCB, during reactor operation and to mitigate a release of materials from the beds. While the reactor was operational, the ACB was intended to filter the offgas from the FDTs, and the MCB was designed to filter the MSRE sampling systems and the reactor fuel loop.

The ACB is made of two 6-in. diameter, 20-ft tall, schedule 10, stainless steel U-tubes (designated ACB-1 and ACB-2) that are located inside of the CBC structure. The ACB inlet is line 561; the outlet is line 562 which runs through a HEPA filter, then a radon trap, and out to the MSRE CVS system. The system is isolated through closing of connecting valves, such as valve 561 isolating the ACB from the FDTs and FFTs. Similar to the ACB, the MCB is made up of four U-tube cells, 1A, 1B, 2A, 2B, that are connected to four inlet lines, 620, 621, 622, and 623, respectively. These lines connect to closed valves in the vent house annex, V-624, V-625, V-626, V-627, that isolate the charcoal beds.

An engineered charcoal canister contains an estimated 4 kg of 233U along with contaminated charcoal from the ACB that was removed after the 1994 uranium migration event. The canister is an ASME 500-psig pressure vessel, with a 4.26-in. inside diameter and a 4-ft height. The canister is seated inside a 4.5-ft inside diameter and 6-ft-high protective concrete and steel outer cask that acts as shielding for the canister. The canister has a valve for venting the canister headspace gas that is protected by two shields to reduce the effects of potential release accidents. The valve is used to vent the canister through a detachable ventilation system. When the headspace gas is vented, it first passes through a charcoal filter, is diluted with helium, and then is routed through the RGRS system downstream of the traps and pumps and upstream of the pre-filters and HEPA filters before being released to atmosphere through the CVS.

The cask is also designed to lower the radiation field to less than 100 mrem/h at the surface. Recent surveys have determined the dose rate to be ~21 mrem/h on contact with the cask.

A nonflammable structure was erected on top of the reinforced concrete walls and the foundation of the CBC during the charcoal bed passivation work. This structure has the ability to be ventilated if the access plugs on top of the cell are removed during work. The primary safety basis concern with the charcoal beds is a fire. Controls, such as Jersey barriers around the vent house, have been put into place to limit the chance of a fire.

6.7.1.1 Current condition evaluation According to UCOR-4777, gas samples were taken in early 1994 from the pipe connecting the drain tanks and the charcoal beds. The sample analysis showed that there was a much higher than expected F2 6-27

concentration as well as the unexpected presence of UF6 in the lines. The MSRE procedure of heating the salt (annealing), done to theoretically restore the salt chemistry, was determined to have likely caused the combination of uranium in the salt with the fluorine gas to create UF6. The UF6 traveled through the line connecting the charcoal beds and the drain tanks into the charcoal beds through a failed valve, valve 561.

As described in DOE/OR/01-1623&D2, valve 561 is a bellows sealed valve, and it had failed in the open position when the threads on the upper bonnet were stripped on the bonnet seat. Once inside the charcoal bed, oxidizing fluorine gases (e.g., F2 and UF6) will react with activated charcoal to produce carbon-fluoride compounds. The major concern with these compounds is the possibility of producing gases that pose a deflagration risk when shocked or exposed at an elevated temperature.

Discovering the uranium migration started a project to passivate the charcoal beds. Before passivation of the charcoal beds, structural modifications would have to be made. These modifications were:

1. Remove water from the CBC.

2. Isolate the offgas system.

3. Isolate the CBC.

4. Eliminate other water sources.

5. Initiate the removal of the reactive gases from the offgas system.

The project successfully removed the water from and isolated the CBC, drained water from the other areas, such as the reactor shield annulus and the vapor condensing tanks, and installed the RGRS system to remove the remaining reactive gases in the lines. One of the more relevant actions from the past projects to this evaluation was sealing valve 561. The valve was sealed by using a steel cable to compress the valve stem to ensure the valve remained closed; this valve remains the isolation boundary mechanism for sealing the FDT headspace from the ACB.

The next phase of the charcoal bed treatment was to passivate the carbon fluoride compounds. As stated in ORNL/TM-13506, ammonia gas was found to be the most effective reagent to convert the fluorinated charcoal into a stable material that could be removed safely. Once the charcoal was passivated, it could be mined out of the ACB and placed into a canister for storage at MSRE.

Following the passivation of the charcoal beds, a project in 2001 removed an estimated 80% of the uranium-contaminated charcoal from the ACB and placed it into a separate canister and the charcoal beds were isolated after the uranium migration. NDA testing performed in 2001 estimated the uranium levels in the charcoal beds and canister to be as follows:

ACB contains 830 g of 233U.

MCB (cells 1A, 1B, 2A, 2B) contains 500 g of 233U.

Charcoal canister contains about 4 kg of 233U.

During the report period, scoping work was started to sample the headspace gases in the MCB, ACB, and charcoal canister. These samples will help to determine if there is continuous gas production in the headspace, as well as the species that make up the headspace gas. The results of these tests will help to determine future actions needed to reduce any potential risks from the beds and the charcoal canister.

During the contaminated charcoal removal, some alterations were made to the ACB system. The top of the ACB was removed and replaced by a lead cap. This cap is not airtight, and it is inside a structure ventilated by the CVS.

6-28

The MCB has an estimated 500 g of 233U. The MCB was sealed off from other systems after the uranium migration by closing the valves. There is not a current system for measuring either the pressure in the MCB or what gases make up the headspace. The MCB is not vented by the CVS like the ACB.

As a result of the charcoal passivation, the charcoal canister is currently generating both H2 and N2 gas through radiolysis, causing the pressure to increase inside the canister. To counter the continual rise in pressure, the canister is vented when the pressure inside the canister approaches 50 psig.

The charcoal canister and shielded cask are considered to be safety significant design features in the DSA, and they are credited in the TSR. The charcoal canister disposition date is fiscal year (FY) 2039.

Accelerating the removal of the canister has been examined, as seen in UCOR 4669, MSRE Charcoal Canister Disposition, Oak Ridge, Tennessee. This paper outlined six options for either continuing storage of the canister or treating the charcoal and disposing of it as waste. The recommended options outlined in the white paper are:

1. Move the cask into a cell in the High Bay.

2. Treat onsite with nitric acid.

3. Treat onsite with oxidation.

4. Furnace oxidation at Y-12.

5. Relocate to Bldg. 3019 for storage and eventual processing.

6. Leave the charcoal canister and shielded cask in place.

The recommendation given in UCOR 4669 is to leave the canister in place; however, this is only a temporary measure that does not reduce the risk the contaminated charcoal poses to the facility.

6.7.1.2 Near term recommendations After discussion with the facility SMEs, the following list gives the short term actions needed for the charcoal beds and the charcoal canister.

1. Develop and execute a plan to move the charcoal canister to a more appropriate facility (e.g.,

Bldg. 3019).

2. Sample and analyze the headspace gases in the ACB, MCB, and the charcoal canister.

Analyze the headspace gases using FTIR, if possible by routing the gas through existing lines.

or Obtain a sample for lab analysis. Work planning was initiated to use MCL to draw gas samples through their sampling cart.

3. Replace degraded Jersey barriers for the CBC.

Of the options presented, the team recommends sending the canister to Building 3019 based on the significant facility modifications required to safely process the material in the High Bay and also the ongoing 233U material storage in Bldg. 3019. Moving the canister to a facility that has a more stringent security and storage system would be beneficial for the project while also significantly reducing the MAR in the facility. In addition to these immediate effects, moving the canister would also be beneficial to the long-term disposal path. The uranium contaminated charcoal cannot be treated inside of the High Bay without extensive upgrades to the facility, so the canister would likely have to be transported to a different 6-29

facility for treatment in the future. If the canister could be moved to a facility that has the capability of treating the charcoal at a future time, the canister should be moved as soon as possible to lessen the cost of operation at the MSRE site.

Since the uranium-laden charcoal was removed in the late 1990s, very little has been done with the charcoal beds. The ACB is thought to be in a stable condition because it is vented through the CVS, but the MCB is isolated from all other systems. In order to verify that both of the beds are currently in a stable state, samples should be taken of the headspace gases to analyze the composition. Compositions of the headspace gases for the ACB and MCB are important given the potential for generation of shock-sensitive compounds. There is a possible piping path that will allow the charcoal bed headspace gases to be analyzed inline by FTIR. If that path is either blocked or contaminated, samples could be collected from sample ports located in the vent house and analyzed by a laboratory.

As stated in the system description, the Jersey barriers are used to protect the CBC from being hit by a vehicle. The current Jersey barriers do not have lifting points for a forklift and the wooden cribbing has deteriorated, so replacing the barriers with new barriers would provide a safer solution for protecting the CBC.

6.7.1.3 Life-extension recommendations The following recommendation is presented to reduce the level of risk associated with the charcoal beds and charcoal canister:

1. Based on the headspace samples, one of the following actions should be taken:

a. If headspace gas samples show no hazardous constituents, air-gap and vent the charcoal beds for complete isolation. Routinely monitor the bed pressure and the headspace gas composition to identify any unusual gas generation until isolation completed.

or

b. If hazardous gases are found, address treatment of charcoal beds.

The best option to reduce the long-term risk to the facility is to isolate the beds to prevent any unwanted migration of uranium or other hazardous materials to the charcoal beds as a result of valve failure and leak through. Valves and lines that are unnecessary should be air gapped. The headspace gases should be tested occasionally to check gas compositions and levels.

6.7.2 References DOE/OR/01-1623&D2. Removal Action Report on the Molten Salt Reactor Experiment Time-Critical Removal Action at Oak Ridge National Laboratory, Oak Ridge, Tennessee, December 1997, Lockheed Martin Energy Research Corp., Oak Ridge, TN.

ORNL/TM-13506. Passivation of Fluorinated Activated Charcoal, October 1997, Oak Ridge National Laboratory, Oak Ridge, TN.

ORNL/TM-2000/296. Conversion of Uranium-Containing Materials Retrieved from the Molten Salt Reactor Experiment into Stable Oxides for Final Storage or Disposition: Results of the First Full-Scale Prototype Test of Burning of Uranium-Laden Charcoal in Oxygen, 2001, UT-Battelle, LLC, Oak Ridge, TN.

6-30

UCOR-4669. MSRE Charcoal Canister Disposition, Oak Ridge, Tennessee, January 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

UCOR-4777. High Risk Review for Building 7503 Molten Salt Reactor Experiment (MSRE) at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, February 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

6.8 STEAM The heat for the MSRE buildings is supplied by steam provided by UT-B through the steam plant at Bldg. 7530, located on the opposite side of HFIR Road. The steam enters the MSRE site through two connections, one on the back side of Bldg. 7516, and the other running to a header outside of the 7509/7503 breezeway. The steam then flows to the various heaters for the High Bay, Bldg. 7503, and Bldg. 7509.

The High Bay is heated by a stack of three steam coils located in the air intake on the northeastern corner of the High Bay. Air is drawn in by the negative pressure created by the CVS. The air then passes over the coils, through a duct, and into the High Bay. Heated air in the High Bay is not only for worker comfort, but is also critical in keeping the fire protection sprinkler pipes in the attic space from freezing.

6.8.1 Current Condition Evaluation Most of the steam lines are original to the building. There have been some recent upgrades to leaking sections. Walkdowns of the system and discussions with facility operations identified leaks in the High Bay and the MSRE basement. The valves outside the breezeway are in need of repair. The main isolation valve needs the packing replaced, and the handles on both of the double isolation valves need to be tightened. The facility has planned to repair the steam lines before the steam is needed.

In winter or other times of high steam demand, there is a possibility that the steam demand for Melton Valley may become too great for the UT-B system to handle. In such an event, the supply to the MSRE buildings will be cut first. The greatest hazard from the resulting heat loss will be the possibility of the sprinkler lines freezing in the High Bay attic. There is a backup heating system for the CVS air at MSRE; it uses a diesel heater to heat the air drawn into the air intake house until the steam supply is restored. The current plan for steam loss is to use the diesel heater to heat the CVS intake air and the air intake house.

After consulting with fire protection SMEs, continuing to rely on this combustible fuel source for 40+ years poses a fire risk from human error or equipment failure. The alternative to the diesel heater is to install electric heaters in the attic space of the High Bay to keep the sprinkler pipes from freezing. These heaters would be turned on only in the case of steam loss to the MSRE facility.

6.8.2 Near-Term Recommendations The evaluation team has the following recommendations for the steam system:

1. Fix the steam leak on the mezzanine in the High Bay.

2. Fix the valves on the steam connection at the 7503/7509 breezeway.

6-31

6.8.3 Life-Extension Recommendations Besides regular inspections and repairs, the action that could prove to be the most beneficial for life extension is as follows:

1. Install electric heaters in the High Bay attic space to provide heat in case of steam loss.

6-32

7. WASTE MANAGEMENT PROGRAM 7.1 PROGRAM DESCRIPTION The waste stored at MSRE is managed by UCOR in accordance with the requirements of PPD-WM-2400, UCOR Waste Management Program Plan, which implements the requirements of DOE Order (O) 435.1, Radioactive Waste Management. The program plan requires newly generated waste to be shipped within 180 days of container closure. Legacy waste, defined as waste generated prior to August 1, 2011, is exempt from this requirement. Specific to MSRE, The Waste Handling Plan for the Molten Salt Reactor Experiment Remediation of Secondary Low-Level Waste Under the Melton Valley Closure Project at Oak Ridge National Laboratory, Oak Ridge Tennessee (DOE/OR/01-2256&D1) provides the execution plan for managing and disposing of waste from the reactor operations and defueling activities performed under the Record of Decision for Interim Action to Remove Fuel and Flush Salts from the Molten Salt Reactor Experiment Facility at the Oak Ridge National Laboratory, Oak

Ridge, Tennessee (DOE/OR/02-1671&D2). The waste handling plan (WHP) includes the list of specific waste at MSRE including container identification, material description, planned disposal site, characterization status and target disposal date.

There is a smaller subset of waste that does not require approval by the regulators in accordance with the WHP. The salt waste stored in FDT-1, FDT-2, and the FFT is not included in the WHP. The MSRE record of decision (ROD) for salt remediation currently identifies a start date of 2032 and a finish date of 2038 in Federal Facility Agreement (FFA) Appendix J for the salt waste.

7.2 CURRENT CONDITION Interviews and a walk down of the MSRE facility were conducted with the SME for waste management and identified the following conditions:

List of waste included in the WHP was well organized.

Characterization and shipping activities with a high priority were on schedule.

Less prioritization was evident for those waste containers not covered by the WHP.

Some waste items (e.g. in Grinder Room) did not appear on either of the lists provided.

Large volume of wood was stored outside at the northeast corner of the High Bay for several years.

Ten large casks (~23,500 pounds) that were never used for salt canister storage stored outside.

Several studies have been performed to identify options for removal of the salt waste currently stored in the FDTs 1 and 2 and the FFT. Many of these studies have assumed that the waste will be transuranic (TRU) waste; however, no formal TRU waste determination has been completed for the salt waste.

Evaluations performed by Los Alamos National Laboratory in 2010 and 2013 concluded the MSRE salt was eligible for disposal at WIPP. The evaluations looked at the radiological, chemical and physical characteristics of the salt and demonstrated that the MSRE salt can be considered TRU waste. The evaluations also describe why the MSRE salt should not be considered as spent fuel or High Level Waste, neither of which can be disposed of at WIPP. In the opinion of the evaluation team, the rationale described in the evaluations is technically correct and reasonable.

7-1

7.3 NEAR-TERM RECOMMENDATIONS

1. Initiate process to complete formal TRU waste determination and defense waste determination.

2. Obtain crane and lift plan to support characterization of the contaminated probes and ship probes for final disposal.

3. Ensure that the working waste inventory lists include all facility waste containers as appropriate.

4. Increase priority to ship out large volume waste (casks and wood inventory).

7.4 REFERENCES

DOE Order 435.1, Chg. 1. Radioactive Waste Management, August 2001, U.S. Department of Energy, Washington, D.C.

DOE/OR-1014, Federal Facility Agreement for the Oak Ridge Reservation, 1992, U.S. Department of Energy, Washington, D.C.

DOE/OR/01-2256&D1. The Waste Handling Plan for the Molten Salt Reactor Experiment Remediation of Secondary Low-Level Waste Under the Melton Valley Closure Project at Oak Ridge National Laboratory, Oak Ridge Tennessee, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR/02-1671&D2. Record of Decision for Interim Action to Remove Fuel and Flush Salts from the Molten Salt Reactor Experiment Facility at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1998, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

LA-UR-10-07278. TRU Waste Determination of MSRE Salt for Disposal Eligibility in WIPP, 2010, Los Alamos National Laboratory, Los Alamos, NM.

LA-UR-13-22367. Evaluation of Molten Salt Reactor Experiment (MSRE) Remnant Salt Properties for Disposal Eligibility at the WIPP, March 2013, Los Alamos National Laboratory, Los Alamos, NM.

PPD-WM-2400/R4. UCOR Waste Management Program Plan, May 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

7-2

8. SALT REMOVAL OPTIONS EVALUATION 8.1 SALT REMOVAL

SUMMARY

Various detailed studies of salt removal and disposal options have been performed with the latest comprehensive study completed in 2010 (reference DOE/OR/01-2496&D1). Conclusions at that time were to leave as is and maintain requisite S&M of the MSRE facility, with a final D&D target date of 2060. However, because of recognized degradation of facility components and consequent challenges to the safety envelope, the 2016 MSRE Engineering Evaluation team was asked to evaluate options for accelerating salt removal/disposition.

As discussed in Sect. 7, previous studies examined the acceptance of the MSRE salt for ultimate disposal at WIPP. The team reviewed these previous studies and concurred that the salt meets TRU and defense waste requirements for WIPP disposal. However, formal approval of the TRU Waste Determination by the DOE Oak Ridge Office (DOE-ORO), DOE Carlsbad Field Office (CBFO), and the headquarters of the DOE Office of Environmental Management (DOE-EM HQ) is necessary. This report assumed disposal at the WIPP in evaluating all options.

The team, assisted by UCOR and AECOM D&D SMEs, evaluated the primary options from the 2010 report, and also developed a chemical dissolution option. Each of these options was then ranked, based on a plus/minus review as well as scoring in areas of technical risk, safety of the process, cost (through disposition), and regulatory barriers.

8.2 SALT REMOVAL OPTIONS ANALYSES The five options evaluated, presented in priority of team recommendation, were as follows:

1. Tank removal/sectioning

2. Thermal salt transfer

3. Chemical dissolution

4. Mechanical removal with needle scaler

5. Entombment in place An earlier option of whole tank disposal was dismissed by this team as an option since packaging and transport of such a large, high-radiation source was deemed impractical.

8.2.1 Tank Removal/Sectioning This option was evaluated to be the top recommendation because it leaves no heel (reducing future D&D complexity), utilizes proven remote handling D&D technologies, avoids heating or chemical additions to the waste, and should offer a reasonable path to regulatory and WIPP waste acceptance criteria (WAC) acceptance. A change to the ROD will be required since this is a change from the accepted thermal transfer option; however, the fact that there will be no heel remaining should be considered a significant positive from the regulatory standpoint.

Fission products in the salt produce a high-radiation environmentover 1,000 R/h internal to the salt, and 25 R/h within 1 ft of the tank exteriorthat requires use of remote tooling for isolation of the tanks from 8-1

structural members and inlet/outlet process lines within the Drain Tank Cell. Such tooling is relatively common in both DOE and commercial applications. Once the tanks are isolated, existing cranes in the MSRE facility are capable of lifting them into shielded containers within the High Bay. Removing the tanks from the Drain Tank Cell is necessary because of the very limited working space in the cell.

The most difficult challenge of this approach is sectioning the tanks and salt into segments that will fit into WIPP-acceptable containers. The high radiation, radioactive contamination, and continued offgassing of fluorine will demand an enclosure that is shielded, inerted, and ventilated via HEPA filtration.

However, although the high radiation rates and the high contamination levels from the tank cutting provide challenges for this option, the remote tooling and filtration capabilities have been demonstrated through D&D of facilities with increasing success over the last few years. Depending on the ultimate size and complexity, such a shielded remote cell might be possible within the MSRE High Bay. If not, a facility adjacent to the MSRE, with the already-filtered ventilation exhausted through the MSRE CVS, may be appropriate. Alternatively, another facility on the Oak Ridge Reservation (ORR) can be considered, but transport of the high-radiation vessel would be a challenge.

The continued fluorine offgassing would be addressed by addition of sodium iodide (NaI) or another fluorine scavenger (getter) material to the waste containers.

The tank removal/sectioning concepts were developed in conjunction with D&D SMEs and determined to be technically feasible. However, no detailed conceptual design was performed, and such a design would take significant time to develop. Relative costs for this option were considered to be the same or less than the thermal transfer option (Sect. 8.2.2). The most significant cost for whole tank removal/sectioning will be the deployment and operation of the shielded remote cell and equipment. Maintaining the salt in solid form provides a significant risk reduction versus high-temperature operations and chemical addition processes involved in the thermal transfer or chemical dissolution options.

8.2.2 Thermal Salt Transfer Of the proposed salt transfer methods, thermal transfer of the salt is the most researched and tested. This method involves melting the salt inside of the FDTs and FFT, and transferring it into disposal containers.

The original design of the facility included various piping paths that could be used to transfer the salt from the drain tanks into disposal containers. During reactor operation, salt was transferred between the reactor primary loop and the tanks via pressurization with helium gas. In addition, pressurized thermal transfer was shown to be effective in 1999, when the coolant salt tank was drained by pressurization with nitrogen and the salt was moved into disposal containers.

As described in a 2010 evaluation of salt disposal options (DOE/PR/01-2496&D1), the salt was defueled between 2004 and 2008. For this project, the tanks were first heated with the existing tank heaters, and then the salt was ultimately melted with a probe. Hydrofluoric acid and hydrogen were introduced to the tanks to restore the salt chemistry (fluorine deficit), and then F2 was pumped into the molten salt to form UF6, which was then captured and dispositioned. After uranium removal, the salt was to be transferred into disposal containers by pressurizing the tanks to 50 psig to push the salt through transfer lines, some of which were original to the MSRE facility. During the initial attempt to transfer salt from the FFT, the transfer line developed a blockage that could not be cleared. Attempts to thermally transfer salt from the FDTs were not made. A formal cause analysis was not performed; however, facility personnel present during the salt transfer attempt have indicated that inadequate heat trace of the transfer lines was the most likely cause for the blockage.

In the 2010 evaluation discussed above, a method for salt transfer was proposed that would eliminate the need to use the existing salt transfer lines. This method would utilize a probe that includes a melting 8-2

element, pump suction line, and an agitator. The tank heaters would be used to bring the salt temperature just below melting, and then the heating element would complete the melting, as was done for defueling the salt. The agitator would be used to mix the salt to create a near homogeneous liquid. It is anticipated that the elemental metals created during past processing would either adhere to the bottom of the tank and be left as a heel or be mixed in suspension and pumped out with the molten salt. An additional advantage of the above method of thermal salt transfer is that the process is not reliant upon one dedicated transfer line for each tank, which proved to be a single-point vulnerability in the previous salt transfer attempt.

That is, should the pump suction line develop a blockage, the assembly could be removed, a spare installed, and the process could continue. It should be noted that the pumped thermal transfer method proposed above takes into account some of the lessons learned from the failed thermal transfer of FFT salt. The above design is conceptual and requires additional development.

As discussed in Sect. 4, the integrity of the FDTs and FFT is not fully known because of the significant difficulties in being able to determine the condition of the tanks. Ultrasonic testing, eddy current testing, borescope, and visual inspections are required to provide an understanding of extent of corrosion. A significant limitation to thermal transfer of the salt is that embrittlement of the FDTs, FFT, and thimbles can only be determined through destructive analysis, which is not feasible. Due to the above tank integrity concerns, previous failed thermal transfer, and the availability of an alternate thermal transfer method, pressurized thermal transfer is not likely to be the preferred means for ultimate salt disposal. If thermal salt transfer is deemed to be the preferred disposal method, pumped thermal transfer would be a superior option.

In the 1998 ROD for interim action at MSRE, defueling and thermal transfer were the selected means to prevent a criticality incident or release of contaminants to the environment. The ROD for MSRE salt has not been revised since 1998, which is an advantage for thermal transfer over other salt disposal options. In addition, recent studies performed by Los Alamos National Laboratory (LANL) on behalf of DOE have demonstrated that the defueled salt in the FDTs and FFT meet the definition of transuranic (TRU) defense waste and would be eligible for disposal at WIPP. The LANL studies also apply to mechanical and chemical salt disposal options; however, acceptable activity density for the non-thermal transfer options will have to be demonstrated, as has been done for thermal transfer. As with other salt disposal options, the RH-72B shipping container could be used for thermal transfer, which is an advantage from a regulatory and cost perspective.

Other benefits of thermal salt transfer relative to other salt disposal options include the following:

Personnel dose likely more favorable than mechanical disposal methods Waste generation more favorable than mechanical methods No particulates generated Fewer salt disposal containers In addition to tank integrity concerns, disadvantages with thermal salt transfer are:

Tank heaters required to be functional Gas management system is required Heel will be left in tanks after thermal transfer High temperature and energy process 8-3

The existing tank heaters have not been maintained since the defueling effort was completed, and the state of the electrical feeds to the heaters within the Drain Tank Cell is unknown. Should the existing heaters be beyond repair, thermal transfer may become less favorable. With regard to gas management, the RGRS will have to be upgraded or replaced to ensure a reliable gas management system is in place during thermal transfer.

8.2.3 Chemical Dissolution Because of questions on the drain tank heating systems and tank integrity being capable of supporting melting and transfer, chemical dissolution and transfer was explored. The concept involves introduction of the solvent to dissolve the salt and pumping the solution through the 3-in. center port of each drain tank. The resulting solution would be pumped into interim containment for drying and/or treatment, then to final containers to meet WIPP waste acceptance requirements.

The positive aspects of this option include avoidance of heating/melting with consequent safety issues and stresses on tank components; the likely ability to use similar containers as designed for the melting option; no particulates generated, limiting radioactive contamination, as occurs in mechanical options; and likely leaving less heel than the melting option.

However, negative aspects include development and deployment of a new solvent/pumping technology; probable need for adding a neutron poison to address criticality concerns; additional volumes of waste generated by solvents (with a follow-up drying technology); and potential WAC issues caused by the addition of solvent, neutron poison and drying agents.

Cost and regulatory barriers were judged to be in the middle of the group of options (Table A.2).

Because of the uncertainties on solvent selection and waste acceptance challenges, this option did not offer clear advantages over the higher ranking tank removal/sectioning and melting options.

Cursory evaluation of potential solvents for salt dissolution was performed to provide a framework for further evaluation, if warranted. Key factors which must be addressed are solubility of the salts in the solvent system, criticality concerns introducing liquids into the drain tanks, and final processing of the resulting solutions prior to waste disposition. Introduction and removal of solutions from the drain tanks would be similar to the melting option. A major drawback of this approach is the significant increase in waste volume that will be generated. A viable solvent system must be identified for this approach to be plausible.

The fuel salt components at MSRE have a wide range of solubilities in water: BeF2 is very soluble, ZrF4 is sparingly soluble, and LiF exhibits poor solubility (Table A.1). The dissolution of the BeF2 fraction of the salt will have an impact on the solubility of LiF. It may enhance the solubility of the LiF due to the increase in the fluoride concentration in solution and the propensity for small fluoride-containing species to undergo polymeric complex formation. Introducing water alone into the drain tanks raises criticality concerns due to the residual uranium and plutonium in the salts.

A boric acid solution would introduce a neutron/reactivity poison to the system, and could provide additional benefits to salt dissolution. Boric acid reacts with fluoride to form boron trifluoride (BF3).

Boron trifluoride is very water soluble and may enhance ZrF4 and LiF solubility. Boron trifluoride also reacts with other fluoride-containing species to form adducts (e.g. CsF + BF3 CsBF4).

8-4

Table A.1. Water solubility of salt components and other species Compound Formula Solubility in water (g/100 mL)

Beryllium fluoride BeF2 Very soluble Lithium fluoride LiF 0.27 (20°C)

Zirconium fluoride ZrF4 1.3 (20°C)

Cesium fluoride CsF Very soluble Strontium fluoride SrF2 Insoluble Boric acid H3BO3 4.7 (20°C)

Boron trifluoride BF3 (as a dimer, B2F6)

Very soluble The use of alkylcarbonates has been investigated for increasing the solubility of LiF (Jones, J., et al.,

2011); however, introduction of organic species into the tanks is of concern because of potential radiation induced degradation of the organic compounds. The removal of cesium fluoride during an initial water wash would remove one of the major radiation sources and might lessen the concern about radiation induced degradation of any organic compounds. Residual organics in the final disposition vessels could be detrimental to vessel integrity due to the potential for continued release of fluorine due to radiolysis and the subsequent reactions, and could be a significant WAC issue.

Resulting solutions from the dissolution process would need to be dried or solidified prior to waste disposition. Aqueous solutions could be mixed with grout immediately prior to transfer to the waste containers. Sodium iodide could be added or pre-mixed with the grout to serve as a fluoride scavenger which would mitigate pressure build up due to continued radiolytic production of fluorine. Mixing of the solutions with grout will essentially dilute the salt which may reduce the production of fluorine from radiolysis. Dilution of the waste will change the profile and may impact suitability for disposal at the WIPP.

Extensive work would be required to determine an acceptable solvent, address criticality concerns, and to develop the fluid transfer and waste stabilization processes necessary for chemical dissolution if the tank removal/sectioning or thermal transfer options were to become untenable. There are potential solutions to the concerns, but more thorough studies would be needed.

8.2.4 Mechanical Removal with Needle Scaler In this option, as discussed in more detail in the 2010 salt removal options evaluation, a type of pneumatic hammer, such as a needle scaler would be used to loosen the salt. Needle scalers have typically been used for removing rust, welding slag, paint or other unwanted coatings from metal parts. Pneumatically driven needle scaler/vacuum tool combinations are commercially available and relatively inexpensive. The needle scaler/vacuum would be placed on the end of a manipulator arm with a camera for visual guidance of the tools. The tool assembly would then be inserted into the tanks through the 3-in. access port. If the fuel and flush salts cannot be easily broken down to granules, the needle scaler could be temporarily replaced on the end of the manipulator arm by a pneumatically driven chisel tool.

Mechanical removal by this method would require a very maneuverable manipulator arm to work around the interferences in the tanks. Interferences in the FDTs include 32 vertical thimbles embedded in the salt.

The manipulator arm, needle scaler, vacuum tool and camera assembly would be inserted into the tanks through the 3-in. access port and would be operated from above the shield floor. The needle scaler and vacuum would remove the salt by breaking it up and vacuum transferring it to disposal containers sized to fit into RH-72B canisters for shipping to WIPP.

8-5

The interferences from the thimbles inside the FDTs make the technical feasibility of this option suspect.

The ability to maneuver the needle scaler and vacuum tool in and around the thimble assemblies would be very difficult. The reliability of this tooling combination in the high radiation environment of the FDTs, the potential for damage to the thimbles from contact to the needle scaler, and damage to the needle scaler from contact with the thimbles are all disadvantages. Also, the risk from the airborne radioactive particulates is considered higher than for whole tank removal and sectioning, since that option would be performed inside a separate filtered enclosure. It is expected that the residual heel left in the FDTs and FFT from this option would be substantial, unlike the tank removal/sectioning option. This removal option would also require a change to the ROD without any significant advantages.

8.2.5 Entombment In this option, as discussed in more detail in the 2010 salt removal options evaluation, the tanks containing the salt are entombed in the Drain Tank Cell by pouring flowable, engineered grout into the cell. The 2010 evaluation includes placing sleeves around the tanks in the cell prior to filling with grout to make the grouted tanks retrievable for disposition at another location if required. The hydrogeologic setting at MSRE is not favorable for long-term (1,000+ yr) disposal of the tanks and salt in the Drain Tank Cell. As discussed in the 2010 evaluation, the groundwater levels are expected to be at an elevation slightly above the bottom of the lower drain tank shield block. This concern is partially addressed by requiring isolation of the tanks (remote tooling required as in the whole tank removal option) and then lifting the tanks with the salt to the level of the lower shield floor to be above the water table level. From a technical feasibility standpoint, multiple grout formulations have been developed for closure of high level waste tanks that could be considered for use in this option. The impacts on the grout from the fluorine radiolysis of the salts would require additional study. This option does not have significant safety or cost risks when compared to the other options. However, final disposal of TRU waste in a location other than the WIPP site is not allowed per the existing requirements. Also, the MSRE ROD assumes final disposal of the salt waste in another location. Therefore, the team recommends that this option not receive further study or consideration.

8.3 SALT REMOVAL OPTIONS EVALUATION AND CONCLUSIONS As discussed above, the current team, with assistance from UCOR SMEs, performed a comparative evaluation and ranked the five salt removal options in four categories:

Technical riskComplexity, proven or new technology SafetySafety off the removal process CostRemoval /transport/disposal Regulatory barriers Positive and negative attributes for each option were first identified and listed as plus or minus under each category above (Table A.1). Then each option was scored as High, Medium, or Low in each category, based on the extent of plus/minus attributes (Table A.2). The scoring was then quantified to determine the relative ranking of the options. Since regulatory barriers were generally outside the control of the facility and often DOE, that category was assigned a weighting of twice the value of the other three.

As shown in Table A.2, tank removal with sectioning of the contents into WIPP-acceptable packaging, led the list of options based on low technical and safety risk (proven, remote technology), and among the lowest regulatory risks (no heel left in facility).

8-6

Thermal transfer was second in ranking with the lowest regulatory barriers (already accepted in the ROD), but slightly higher technical and Safety risk because of the tank integrity concern, particularly the potential for embrittlement, high temperature and higher energy process, and the heel that will be left in the tanks.

Other options were more difficult from both technical and regulatory perspectives and offered no appreciable advantage to the first two options. Entombment was not considered acceptable due to obvious Regulatory obstacles.

The more quickly the facility can further the concept and design of the tank removal/sectioning and minimize or eliminate further studies or maintenance of equipment for the melting option, the earlier cost savings/avoidance can be realized.

8.4 REFERENCES

BJC/OR-3301. Engineering Evaluation of the Proposed Alternative Salt Transfer Method for the Molten Salt Reactor Experiment Oak Ridge National Laboratory Oak Ridge, Tennessee, October 2009, Bechtel Jacobs Company LLC, Oak Ridge, TN.

DOE/OR/02-1671&D2. Record of Decision for Interim Action to Remove Fuel and Flush Salts from the Molten Salt Reactor Experiment Facility at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, 1998, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR/01-2496&D1. Engineering Evaluation of Options for Molten Salt Reactor Experiment Defueled Coolant Salts, Oak Ridge, Tennessee, 2010, U.S. Department of Energy, Office of Environmental Management, Oak Ridge, TN.

DOE/OR/01-2560&D2. Remediation Strategy Plan for the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, January 2013, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

Jones, J., et. al. Solubilization of SEI lithium salts in alkylcarbonate solvents, Fluid Phase Equilibria, 2011, Vol. 305, pp. 121-126, Elsevier B. V., Amsterdam, Netherlands.

LA-UR-10-07278. TRU Waste Determination of MSRE Salt for Disposal Eligibility in WIPP, 2010, Los Alamos National Laboratory, Los Alamos, NM.

LA-UR-13-22367. Evaluation of Molten Salt Reactor Experiment (MSRE) Remnant Salt Properties for Disposal Eligibility at the WIPP, March 2013, Los Alamos National Laboratory, Los Alamos, NM.

8-7

8-8 Table A.1. Evaluation of salt removal options

+/-

Thermal Chemical Mechanical (Tank Removal &

Sectioning)

Mechanical (Needle Scaler)

Entombment Environmental, Safety, and Health Plus

• No particulates generated

• Personnel dose likely more favorable than mechanical disposal methods

• Waste generation more favorable than mechanical methods

• No particulates generated

• No melting of tank salts required

• Should minimize heel left in drain tanks

• Personnel dose likely more favorable than mechanical disposal methods

• Salt Waste remains in solid form

• No melting of tank salt required

• No melting required

• Preparation work for grout pours can be done remotely (low dose)

• No offsite waste shipments Minus

• Will require off gas monitoring during processing

• High energy process

• Additional criticality controls required

• Will require off gas monitoring during processing

• High rad environment must be controlled

• Rad waste generated from evolutions

• Particulates generated during process

• Airborne rad particulates likely very high

• Water table level above Drain Tanks

• Performance assessment over minimum of 1000 yr required Technical Feasibility Plus

• Well-studied process. Previous salt melts with success

• Pre-defueling (1999) drain tank integrity measurements available for comparison

• New piping system, partially due to lessons learned from previous salt transfer attempt

• Does not rely on tank heaters or additional heat trace on transfer lines

• Does not increase thermal stress on drain tanks

• Resulting solutions will fit any container required for disposition

• Similar to proven D&D technology

• Lifting cranes and equipment intact

• Long term stop/start capability

• Grouting of rad waste well-studied Minus

• Tank integrity concerns o HF, F2, and salt corrosion o Possible tellurium and radiation embrittlement

• External heaters required, which have not been maintained since

~2008

• Elemental metals in tank bottoms

• Failed FFT salt transfer in 2005

• Cannot destructively test for INOR-8 embrittlement

• Optimal aqueous solution must be found

• Increased mechanical loading on the tank

• Boron or other neutron poison required

• Reliability of remote tools in high rad environment

• Tank internals provide difficult mechanical environment

• Likely leaves large heel

• Reliability of remote tools in high rad environment

• None

Table 8.2. Evaluation of salt removal options (cont.)

8-9

+/-

Thermal Chemical Mechanical (Tank Removal &

Sectioning)

Mechanical (Needle Scaler)

Entombment Regulatory Plus

• 2010 LANL evaluation demonstrated WIPP eligibility possible

• Can use RH-72B container

• Identified option in the ROD

• Can use RH-72B container

• No tank heel left on site for disposal

• Minimum WIPP eligibility evaluation required

• Can use RH-72B container

• Minimum WIPP eligibility evaluation required

• None Minus

• None

• Additional waste characterization for WIPP eligibility required

• Potential for non-homogeneous waste profiles

• Requires change in the ROD

• Requires change to ROD

• Requires change to ROD

• Current ROD does not allow disposal on-site

• If salt determined to be TRU waste cannot be disposed of in state

• Interaction with ORNL final closure plans Cost Plus

• Fewer salt disposal containers relative to other methods

• Minimal developmental testing required

• Eliminates need for maintenance of melting equipment and evolution

• Eliminates need for maintenance of melting equipment and evolution

• Eliminates need for maintenance of melting equipment and evolution

• Low cost Minus

• Design and installation of new salt transfer system

• Modifications to gas removal system may be required

• Design, and installation of new salt dissolution system

• Will require off gas monitoring during processing

• Greater volume of waste increases transportation cost

• Costly deployment of remote cutting facility and equipment

• Costly deployment and maintenance of remote cutting facility and equipment

• None

8-10 Table A.2. Salt removal options ranking results Method Technical Risk Safety Cost Regulatory Barriers Entombment L

M L

H Thermal Transfer M

M+

M L

Needle Scaler Removal H

H H

M Chemical Dissolution M+

M+

M M

Tank Removal/Sectioning L+

L+

M+

L Column Multiplier 1

1 1

2

9. RECOMMENDATIONS 9.1

SUMMARY

The recommendations for each of the MSRE systems were graded on four criteria: For the near-term recommendations, the criteria used were technical feasibility of the project, project cost, impact on facility safety, and integration of the recommendations into other systems. For the life-extension recommendations, the criteria were technical feasibility of the project, safety impact, initial capital cost, and the possible cost savings over project life. The recommendations were assigned a grade of High, Medium, or Low in each category. Based on these grades, the recommendations were ordered according to priority, and are presented in the next section.

9.2 NEAR-TERM RECOMMENDATIONS The near-term recommendations for all of the systems were collected and prioritized based on the technical feasibility of the project, project cost, impact on facility safety, and integration of the recommendations with other systems. The items listed below are in the prioritized order.

1. Provide timely source of backup power capability for facility to accomplish the following:

a. Provide a reliable connection for a backup power system along with a plan/procedure to obtain a backup diesel generator.

b. Provide reliable source of dry pipe air via station air compressors.

c. Backup diesel generator power for facilityemphasis on CVS.

2. Immediately return both fans to operable conditions:

a. Fix the base and replace the motor of Fan #1. Completed August 2016.

b. Investigate the bearing issue and repair Fan #2.

3. Implement routine and more frequent pumpdowns to minimize corrosion and safety risks from gas headspace buildup, and to improve system reliability.

4. Obtain gas samples directly from the tank headspace (FDT-1 and FDT-2). This will allow evaluation of chemicals that cannot be detected using the FTIR alone.

5. Initiate actions to evaluate feasibility of continuous-purge life-extension option for RGRS using CVS with no traps.

6. Implement improved procedural controls for flow rates for October 2016 pumpdown to minimize the risk of channeling through traps.

7. Document system performance after each pumpdown to use in trending and analysis of system issues.

Begin this process with the headspace pumpdown in the fall of 2016. Compare the condition to that found following the April 2016 pumpdown.

a. Obtain intact section of stainless steel piping with compression fitting for embrittlement evaluation.

b. Collect samples of solid residues if system plugging is found.

8. Determine the integrity of the Monel tubing in the RGRS by removing a section in an area upstream of the NaF traps where air in-leakage is suspected, and examine the tubing for evidence of stress 9-1

corrosion cracking. A similar evaluation should be performed on a section of stainless steel tubing between the molecular sieve and the vacuum pump.

9. Test the integrity of the underground electrical connections to CVS and any other critical equipment, and address deficiencies.

10. Calibrate the 10-cm and 4-m FTIR for the primary headspace gases of interest (HF, UF6. MoF6. SiF4, ClF4, ClF3 and SF6).

11. Perform UT, eddy current testing, and borescope examination on all 32 thimbles and one thermowell for both FDT-1 and FDT-2.

a. UT thimbles and FDT tanks for wall thicknesses and evidence of pitting.

b. Eddy-current test full circumference of thimbles for entire length (if possible, ECT FDTs).

c. Borescope thimbles, focusing on tank head-thimble interface at welds.

d. Via remote camera, inspect salt, internal tank welds, tank supports, and external tank welds (if possible).

12. Initiate process to complete formal TRU waste determination and defense waste determination.

13. Proceed with the current operations team plan for installing a second alumina trap and replace molecular sieve to increase the capacity for fluorine capture. Completed September 2016.

14. Initiate MSRE structural inspection program.

15. Purchase spare parts for the fan system.

16. Proceed with the current operations team plan for installing sacrificial nickel wool traps to consume corrosive gases that pass through the entire RGRS. These traps serve two purposes: (1) protection of the vacuum pumps, and (2) reduction of possible emission of reactive gases via the CVS.

17. Increase rigor of FTIR DAQ system use through development and use of a startup checklist and performance testing.

18. Configure the FTIR systems to identify other components in the headspaces gases and collect the data. Additional regions of interest can be set up for several oxyhalides believed to be present in the gas mixture.

19. Proceed with the current operations team plan for installing a second vacuum pump to provide redundancy.

20. PM overcurrent protection devices. Perform current PMs for needed electrical components. Ensure electrical PM program is in compliance with regulatory and contractual requirements.

21. Obtain spares for overcurrent protection devices and motor starters required for facility operation.

22. Replace PT-576 for pressure indication of the FFT.

23. Develop and execute a plan to move the charcoal canister to a more appropriate facility (e.g.

Bldg. 3019).

24. Obtain crane and lift plan to support characterization of contaminated probes, and ship off probes.

25. Sample and analyze the headspace gases in the ACB, MCB, and charcoal canister by one of the following options:

a. Analyze the headspace with the FTIR by routing the gas through existing lines, if possible.

b. Obtain sample for lab analysis. Work package was initiated to identify sampling points to use Materials and Chemistry Laboratory, Inc. (MCL) to obtain gas samples.

9-2

26. Address air-to-close valves power-and air-loss issues by developing procedures/equipment to maintain air supply after the current 24-h backup is exhausted.

27. Check functionality of the FDT and FFT heaters with resistance testing. This does not include turning on the heaters.

28. Update the Nuclear Criticality Safety Evaluation to reflect the defueled state of the salt in the FDTs and the FFT and potentially reduce some controls currently in place for RGRS operations.

29. Install 5-zone thermocouple assemblies in the molecular sieve and the second alumina trap to monitor the reaction front during pumpdowns.

30. Identify and purchase critical spare parts for the crane systems.

31. Replace either the limit switches or install new bumpers on the cranes.

32. Make the 30-ton crane operational:

a. Install a new bus bar on the 30-ton crane and purchase a spare bus bar for the 6-ton crane.

b. Replace wire rope.

33. Improve execution of the combustible control program.

34. Correct labeling issues with in-service electrical components, including the following:

a. Distinguish labeling for active SSCs described in the DSA.

b. Label the CVS fan start/stop buttons in the motor control system.

c. Label electrical components appropriately in accordance with UCOR procedure PPD-FO-1036, Conduct of Operations Program Description.

35. Protect the vents and structural supports outside the HEPA house and CBC (e.g. Jersey barriers, road closure).

36. Repair identified steam system leaks/deficiencies:

a. Steam leak on the mezzanine in the High Bay.

b. Fix the valves on the steam connection at the 7503/7509 breezeway.

37. Replace the fire alarm control panel.

38. Ensure that the working waste inventory lists include all facility waste containers as appropriate.

39. Increase priority to ship out large-volume waste, e.g., casks and wood inventory.

40. Replace HV-59 with a valve better suited for regulating RGRS flow rate 9.3 LIFE-EXTENSION RECOMMENDATIONS The life-extension actions have been listed below in order of importance per the analysis by the evaluation team.

1. Establish project to implement continuous purge option for RGRS using CVS with no traps.

a. Alternatively, implement the purge option with a NaF trap, fluorine trap, and booster pump with interlocks if the continuous purge with no traps is not an acceptable option 9-3

2. Establish project to design and build fluorine gas removal system replacement for RGRS:

a. New tubing alloy.

b. Shorter tubing runs.

c. Optimized trap size and media.

3. Install a UV detection system to measure F2 levels being released from the tanks.

4. Perform a smoke test in the South Truck Bay and High Bay and address the deficiencies.

5. Install electric heaters in the High Bay attic space to provide heat in case of steam loss.

6. Design and install a consolidated electrical distribution system to support continued operation of facility systems, such as:

a. CVS.

b. RGRS.

c. Support systems (cranes, air compressors, etc.).

d. Fire protection and LSS.

7. Air gap electrical feeds associated with unneeded loads.

8. Restore functionality of Drain Tank Cell sump level indication, alarm, and transfer jet.

9. Evaluate elimination of HEPA filters.

10. Install manual valves downstream of the air-operated drain tank valves. Permanently isolate FDTs from those process lines not related to RGRS:

a. Failure of air-to-close AOVs HCV-573, -575, -577, -544, -454, and -456 will extend the headspace gas confinement boundary to valve V-561 and to process lines not permanently isolated from the Off-Gas System, with potential adverse effects to RGRS operations.

b. Station air load will decrease, eliminating reliance on an active component to maintain confinement boundary.

c. Length of process gas lines will be decreased, reducing probability of leakage and overall risk of releasing headspace gas.

11. Air-gap the charcoal beds for complete isolation if headspace gas samples show no hazardous constituents. Routinely monitor the bed pressure and the headspace gas composition to identify any unusual gas generation until isolation is completed.

9.4 REFERENCES

PPD-FO-1036/R2. Conduct of Operations Program Description, August 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

9-4

APPENDIX A.

GENERAL CORROSION INFORMATION A-1

This page intentionally left blank.

A-2

The fluoride salt in the MSRE along with the variety of treatments that have been used, create conditions where a wide range of degradation mechanisms could be active. General or uniform corrosion is often seen as the result of the reaction of a metallic material with gaseous or liquid corrodents. General or uniform removal of material is more easily predicted and accommodated in the selection of materials for application in a corrosive environment. Localized corrosion is generally manifested as pitting or crevice corrosion and it is more difficult to predict and detect. Stress corrosion cracking is another type of corrosion that only occurs under certain environments and to vulnerable materials. Sometimes the cracks will progress in an intergranular manner, i.e. along grain boundaries, while another mechanism has the cracks progressing in a transgranular manner. Degradation of this type, whether intergranular or transgranular, will often not be detected until a fairly catastrophic event occurs.

A few other less common types of corrosion also have to be considered. Liquid metal embrittlement involves penetration of very specific metal atoms along the grain boundaries of a structural material. The alkali metals with their relatively low melting points are sometimes the corrodent in liquid metal embrittlement. Loss of ductility and subsequent cracking can also be caused by radiation and the products that are produced in the material being irradiated. Often a gaseous species will be produced by a decay reaction, and the gaseous product can accumulate, sometimes on grain boundaries, and cause a loss of strength and ductility. Unfortunately, all the corrosion mechanisms just described can occur in the environments of the MSRE and the associated gas handling systems, such as RGRS.

The original version of INOR-8 (the commercial product was marketed as Hastelloy N) was developed for elevated temperature service in molten fluoride salts. It was a nickel base alloy containing 16 wt%

molybdenum for solid solution strengthening, 7 wt% chromium for moderate oxidation resistance, up to 4 wt% iron to allow the chromium to be added as ferrochrome, 0.05% carbon and up to 1 wt% each manganese and silicon. Typically, alloys intended to have corrosion resistance in air at elevated temperature have considerably higher chromium content, but the fact that the fluoride salt prevents formation of a protective chromium oxide results means the primary corrosion mechanism in molten fluoride salts is selective removal of chromium. This necessitated limiting the chromium content in INOR-8/Hastelloy N. The 7% concentration provided chromium to form a protective chromium oxide layer with sufficient resistance to air oxidation at anticipated reactor temperatures while also limiting the amount of chromium available for reaction with the molten fluoride salts. This composition of the alloy was used for fabrication of the components for the MSRE but it was subsequently found to have some shortcomings, in particular radiation induced embrittlement as well as shallow grain boundary penetration by fission product tellurium. Subsequent modifications of the Hastelloy N composition addressed these shortcomings, but these modified compositions were not developed until after construction and operation of the MSRE (ORNL/TM-5920). Consequently, the possibility of these degradation mechanisms affecting the integrity of MSRE components and particularly the drain tanks will need to be addressed.

For the present task, the compatibility of Hastelloy N as well as other materials used in construction of the fuel drain tanks, associated piping and the reactive gas removal system needs to be determined to assess or at least make an estimate of the condition of these metallic components after service during reactor operation, the subsequent defueling steps and the fluorine removal operations. The alloys whose performance needs to be assessed are listed in Table A.1 along with their compositions, and the non-salt environments to which the alloys were potentially exposed are listed in Table A.2. Ideally, the RGRS and drain tank environments would have only been exposed to fluorine except during the hydrofluorination process used for defueling. The post-defueling gases in the drain tanks and the RGRS were expected to be fluorine and argon; however, HF is routinely detected on the 10-cm FTIR at the beginning of each pumpdown. The presence of HF and the apparent deposition in transfer lines of uranium containing species provide strong evidence of in-leakage of moist air. This apparent moist air in-leakage results in the formation of hydrogen fluoride which increases the range of environments that need to be considered.

The concentrations of silicon tetrafluoride and chlorine trifluoride would be expected to be quite low, but A-3

they are included in the table since they will be present and could affect the performance of some materials. Table A.1 also includes information for Hastelloy C-276, a nickel base alloy that should be considered an alternative for future installations in the reactive gas removal system. Since the fuel drain tanks and their internal components as well as the fuel flush tank were exposed to molten salt during some of the fluorine and hydrogen fluoride treatments, the performance of these Hastelloy N components under molten salt/hydrogen fluoride and molten salt/fluorine conditions needs to be addressed.

A number of publications by professional societies, trade organizations and open literature journals have provided values for the expected corrosion rates of most of the alloys under consideration in the non-salt environments of interest. Corrosion rates from a number of these sources are provided in the following text. Most of these compilations only address some of the most commonly used alloys, so Hastelloy N, being an alloy designed for a very specific application, is not included in a lot of these compilations.

A.1 PUBLISHED CORROSION RATE INFORMATION To convey an indication of the extent of corrosion that might be encountered for each of the alloys in the possible RGRS and drain tank environments, we have collected data from a number of sources, including industrial guidelines, corrosion tables, and open literature publications. Results of this survey are presented in the following tables.

Table A.1. Nominal composition of alloys used in the MSRE FDTs and RGRS Alloy UNS No Ni Cr Fe Mo Cu Mn C

Other wt%

wt%

wt%

wt%

wt%

wt%

wt%

wt%

INOR-8 (Hastelloy N)

N10003 71 7

4M 16 0.35M 0.8M 0.06 Co 0.02M, W 0.5M Monel 400 N04400 66 1.5 31 1.0 0.1 304L stainless steel S30403 9.0 18.3 70 1.7 0.02 316L stainless steel S31603 10.2 16.4 69 2.1 1.6 0.02 Hastelloy C-276 N10276 57 16 5

16 0.5M 1M 0.01 4 W, Co 2.5M M = maximum Table A.2. Environments to which the components of the drain tanks and RGRS have been exposed Compound Chemical formula Melting point Boiling point Fluorine F2

-220°C (-363°F)

-188°C (-307°F)

Anhydrous hydrogen fluoride HF

-83.6°C (-118°F) 19.5°C (67°F) 49% hydrofluoric acid HF in H2O

-37°C (-34°F) 106°C (224°F)

Silicon tetrafluoride SiF4

-90.2°C (-130.4°F)

-86°C (-123°F)

Chlorine trifluoride ClF3

-83°C (-117°F) 11.3°C (52.3°F)

Fluorosilicic acid H2SiF6*2H2O 108.5°C (227.3°F)

Depending on the expected service life, corrosion rates of 0.010 in./yr would be the most that could be tolerated for service of no more than 2-3 years, and the limit would need to be considerably lower if the A-4

system were expected to last for at least an additional 40 years which is the possible additional lifetime proposed for MSRE.

A.1.1 Materials of Construction Guideline for Anhydrous Hydrogen Fluoride Source: Howells, M., et al. Materials of Construction Guideline for Anhydrous Hydrogen Fluoride, 2004, Hydrogen Fluoride Industry Practices Institute, Washington, D.C.

This guideline says that for anhydrous hydrogen fluoride (AHF), which is the term used to describe hydrogen fluoride that contains less than 400 ppm of water, many alloys are suitable for use in containment vessels. Carbon steel, so long as it contains less than a total of 0.2 wt% of Cu, Ni and Cr, can be used at temperatures up to 66°C (150°F). Longer term corrosion rates for carbon steel are on the order of 0.001 in./yr (0.025 mm/yr) but initial corrosion rates can be up to 10 times higher until a protective fluoride film is formed.

Austenitic stainless steels, including types 304/304L and 316/316L, are resistant to AHF up to at least 100°C (212°F). However, duplex alloys, ferritic grades, and severely cold-worked austenitic stainless steels, that have hardness values greater than a critical level (Rockwell C22), should be avoided because of potential problems with hydrogen assisted stress corrosion cracking.

Nickel base alloys such as Monel 400, Hastelloy C-276, and Inconel 600 are suitable for use in liquid AHF at temperatures up to at least 125°C (257°F) and in vapor AHF at temperatures up to at least 300°C (572°F). However, if moisture and oxygen are present, Monel 400 is susceptible to stress corrosion cracking in the vapor.

A.1.2 Corrosion Resistance Tables Source: Phillip A. Schweitzer. Corrosion Resistance Tables, Second Edition, February 1986, Marcel Dekker, Inc., New York, NY.

Reported corrosion rates (Schweitzer) are only given as being greater or less than certain levels for specific temperature ranges. Consequently, only corrosion rate ranges are available from this document.

In wet fluorine gas at 100°F (38°C), anticipated corrosion rates would be:

304/304L stainless steel

<0.002 in./yr 316/316L stainless steel

<0.002 in./yr Monel 400

<0.002 in./yr Hastelloy N No data Hastelloy C-276

<0.002 in./yr In hydrogen fluoride at 100°F (38°C), anticipated corrosion rates would be:

304/304L stainless steel

<0.002 in./yr 316/316L stainless steel

<0.002 in./yr Monel 400

<0.020 in./yr Hastelloy N No data Hastelloy C-276

<0.020 in./yr A-5

In 100% hydrofluoric acid at 100°F (38°C), anticipated corrosion rates would be:

304/304L stainless steel

>0.050 in./yr 316/316L stainless steel

<0.020 in./yr Monel 400

<0.020 in./yrmaterial is subject to stress cracking Hastelloy N No data Hastelloy C-276

<0.020 in./yr In fluorosilicic acid at 100°F (38°C), anticipated corrosion rates would be:

304/304L stainless steel

>0.050 in./yr 316/316L stainless steel

<0.020 in./yrmaterial is subject to pitting Monel 400

<0.002 in./yrmaterial is subject to stress cracking Hastelloy N No data Hastelloy C-276

<0.020 in./yr A.1.3 Corrosion Resistance of Nickel-Containing Alloys to Fluorine Compounds Source: Publication No. 433. Corrosion Resistance of Nickel-Containing Alloys in Hydrofluoric Acid, Hydrogen Fluoride and Fluorine, 1968, Nickel Institute, Brussels, Belgium.

Nickel Development Institute Publication No. 443 provides corrosion rate information for a very wide range of alloys and emphasizes the performance of materials in hydrofluoric acid. The document addresses exposure of alloys in a wide range of fluoride-containing compounds at elevated temperature in addition to hydrofluoric acid.

For gaseous fluorine at room temperature, a corrosion rate of 0.0002 in./yr was reported for Monel 400 while 304L stainless steel showed no corrosion both of which are consistent with results reported by other researchers. In anhydrous chlorine trifluoride, corrosion rates at 30°C of nickel-containing alloys were negligible measuring less than 0.00001 in./yr. Alloys tested included Monel 400 and type 347 stainless steel. Fluorosilicic acid, which is formed by reaction of silicon tetrafluoride with water, can be contained by Monel 400 and Hastelloy C provided no strong oxidizer is present. However, fluorosilicic acid, like hydrofluoric acid, may cause stress corrosion cracking of Monel 400. Under the same conditions, austenitic stainless steels may suffer pitting.

A.1.4 Corrosion Characteristics of Stainless Steel Alloys Exposed to F2 Source: Craig, B. D.. Handbook of Corrosion Data, 1989, ASM International, Metals Park, OH.

Most austenitic stainless steel alloys are reported to be resistant to completely dry fluorine gas. However, the presence of small amounts of moisture results in accelerated attack with a possibility of pitting and stress corrosion cracking. Likewise, austenitic stainless steels have good resistance to anhydrous hydrogen fluoride but limited resistance to hydrofluoric acid. Type 316 stainless has useful resistance to concentrations below 10% while 304 stainless has poor resistance to any significant hydrofluoric acid concentration. Annealed austenitic stainless steels are resistant but not immune to stress corrosion cracking by hydrofluoric acid. Severe pitting of 304 and 316 stainless steels was reported for some concentrations of fluorosilicic acid. No data was provided for nickel-base alloys.

A-6

A.1.5 Corrosion Rates for Selected Alloys in Hydrofluoric Acid Source: Pawel, S. J. Corrosion of High-Alloy Metals in Aqueous Hydrofluoric Acid Environment, 1994, CORROSION, Vol. 50, No. 12, pp. 963-976.

Immersion tests were conducted with eight materials in a variety of hydrofluoric acid environments including concentrations of 0.05% HF, 1.0% HF, 10% HF, 24% HF, 48% HF with exposures at temperatures of 24°C, 50°C, and 76°C. Alloys of interest included 316L stainless steel, Monel 400 and Hastelloy C-276. Results of these studies for the three lowest acid concentrations and the two lowest temperatures are shown in Table A.3. Pawel also noted that previous authors had reported Monel 400 to be susceptible to stress corrosion cracking in moist, aerated hydrogen fluoride vapors.

Table A.3. Corrosion rates for selected alloys in 3 hydrofluoric acid concentrations at 24°C and 50°C Alloy Acid concentration 24°C immersed 24°C vapor 50°C immersed 50°C vapor in./yr in./yr in./yr in./yr 316L stainless steel 0.05%

<0.001

<0.001 0.001

<0.001 1.0%

0.022 0.015 0.114 0.044 10.0%

0.333 0.246 Monel 400 0.05%

0.002 0.006 0.007 0.013 1.0%

0.002 0.036 0.007 0.133 10.0%

0.002 0.088*

0.006 0.231*

Hastelloy C-276 0.05%

<0.001

<0.001 0.001

<0.001 1.0%

<0.001 0.002 0.005 0.013 10.0%

0.002 0.014 0.008 0.093

• Indicates extreme attack at vapor-solution interface The conclusion from these data compilations is that, at ambient temperatures, dry, oxygen free fluorine gas can be contained by essentially all the structural materials being considered. Presumably the RGRS designers considered this information when making decisions on the selection of alloys for system fabrication, and they did not expect that the system would be subject to in-leakage of moist air. As shown in this compilation of corrosion data, if the fluorine is exposed to a very limited amount of moisture and anhydrous hydrogen fluoride is able to form, corrosion rates will be significant but tolerable for the current application. However, if small amounts of moisture (> 400 ppm) and oxygen are added to the anhydrous hydrogen fluoride, Monel 400 is susceptible to stress corrosion cracking. Severely cold worked stainless steels could be susceptible to hydrogen assisted stress corrosion cracking in anhydrous hydrogen fluoride. If sufficient moisture is present that hydrofluoric acid can be formed, then the corrosion rate of 304L stainless steel reaches an unacceptable level. In addition, Monel 400 is reported to be susceptible to stress corrosion cracking in this environment. As described in this data compilation, fluorosilicic acid is also reported to be capable of causing severe pitting in the 300 series stainless steels and stress corrosion cracking of Monel 400.

These data raise serious concerns about the use of Monel 400 in environments where moist hydrogen fluoride is present which includes the RGRS gas lines and the containers for the three types of traps that are utilized in the RGRS. It is also important to note that the behavior of Hastelloy N when exposed to molten salt and these gaseous components is not addressed in any of the conventional data bases.

Corrosion in the RGRS would not be expected to be impacted by the presence of molten salts.

A-7

A.2 ADDITIONAL REFERENCES ORNL/TM-5920. Status of Materials Development for the Molten Salt Reactor, 1978, Union Carbide Corporation, Oak Ridge, TN.

A-8

APPENDIX B.

ULTRASONIC THICKNESS AND X-RAY FLUORESCENCE MEASUREMENTS OF RGRS TUBING B-1

This page intentionally left blank.

B-2

Subsequent to the Spring 2016 pumpdown of the FDTs and FFT, the scroll-type vacuum pump was removed for examination and cleaning. This pump had been installed and placed in service during the pumpdown as a result of complete plugging of the vacuum pump discharge line that was installed at the beginning of the Spring 2016 evolution (RPJ-027503). During removal of the vacuum pump after pumpdown was complete, the stainless steel (SS-304) discharge tubing experienced a circumferential failure at the pump discharge connection fitting (Figure B.1). See Appendix C for additional discussion of the metallurgical analysis of the tubing and evaluation of the deposits recovered from the inside of the failed RGRS discharge line.

As a result of the failure experienced during the Spring 2016 pumpdown, as well as previous plugging events experienced in RGRS tubing at MSRE, the 2016 MSRE Engineering Evaluation Team investigated possible contributing causes for the plugging and tubing failure that have not previously been identified. In addition to a review of relevant literature on material compatibility of alloys deployed in the RGRS with expected corrosive gases from the drain tanks headspace (see Appendix A), ultrasonic testing (UT) and x-ray fluorescence (XRF) measurements were made to determine the wall thickness and material composition of the tubing. It should be noted that neither UT nor XRF measurements will provide an indication of embrittlement or stress corrosion cracking.

Based on available references and design drawings, the RGRS tubing was apparently assembled using the nickel-copper alloy Monel 400 from the beginning of RGRS up to the discharge of the molecular sieve (i.e., upstream of traps), at which point the tubing was expected to be made of SS-304L for the remaining tubing runs through the vacuum pump and out to the CVS (i.e., downstream of traps). Due to the recurring plugging of RGRS tubing and components during and subsequent to defueling of the fuel and flush salts, corrosion and thinning of the RGRS tubing was expected, as previous analysis indicated that RGRS tubing materials may be present in the plugging material (RPJ-027503). Nominal wall thicknesses for the Monel and stainless steel tubing could not be confirmed from review of historical design drawings, but the majority of accessible RGRS tubing is 0.5 in. in diameter.

For the RGRS tubing upstream of the traps, sample points were taken in the Transmitter Room and on line 1567 in the High Bay by pulling back the 3-in. neoprene flex hose and cleaning the sample locations prior to taking measurements. Availability of many runs of Monel RGRS tubing was limited due to their locations in high-radiation or difficult-to-access areas. For the tubing downstream of the traps, the team attempted to take measurements at locations representing as many tubing configurations as could be observed (i.e., 0.25-in. vs 0.5-in., in between fittings, near inlet and outlet of the vacuum pump, and near the tubing failure location discussed above).

The results of the UT and XRF measurements are depicted in Figure B.2 and Figure B.3. Below are the conclusions that can be taken from these results:

The downstream tubing was constructed primarily of SS-304 but one recently replaced section was constructed of SS-316, with a mix of 0.035 and 0.065-in. nominal wall thickness.

In discussions with pipefitters who have previously installed replacement runs of RGRS downstream stainless steel tubing, it is the teams understanding that the stainless steel alloy and nominal wall thickness for replacement tubing are not specified. The alloys and wall thicknesses currently installed have been deemed to be acceptable by facility personnel.

The upstream tubing was confirmed to be made of Monel 400.

B-3

Figure B.1. RGRS tubing failure, discharge of vacuum pumpSpring 2016.

B-4

The wall thicknesses of both the Monel 400 and stainless steel tubing were measured to be very close to nominal values.

Since the UT-measured wall thicknesses tended to group around standard tubing nominal wall thicknesses of 0.035 in. and 0.065 in., the team reasoned that the as-installed tube wall thicknesses were either 0.035 in. or 0.065 in. (whichever was closer to the UT-measured value).

For those measurements that read less than nominal, the difference between nominal was too small to draw conclusions regarding the extent of corrosion.

UT-measured wall thicknesses measuring greater than nominal appeared to be within instrument error, and are not a result of corrosion product buildup (UT instrumentation cannot see such buildup).

Subsequent to the final UT and XRF measurements, an additional SS-304 tubing failure was experienced on the discharge line of the RGRS vacuum pump during pump replacement. The failure location was immediately upstream of sample location #10 shown in Figure B.2 on the horizontal run of 0.25-in.

tubing within the fitting. As with the Spring 2016 failure, the failure occurred at the fitting ferrule and was circumferential. As pipefitters began repair activities, the 0.5-in. tubing section immediately upstream of sample point #10 on the same tee experienced an identical circumferential failure (Figure B.4). Note the presence of deposits on the outer surface of the tube indicating the existence of the tube failure prior to flow of the process gas/fluid through the tube. This segment of tubing was repaired and RGRS vacuum pump declared functional. The pipefitters who identified the failures indicated that they were identical to the Spring 2016 failure. They also noted that the internal deposits which they observed in the Spring 2016 tube blockage was seen several feet downstream of the vacuum pump outlet, and that the amount of material is less as the run gets farther from the pump outlet. The pipefitters also noted that they consistently see a green coating on the inside diameter of the stainless steel RGRS tubing.

As discussed above and in Appendix C, the failures of stainless steel RGRS tubing in 2016 appear to be localized brittle failures immediately behind ferrule locations. Several recommendations in this report support additional testing and analysis to aid in identification of the degradation and failure mechanisms causing the failures, and to investigate possible similar effects on the Monel 400 RGRS tubing. The conclusion that the embrittlement is localized is further supported by bending of tubing near the Spring 2016 failure location, as seen in Figure B.1, which demonstrated that the freespan region of the tubing is ductile.

B.1 REFERENCES RPJ-027503-D842. Engineering Evaluation of the Plugging Event of April 20, 2016 at MSRE, May 2016, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

B-5

B-6 Figure B.2. UT and XRF measurements of stainless steel RGRS tubing downstream of Al2O3 and molecular sieve traps.

B-7 Figure B.3. UT and XRF measurements of Monel RGRS tubing upstream of Al2O3 and molecular sieve traps.

Figure B.4. RGRS discharge tubing failure, downstream of vacuum pumpSeptember 2016.

B-8

Figure B.5. Calibration data and NDE technician certification.

B-9

This page intentionally left blank.

B-10

APPENDIX C.

MCL METALLURGICAL AND ANALYTICAL RESULTS ON RGRS PUMP TUBING/DEPOSITS C-1

This page intentionally left blank.

C-2

To get a better idea of the degree of corrosion that has occurred in the drain tanks to both the tank wall and the outer diameter of the 33 thimbles in each tank, a determination of the source of the molybdenum found in traps, on plugs and deposited on tube surfaces should help provide an understanding. Significant amounts of molybdenum have been detected and reported in several previous publications, and a conservative assumption might be to assume all the molybdenum is a result of corrosion of the INOR-8/Hastelloy N. However, there is good evidence that molybdenum detected and identified prior to 233U defueling was not a result of corrosion.

Molybdenum is somewhat unique because there is a range of stable isotopes two of which are decay products of 233U and 235U. More specifically, 99Mo is produced when 235U is the fissile component while 100Mo is produced when 233U is the fissile component of the fuel salt. In contrast, naturally occurring molybdenum, like that used in Hastelloy N, contains a distribution of 92Mo, 94Mo, 95Mo, 96Mo, 97Mo, 98Mo and 100Mo. Consequently, if the specific molybdenum isotope(s) can be identified, then it can be determined whether the molybdenum was a product of corrosion or radioactive decay.

In a June 1967 publication (ORNL/TM-1853), W. R. Grimes reported on the identity of isotopes deposited on the graphite and Hastelloy N reactor shell after initial operation of the MSRE with 235U fuel.

The 99Mo isotope was reported to be one of the noble metal isotopes found on the surfaces of the MSRE reactor, and this was the only molybdenum isotope reported. No information was provided on the method used to identify the isotope, so it cannot be confirmed that the isotope was positively identified rather than the identity having been assumed since it is known that 99Mo is the product of 235U.

Similar information was reported by R. J. Kedl in his December 1972 publication (ORNL/TM-3884).

Analysis of these noble metal isotopes after the final MSRE run with 235U identified 99Mo as the molybdenum isotope collected on reactor vessel surfaces. Both these publications identified 99Mo which is not present in natural molybdenum but is a fission product of 235U. Again, the method used to identify the isotope is not provided in the report. If it is assumed that a positive identification of the isotope was made, it is a logical conclusion that the molybdenum recovered from the reactor vessel during and after operation with 235U is entirely a fission product and not a result of corrosion of molybdenum containing alloys, i.e., Hastelloy N.

Publications have not been found that provide direct results from measurement of the molybdenum isotope recovered after operation with 233U. However, the report Isotopic Distribution in MSRE Chemical Traps (UCOR-4610/R0) describes how one isotope, typically 137Cs, was quantified for a single trap or a group of traps using gamma spectroscopy, then the activity of other isotopes was proportioned to the measured reference isotope. This publication further estimates that the 100Mo isotope is present and is the isotope present in the highest concentration. If these assumptions and modeling efforts are correct, then all the molybdenum collected through all the MSRE operation was a fission product and not a corrosion product. However, no information was given as to why 100Mo was assumed to be the sole molybdenum isotope present nor was it clear why a very large amount of molybdenum was projected to be produced.

The period when corrosion of Hastelloy N, which comprises the salt drain tanks and the thimbles in these tanks, was expected to be greatest was during the 233U defueling effort. This defueling was done in the drain tanks rather than in the fuel processing tank as was the case during the 235U defueling, so the tank wall and thimbles were exposed to molten salt as well as periodic injections of hydrogen fluoride and fluorine and this was, almost certainly, a very corrosive environment to the Hastelloy N and one far more corrosive than encountered during normal reactor operations. If that were the case, then molybdenum collected in traps, filters and plugs would be expected to contain the molybdenum isotopes characteristic of natural molybdenum.

C-3

Although there are a number of samples that, potentially, contain significant amounts of molybdenum collected after the 233U defueling effort, none had been analyzed for the molybdenum isotopic content until August 2016. As part of the examination of a section of cracked stainless steel tube, a section of deposit was removed from the tube and analyzed for molybdenum isotopes. Results of that analysis are included in the description of the examination of a piece of stainless tubing that was reported to have failed in a brittle manner by the pipefitters who were removing the vacuum pump. Analysis of additional, and hopefully older, samples of deposits may provide additional information on the origin of molybdenum collected at various locations over an extended period of time.

To gain a better description of the situation when the stainless tube section was removed, the pipefitters who discovered the failed tube were interviewed. The pipefitters explained that they were in the process of disconnecting the discharge line from the pump and had made a single turn of the Swagelok nut when they realized that the tube was broken and could be pulled out of the nut without loosening the nut any further. With the tube removed, the pipefitter reported he removed the nut to expose the tube stub and the ferrule which was still attached to the tube stub. Unfortunately, the broken tube section was discarded and put into a box accumulating radioactive waste. When an interest was shown in analyzing the fracture surface of the broken tube, the tube section was recovered from the waste container. The manufacturers marking on the tube indicated it was 304/304L stainless steel. An effort was made to break off a piece from the fracture surface of this recovered tube, but away from the fractured edge the tube showed the degree of ductility expected of a 300-series stainless steel. Consequently, the broken end of the tube suffered quite a bit of plastic deformation. In order to examine the fracture surface as well as the adjacent tube sections, the end of the tube containing the fracture surface, the small piece broken from the fracture surface as well as several 0.5-in.-long sections were submitted to an analytical laboratory for metallographic examination. The tubing section containing the fracture surface as well as an adjacent tube section are shown in Figure C.1 and Figure C.2. These tube sections as well as deposits from inside the tube sections were examined using a Hitachi S-4500 scanning electron microscope (SEM).

The SEM examination of the fracture surface, the inner diameter of the tube samples, the cut surface of the second tube section, as well as the deposits taken from the tube sections was conducted by the laboratory; and a report was issued summarizing the results. Information presented in the report SEM Analysis of 304 Stainless Steel Tubing and Residues, MCLinc Project UCO003140A, provided a considerable amount of useful information, which will be summarized here. The two tube sections as delivered to the laboratory are shown in Figure C.1 and Figure C.2. After these photos were taken, a portion was cut off the longer tube section on the left so that it would fit in the SEM. About a quarter of the tube section on the right was removed to expose the internal surface of the tube section as well as exposing a cleanly cut surface that could be used to determine the composition of the alloy. The energy dispersive spectrometry (EDS) system on the SEM was used to identify the elements present in selected areas as well as a semiquantitative measure of the concentration of the elements present in the areas being examined.

The micrograph in Figure C.3 shows a fractured area of the broken tube, but the edges of the area being examined show obvious evidence of mechanical deformation that quite possibly occurred when the small sample was being broken off the tube. The bright areas in the micrographs are indicative of nonconductive material on the surface of the sample.

C-4

Figure C.1. Section of type 304L stainless steel tube with the fracture surface on the upper right.

Figure C.2. Section of type 304L stainless steel tube cut from the left end of the tube sample in Figure C.1.

Figure C.3. Secondary electron micrograph showing a portion of the fracture surface of the tube sample in Figure C.1.

The EDS spectra of two areas on the fracture surface, an area where charging occurred and an area where no charging occurred are shown in Figure C.4 and Figure C.5, respectively; and the results of the semiquantitative analysis are summarized in Table C.1. These results indicate a considerable amount of carbon was deposited on the surface and the examined areas contain molybdenum and silicon in addition to considerable amounts of oxygen. The expected concentrations of iron, chromium and nickel are present.

C-5

Figure C.4. EDS spectra from an area of the fracture surface where charging occurred.

Figure C.5. EDS spectra from an area of the fracture surface where no charging occurred.

Figure C.6 shows the fracture surface of another area of the tube along with the internal surface of the tube. The edge of the fracture does not show any evidence of plastic deformation as might be expected with a ductile fracture. EDS spectra of the fracture surface and the ID of the sample shown in Figure C.6 are displayed in Figure C.7 and Figure C.8, and Table C.2 summarizes the elemental analysis of the areas where the spectra were taken. The spectra of the fracture surface show significant amounts of oxygen and sodium with lesser amounts of silicon and molybdenum and the expected concentrations of the stainless steel components, iron, chromium and nickel. The internal surface of the tube also shows significant amounts of oxygen and sodium and moderate amounts of silicon and molybdenum and the expected components of the stainless steel. The presence of oxygen, sodium, silicon and molybdenum on the fracture surface is a strong indication that the tube had been broken for some time, and the components of the environment that deposited on the tube interior also deposited on the fracture surface.

C-6

Table C.1. Semiquantitative results of EDS examination of selected areas on the sample in Figure C.3 Element Edge of fracture surface of sample in Figure C.3 Fracture surface of sample in Figure C.3 (atomic %)

(atomic %)

Carbon 48.26 28.46 Oxygen 14.93 48.82 Sodium 3.54 Silicon 2.12 1.73 Molybdenum 1.01 0.5 Chlorine 1.01 Potassium 0.75 Chromium 6.29 3.85 Iron 20.0 15.14 Nickel 2.09 1.5 Figure C.6. Secondary electron image of the fracture surface and the internal surface (top left) of the tube section in Figure C.2.

C-7

Figure C.7. EDS spectra from the fracture surface shown in Figure C.6.

Figure C.8. EDS spectra from the inner surface of the sample shown in Figure C.6.

Table C.2. Semiquantitative results of EDS examination of selected areas on the sample in Figure C.6 Element Fracture surface of sample in Figure C.6 Interior surface of sample in Figure C.6 (atomic %)

(atomic %)

Oxygen 34.63 28.57 Sodium 9.06 11.13 Silicon 1.78 2.6 Molybdenum 0.9 2.64 Potassium 0.48 Chromium 11.46 13.49 Iron 36.83 36.58 Nickel 5.34 4.5 C-8

In order to confirm the composition of the tube and the identification of the alloy, an EDS spectra was taken on a freshly cut surface of the shorter tube section. The spectrum of that area is shown in Figure C.9, and the composition shown in Table C.3 confirms that the alloy is 304 or 304L stainless steel.

The absence of molybdenum confirms the alloy is not 316/316L stainless and the presence of oxygen suggests oxidation of the surface occurred as a result of heating during the cutting.

Figure C.9. EDS spectra collected from a cut surface of the tube section in Figure C.6.

Table C.3. Semiquantitative results of EDS examination of freshly cut tube surface Element Fracture surface of sample in Figure C.6 (atomic %)

Oxygen 23.17 Silicon 1.24 Chromium 15.42 Iron 53.04 Nickel 7.13 Micrographs of two areas of the deposit taken from the internal surface of the shorter tube sample are shown in Figure C.10 and Figure C.11. There are two distinct microstructures on the surface of this deposit, and spectra taken from each of the areas are shown in Figure C.12 and Figure C.13.

C-9

Figure C.10. Secondary electron micrograph of the deposit removed from the internal surface of the short section of tubing shown in Figure C.2Area 1.

Figure C.11. Secondary electron micrograph of the deposit removed from the internal surface of the short section of tubing shown in Figure C.2Area 2.

Figure C.12. EDS spectra of the scale shown in Figure C.10.

C-10

Figure C.13. EDS spectra of the cubic shaped material in the scale shown in Figure C.11.

This examination provides a good deal of information about the tube failure, but there is not enough information to determine a cause for the failure. The absence of plastic deformation and the characteristic microstructural features associated with ductile failure provide strong evidence that this was a brittle tube failure or a failure by some stress corrosion mechanism. The unavailability of a clean fracture surface made it impossible to search for elements or compounds that caused the failure. Consequently, it would be worthwhile to examine other failed tubes, but it might be more informative to examine cross sections of tubes exposed in equivalent environments and in which cracks have initiated but not progressed through the entire tube wall. It is quite possible that a tube might be found containing cracks that have not progressed completely through the tube. A study of the morphology of any such cracks with an SEM/EDS identification of elements at the tip(s) of the crack(s) might provide enough information to identify the cracking mechanism. See Table C.4 for the composition of the deposits shown in Figure C.10 and Figure C.11.

Table C.4. Semiquantitative results of the EDS examination of the deposit shown in Figure C.10 and Figure C.11 Element Surface of deposit shown in Figure C.10 Surface of deposit shown in Figure C.11 (atomic %)

(atomic %)

Oxygen 39.83 44.1 Fluorine 41.14 29.93 Aluminum 0.32 0.5 Silicon 0.53 5.16 Chromium 1,43 2.58 Manganese 0.39 0.66 Iron 15.31 8.46 Nickel 0.13 6.33 Selenium 0.09 0.16 Molybdenum 0.83 2.13 C-11

As noted earlier in this text, previous reports on identification of molybdenum isotopes from material collected on MSRE components prior to 233U defueling indicate the molybdenum was a fission product and not from corrosion of Hastelloy N. However, no reports were found that identified the molybdenum isotopes in traps, filters and deposits after the defueling operation was completed in 2008. There are filter and tube sections that could be studied to determine the source of any molybdenum in those components, but no such studies have been reported. Since deposits were readily available in the sections of tubing examined because of the brittle tube failure, a determination of the molybdenum isotopes in one of the deposits was done. The result of that examination, shown in Table C.5, indicates the molybdenum in the deposit taken from the failed tube section was all naturally occurring molybdenum, i.e., molybdenum from corrosion of some component of the MSRE, most likely the drain tanks or their Hastelloy N internal components. It would be very meaningful to analyze additional samples, particularly samples that accumulated deposits after the 233U defueling operation was completed.

Table C.5. Identification of molybdenum isotopes in deposits collected from failed stainless steel tube. The isotopic composition of naturally occurring molybdenum is included for reference Molybdenum source 92Mo 94Mo 95Mo 96Mo 97Mo 98Mo 99Mo 100Mo Isotope distribution in naturally occurring molybdenum 14.84%

9.25%

15.92%

16.68%

9.55%

24.13%

None detected 9.63%

Isotope distribution in deposit collected Aug 2016 14.68%

9.24%

16.11%

16.67%

9.51%

24.09%

None detected 9.70%

C.1 REFERENCES ASTM STP 457. Development of Fuel Cladding for Fast Reactors, 1969, G. W. Cunningham, American Society for Testing and Materials, Philadelphia, PA.

MDO-723-0043. Assessing the Effects of Radiation Damage on Ni-Based Alloys for the Prometheus Space Reactor System, T. Angeliu, 2006, Knolls Atomic Power Laboratory, Schenectady, NY.

ORNL-TM-728. MSRE Design and Operations Report, Part I, Description of Reactor Design, R. C. Robertson, January 1965, Union Carbide Corporation, Oak Ridge, TN.

ORNL/TM-3884. The Migration of a Class of Fission Products (Noble Metals) in the Molten-Salt Reactor Program, R. J. Kedl, December 1972, Union Carbide Corporation, Oak Ridge, TN.

ORNL/TM-1853. Chemical Research and Development for Molten-Salt Breeder Reactors, W. R. Grimes, 1967, Union Carbide Corporation, Oak Ridge, TN.

ORNL/TM-2578. Processing of the MSRE Fuel and Flush Tanks, R. B. Lindauer, 1969, Union Carbide Corporation, Oak Ridge, TN.

UCOR-4807. Engineering Evaluation of Plugging Thermal Wells and Drying Drain Tank Cell Air in the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, December 2015, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

C-12

APPENDIX D.

CONCENTRATION OF HF RELEASED UNDER CONTINUOUS RGRS SYSTEM D-1

This page intentionally left blank.

D-2

D.1 EXECUTIVE

SUMMARY

The calculation below shows that regardless of the variation in the N2 purge flowrate (1 L/min. to 25 L/min.), the concentration of HF under a worst case scenario (all F2 produced converted to HF upon generation) remains below the 0.02-ppm odor threshold under a continuous purge system. This scoping calculation shows that implementing a continuous purge process is plausible, but more in depth analysis for the system would be required to verify HF release concentrations after a prolonged outage of the system or during the initial system startup. Also, using the data collected from the future gas headspace samples for a HF release calculation will provide a more accurate determination of possible release concentrations.

D.2 PURPOSE Before a continuous purge offgas treatment system can be implemented for the FDTs and FFT at MSRE, the HF concentration of the gas released through the CVS must be determined. The calculation below gives a preliminary estimation for the HF concentration in the CVS release from MSRE. The bounding case of HF concentration chosen for this calculation is a HF concentration at the lower range of detection by smell, 0.02 ppm (mg/kg), in the CVS output (air). The concentration of the HF in the CVS output will be dependent on the flow rate of the purge system, so the equation was set in Excel to calculate the emission data with a range of flowrates from 1 L/min. to 25 L/min. to show the different HF concentrations released from the stack at different flow rates.

D.3 CALCULATIONS D.3.1 Assumptions All F2 produced in the salt is converted to HF.

The fluorine inventory has been depleted significantly by repeated pumpdowns before the continuous purge was started.

Tanks are pressurized to 14.7 psig (1.0135 bar) with N2.

F2 generation rates:

2.87E-4 psig/h in FDT-1 2.17E-4 psig/h in FDT-2 Tank is at 298 K.

Gases in tank headspace behave as ideal gases.

F2 is generated at a constant rate that is independent of HF concentration in all of the tanks through radiolysis of the salt.

CVS is running at 16,000 SCFM.

Air density of CVS air is the same as ambient air.

The system is at steady-state. There is no change in HF concentration with respect to time.

D-3

The continuous purge system has:

Perfectly mixed headspace Concentration of HF in the effluent stream equals concentration in the headspace No accumulation of HF in the headspace N2 purge volumetric flow is much greater than HF volumetric generation.

D.3.2 Diagram Figure D.1 shows a diagram of the system.

Figure D.1. Drawing of the tank headspace and streams.

D.3.3 Variables 1

2 2

3 =

= 2

=

D.3.4 Equations 1

+ 1

= 1 (1) 2

+ 2

= 1 (2) 3

+ 3

= 1 (3) 1 1

+ 2 2

= 3 3 (4) 1 1

+ 2 2

= 3 3 (5)

D-4

D.3.5 Conversions 1 = 0.0689

=

() =

D.3.6 Methodology This calculation is designed to show the HF release concentrations from a continuously purged RGRS system through the CVS stack with respect to various flow rates. To do this, an overall mass balance was set up with the Fuel Drain Tank Headspace as the control volume.

=

The continuous purge system was assumed to be operating at steady state. Steady state is an appropriate assumption if the system is continuous and has been running for an extended period of time. The first step was to find the concentration of HF in the FDT headspace effluent stream. To do this, M1 was designated as the N2 purge, M2 was F2 generation rate, and M3 was the tank effluent stream. M1 and M2 were assumed to be pure streams, so the mole fraction for these streams is 1.

1

= 1 1

= 0 2

= 0 2

= 1 The F2 generation rate was treated as a constant flowrate into the tank headspace because the generation rate through radiolysis is not a chemical reaction that is dependent upon F2 concentration in the headspace. The pressure increase of F2 in each tank, calculated in UCOR-4804, was converted to a molar flow rate by the ideal gas law.

()=

2

(6)

All of the F2 generated was assumed to be converted to HF to create a worst-case scenario for HF release.

In the reaction producing HF, 1 mole of F2 creates 2 moles of HF, so:

2 2

=

= 2 (7)

The flow rates of HF into both FDTs, FDT-1 and FDT-2, were then added to the flow rate of the N2 purge to find the total mass flowing in the system. Using Equations 5, the mole fractions of the effluent stream were calculated to find the concentration of HF (ppm) in the effluent stream:

1 1

+ 2 2

= 3 3 D-5

1 1

+ 2 2

= 3 3 0 + 2 2

= 3 3 (3.37 105 + 2.54 105) 1 = 3

3 (5.91 105) 1 = 3 3 Equation 5 is the general equation for finding the mole fraction of HF. The different values for the flow rates were plugged into this equation (as M3) to find the mole fraction. Once it was found, the concentration of HF in the tank effluent stream was determined using Equation 8.

=

3 3 1000 3 2 (8)

To find the concentration of HF released from the CVS stack:

(

) =

3 3

(

)

3(

)(

) 2+ ( 3

)(

3)

(9)

The results are presented in the following section.

D.4 RESULTS The measurements and constants used in this calculation are shown in Table D.1 and Table D.2. The results of the calculation are in Table D.3, and Table D.4 shows the formulas used in Microsoft Excel to obtain the results.

In all of the Excel calculations, the release concentration was uniformly in the 0.002-ppm range. The term impacted the most by the flowrates was the concentration of the stream immediately out of the tank.

Table D.1. CVS measurements used in this calculation CVS Air generation (CFM) 16000 Volume of CVS air (m3/min) 453.0695454 F2 generation FDT-1 (psi/h) 2.87E-04 F2 generation FDT-2 (psi/h) 2.17E-04 HF concentration out of the stack limit (mg/m3) 0.01 Volumetric flow (L/h) 2265.00000 Volumetric flow (m3/h) 2.265 Molar flow HF FDT-1 (mol/min) 3.37E-05 Molar flow HF FDT-2 (mol/min) 2.54E-05 Volumetric flow HF FDT-1 (m3/min) 5.8559E-07 Volumetric flow HF FDT-2 (m3/min 4.42763E-07 D-6

Table D.2. Constants used in this calculation Constants N2 MW (g/mol) 2.80E+01 HF MW (g/mol) 20.01 R (L*bar*K-1*mol-1) 8.31E-02 N2 density (kg/m3) 1.165 HF density (kg/m3) 1.15 FDT-1 Headspace volume (L) 1265 FDT-2 Headspace volume (L) 1390 Volume of FDT-1 headspace (m3) 1.265 Volume of FD2-1 headspace (m3) 1.39 Tank temperature (K) 298 Tank pressure (bar) 1.01E+00 Air density (kg/m3) 1.225 D.5 DISCUSSION Based on the release concentration of HF from the CVS stack, a continuous purge RGRS system is feasible. The release limits for hazardous substances such as HF are 1,000 lb per species. The amount of HF released from the continuous purge from the FDTs and FFT is well below this limit. Also, the concentration at the stack is well below the 0.02-ppm odor threshold and the action level of 3 ppm, so there should not be any detrimental effects from releasing the HF into the atmosphere.

This is a basic calculation that makes many conservative assumptions. The limiting factor on this calculation is it does not account for residual fluorine in the tank or for restarting the process after an equipment failure. Because of the extremely low generation rate of F2 in the tank headspace, there is an allowance of time for the purge system to be non-operational.

Gas headspace samples from the FDTs and FFT will be crucial in forming a model to test the feasibility of a continuous purge fluorine removal system. These samples will give critical information, such as exact HF concentration in the tanks, which are needed to build an accurate model for the system and will provide insight into the conservatism related to the fluorine generation rate.

D.6 REFERENCES UCOR-4804. Fluorine Inventory Control in the MSRE Defueled Salt Drain Tanks at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, November 2015, URS l CH2M Oak Ridge, LLC, Oak Ridge, TN.

D-7

D-8 Table D.3. Results of HF concentration calculations N2 Purge Flow Rate (L/min)

N2 Purge Flow Rate (m3/min)

Molar Flow N2 (moles/min)

Molar Flow Out of Tanks Mole Fraction N2 in Effluent Stream Mole Fraction HF in Effluent Stream HF Concentration in Tank Effluent Stream (mol/m3)

HF Concentration in Tank Effluent Stream ppm Stack Release Concentration (ppm)

A B

C D

E F

G H

I 18 1

0.001 4.16E-02 4.16E-02 0.9986 0.0014 0.059 1015.111875 0.002130777 19 2

0.002 8.32E-02 8.32E-02 0.9993 0.0007 0.030 507.5559374 0.002130772 20 3

0.003 1.25E-01 1.25E-01 0.9995 0.0005 0.020 338.3706249 0.002130768 21 4

0.004 1.66E-01 1.66E-01 0.9996 0.0004 0.015 253.7779687 0.002130763 22 5

0.005 2.08E-01 2.08E-01 0.9997 0.0003 0.012 203.022375 0.002130759 23 6

0.006 2.49E-01 2.50E-01 0.9998 0.0002 0.010 169.1853125 0.002130754 24 7

0.007 2.91E-01 2.91E-01 0.9998 0.0002 0.008 145.0159821 0.00213075 25 8

0.008 3.33E-01 3.33E-01 0.9998 0.0002 0.007 126.8889843 0.002130746 26 9

0.009 3.74E-01 3.74E-01 0.9998 0.0002 0.007 112.7902083 0.002130741 27 10 0.01 4.16E-01 4.16E-01 0.9999 0.0001 0.006 101.5111875 0.002130737 28 11 0.011 4.57E-01 4.57E-01 0.9999 0.0001 0.005 92.28289771 0.002130732 29 12 0.012 4.99E-01 4.99E-01 0.9999 0.0001 0.005 84.59265623 0.002130728 30 13 0.013 5.41E-01 5.41E-01 0.9999 0.0001 0.005 78.08552883 0.002130723 31 14 0.014 5.82E-01 5.82E-01 0.9999 0.0001 0.004 72.50799105 0.002130719 32 15 0.015 6.24E-01 6.24E-01 0.9999 0.0001 0.004 67.67412498 0.002130714 33 16 0.016 6.65E-01 6.65E-01 0.9999 0.0001 0.004 63.44449217 0.00213071 34 17 0.017 7.07E-01 7.07E-01 0.9999 0.0001 0.003 59.71246322 0.002130705 35 18 0.018 7.48E-01 7.48E-01 0.9999 0.0001 0.003 56.39510415 0.002130701 36 19 0.019 7.90E-01 7.90E-01 0.9999 0.0001 0.003 53.42694078 0.002130696 37 20 0.02 8.32E-01 8.32E-01 0.9999 0.0001 0.003 50.75559374 0.002130692 38 21 0.021 8.73E-01 8.73E-01 0.9999 0.0001 0.003 48.3386607 0.002130687 39 22 0.022 9.15E-01 9.15E-01 0.9999 0.0001 0.003 46.14144885 0.002130683 40 23 0.023 9.56E-01 9.56E-01 0.9999 0.0001 0.003 44.1352989 0.002130678 41 24 0.024 9.98E-01 9.98E-01 0.9999 0.0001 0.002 42.29632811 0.002130674 42 25 0.025 1.04E+00 1.04E+00 0.9999 0.0001 0.002 40.60447499 0.002130669

• Calculation spreadsheet column and row labels

D-9 Table D.4. Spreadsheet formulas N2 Purge Flow Rate (L/min)

N2 Purge Flow Rate (m3/min)

Molar Flow N2 (moles/min)

Molar Flow Out of Tanks Mole Fraction N2 in Effluent Stream Mole Fraction HF in Effluent Stream HF Concentration in Tank Effluent Stream (mol/m3)

HF Concentration in Tank Effluent Stream (ppm)

Stack Release Concentration (ppm)

A B

C D

E F

G H

I 18 1

=A18*0.001 =(B18*$E$7*1000)/

$E$4

=C18+$B$11+

$B$12

=C18/D18

=($B$11+$B$12)/

D18

=($B$11+$B$12)/

B18

=G18*$E$5*1000*

(1/$E$7)

=((F18*($B$11+$B$12+C18))*$E$5*1000)/

((B18*$E$7)+($B$5*$E$15))

19 2

=A19*0.001 =(B19*$E$7*1000)/

$E$4

=C19+$B$11+

$B$12

=C19/D19

=($B$11+$B$12)/

D19

=($B$11+$B$12)/

B19

=G19*$E$5*1000*

(1/$E$7)

=((F19*($B$11+$B$12+C19))*$E$5*1000)/

((B19*$E$7)+($B$5*$E$15))

20 3

=A20*0.001 =(B20*$E$7*1000)/

$E$4

=C20+$B$11+

$B$12

=C20/D20

=($B$11+$B$12)/

D20

=($B$11+$B$12)/

B20

=G20*$E$5*1000*

(1/$E$7)

=((F20*($B$11+$B$12+C20))*$E$5*1000)/

((B20*$E$7)+($B$5*$E$15))

21 4

=A21*0.001 =(B21*$E$7*1000)/

$E$4

=C21+$B$11+

$B$12

=C21/D21

=($B$11+$B$12)/

D21

=($B$11+$B$12)/

B21

=G21*$E$5*1000*

(1/$E$7)

=((F21*($B$11+$B$12+C21))*$E$5*1000)/

((B21*$E$7)+($B$5*$E$15))

22 5

=A22*0.001 =(B22*$E$7*1000)/

$E$4

=C22+$B$11+

$B$12

=C22/D22

=($B$11+$B$12)/

D22

=($B$11+$B$12)/

B22

=G22*$E$5*1000*

(1/$E$7)

=((F22*($B$11+$B$12+C22))*$E$5*1000)/

((B22*$E$7)+($B$5*$E$15))

23 6

=A23*0.001 =(B23*$E$7*1000)/

$E$4

=C23+$B$11+

$B$12

=C23/D23

=($B$11+$B$12)/

D23

=($B$11+$B$12)/

B23

=G23*$E$5*1000*

(1/$E$7)

=((F23*($B$11+$B$12+C23))*$E$5*1000)/

((B23*$E$7)+($B$5*$E$15))

24 7

=A24*0.001 =(B24*$E$7*1000)/

$E$4

=C24+$B$11+

$B$12

=C24/D24

=($B$11+$B$12)/

D24

=($B$11+$B$12)/

B24

=G24*$E$5*1000*

(1/$E$7)

=((F24*($B$11+$B$12+C24))*$E$5*1000)/

((B24*$E$7)+($B$5*$E$15))

25 8

=A25*0.001 =(B25*$E$7*1000)/

$E$4

=C25+$B$11+

$B$12

=C25/D25

=($B$11+$B$12)/

D25

=($B$11+$B$12)/

B25

=G25*$E$5*1000*

(1/$E$7)

=((F25*($B$11+$B$12+C25))*$E$5*1000)/

((B25*$E$7)+($B$5*$E$15))

26 9

=A26*0.001 =(B26*$E$7*1000)/

$E$4

=C26+$B$11+

$B$12

=C26/D26

=($B$11+$B$12)/

D26

=($B$11+$B$12)/

B26

=G26*$E$5*1000*

(1/$E$7)

=((F26*($B$11+$B$12+C26))*$E$5*1000)/

((B26*$E$7)+($B$5*$E$15))

27 10

=A27*0.001 =(B27*$E$7*1000)/

$E$4

=C27+$B$11+

$B$12

=C27/D27

=($B$11+$B$12)/

D27

=($B$11+$B$12)/

B27

=G27*$E$5*1000*

(1/$E$7)

=((F27*($B$11+$B$12+C27))*$E$5*1000)/

((B27*$E$7)+($B$5*$E$15))

28 11

=A28*0.001 =(B28*$E$7*1000)/

$E$4

=C28+$B$11+

$B$12

=C28/D28

=($B$11+$B$12)/

D28

=($B$11+$B$12)/

B28

=G28*$E$5*1000*

(1/$E$7)

=((F28*($B$11+$B$12+C28))*$E$5*1000)/

((B28*$E$7)+($B$5*$E$15))

29 12

=A29*0.001 =(B29*$E$7*1000)/

$E$4

=C29+$B$11+

$B$12

=C29/D29

=($B$11+$B$12)/

D29

=($B$11+$B$12)/

B29

=G29*$E$5*1000*

(1/$E$7)

=((F29*($B$11+$B$12+C29))*$E$5*1000)/

((B29*$E$7)+($B$5*$E$15))

30 13

=A30*0.001 =(B30*$E$7*1000)/

$E$4

=C30+$B$11+

$B$12

=C30/D30

=($B$11+$B$12)/

D30

=($B$11+$B$12)/

B30

=G30*$E$5*1000*

(1/$E$7)

=((F30*($B$11+$B$12+C30))*$E$5*1000)/

((B30*$E$7)+($B$5*$E$15))

31 14

=A31*0.001 =(B31*$E$7*1000)/

$E$4

=C31+$B$11+

$B$12

=C31/D31

=($B$11+$B$12)/

D31

=($B$11+$B$12)/

B31

=G31*$E$5*1000*

(1/$E$7)

=((F31*($B$11+$B$12+C31))*$E$5*1000)/

((B31*$E$7)+($B$5*$E$15))

32 15

=A32*0.001 =(B32*$E$7*1000)/

$E$4

=C32+$B$11+

$B$12

=C32/D32

=($B$11+$B$12)/

D32

=($B$11+$B$12)/

B32

=G32*$E$5*1000*

(1/$E$7)

=((F32*($B$11+$B$12+C32))*$E$5*1000)/

((B32*$E$7)+($B$5*$E$15))

33 16

=A33*0.001 =(B33*$E$7*1000)/

$E$4

=C33+$B$11+

$B$12

=C33/D33

=($B$11+$B$12)/

D33

=($B$11+$B$12)/

B33

=G33*$E$5*1000*

(1/$E$7)

=((F33*($B$11+$B$12+C33))*$E$5*1000)/

((B33*$E$7)+($B$5*$E$15))

34 17

=A34*0.001 =(B34*$E$7*1000)/

$E$4

=C34+$B$11+

$B$12

=C34/D34

=($B$11+$B$12)/

D34

=($B$11+$B$12)/

B34

=G34*$E$5*1000*

(1/$E$7)

=((F34*($B$11+$B$12+C34))*$E$5*1000)/

((B34*$E$7)+($B$5*$E$15))

Table D.4. Spreadsheet formulas (cont.)

D-10 N2 Purge Flow Rate (L/min)

N2 Purge Flow Rate (m3/min)

Molar Flow N2 (moles/min)

Molar Flow Out of Tanks Mole Fraction N2 in Effluent Stream Mole Fraction HF in Effluent Stream HF Concentration in Tank Effluent Stream (mol/m3)

HF Concentration in Tank Effluent Stream (ppm)

Stack Release Concentration (ppm)

A B

C D

E F

G H

I 35 18

=A35*0.001 =(B35*$E$7*1000)/

$E$4

=C35+$B$11+

$B$12

=C35/D35

=($B$11+$B$12)/

D35

=($B$11+$B$12)/

B35

=G35*$E$5*1000*

(1/$E$7)

=((F35*($B$11+$B$12+C35))*$E$5*1000)/

((B35*$E$7)+($B$5*$E$15))

36 19

=A36*0.001 =(B36*$E$7*1000)/

$E$4

=C36+$B$11+$

B$12

=C36/D36

=($B$11+$B$12)/

D36

=($B$11+$B$12)/

B36

=G36*$E$5*1000*

(1/$E$7)

=((F36*($B$11+$B$12+C36))*$E$5*1000)/

((B36*$E$7)+($B$5*$E$15))

37 20

=A37*0.001 =(B37*$E$7*1000)/

$E$4

=C37+$B$11+$

B$12

=C37/D37

=($B$11+$B$12)/

D37

=($B$11+$B$12)/

B37

=G37*$E$5*1000*

(1/$E$7)

=((F37*($B$11+$B$12+C37))*$E$5*1000)/

((B37*$E$7)+($B$5*$E$15))

38 21

=A38*0.001 =(B38*$E$7*1000)/

$E$4

=C38+$B$11+$

B$12

=C38/D38

=($B$11+$B$12)/

D38

=($B$11+$B$12)/

B38

=G38*$E$5*1000*

(1/$E$7)

=((F38*($B$11+$B$12+C38))*$E$5*1000)/

((B38*$E$7)+($B$5*$E$15))

39 22

=A39*0.001 =(B39*$E$7*1000)/

$E$4

=C39+$B$11+$

B$12

=C39/D39

=($B$11+$B$12)/

D39

=($B$11+$B$12)/

B39

=G39*$E$5*1000*

(1/$E$7)

=((F39*($B$11+$B$12+C39))*$E$5*1000)/

((B39*$E$7)+($B$5*$E$15))

40 23

=A40*0.001 =(B40*$E$7*1000)/

$E$4

=C40+$B$11+$

B$12

=C40/D40

=($B$11+$B$12)/

D40

=($B$11+$B$12)/

B40

=G40*$E$5*1000*

(1/$E$7)

=((F40*($B$11+$B$12+C40))*$E$5*1000)/

((B40*$E$7)+($B$5*$E$15))

41 24

=A41*0.001 =(B41*$E$7*1000)/

$E$4

=C41+$B$11+$

B$12

=C41/D41

=($B$11+$B$12)/

D41

=($B$11+$B$12)/

B41

=G41*$E$5*1000*

(1/$E$7)

=((F41*($B$11+$B$12+C41))*$E$5*1000)/

((B41*$E$7)+($B$5*$E$15))

42 25

=A42*0.001 =(B42*$E$7*1000)/

$E$4

=C42+$B$11+$

B$12

=C42/D42

=($B$11+$B$12)/

D42

=($B$11+$B$12)/

B42

=G42*$E$5*1000*

(1/$E$7)

=((F42*($B$11+$B$12+C42))*$E$5*1000)/

((B42*$E$7)+($B$5*$E$15))

• Calculation spreadsheet column and row labels

APPENDIX E.

PREVIOUS MSRE SALT REMOVAL STUDIES E-1

This page intentionally left blank.

E-2

BJC/OR-2011/R1. Molten Salt Reactor Experiment Post-Fuel Salt Disposition Inventory for Stabilization/

Deactivation, Oak Ridge National Laboratory, April 2005, Bechtel Jacobs Company, LLC, Oak Ridge, TN.

BJC/OR-2883. MSRE Salt Removal Planning Roadmap at Oak Ridge National Laboratory, January 2008, Bechtel Jacobs Company LLC, Oak Ridge, TN.

BJC/OR-3301. Engineering Evaluation of the Proposed Alternative Salt Transfer Method for the Molten Salt Reactor Experiment, Oak Ridge National Laboratory, Oak Ridge, Tennessee, September 2009, Bechtel Jacobs Company, LLC, Oak Ridge, TN.

CONF-960271-2. Disposition of the Fluoride Fuel and Flush Salts from the Molten Salt Reactor Experiment at Oak Ridge National Laboratory, Prepared for Presentation at the AIChE 1996 Spring Meeting in New Orleans, LA, Lockheed Martin Energy Systems, Inc., Oak Ridge, TN.

CONF-96061216-7. Characterization of the Molten Salt Reactor Experiment Fuel and Flush Salts, Prepared for Presentation at the 1996 American Nuclear Society Meeting in Reno, Nevada, June 1996, Lockheed Martin Energy Systems, Inc., Oak Ridge, TN.

DOE/OR/01-2496&D1. Engineering Evaluation of Options for the Molten Salt Reactor Experiment Defueled Coolant Salts, Oak Ridge, Tennessee, December 2010, Bechtel Jacobs Company LLC, Oak Ridge, TN.

DOE/OR/01-2560&D2. Remediation Strategy Plan for the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, Oak Ridge, Tennessee, January 2013, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

DOE/OR/02-1671&D2. Record of Decision for Interim Action to Remove Fuel and Flush Salts from the Molten Salt Reactor Experiment Facility at the Oak Ridge National Laboratory, June 1998, Bechtel Jacobs Company LLC, Oak Ridge, TN.

ORNL/CP-98146. Removal of Uranium and Salt from the Molten Salt Reactor Experiment, 1998, Bechtel Jacobs Company, LLC, Oak Ridge, TN.

ORNL-TM-2578. Processing of the MSRE Flush and Fuel Salts, August 1969, Union Carbide Corporation, Oak Ridge, TN.

ORNL/RAP-17. Decommissioning of the Molten Salt Reactor ExperimentA Technical Evaluation, January 1988, Martin Marietta Energy Systems, Inc., Oak Ridge, TN.

LA-UR-10-07278. TRU Waste Determination of MSRE Salt for Disposal Eligibility in WIPP, 2010, Los Alamos National Laboratory, Los Alamos, NM.

LA-UR-13-22367. Evaluation of Molten Salt Reactor Experiment (MSRE) Remnant Salt Properties for Disposal Eligibility at the WIPP, March 2013, Los Alamos National Laboratory, Los Alamos, NM.

UCOR-4424. Potential Cost Savings During Salt Disposition from the Molten Salt Reactor Experiment at the Oak Ridge National Laboratory, issued August 2013, URS l CH2M Oak Ridge LLC, Oak Ridge, TN.

E-3

This page intentionally left blank.

E-4

STJ-02MSRE-D992 RECORD COPY DISTRIBUTION FileDMCRC