Hydrogen Hazards in Nuclear Facilities: Research and Testing Needs CNS Perspective American Nuclear Society Conferenc

1. Hydrogen Hazards in Nuclear Facilities: Research and Testing NeedsCNS PerspectiveAmerican Nuclear Society ConferenceJune 2010 Richard H. Lagdon, Jr. Chief of Nuclear Safety U.S. Department of Energy Information for this presentation gathered from personal communication with Vic Callahan and Greg Jones (Hanford), Jeff Woody (Oak Ridge), Arnie Preece and Gordon McClellan (Idaho), Neal Askew, Jay Lavender, and John Schwenker (Savannah River), Roger Nelson (WIPP), Joe Shepherd (Caltech).Information for this presentation gathered from personal communication with Vic Callahan and Greg Jones (Hanford), Jeff Woody (Oak Ridge), Arnie Preece and Gordon McClellan (Idaho), Neal Askew, Jay Lavender, and John Schwenker (Savannah River), Roger Nelson (WIPP), Joe Shepherd (Caltech).

2. 2

3. Facilities with Hydrogen Issues Liquid Waste Storage Tank Farms (Hanford, Savannah River) Waste Treatment and Processing Defense Waste Processing Facility (DWPF), Salt Waste Processing Facility (SWPF), Waste Treatment and Immobilization Plant (WTP), Integrated Waste Treatment Unit (IWTU) Packaging/Temporary Storage Burial Grounds, TRU drums Long Term Storage/Permanent Disposal WIPP, NTS Reactors 3 High level waste is stored in million gallon tanks, then sent to other facilities for processing. Some solid waste is stored temporarily on site before being processed and sent to permanent disposal sites. In nuclear reactors, hydrogen generation is also an issue after accidents (e.g. cladding+H2O reaction post LOCA). High level waste is stored in million gallon tanks, then sent to other facilities for processing. Some solid waste is stored temporarily on site before being processed and sent to permanent disposal sites. In nuclear reactors, hydrogen generation is also an issue after accidents (e.g. cladding+H2O reaction post LOCA).

4. Hydrogen Generation at DOE Nuclear Facilities Radiolytic Decomposition Thermolytic Decomposition Catalytic Decomposition Biological Hydrogen Production Use of Hydrogen in Process 4 Radiolytic and thermolytic decomposition of hydrogen bearing compounds (e.g., water, hydrogen peroxide, organics) Catalytic decomposition of compounds occur when using chemical cleaning, for example, oxalic acid and hydrogen peroxide when removing heel in waste tanks Hydrogen used as part of process at HFIR to produce cold neutrons, at IWTU in reduction reactions This spent fuel is usually stored in water pools, awaiting permanent disposal or reprocessing. The yield of hydrogen resulting from the irradiation of water with � and ? radiation is low (G-values = <1 molecule per 100 electronvolts of absorbed energy) but this is largely due to the rapid reassociation of the species arising during the initial radiolysis. If impurities are present or if physical conditions are created that prevent the establishment of a chemical equilibrium, the net production of hydrogen can be greatly enhanced.Radiolytic and thermolytic decomposition of hydrogen bearing compounds (e.g., water, hydrogen peroxide, organics) Catalytic decomposition of compounds occur when using chemical cleaning, for example, oxalic acid and hydrogen peroxide when removing heel in waste tanks Hydrogen used as part of process at HFIR to produce cold neutrons, at IWTU in reduction reactions This spent fuel is usually stored in water pools, awaiting permanent disposal or reprocessing. The yield of hydrogen resulting from the irradiation of water with � and ? radiation is low (G-values = <1 molecule per 100 electronvolts of absorbed energy) but this is largely due to the rapid reassociation of the species arising during the initial radiolysis. If impurities are present or if physical conditions are created that prevent the establishment of a chemical equilibrium, the net production of hydrogen can be greatly enhanced.

5. Hydrogen Generation Rates 5 6M containers had U, Pu, and �other debris� including soil samples, metals, plastic, oxides, uranium fissium alloy. (Fissium = �An equilibrium mixture of fission products in reactor fuel that can improve the stability of uranium and uranium-plutonium fuel alloys under fast-neutron irradiation.�) WIPP does not have a problem. Samples from closed panels (~80,000 containers) show hydrogen concentration << LFL. Oxygen consumed in panel, so O2 unavailable for reactions. SRS has �rapid� and �slow� generation tanks, depending on activity, chemistry, etc. Measured hydrogen concentration has been up to 25% LFL. Data come from controlled experiments on actual and simulated Hanford waste. Historical samples drawn from DWPF vessels have been less than 10% LFL of H2.6M containers had U, Pu, and �other debris� including soil samples, metals, plastic, oxides, uranium fissium alloy. (Fissium = �An equilibrium mixture of fission products in reactor fuel that can improve the stability of uranium and uranium-plutonium fuel alloys under fast-neutron irradiation.�) WIPP does not have a problem. Samples from closed panels (~80,000 containers) show hydrogen concentration << LFL. Oxygen consumed in panel, so O2 unavailable for reactions. SRS has �rapid� and �slow� generation tanks, depending on activity, chemistry, etc. Measured hydrogen concentration has been up to 25% LFL. Data come from controlled experiments on actual and simulated Hanford waste. Historical samples drawn from DWPF vessels have been less than 10% LFL of H2.

6. Hanford Tank Wastes H2 Generation Rates in Tanks S-102 and SY-103 6 Using the thermal and radiolytic activation parameters for gas generation in actual tank waste, the rate of hydrogen generation in the entire tank can be estimated for S-102. The rate of hydrogen generation in tank material is a sum of thermal and radiolytic rates. The thermal rate, at 4 1 "C, is 8.6E-8 mol/kg/day. The radiolytic rate is 1.6E-7 mol/kg/day, using a G(H,) value of 0.01 7 molecules/l 00 eV and a tank dose rate of 207 R/hr. The sum of the thermal and radiolytic rates is 2.5E-07 mol/kg/day, which translates to a hydrogen generation rate for the entire tank of 1 .O mol/day. An independent estimate of the rate of hydrogen formation is available.'") The rate of hydrogen formation in Tank S-102 has been estimated by assuming that the tank generation rate is equal to the amount of hydrogen released fiom the tank. This gives an estimated hydrogen generation rate of 3 ft3/day. Assuming this gas is at 25OC and 1atmosphere gives an observed rate of 3.8 mol H,/day fiom the entire tank. The waste is being analyzed for specific organic components. Separate organic analysis samples were taken before and after heating and radiolysis to help identify the organic species responsible for gas generation. By following the specific organic species present and their concentration changes as a function of heating and irradiation, together with the results of measurements of the gases formed during the heating and irradiation treatments, a better under- standing of the organics responsible for gas generation is possible. The organic analysis of the waste will be reported in a subsequent document. A long-term gas generation test using S-102 waste conducted under thermal and radiolytic conditions that best match the tank waste temperature (43�C) and dose-rate (207 R/h) is being conducted. The results of this experiment will be compared with the predicted gas generation behavior using higher temperatures and dose rates. This experiment is still in progress, and results will be reported in a subsequent gas generation report. Using the thermal and radiolytic activation parameters for gas generation in actual tank waste, the rate of hydrogen generation in the entire tank can be estimated for S-102. The rate of hydrogen generation in tank material is a sum of thermal and radiolytic rates. The thermal rate, at 4 1 "C, is 8.6E-8 mol/kg/day. The radiolytic rate is 1.6E-7 mol/kg/day, using a G(H,) value of 0.01 7 molecules/l 00 eV and a tank dose rate of 207 R/hr. The sum of the thermal and radiolytic rates is 2.5E-07 mol/kg/day, which translates to a hydrogen generation rate for the entire tank of 1 .O mol/day. An independent estimate of the rate of hydrogen formation is available.'") The rate of hydrogen formation in Tank S-102 has been estimated by assuming that the tank generation rate is equal to the amount of hydrogen released fiom the tank. This gives an estimated hydrogen generation rate of 3 ft3/day. Assuming this gas is at 25OC and 1atmosphere gives an observed rate of 3.8 mol H,/day fiom the entire tank. The waste is being analyzed for specific organic components. Separate organic analysis samples were taken before and after heating and radiolysis to help identify the organic species responsible for gas generation. By following the specific organic species present and their concentration changes as a function of heating and irradiation, together with the results of measurements of the gases formed during the heating and irradiation treatments, a better under- standing of the organics responsible for gas generation is possible. The organic analysis of the waste will be reported in a subsequent document. A long-term gas generation test using S-102 waste conducted under thermal and radiolytic conditions that best match the tank waste temperature (43�C) and dose-rate (207 R/h) is being conducted. The results of this experiment will be compared with the predicted gas generation behavior using higher temperatures and dose rates. This experiment is still in progress, and results will be reported in a subsequent gas generation report.

7. Quantifying Hydrogen Hazards Samples Tank, vessel, and drum head space Monitor hydrogen levels Calculations/Models/Correlations H2 generation rate H2 deflagration/detonation events Test Programs/Experiments Prevention Mitigation Historical Knowledge These are largely done on case-by-case basis 7 Samples taken per surveillance requirements and as needed, as in drums prior to shipping. Calculations, models, and correlations for generation rates based on tests and experiments done on materials of interest (waste, water, acids, plastic, etc.). Calculations, models, and correlations for deflagration/detonation events based on tests and experiments and the best engineering knowledge available. The multi-physics models necessary are complicated. Always uncertainties associated with models, calculations, correlations, and other numerical methods. Enough margin is used to ensure safety. 3. Tests and experiments for prevention purposes is to understand the H2 generation and depletion, and for mitigation is to understand how SSC�s will handle deflagrations and detonations. Usually these are very narrow in scope and address only specific immediate needs. Historical knowledge of contents is sometimes the best we can get, but not always 100% reliableSamples taken per surveillance requirements and as needed, as in drums prior to shipping. Calculations, models, and correlations for generation rates based on tests and experiments done on materials of interest (waste, water, acids, plastic, etc.). Calculations, models, and correlations for deflagration/detonation events based on tests and experiments and the best engineering knowledge available. The multi-physics models necessary are complicated. Always uncertainties associated with models, calculations, correlations, and other numerical methods. Enough margin is used to ensure safety. 3. Tests and experiments for prevention purposes is to understand the H2 generation and depletion, and for mitigation is to understand how SSC�s will handle deflagrations and detonations. Usually these are very narrow in scope and address only specific immediate needs. Historical knowledge of contents is sometimes the best we can get, but not always 100% reliable

8. Managing Hydrogen Hazards�Current Methods Combination of engineered and administrative controls Prevent hydrogen buildup above LFL Mitigate in case of detonation/deflagration Graded approach based on hydrogen generation rates Uncertainties lead to varying degrees of conservatisms 8

9. Managing Hydrogen Hazards�Current Methods Monitor hydrogen levels Vent/Purge Manage agitation and mixing Monitor and manage chemistry and radioactivity (Waste Acceptance Criteria) Hydrogen Generation Rate Control Programs Engineered SSCs to withstand pressure of detonation/deflagration events 9 Operational parameters (e.g., flushing, venting, monitoring) may be more practical in the long run for many DOE facilities if too many uncertainties exist in the calculationsOperational parameters (e.g., flushing, venting, monitoring) may be more practical in the long run for many DOE facilities if too many uncertainties exist in the calculations

10. Research Needs Maintain expertise with core research program Better understanding of basic phenomena Ignition Radiolysis/Thermolysis Release of Confined Hydrogen Better understanding of impacts and outcomes of detonation/deflagration events Better database of empirical data 10 Cannot forget lessons learned from past experience. Workforce attrition is happening. Need consistent long-term effort. Basic Phenomena Ignition�currently assumption is that as soon as LFL reached, ignition occurs. What is probability that ignition will happen? Understanding basic phenomena of radiolysis and thermolysis will reduce burden of investigating every case. Basic reactions, reverse reactions, adsorbed moisture, fate of oxygen, degradation of material. How much is confined, what basic mechanisms release it? Multi-physics problems like deflagration/detonation in pipes and vessels are very complicated. Using best data and engineering knowledge available, but not every case is bound by information (e.g., stress/strain over short period vs. long periods). Currently, lots of time and money spent on looking at specific problems at specific facilities. Need consistent-long term effort in basic research. Cannot forget lessons learned from past experience. Workforce attrition is happening. Need consistent long-term effort. Basic Phenomena Ignition�currently assumption is that as soon as LFL reached, ignition occurs. What is probability that ignition will happen? Understanding basic phenomena of radiolysis and thermolysis will reduce burden of investigating every case. Basic reactions, reverse reactions, adsorbed moisture, fate of oxygen, degradation of material. How much is confined, what basic mechanisms release it? Multi-physics problems like deflagration/detonation in pipes and vessels are very complicated. Using best data and engineering knowledge available, but not every case is bound by information (e.g., stress/strain over short period vs. long periods). Currently, lots of time and money spent on looking at specific problems at specific facilities. Need consistent-long term effort in basic research.

11. Conclusion Hydrogen generation can create safety issues at its nuclear facilities DOE is managing hydrogen, but largely done on case-by-case bases�labor intensive effort Not a simple solution Cannot lose core knowledge Better understanding of basic phenomena will streamline efforts 11

12. References WSRC-TR-2007-00200, Evaluation of Radiolysis-Induced Hydrogen Generation in DOT 6M Drums from INTEC, June 2007 WSRC-TR-2004-00468, Radiolytic Hydrogen Generation in Savannah River Site (SRS) High Level Waste Tanks�Comparison of SRS and Hanford Modeling Predictions, August 2004 M. Devarakonda, Letter to D. Mercer, Estimation of Hydrogen Generation Rates from Radiolysis in WIPP Panels, 26 July 2006 WTP-RPT-115, Gas Generation Testing and Support for the Hanford Waste Treatment and Immobilization Plant, June 2004 WSRC-SA-2002-00007 Rev. 10, Concentration, Storage, and Transfer Facilities Documented Safety Analysis, March 2009 12

13. References WSRC-TR-2003-00401, Waste Tank Heel Chemical Cleaning Summary, 9 September 2003 24590-WTP-PSAR-ESH-0 1-002-02 Rev. 4e, Preliminary Documented Safety Analysis to Support Construction Authorization; PT Facility Specific Information, 2 July 2009 WSRC-SA-6 Rev. 28, Final Safety Analysis Report Savannah River Site Defense Waste Processing Facility, July 2009 Kline, D. A. and Wentink, M. G., Streamlining Hydrogen Controls in Piping Systems, EFCOG Conference, Knoxville, TN 24 � 29 April 2010 Askew, N. M., Radiolytically-generated Hydrogen: SRS Experience, JOWOG Technical Meeting, 21 April 2009 13