ML24142A002
| ML24142A002 | |
| Person / Time | |
|---|---|
| Site: | 07109362 |
| Issue date: | 03/05/2024 |
| From: | Orano NCS GmbH |
| To: | Office of Nuclear Material Safety and Safeguards |
| Shared Package | |
| ML24142A001 | List: |
| References | |
| 0023-SVK-2024-028, EPID L-2024-LLA-0010 | |
| Download: ML24142A002 (1) | |
Text
Safety Analysis Report
Thermal Analysis of the DN30 package for the Transport of Uranium Hexafluoride 0023-BSH-2016-002-Appendix-2.2.2.3-Rev. 5
Prepared Checked Released
F. Schner M. Hennebach M. Hennebach
30.04.2024 03.05.2024 03.05.2024
1 / 100
0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
List of Revisions
Rev. Revision No.
0 Original
1 Full revision due to design change of added Microtherm
2 Full revision due to change of calculation program to ANSYS Workbench
3 Changes due to remarks from IRSN, as well as updated material properties for the intu-mescent material
4 Chapter 11 added for the design change to remove the housing Updated the report layout to the Orano corporate design Updated regulations Description of geometry in section 6.2 updated to improve clarity More differentiated definition of the admissible temperatures for the 30B cylinder and its contents, calculations for admissible te mperature from main part of the SAR added to section 9.7 as well Changed reference [BAILEY] to [FRANSSEN] as old reference is not available anymore Changes aside from layout and regulation updates are marked with a line on the right
5 Updated Figure 6-3 Updated calculations for Table 9-11 and Table 9-12
2 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table of Contents
List of Figures 6
List of Tables 9
1 Introduction 11
2 Package contents 12
3 Objective of proof 12 3.1 Verification for all types of packages........................................................................... 12 3.2 Verification for packages containing uranium hexafluoride......................................... 12 3.3 Verification for type B(U) packages............................................................................. 12 3.4 Verification for packages containing fissile material.................................................... 12 3.5 Admissible temperatures of the DN30 package.......................................................... 13 3.5.1 Admissible temperatures of the 30B cylinder............................................................... 13 3.5.2 Admissible temperatures of the contents of the 30B cylinder...................................... 13 3.5.3 Admissible component temperatures of the DN30 package........................................ 14
4 Assumptions for the thermal analysis 15 4.1 Basic assumptions...................................................................................................... 15 4.2 Assumptions for RCT and NCT................................................................................... 15 4.3 Assumptions for ACT.................................................................................................. 16
5 Software 16 5.1 Benchmarking/Validation............................................................................................ 16
6 Calculation model 17 6.1 Coordinate system...................................................................................................... 17 6.2 Geometry.................................................................................................................... 17 6.3 Thermal contacts......................................................................................................... 22 6.4 Material properties...................................................................................................... 23 6.4.1 UF6................................................................................................................................ 23 6.4.2 Carbon Steel (30B cylinder)......................................................................................... 24 6.4.3 Stainless steel (DN30 PSP).......................................................................................... 25 6.4.4 Foam............................................................................................................................. 26 6.4.5 Intumescent material.................................................................................................... 29 6.4.6 Air.................................................................................................................................. 32 6.4.7 Microporous thermal insulation Microtherm.................................................................. 34 6.5 Initial temperatures...................................................................................................... 34
3 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.6 Heat generation........................................................................................................... 35 6.6.1 Heat generation of the radioactive content................................................................... 35 6.6.2 Heat generation because of incineration of the foam................................................... 35 6.7 Solar insolation............................................................................................................ 37 6.8 Heat transfer to ambient.............................................................................................. 38 6.8.1 Ambient temperature.................................................................................................... 38 6.8.2 Radiation....................................................................................................................... 38 6.8.3 Convection.................................................................................................................... 38 6.9 Heat transfer in gaps................................................................................................... 40 6.9.1 Radiation....................................................................................................................... 40 6.9.2 Conduc tion.................................................................................................................... 40 6.9.3 Convection.................................................................................................................... 40
7 Benchmark calculations 42 7.1 Parameters for the benchmark calculation.................................................................. 42 7.1.1 Geometry...................................................................................................................... 42 7.1.2 Material properties........................................................................................................ 42 7.1.3 Initial temperatures....................................................................................................... 42 7.1.4 Heat generation............................................................................................................ 42 7.1.5 Solar insolation............................................................................................................. 42 7.1.6 Heat transfer to the ambient......................................................................................... 43 7.1.7 Heat transfer in gaps.................................................................................................... 48 7.2 Results of the analysis for Benchmark 1..................................................................... 49 7.3 Results of the analysis for Benchmark 2..................................................................... 56 7.4 Conclusion for the benchmark calculations................................................................. 59
8 Calculation for RCT and NCT 60 8.1 Results without solar insolation................................................................................... 60 8.2 Results with solar insolation........................................................................................ 60
9 Calculations for ACT 62 9.1 Empty 30B cylinder..................................................................................................... 62 9.2 Filled 30B cylinder....................................................................................................... 66 9.3 Partially filled 30B cylinder.......................................................................................... 70 9.4 Sensitivity analysis of reduced thi ckness of the Microtherm layer.............................. 72 9.5 Sensitivity analysis of extrapolated material properties of the foam........................... 78 9.6 Sensitivity analysis of parameters controlling the burning of the foam....................... 81
4 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.7 Pressure build-up in the 30B cylinder......................................................................... 82
10 Proof of the package DN30 to meet the requirements of ADR and IAEA regulations 88 10.1 Ambient temperatures and pressures......................................................................... 88 10.2 Rupture of the containment system............................................................................ 88 10.3 Temperature of the accessible surface....................................................................... 88 10.4 Influence of the thermal test on the shielding analysis................................................ 88 10.5 Influence of the thermal test on the containment analysis.......................................... 88 10.6 Influence of the thermal test on the criticality safety analysis..................................... 88 10.7 Component temperatures of the DN30 package......................................................... 89
11 Thermal Analysis without the valve housing 90 11.1 Evaluation for RCT and NCT...................................................................................... 91 11.2 Evaluation for ACT based on the results of the experimental fire tests...................... 91 11.2.1 Evaluation based on thermocouples near the valve not covered by the housing........ 91 11.2.2 Evaluation based on the behavior of the intumescent material.................................... 93 11.2.3 Evaluation based on a comparison between valve and plug side................................ 94 11.3 Evaluation for ACT using numerical analyses............................................................. 95 11.4 Conclusion.................................................................................................................. 98
12 Summary 99
References 100
5 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
List of Figures Figure 6-1: Geometry of the calculation model for the DN30 package, full view........................ 17 Figure 6-2: Geometry of the calculation model for the DN30 package, detailed view................ 18 Figure 6-3: Valve end of the calculation model for the DN30 package...................................... 19 Figure 6-4: Plug end of the calculation model for the DN30 package........................................ 20 Figure 6-5: Detailed view of the plug end of the calculation model for the DN30 package........ 21 Figure 6-6: Mesh of the calculation model of the DN30 package............................................... 22 Figure 6-7: Detailed view of the mesh at the plug end of the calculation model........................ 22 Figure 6-8: Resulting heat and diffusion generation curves....................................................... 36 Figure 6-9: Measured loss of mass of RTS 120 and RTS 320 foam samples........................... 37 Figure 7-1: Flame temperatures in the fire phase (A to F = sensors in the flames) during the thermal test with the prototype of the DN30 package (Benchmark 1)........................................ 44 Figure 7-2: Flame temperatures in the fire phase (A to F = sensors in the flames) during the thermal test with the prototype of the DN30 package (Benchmark 2)........................................ 45 Figure 7-3: Measured temperature (in red) and wind speed (in blue) during the thermal test with the prototype of the DN30 package (Benchmark 1)................................................................... 46 Figure 7-4: Measured temperature (in red) and wind speed (in blue) during the thermal test of the prototype of the DN30 package (Benchmark 2)......................................................................... 47 Figure 7-5: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 1 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down................................................................ 49 Figure 7-6: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 1 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down; detailed view.......................................... 50 Figure 7-7: Temperature distribution at the DN30 package for Benchmark 1 at fire end (t = 1810 s)................................................................................................................................. 51 Figure 7-8: Temperature distribution at the DN30 package for Benchmark 1 at the end of the foam burning (t = 3630 s).................................................................................................................... 52 Figure 7-9: Temperature distribution at the DN30 package for Benchmark 1 around the time of maximal valve temperature (t = 11720 s)................................................................................... 52 Figure 7-10: Heat generation caused by the burning of the foam (t = 598 s)............................. 53 Figure 7-11: Heat generation caused by the burning of the foam (t = 1810 s)........................... 53 Figure 7-12: Remaining foam concentration for RTS 120 calculated for the thermal test for Benchmark 1.............................................................................................................................. 54 Figure 7-13: Remaining foam concentration for RTS 320 calculated for the thermal test for Benchmark 1.............................................................................................................................. 54 Figure 7-14: Remaining foam for RTS 120 mantle side, top shell after the prototype fire test for Benchmark 1.............................................................................................................................. 55 Figure 7-15: Remaining foam for the valve side, bottom shell after the prototype fire test for Benchmark 1.............................................................................................................................. 55 Figure 7-16: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 2 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down................................................................ 56 Figure 7-17: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 2 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down; detailed view.......................................... 57
6 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-18: Temperature distribution at the DN30 package for Benchmark 2 at fire end (t = 1930 s)................................................................................................................................. 58 Figure 7-19: Temperature distribution at the DN30 package for Benchmark 2 at the end of the foam burning (t = 3740 s)........................................................................................................... 59 Figure 7-20: Temperature distribution at the DN30 package for Benchmark 2 around the time of maximal valve temperature (t = 5720 s)..................................................................................... 59 Figure 8-1: Temperatures at the DN30 package loaded with a filled 30B cylinder, RCT and NCT, with 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of insolation / no insolation................................................................................... 61 Figure 8-2: Temperature distribution for the steady-state analysis for constant solar insolation 61 Figure 9-1: Temperatures at the DN30 package loaded with an empty 30B cylinder - all temperatures for fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down.......................................................... 63 Figure 9-2: Temperatures at the DN30 package loaded with an empty 30B cylinder - mantle of the 30B cylinder as well as valve and plug thread for the fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down
................................................................................................................................................... 64 Figure 9-3: Temperature distribution at the DN30 package loaded with an empty 30B cylinder at the end of the fire (t = 1810 s).................................................................................................... 65 Figure 9-4: Temperature distribution at the DN30 package loaded with an empty 30B cylinder at the end of the burning of the foam (t = 3620 s).......................................................................... 65 Figure 9-5: Temperature distribution at the DN30 package loaded with an empty 30B cylinder around the time of the maximum valve temperature (t = 12710 s)............................................. 65 Figure 9-6: Temperatures at the DN30 package loaded with a filled 30B cylinder - all temperatures; for the fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down................................................... 66 Figure 9-7: Temperatures at the DN30 package loaded with a filled 30B cylinder - mantle of the 30B cylinder, valve and plug thread, for the fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down................ 67 Figure 9-8: Temperature distribution at the DN30 package loaded with a filled 30B cylinder at the end of the fire (t = 1810 s).......................................................................................................... 68 Figure 9-9: Temperature distribution at the DN30 package loaded with a filled 30B cylinder at the end of the burning of the foam (t = 3620 s)................................................................................ 68 Figure 9-10: Temperature distribution at the DN30 package loaded with a filled 30B cylinder around the time of the maximum valve temperature (t = 10910 s)............................................. 69 Figure 9-11: Temperature distribution of the volume (UF 6) inside the 30B cylinder at the maximum temperature (t = 10910 s)........................................................................................................... 69 Figure 9-12: Maximum temperatures at the DN30 package loaded with a partially filled 30B cylinder................................................................................................................................ 70 Figure 9-13: Comparison of valve and 30B cylinder cavity temperatures for different filling ratios
................................................................................................................................................... 71 Figure 9-14: Maximum temperatures at the DN30 package loaded with an empty cylinder, thickness of Microtherm reduced to 9.5 mm.............................................................................. 72 Figure 9-15: Maximum temperatures at the valve and plug of the 30B cylinder for different thicknesses of the Microtherm insulation layer........................................................................... 74 Figure 9-16: Gap in the Microtherm insulation layer (pictured is the gap for four times the area of drop test bar).............................................................................................................................. 75 Figure 9-17: Comparison of the area of a drop test bar with Ø 150 mm and the gap with an equal area resp. four times the area.................................................................................................... 75 Figure 9-18: Maximum temperatures at the valve of the 30B cylinder for different gaps in the Microtherm insulation layer........................................................................................................ 76
7 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 9-19: Maximum temperatures at the plug of the 30B cylinder for different gaps in the Microtherm insulation layer........................................................................................................ 77 Figure 9-20: Maximum temperatures at the valve and plug of the 30B cylinder for a variation of thermal properties of the foam................................................................................................... 80 Figure 9-21: Maximum temperatures at the valve and plug of the 30B cylinder for a variation of the parameters controlling the burning of the foam.................................................................... 81 Figure 11-1: Positions of temperature sensors during the Benchmark 1 fire test...................... 91 Figure 11-2: Maximum temperatures at the 30B cylinder during the cooldown phase of the Benchmark 1 fire test................................................................................................................. 92 Figure 11-3: Housing with intumescent material after the Benchmark 1 and 2 fire tests........... 93 Figure 11-4: Crack around the plug protecting device at the inner shell after the pre-damaging drop tests for the Benchmark 1 fire test..................................................................................... 94 Figure 11-5: Damages to the outer shell after the pre-damaging drop tests for the Benchmark 1 fire test........................................................................................................................................ 95 Figure 11-6: Maximum temperatures at the valve and the plug with and without the housing of the valve protecting device............................................................................................................... 96 Figure 11-7: Maximum temperatures at the valve and the plug for the modified model compared to the standard model................................................................................................................. 98
8 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
List of Tables Table 3-1: Admissible component temperatures of the package DN30..................................... 14 Table 6-1: Thermal properties of UF 6......................................................................................... 23 Table 6-2: Thermal properties of carbon steel........................................................................... 24 Table 6-3: Thermal properties of stainless steel........................................................................ 25 Table 6-4: Thermal properties of pure RTS 120 and RTS 320 foam......................................... 26 Table 6-5: Thermal properties of RTS 120 foam / stainless steel mixture used in the analysis. 27 Table 6-6: Thermal properties of RTS 320 foam used in the analysis....................................... 28 Table 6-7: Thermal properties of intumescent material attached radially to the DN30 PSP inner shell............................................................................................................................................ 30 Table 6-8: Thermal properties of intumescent material attached axially to the DN30 PSP inner shell............................................................................................................................................ 31 Table 6-9: Thermal properties of the air gap between intumescent material and 30B cylinder. 32 Table 6-10: Thermal properties of air inside the 30B cylinder cavity for empty 30B cylinders... 33 Table 6-11: Thermal properties of Microtherm 1000R HY thermal insulation............................ 34 Table 6-12: Factors used for the calculation of the heat generation of the foam....................... 35 Table 6-13: Solar insolation data................................................................................................ 37 Table 6-14: Ambient temperature............................................................................................... 38 Table 6-15: Heat transfer by radiation at the surface of the DN30 package.............................. 38 Table 6-16: Heat transfer by convection at the outer surface of the DN30 package for RCT, NCT, and ACT..................................................................................................................................... 39 Table 6-17: Heat transfer by radiation in the ca vities between the skirts of the 30B cylinder valve side............................................................................................................................................. 40 Table 6-18: Heat transfer by convection in the gap between the skirts of the 30B cylinder valve side and plug side...................................................................................................................... 41 Table 7-1: Ambient temperatures and time steps for the Benchmark 1 calculation................... 43 Table 7-2: Ambient temperatures and time steps for the Benchmark 2 calculation................... 43 Table 7-3: Heat transfer by convection at the outer surface of the prototype of the DN30 package used for the benchmark analysis................................................................................................ 48 Table 7-4: Comparison of measured and calc ulated maximum temperatures during the thermal test with the prototype of the DN30 package - Benchmark 1.................................................... 51 Table 7-5: Comparison of measured and calc ulated maximum temperatures during the thermal test with the prototype of the DN30 package - Benchmark 2.................................................... 58 Table 8-1: Temperatures at the DN30 package loaded with a filled 30B cylinder under RCT and NCT............................................................................................................................................ 60 Table 9-1: Ambient temperatures and time steps for the ACT calculations............................... 62 Table 9-2: Temperatures at the DN30 package loaded with an empty 30B cylinder................. 64 Table 9-3: Temperatures at the DN30 package loaded with a filled 30B cylinder...................... 67 Table 9-4: Maximum temperatures at the DN30 package loaded with a partially filled 30B cylinder
................................................................................................................................................... 71 Table 9-5: Temperatures at the DN30 package loaded with an empty 30B cylinder, thickness of Microtherm reduced to 9.5 mm.................................................................................................. 73
9 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 9-6: Position of the gap for the damaged Microtherm insulation layer............................. 76 Table 9-7: Thermal properties of RTS 120 foam / stainless-steel mixture used in the sensitivity analysis...................................................................................................................................... 78 Table 9-8: Thermal properties of RTS 320 used in the sensitivity analysis................................ 79 Table 9-9: Factors used for the calculation of the heat generation of the foam......................... 81 Table 9-10: Vapor pressure of UF 6 extracted from [DeWITT].................................................... 82 Table 9-11: Required wall thickness for the cylinder for internal pressure................................. 85 Table 9-12: Required wall thickness for the elliptical heads for internal pressure...................... 86
10 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
1 Introduction The packaging DN30, consisting of the DN30 Protective Structural Packaging (PSP) and the 30B cylinder, is designed for the transport of uranium in the chemical form of UF 6 (Uranium Hex-afluoride). The uranium could be enriched natural uranium or reprocessed (enriched) uranium with a maximal enrichment of 5 wt.% U-235 in uranium.
The UF6 is contained in a 30B cylinder according to [ISO 7195] 1. The DN30 PSP accommodates the 30B cylinder and provides mechanical and thermal protection during routine conditions of transport (RCT), normal conditions of transport (NCT) and accident conditions of transport (ACT).
The report at hand contains the thermal analysis of the package under RCT, NCT and ACT con-sidering a thermal power of the radioactive content of 3 W.
After a brief definition of the package contents in chapter 2, the objective of proof and the as-sumptions on which the proof is based are specified in chapters 3 and 4. Then, the calculation methods are described and justified in chapter 5.
Next, the model used for the analyses is described with geometry, materials, initial temperatures, thermal power, heat generation and solar insolation as well as heat transfer to and from the am-bient and within gaps between the DN30 PSP and th e 30B cylinder is described in chapter 6.
The first step of the analysis is detailed in chapter 7 with the comparison of calculated tempera-tures of the DN30 package to the temperatures measured during experimental fire tests. The boundary conditions for the calculation comply with the boundary conditions present during the experimental tests.
The analysis of the temperatures of the DN30 package under RCT, NCT and ACT is the main part of this report and detailed in chapter 8 for RCT and NCT, and chapter 9 for ACT.
The temperatures of the DN30 package during RCT and NCT are a result of the presence or absence of solar insolation due to the very low thermal power of the content. Two cases are considered here. The first case is the repetition of an insolation cycle of 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> insolation and 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> without insolation until a stable temperature pattern is reached. The second case is a constant solar insolation with 100 % of the heat flux used for the 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> insolation/no insolation cycle as bounding case.
For ACT, different filling ratios of the 30B cylinder from empty to 2277 kg UF 6 are analyzed. The boundary and initial conditions comply in all cases with the requirements of [ADR] and [SSR-6].
The influence of a reduction of the thickness of the microporous insulation layer Microtherm caused by structural damages following the free drop tests for ACT is investigated with these boundary conditions, too.
In the DN30 PSP, two types of foam (RTS 120 and RTS 320) are included as shock absorbers.
A sensitivity study is conducted for the thermal conductivity and specific heat of the foam for high temperatures as well as for the parameters controlling the burning of the foam.
After the conclusion of the extensive thermal analysis, chapter 10 shows that the objectives of the proof are met and that the DN30 package fulfils the requirements of [ADR] and [SSR-6] with re-spect to the thermal requirements.
Finally, in chapter 11, the design change of the removal of the housing is investigated using nu-merical analyses and the data from the experimental fire test to show that the removal of the housing does not impede the thermal safety of the packaging, as its designed safety function is no longer required due to the effectiveness of the microporous insulation layer.
1 Cylinders may be also manufactured according to the cu rrent or earlier valid revisions of [ANSI N14.1]
11 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
2 Package contents The content of the DN30 package is UF 6. The minimal mass of the content is zero (empty cylinder) and the maximal content 2277 kg. The thermal properties of UF 6 are specified in subsection 6.4.1.
3 Objective of proof
3.1 Verification for all types of packages It is verified that the design of the package cons iders ambient temperatures and pressures that are likely to be encountered in RCT ([ADR], No 6.4.2.10 or [SSR-6] para. 616).
3.2 Verification for packages containing uranium hexafluoride It is verified that the package can withstand, without rupture of the containment system, the ther-mal test specified in [ADR], No 6.4.17.3 or [SSR-6] para. 728.
3.3 Verification for type B(U) packages It is verified that under ambient conditions as specified in [ADR], No 6.4.8.5 or [SSR-6] para. 656 (ambient temperature 38 °C) and in absence of solar insolation, the temperature of the accessible surfaces of the package is below 50 °C ([ADR], No 6.4.8.3 or [SSR-6] para. 654).
With this proof, it is also verified that the re quirements of [ADR], No 6.4.8.4 or [SSR-6] para. 655 are met.
Furthermore, the consequences of the thermal test specified in [ADR], No 6.4.17.3 or [SSR-6]
para. 728 on the package subjected to the mechanical tests specified in [ADR] No. 6.4.17.1, 6.4.17.2 a), 6.4.17.2 b) and 6.4.17.4 or [SSR-6] paras 726, 727 (a), 727 (b) and 729 are analyzed to verify the requirements of [ADR] No. 6.4.8.8 or [SSR-6] para. 659:
Retain sufficient shielding to ensure that the radiation level 1 m from the surface of the package would not exceed 10 mSv/h - verified in subsection 2.2.4 of the main part of the SAR.
It would restrict the accumulated loss of radioactive material to not more than A 2 per week
- verified in subsection 2.2.3 of the main part of the SAR.
3.4 Verification for packages containing fissile material The consequences of the thermal test specified in [ADR], No 6.4.17.3 or [SSR-6] para. 728 on the package subjected to the mechanical tests specified in [ADR] 6.4.17.1, 6.4.17.2 a), 6.4.17.2 b) and 6.4.17.4 or [SSR-6] paras 726, 727 (a), 727 (b) and 729 are analyzed to verify the require-ments of [ADR] No. 6.4.11.10 and 6.4.11.13 or [SSR-6] para. 682 and 685. The verification is contained in subsection 2.2.5 of the main part of the SAR.
12 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
3.5 Admissible temperatures of the DN30 package
3.5.1 Admissible temperatures of the 30B cylinder For the containment system of the DN30 package, which consists of the 30B cylinder with in-stalled valve and plug as specified in [ANSI N14.1] and [ISO 7195], a covering admissible tem-perature is calculated that considers the admissible temperatures of the materials with regard to their thermal properties.
The materials used for the cylinder shell, the valve and the plug are according to [ANSI N14.1]:
30B cylinder shell: the maximum temperature defined in [ASME BPVC] for SA516 steel grade 55/60 is 371.11 °C or 700 °F valve/plug body (aluminum bronze UNS C63600): the melting point is 1030 °C, the hot-working temperature is 760 - 875 °C valve/plug stem (nickel copper alloy UNS N04400): the melting point is 1300 - 1350 °C, the hot-working temperature is 648 - 1176 °C valve/plug solder (tin-lead alloy): the solidus temperature of a tin lead solder compliant with ASTM B32 alloy grade Sn 50 is 183 °C, the liquidus temperature is 216 °C (see [HAR-RIS SN50])
The admissible temperature for the 30B cylinder and its components is therefore set to 183 °C.
3.5.2 Admissible temperatures of the contents of the 30B cylinder The admissible temperature for the contents of the 30B cylinder is set to 131 °C with regard to a possible pressure build-up from melted UF 6 content according to [USEC 651] and [ASME BPVC].
The detailed calculation for this admissible temperature is listed in section 9.7.
13 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
3.5.3 Admissible component temperatures of the DN30 package Admissible component temperatures at the package DN30 are specified for RCT, NCT and ACT in Table 3-1.
Table 3-1: Admissible component temperatures of the package DN30
Admissible temperature [°C]
Component Material Remark/Reference RCT and NCT ACT
70 °C for handling and Outer shell of RCT, 100 °C for the the DN30 PSP 1.4301 70/1001) 9004) lifting lugs at the top half Inner shell of the DN30 PSP 1.4301 601) 9004)
Foam insulation Foam 60 2) -
Microporous Technical data sheet Thermal insula-material, e.g., 602) 1000 (Appendix 1.4.4 of the tion Microtherm or main part of the SAR)
Multiflex Intumescent Intumescent Technical data sheet material material, e.g., 80 4005) (Appendix 1.4.3 of the Promaseal main part of the SAR) 30B cylinder Pressure shell vessel steel 643) 4006) [DIN EN 10028-3]
Valve and plug thread Tin 643) 1837)
Contents of the 30B cylinder Air/UF6 643) 1318)
- 1) Calculation temperature
- 2) Identical to temperature of shells
- 3) Triple point temperature of UF6
- 4) The hot forming of material 1.4301 is carried out at temperatures of 950 - 1200 °C. At 900 °C, sufficient strength remains to prevent a deformation of the material by its own weight. The strength of the outer shell is not relevant for the containment system, the shielding, nor the criticality safety.
- 5) Temperature when the degradation of the material starts (see Appendix 1.4.3 of the main part of the SAR)
- 6) Max. temperature defined in DIN EN 10028-3 for the similar steel grades P275NH and P355NH
- 7) See section 3.5.1
- 8) See section 3.5.2 and 9.7
14 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
4 Assumptions for the thermal analysis The thermal analysis of the DN30 package is based on the assumptions listed in the following sections.
4.1 Basic assumptions The following assumptions are valid for all calculations carried out in the following chapters:
For the thermal analysis, a simplified model of the DN30 PSP is used:
o The DN30 PSP is modelled as a cylindrical shell consisting of the outer stainless-steel shell, the foam, the insulation layer between the RTS 120 foam and the inner shell consisting of a Microtherm layer, the inner stainless-steel shell, and the layer of intumescent material.
o The steel flange between top half and bottom half is not modelled but accounted for in the thermal properties of the foam used in the analysis.
o On the outer surface, the feet, the closure systems, the name plate, the thermal plugs etc. are neglected.
o Structural steel parts separating the foam blocks are considered.
o The valve and the plug protecting devices are neglected; both ends of the DN30 PSP are modelled as a simple wall consisting of the layer of intumescent material, the inner stainless-steel shell, the Microtherm layer, the foam, and the outer steel shell with uniform thickness over the whole area.
The model of the 30B cylinder is simplified as well, o The valve and plug end are modeled with straight heads instead of curved heads.
o The total cavity volume is modeled with 0.736 m 3 (minimal volume of a 30B cylin-der according to [ISO 7195]).
o Valve, plug, nameplate, holes in skirt are neglected.
The 30B cylinder is positioned centrally in the DN30 PSP, i.e., there is a uniform gap be-tween DN30 PSP and 30B cylinder.
For the UF6 a simplified model is used for the analysis. The UF 6 completely fills the cavity of the 30B cylinder model, i.e., the density of the material is adjusted so that the mass of the UF6 in the model is preserved.
An empty 30B cylinder filled with air instead of UF 6 is used for the Benchmark calculations in accordance with the Benchmark fire tests. For most calculations for the sensitivity anal-yses an empty 30B cylinder is used as well because it is conservative compared to a filled or partially filled cylinder.
The dimensions of all the components are adapted to the two-dimensional environment of the numerical simulation. All components are modelled to be in direct contact with their neighboring components.
All these changes are validated during a benchmarking process using the data of two experi-mental fire tests (see chapter 7 for more information).
4.2 Assumptions for RCT and NCT The following assumptions are valid for the thermal analysis of the package under RCT and NCT:
The geometry of the model complies with an undamaged DN30 package.
There is no phase change of the UF 6 (justified by the result of the calculations).
15 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
4.3 Assumptions for ACT The following assumptions are made for ACT, considering the tests mentioned in [ADR] No.
6.4.11.13 (b) or [SSR-6], para. 685 (b):
The geometry of the model complies with an undamaged DN30 package.
The influence of damages due to ACT is taken into account with the benchmark analysis and the adjustment of the material and boundary conditions to the results of the experi-mental thermal tests carried out with DN30 prototypes which had passed double ACT requirements (two full drop test sequences).
Phase change of the UF 6 is considered.
5 Software The thermal analysis is carried out with ANSYS Workbench 19 [ANSYS].
ANSYS Workbench is a software environment for performing linear and non-linear structural, thermal, and electromagnetic analysis using the finite element method (FEM). The capabilities of ANSYS Workbench encompass geometry creation or optimization, meshing, setting up the finite element model, solving and reviewing the resu lts. For thermal problems, ANSYS Workbench can solve steady-state as well as transient problems using two-or three-dimensional models. A model may include multiple materials, and the thermal conductivity, density, and specific heat of each material may be temperature dependent. Materials may undergo change of phase. Thermal prop-erties of materials may be entered as data or may be extracted from a material properties library.
The boundary conditions, which may be surface-to-environment or surface-to-surface, may be specified temperatures or any combination of prescribed heat flux, forced convection, natural convection, and radiation. The boundary condition parameters may be time-and/or temperature-dependent. General grey body radiation problems may be modelled with user-defined factors for radiant exchange. Heat-generation rates may be dependent on time and position, and boundary temperatures may be time-and position dependent. Coupled-physics analysis are supported, e.g., structural-thermal, thermal-electric or thermal-diffusion. Additional functions, that are not available via the Workbench user interface can be included in the model using APDL (ANSYS parametric design language) command scripts.
ANSYS Workbench uses linear interpolation for all tabular values. Extrapolation is not used for calculated values outside the range of defined tabular values, instead the last grid point is used.
5.1 Benchmarking/Validation For the benchmarking and validation of the calculation model the results of two real world fire tests with DN30 prototype packages are used:
The DN30 prototype is modelled as described below in chapter 6.
The thermal conductivity of the foam is adjusted so that the temperatures at the 30B cyl-inder are close to the measured temperatures in the prototype.
As the foam of the prototype burned during the fire test and in the cooling phase, it was assumed that the foam acts as a thermal power source during that time. This thermal power is adjusted such that the maximal temperatures reached at the 30B cylinder and its components comply with the measured temperatures for the prototype.
The thermal analysis of the DN30 package according to IAEA/ADR guidelines is then carried out with the validated calculation model.
16 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6 Calculation model
6.1 Coordinate system For the calculation model, an axisymmetric two-dimensional model is used.
The axis of symmetry is the global y-axis which runs along the inner side of the model longitudi-nally through the center of the cavity. The radial direction is the global x-axis.
6.2 Geometry The geometry of the calculation model is shown in Figure 6-1 to Figure 6-5.
Figure 6-1 shows a general view of the model with the different components. Details of the design such as the flange between top and bottom half or rotation preventing devices are not modelled.
Figure 6-1: Geometry of the calculation model for the DN30 package, full view
The dimensions of all the components are adapted to the two-dimensional environment of the numerical simulation. For example, the dimensions of the outer shell of the DN30 PSP with a height of 1125 mm and a width of 1104 mm are modelled with a constant diameter of 1116 mm.
All components are modelled to be in direct contact with their neighboring components. Small gaps needed for fitting during manufacturing are conservatively neglected. These changes do not have an impact on the validity of the thermal simulations, as the model with these adaptations is benchmarked and validated using the data of two experimental fire tests (see chapter 7 for more information).
The outer diameter, the total length including the skirts and the wall thickness of 13 mm of the 30B cylinder are identical to the dimensions specified in [ISO 7195]. The cavity volume amounts to 0.736 m³, which is the minimum volume given in [ISO 7195].
The inner and outer stainless-steel shells of the DN30 PSP are modelled with their original thick-nesses:
Thickness of the outer shell: 4 mm in radial direction, 3 mm in axial direction.
Thickness of the inner shell: 10 mm in radial direction, 10 mm in axial direction at the valve side, and 2 mm in axial direction at the mantle side and plug side.
The reinforcements separating the RTS 320 foam parts and the RTS 120 parts at the mantle side are 2 mm in thickness.
Figure 6-2 shows the DN30 package with added measurements in millimeter for the sizes of the foam components and the 30B cylinder as well as the overall dimensions and the length of the cavity.
17 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 6-2: Geometry of the calculation model for the DN30 package, detailed view
Figure 6-3 and Figure 6-4 show the details of the valve side and the plug side of the DN30 PSP (all measurements in millimeter).
18 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 6-3: Valve end of the calculation model for the DN30 package
19 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 6-4: Plug end of the calculation model for the DN30 package
The microporous insulation layer Microtherm (orange) is 10 mm thick and located in axial and radial direction between the RTS 120 foam and the inner shell. The length of the microporous insulation layer is equal to the length of the RTS 120 foam parts next to them.
The intumescent material fixed to the inside of the inner shell is modelled with a thickness of 2.5 mm. The air/intumescent part is modelled with a thickness of 10.5 mm and runs along the complete length of the cavity.
Figure 6-5 shows the magnified detail of the plug end with the intumescent material (grey) at-tached to the inner shell of the DN30 PSP (dark blue). Between the intumescent material and the 30B cylinder skirts there is an air gap without elements; this is a necessary requirement to allow for the consideration of radiation in that gap.
20 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 6-5: Detailed view of the plug end of the calculation model for the DN30 package
21 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
The mesh is shown in Figure 6-6 below. Elements with linear shape functions are used in the model; the mesh consists of mostly rectangular elements with some triangular elements. The total count is approximately 54,000 nodes and 44,000 elements.
Figure 6-6: Mesh of the calculation model of the DN30 package
There are at a minimum two elements over the cross section of every component as shown in the detailed view in Figure 6-7.
Figure 6-7: Detailed view of the mesh at the plug end of the calculation model
6.3 Thermal contacts The contacts between different components are modelled using standard bonded contacts, which use an augmented Lagrange formulation, and symmetric behavior. The thermal conductance of the contacts is calculated by the program based on the material properties.
22 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4 Material properties The material properties are given for three different conditions. Material properties are identical for RCT and NCT, while ACT is split up between the fire phase and the cooling phase. For the foam parts, the split is different to account for the complete duration of the incineration of the foam as observed in the fire test. The properties for the foam parts are therefore divided into the fire phase, including the first 30 minutes of the cooling phase, and the remainder of the cooling phase.
ANSYS Workbench uses linear interpolation for all tabular values. For temperatures outside the defined range, the last defined grid point is used.
6.4.1 UF6 The thermal properties of UF6 are extracted from [DeWITT] and listed in Table 6-1. The heat of fusion as well as the heat of vaporization are conservatively neglected.
Table 6-1: Thermal properties of UF 6
Temperature [°C] RCT + NCT Fire phase Cooling down
Thermal conductivity solid [W/(mK)]
0 0.481 0.481 0.481 50 0.576 0.576 0.576 64.1 0.6025 0.6025 0.6025
Thermal conductivity liquid [W/(mK)]
64.11 - 0.15634 0.15634 72 - 0.16 0.16 100 - 0.173 0.173 150 - 0.196 0.196 200 - 0.22 0.22
Density [kg/m3]
- 3093.81) 3093.81) 3093.81)
Specific heat solid/liquid [J/(kg K)]
0 497 497 497 60 529 529 529 70 564 564 564 100 - 588 588 225 - 588 588
- 1) Calculated from 2277 kg in a volume of 0.736 m 3
23 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4.2 Carbon Steel (30B cylinder)
For the carbon steel of the 30B cylinder, the thermal properties listed in Table 6-2 are used.
Table 6-2: Thermal properties of carbon steel
Temperature [°C] RCT + NCT Fire phase Cooling down
Thermal conductivity [W/(mK)]
- 50 50 50
Density [kg/m3]
- 7821.21) 7821.21) 7821.21)
Specific heat [J/(kg K)]
- 500 500 500
- 1) standard value as used for the analysis of the external dose rates
24 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4.3 Stainless steel (DN30 PSP)
For the stainless steel of the DN30 PSP, the thermal properties listed in Table 6-3 are used.
These values are based on following formula [FRANSSEN].
14.6 1.27 10 With
= thermal conductivity [W/(mK)]
T = temperature [°C]
Table 6-3: Thermal properties of stainless steel
Temperature [°C] RCT + NCT Fire phase Cooling down
Thermal conductivity [W/(mK)]
0 14.6 14.6 14.6
100 15.9 15.9 15.9
200 17.2 17.2 17.2
300 - 18.4 18.4
400 - 19.7 19.7
500 - 21.0 21.0
600 - 22.2 22.2
700 - 23.5 23.5
800 - 24.8 24.8
Density [kg/m3]
- 79401) 79401) 79401)
Specific heat [J/(kgK)]
- 460 460 460
- 1) standard value as used for the analysis of the external dose rates
25 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4.4 Foam In the DN30 PSP, two types of polyisocyanurate rigid foams (PIR foams), RTS 120 and RTS 320, are included. These foams serve as shock absorbers and as thermal protection under RCT and NCT. The thermal properties listed in Table 6-5 for RTS 120 and in Table 6-6 for RTS 320 are used.
The thermal conductivity of the RTS 120 compone nts is calculated as a mixture of the thermal conductivity of RTS 120 and the stainless steel of the flanges of the top and bottom half. The thermal conductivity of RTS 120 and RTS 320 as a function of the temperature is given in Ta-ble 6-4. The values are extracted from Appendix 1.4.2 of the main part of the SAR. The thermal conductivity of stainless steel is given in Table 6-3.
The total volume of the steel flanges not included in the two-dimensional calculation model amounts to 1.481510-2 m³, the total volume of the RTS 120 components is 9.2035 10-1 m³. The thermal conductivity of the mixture can then be calculated as follows:
1.4815 10 9.2035 101.4815 10 9.2035 10
The dimensions and positions of the foam parts are shown in Figure 6-2 to Figure 6-4.
It is assumed that the foam is present during the fire phase and 30 min of the cooling phase; it was observed during the thermal test that the burning of the foam stops after that time. The re-maining foam is assumed to be burned/charred after that time. Therefore, the material is changed using an APDL command to switch the material from RTS 120 resp. RTS 320 to the burned/charred variant COAL 120 resp. COAL 320.
This assumption of the density change of the foam is derived from Appendix 1.4.2 of the main part of the SAR. The values of the specific heat are kept at the same values in the fire and in the cooling phase. With this approach the model of the cooling phase is conservative as the heat of the foam is kept in the model.
The benchmark calculation documented in chapter 7 for the second thermal test showed that the derived thermal conductivity for RTS 120 foam was not sufficient at the plug side to transfer enough heat into the 30B cylinder. The higher heat input at the plug side of the prototype for Benchmark 1 is caused by a direct steel-steel contact between the outer shell and the plug pro-tection device. To represent this increased heat transfer, the thermal conductivity of RTS 120 foam at the plug side is increased by a factor of 5 (labelled as Modified 1 in Table 6-5). An increased heat transfer caused by the ingress of hot gases is present for both benchmark fire tests. Therefore, the thermal conductivity of RTS 320 foam is increased by a factor of 4 for both models and additionally, the thermal conductivity of RTS 120 foam at the mantle side was in-creased by a factor of 2 for Benchmark 2 (labelled as Modified 2 in Table 6-5).
Table 6-4: Thermal properties of pure RTS 120 and RTS 320 foam
Temperature [°C] Thermal conductivity [W/(m K)] RTS 120 RTS 320
23 0.035 0.049 50 0.039 0.054 75 0.043 0.059 150 0.052 0.073 200 0.059 0.080 250 0.068 0.106
26 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-5: Thermal properties of RTS 120 foam / stainless steel mixture used in the analy-sis
Temperature RCT + NCT Fire phase + first 30 min of cooling down Rest of cooling
[°C] down RTS 120 RTS 120 Coal 120 Standard Modified 1 Modified 2 Thermal conductivity [W/(m K)]
0 0.266 0.266 1.328 0.531 100 0.294 0.294 1.472 0.589 200 0.330 0.330 1.651 0.660 Equal to the 300 - 0.458 2.288 0.915 values defined 400 - 0.792 3.961 1.585 for Fire phase +
500 - 0.873 4.367 1.747 first 30 min of 600 - 0.967 4.834 1.934 cooling down
700 - 1.062 5.309 2.123 800 - 1.157 5.784 2.313 Density [kg/m3]
23 120 120 84 250 - 120 -
300 - 89 -
400 - 86.5 1) -
500 - 84 -
800 - 84 -
Specific heat [J/(kg K)]
23 1323 1323 1323 50 1466 1466 1466 75 1633 1633 1633 150 1954 1954 1954 200 2219 2219 2219 250 2571 2571 2571 300 - 2789 2789 400 - 3318 3318 500 - 3847 3847 600 - 4377 4377 700 - 4906 4906 800 - 5435 5435
- 1) value adjusted for numerical stability/convergence
27 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-6: Thermal properties of RTS 320 foam used in the analysis
Temperature [°C] RCT + NCT Fire phase + first Rest of cooling 30 min of cooling down down RTS 320 RTS 320 Coal 320 Thermal conductivity [W/(m K)]
0 0.049 0.197 0.197 100 0.064 0.256 0.256 200 0.080 0.320 0.320 300 - 0.451 0.451 400 - 0.465 0.465 500 - 0.487 0.487 600 - 0.558 0.558 700 - 0.629 0.629 800 - 0.701 0.701 Density [kg/m3]
23 320 320 162 250 - 320 -
300 - 212 -
400 - 181 -
500 - 162 -
800 - 162 -
Specific heat [J/(kg K)]
23 1284 1284 1284 50 1415 1415 1415 75 1542 1542 1542 150 1912 1912 1912 200 2082 2082 2082 250 2772 2772 2772 300 - 2876 2876 400 - 3470 3470 500 - 4063 4063 600 - 4657 4657 700 - 5251 5251 800 - 5845 5845
28 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4.5 Intumescent material For the intumescent material, the thermal properties are extracted from Appendix 1.4.3 of the main part of the SAR and are listed in Table 6-7.
The dimensions and positions of the intumescent material can be derived from Figure 6-2 to Fig-ure 6-5.
The expansion of the intumescent material starts at a temperature of 150 °C. For the thermal analysis, it is assumed that the expansion is finished at a temperature of 300 °C.
The thermal conductivity is taken from Appendix 1.4.3 of the main part of the SAR. The thickness of the intumescent material is 2.5 mm, the air gap between intumescent material and the 30B cyl-inder is 10.5 mm in radial direction and greater than 30 mm for most parts in axial direction.
Therefore, the thermal conductivity for intumescent ma terial in radial direction is derived from the measurement for the 10 mm-gap resp. the 30 mm-gap for intumescent material in axial direction.
This leads to three different values for the thermal conductivity:
The initial value for the base material (not expanded) at room temperature is derived from the technical data sheet.
The conductivity for temperatures above 150 °C for the expanded material is derived from Appendix 1.4.3 of the main part of the SAR.
The thermal conductivity for the expanded mate rial at ambient temperature is listed under Intumescent Exp. in Table 6-7 and Table 6-8 and is derived from Appendix 1.4.3 of the main part of the SAR.
The density of the intumescent material is adjusted to the expansion ratio. The air gaps allow for the full thermal expansion of the intumescent material of approximately 272 % on top of the initial thickness. Hence the density changes from 1000 kg/m 3 to the following density at maximum ex-pansion ratio:
1000 100 % 272 % 269
29 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-7: Thermal properties of intumescent material attached radially to the DN30 PSP inner shell
Temperature [°C] RCT + NCT Fire phase Cooling down
Intumescent Intumescent Intumescent Exp.
Thermal conductivity [W/(m K)]
20 0.19 0.19 1.47 150 - 0.19 -
300 - 1.11 1.11 800 - 1.11 -
Density [kg/m3]
0 1000 1000 269 180 - 1017 -
195 - 579 -
210 - 467 -
225 - 426 -
240 - 360 -
255 - 279 -
270 - 272 -
285 - 269 -
300 - 269 -
405 - 269 -
Specific heat [J/(kg K)]
30.3 1232 1232 1232 84.7 1359 1359 1359 149.0 1559 1559 1559 208.4 1680 1680 1680 287.5 1662 1662 1662
30 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-8: Thermal properties of intumescent material attached axially to the DN30 PSP inner shell
Temperature [°C] RCT + NCT Fire phase Cooling down
Intumescent Intumescent Intumescent Exp.
Thermal conductivity [W/(m K)]
20 0.19 0.19 0.41 150 - 0.19 -
300 - 0.52 0.52 800 - 0.52 -
Density [kg/m3]
0 1000 1000 269 180 - 1017 -
195 - 579 -
210 - 467 -
225 - 426 -
240 - 360 -
255 - 279 -
270 - 272 -
285 - 269 -
300 - 269 -
405 - 269 -
Specific heat [J/(kg K)]
30.3 1232 1232 1232 84.7 1359 1359 1359 149.0 1559 1559 1559 208.4 1680 1680 1680 287.5 1662 1662 1662
31 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4.6 Air In the cavity of the DN30 package, there are air gaps between the intumescent material and the 30B cylinder, which will be filled during the fire test by the expanding intumescent material. This air gap is modelled as the component Air/Intumescent with a thickness of 10.5 mm (see Figure 6-2 and Figure 6-5) and its temperature dependent thermal properties are defined in Table 6-9.
These values are based on the same assumptions as the values defined for the intumescent material in Table 6-7.
Table 6-10 shows the values used for the cavity of empty 30B cylinders.
Table 6-9: Thermal properties of the air gap between intumescent material and 30B cylin-der
Temperature [°C] RCT + NCT Fire phase Cooling down Thermal conductivity [W/(m K)]
20 0.0258 0.0258 1.47 40 0.0274 0.0274 -
60 0.0281 0.0281 -
80 0.0302 0.0302 -
100 0.0316 0.0316 -
120 0.0330 0.0330 -
140 0.0343 0.0343 -
150 0.0350 0.0350 -
300 - 1.11 1.11 800 - 1.11 -
Density [kg/m3]
0 0.8 0.8 269 150 - 0.8 -
300 - 269 -
800 - 269 -
Specific heat [J/(kg K)]
0 1000 1000 150 - 1000 see Table 6-7 or Table 6-8:
300 - 1662 Cooling Down, Intumescent Exp.
800 1662
32 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-10: Thermal properties of air inside the 30B cylinder cavity for empty 30B cylin-ders
Temperature [°C] RCT + NCT Fire phase Cooling down Thermal conductivity [W/(m K)]
20 0.0257 40 0.0271 60 0.0285 80 0.0299 100 0.0314 120 0.0328 140 0.0343 160 0.0358 180 0.0372 200 0.0386 250 0.0421 300 0.0454 350 0.0485 400 0.0516 450 0.0543 500 0.0570 600 0.0621 700 0.0667 800 0.0706
Density [kg/m3]
- 0.8 Specific heat [J/(kg K)]
- 1000
33 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.4.7 Microporous thermal insulation Microtherm For the thermal insulation between the inner shell of the DN30 PSP and the RTS 120 foam the thermal properties listed in Table 6-11 are used. These values are based on the technical data sheet for Microtherm 1000R HY as provided in Appendix 1.4.4 of the main part of the SAR.
The dimensions and positions of the microporous insulation can be derived from Figure 6-2 to Figure 6-5.
Table 6-11: Thermal properties of Microtherm 1000R HY thermal insulation
Temperature [°C] RCT + NCT Fire phase Cooling down Thermal conductivity [W/(m K)]
0 0.026 0.026 0.026 200 - 0.026 0.026 400 - 0.030 0.030 600 - 0.038 0.038 800 - 0.049 0.049
Density [kg/m3]
- 260 260 260
Specific heat [J/(kg K)]
0 920 920 920 200 - 920 920 400 - 1000 1000 600 - 1040 1040 800 - 1080 1080
6.5 Initial temperatures The initial temperatures for the steady state and transient calculations for RCT and NCT are uni-formly 38 °C.
The initial temperatures for the transient calculations for ACT are:
For the fire phase: uniformly 63 °C, the maximum temperature calculated for RCT and NCT boundary conditions.
For the cooling phase: the temperatures at the end of the fire phase.
34 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.6 Heat generation
6.6.1 Heat generation of the radioactive content The thermal power of the content is in all cases 3 W, as derived in the main part of the SAR.
It is assumed that the UF6 completely fills the cavity of the 30B cylinder. The volumetric heat source is hence calculated with a volume of 0.736 m 3 (minimal certified volume of a 30B cylinder):
3 0.736 4.1
6.6.2 Heat generation because of incineration of the foam During the thermal test with the DN30 prototype, an incineration of the foam was observed.
Hence, heat generation in the foam is included in the calculation model for ACT. Flames were visible for half an hour after the end of the fire phase, therefore the heat generation in the foam is active during the fire phase and during the first 30 min of the cooling phase.
To model the incineration front passing through the foam and to account for the loss of already incinerated foam not available for burning any more, a coupled thermal-diffusion analysis is used.
The incineration of the foam is modelled using the formula below, which is based on the Arrhenius equation:
/
With rate = rate constant A = rate of the reaction B = activation temperature C = foam concentration T = foam temperature The heat generation in the foam parts is then calculated with the enthalpy of the reaction ENT:
The rate of foam consumed by incineration is calculated as a negative diffusion generation:
With the benchmark analysis documented in chapter 7, the heat generation rate of the foam is determined. The best fit to the measured temperature curves is reached for the factors listed in Table 6-12.
Table 6-12: Factors used for the calculation of the heat generation of the foam
Factor Value
A 1.010-3 s-1
B 300 °C ENT 2.0107 J/m³
The resulting heat and diffusion generation curves for these factors are shown in Figure 6-8 for a foam concentration of 1.0 and 0.5.
35 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
1,60E+04 0,00E+00
1,40E+04 -1,00E-04
1,20E+04 -2,00E-04
1,00E+04 -3,00E-04
8,00E+03 -4,00E-04
6,00E+03 -5,00E-04
4,00E+03 -6,00E-04 HGEN (C = 1.0) 2,00E+03 HGEN (C = 0.5) -7,00E-04 DGEN (C = 1.0)
DGEN (C = 0.5) 0,00E+00 -8,00E-04 0 200 400 600 800 1000 Temperature [°C]
Figure 6-8: Resulting heat and diffusion generation curves
The diffusion coefficient is determined by taking the squared distance of the incineration front divided by twice the time measured for the incineration front to travel that distance. The resulting coefficient for RTS 120 is 2.410-7; the coefficient for RTS 320 is 8.010-8.
In Appendix 1.4.2 of the main part of the SAR, the behavior of RTS 120 and RTS 320 at elevated temperatures was investigated. Figure 6-9 shows the loss of mass as function of the temperature.
For RTS 120 foam the loss of mass starts at 200 °C and is fully developed at about 250 °C. For RTS 320 foam, the loss of mass starts at 300 °C and is fully developed at about 330 °C. The loss of mass is considered in the material properties of the foam.
36 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
4
2
0
-2
-4
-6
-10 0 50 100 150 200 250 300 350 400 450 500 Temperature [°C]
Figure 6-9: Measured loss of mass of RTS 120 and RTS 320 foam samples
6.7 Solar insolation The solar insolation data is specified in Table 6-13. The data complies with [ADR], No 6.4.8.6 or
[SSR-6] para. 657. For the calculation of the initial temperatures for ACT, a 12-hour insolation /
no insolation cycle as well as a constant solar insolation with the values given in Table 6-13 was investigated.
Table 6-13: Solar insolation data
Surface orientation RCT + NCT Fire phase Cooling down
Insolation for 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> per day [W/m 2]
Vertical surfaces (valve and plug end) 200 - 200
All other surfaces (cylindrical surfaces) 400 400
37 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.8 Heat transfer to ambient
6.8.1 Ambient temperature The ambient temperature is defined in Table 6-14. For RCT, NCT and the cooling phase of ACT, the ambient temperature is 38 °C according to [ADR], No 6.4.8.5 or [SSR-6] para. 656. For the fire phase of ACT, the temperature is set to 800 °C according to [ADR], No 6.4.17.3 or [SSR-6]
para. 728.
Table 6-14: Ambient temperature
RCT + NCT Fire phase Cooling down
Ambient temperature [°C]
38 800 38
6.8.2 Radiation The radiation coefficient of the outer surface of the DN30 package is 0.44 for RCT and NCT (stainless steel, rough surface).
During the fire, the surface absorptivity is set to 0.8 ([SSG-26] para. 728.29) and the flame emis-sivity to 0.9 ([SSG-26] para. 728.28). In the cooling phase, the emissivity is set to the same value of 0.8 as the absorptivity in the fire.
The radiation coefficients are listed in Table 6-15.
Table 6-15: Heat transfer by radiation at the surface of the DN30 package
Temperature [°C] RCT + NCT Fire phase Cooling down
Emissivity/absorptivity [-]
- 0.44 0.72 0.8
6.8.3 Convection For the convective heat transfer at the outer surface of the DN30 package for RCT and NCT as well as the cooling phase of ACT, the formula given in [SSG-26] para. 728.31 is used.
0.13
For the convective heat transfer for the fire phase of ACT, the formula given in [SSG-26] para.
728.30 is used.
0.036.
The characteristic length is in all cases the diameter of the DN30 package of 1.104 m. The pool fire gas velocity is assumed with 7.5 m/s.
The formulas are evaluated as function of the temperature and listed in Table 6-16.
38 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-16: Heat transfer by convection at the outer surface of the DN30 package for RCT, NCT, and ACT
Temperature Heat transfer coefficient [W/(m 2 K)]
[°C] RCT + NCT Fire phase Cooling down
38.1 0.69 28.0 0.69
39 1.50 27.9 1.50
40 1.89 27.9 1.89
50 3.35 27.3 3.35
60 4.02 26.9 4.02
70 4.46 26.4 4.46
80 4.79 25.9 4.79
90 5.04 25.5 5.04
100 5.24 25.1 5.24
150 5.85 23.3 5.85
200 6.13 21.8 6.13
250 6.26 20.7 6.26
300 6.31 19.5 6.31
350 6.32 18.7 6.32
400 6.31 17.8 6.31
450 6.28 17.2 6.28
500 6.24 16.5 6.24
550 6.25 16.1 6.25
600 6.15 15.6 6.15
650 6.09 15.1 6.09
700 6.04 14.5 6.04
750 5.99 14.2 5.99
800 5.93 13.8 5.93
850 5.88 13.4 5.88
900 5.83 13.1 5.83
950 5.77 12.8 5.77
1000 5.72 12.5 5.72
39 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
6.9 Heat transfer in gaps Heat transfer in the two cavities between the skirts of the 30B cylinder at the valve and at the plug side is modelled using radiation, conduction, and convection.
6.9.1 Radiation For the painted surface of the 30B cylinder, an emissivity of 0.9 is assumed. For the emissivity of the intumescent material, a value of 0.8 is assumed.
These cavities are the air gaps between the outer envelope of the 30B cylinder at the valve and plug side and the intumescent material. Entering of smoke and fumes is possible if a gap between the upper and lower half of the DN30 PSP is present, hence a complete black radiation is as-sumed for ACT. Due to the large air gap, it is assumed that these cavities are not closed by the intumescent material in the fire phase. The values for the emissivity are listed in Table 6-17.
Table 6-17: Heat transfer by radiation in the cavities between the skirts of the 30B cylin-der valve side
RCT + NCT Fire phase Cooling down
Emissivity [-]
0.9 / 0.8 1.0 / 1.0 1.0 / 1.0
6.9.2 Conduction Conduction in the gap between the 30B cylinder and the DN30 PSP is considered. The conduc-tivity of the thermal contact as determined in the benchmark calculation is 30 W/(m² °C) for the gap at the valve side and 45 W/(m²°C) for the gap at the plug side.
6.9.3 Convection Due to the dimensions of the cavities, convection is considered. For these vertical gaps, the for-mula from [VDI Heat] is used.
0.42..
With s = 0.114 m h = 0.736 m
= see Table 6-10 All other values for air are extracted from [VDI Heat].
40 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 6-18: Heat transfer by convection in the gap between the skirts of the 30B cylinder valve side and plug side
Temperature [°C] RCT + NCT Fire phase Cooling down
Convective heat transfer t0.25 [W/(m2 K)]
38.1 1.14 1.14 1.14
100 1.09 1.09 1.09
200 1.01 1.01 1.01
300 0.95 0.95 0.95
400 0.90 0.90 0.90
500 0.86 0.86 0.86
600 0.83 0.83 0.83
700 0.80 0.80 0.80
800 0.78 0.78 0.78
900 0.76 0.76 0.76
1000 0.74 0.74 0.74
The values listed in Table 6-18 are approximated with a polynomial of fourth order and included in the calculation model using an APDL command. With this command, a point mass is created and linked to the nodes of the cavity with LINK 34 convection links.
The polynomial of fourth order is:
5 10 3 10 9 10 0.001 1.1779
The volumes of the cavities are 0.028 m³ for the valve side and 0.015 m³ for the plug side.
41 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
7 Benchmark calculations For the benchmark calculations, the results of two thermal tests with a prototype of the DN30 package are compared with the results of the analyses with ANSYS Workbench used in this report. These calculations are also used to adapt parameters of the model, which cannot be derived from standards or literature so that the temperatures at safety relevant parts calculated with the benchmark model comply with measured values. In doing this, the time-temperature curves of the calculations with the numerical model are compared with the results from the ex-perimental fire test as well.
The experimental fire tests are described in detail in Appendix 2.2.2.2 of the main part of the SAR.
For the benchmark calculations, the ambient conditions recorded during the experimental fire tests are used. These are specified in subsection 7.1.6 below.
The experimental fire test designated Benchmark 1 was conducted in November 2017 using a prototype similar to the production model with the Microtherm thermal insulation layer. The ex-perimental fire test labelled Benchmark 2 was conducted in September 2016 with a prototype without the Microtherm thermal insulation layer.
7.1 Parameters for the benchmark calculation
7.1.1 Geometry The geometry is the same as described in section 6.2.
7.1.2 Material properties The data for the materials are identical to the data specified in section 6.4.
For the foam parts made of RTS 120, Standard RTS 120 data as listed in Table 6-5 is used for the Benchmark 1 and 2 models. Exceptions are the foam part at the plug side for Benchmark 1, for which the Modified 1 data is used and the foam parts at the mantle side for Benchmark 2, for which the Modified 2 data is used.
7.1.3 Initial temperatures The initial temperatures for the benchmark anal yses are uniformly 63 °C. The prototypes of the DN30 package were pre-heated by a heated sleeve to that temperature. Due to the short time required to remove the sleeve between the end of the pre-heating and the start of the fire, the temperature of the outer shell of the DN30 package was below this value at the start of the fire.
This had no consequences for the temperatures of the 30B cylinder, which was still 63 °C.
7.1.4 Heat generation
7.1.4.1 Heat generation of the radioactive content In the benchmarks, the 30B cylinder is empty and contains only air as in the thermal test. The thermal power is zero.
7.1.4.2 Heat generation due to incineration of the foam For the benchmark calculations, the same heat generation values as specified in subsection 6.6.2 are used.
7.1.5 Solar insolation Solar insolation before the thermal test is considered by heating the DN30 prototype to a uniform temperature of 63 °C.
Solar insolation is not considered in the cooling phase as there was no significant insolation during the cooling phase of the thermal tests.
42 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
7.1.6 Heat transfer to the ambient
7.1.6.1 Ambient temperature The ambient temperatures used in the calculation are derived from sensor data listed in Appendix 2.2.2.2. The temperatures are averaged for each time step of the transient analysis. For Bench-mark 1, the longer fire duration on the plug side is considered as well as the overall higher tem-peratures, especially at the plug side. For Benchmark 2, the longer fire duration and overall higher temperature is considered.
The average flame temperature during the thermal tests were between 900 °C and 1000 °C and thus higher than the value of 800 °C specified in [ADR] or [SSR-6]. The duration of the fire was 30 min at the valve and mantle side and 34 min at the plug side for Benchmark 1, and 32 min for Benchmark 2. The duration of the fire therefore exceeded the time of 30 min specified in [ADR]
or [SSR-6]. During the cooling phase the temperature was on average around 10 °C to 11 °C for Benchmark 1 and between 18 °C to 22 °C for Benchmark 2, which is below the value of 38 °C specified in [ADR] or [SSR-6].
The duration of the time steps in the benchmark analysis and corresponding temperatures are listed in Table 7-1 and Table 7-2 below.
Table 7-1: Ambient temperatures and time steps for the Benchmark 1 calculation
Time Step End time of Temperature [°C] Phase time step [s] Valve/Mantle Side Plug Side
1 0 63 63 2 10 920 996 Fire phase 3 1810 920 996 4 1820 10 996 5 2060 10 996 6 2070 10 10 Cooling down 7 3630 10 10 8 25230 12 12
Table 7-2: Ambient temperatures and time steps for the Benchmark 2 calculation
Time Step End time of time step [s] Temperature [°C] Phase
1 0 63 2 10 977 Fire phase 3 1930 977 4 1940 19 5 3740 19 Cooling down 6 25340 26
43 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-1 and Figure 7-2 show the graph of the measured temperatures for the fire phase. Sen-sor A is positioned at the valve side, sensor D at the plug side and sensors B, C, E and F at the mantle of the DN30 package.
1200
1000
800
600
A 400 B C
D 200 E F
AVG 0
-4 -2 0 2 4 6 8 101214161820222426283032343638 Tim e [m in]
Figure 7-1: Flame temperatures in the fire phase (A to F = sensors in the flames) during the thermal test with the prototype of the DN30 package (Benchmark 1)
44 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
1200
1000
800
600
A 400 B C
D 200 E F
AVG 0
- 4 -2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 Tim e [m in]
Figure 7-2: Flame temperatures in the fire phase (A to F = sensors in the flames) during the thermal test with the prototype of the DN30 package (Benchmark 2)
7.1.6.2 Radiation For radiation, the values defined in subsection 6.8.2 are used.
7.1.6.3 Convection For the heat transfer by convection, the values listed in Table 7-3 were used. For the fire phase, these values comply with the values given in Table 6-16.
For the cooling phase, the influence of wind was taken into account. The measured wind speed is shown in Figure 7-3 and Figure 7-4. Best fit with the measured temperature curve of the surface could be reached by assuming a wind speed of 2.8 m/s, which is somewhat higher than the av-erage measured wind speed.
Figure 7-3 and Figure 7-4 show the graph of the ambient temperature and the wind speeds over time during cooling down.
45 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
20 5,0 18 Temp eratur e 4,5 Wind Spe ed 16 4,0
14 3,5
12 3,0
10 2,5
8 2,0
6 1,5
4 1,0
2 0,5
0 0,0 40 90 140 190 240 290 340 390 440 Tim e [m in]
Figure 7-3: Measured temperature (in red) and wind speed (in blue) during the thermal test with the prototype of the DN30 package (Benchmark 1)
46 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
40 5,0 36 Temp eratur e 4,5 Wind Spe ed 32 4,0
28 3,5
24 3,0
20 2,5
16 2,0
12 1,5
8 1,0
4 0,5
0 0,0 40 90 140 190 240 290 340 390 440 Tim e [m in]
Figure 7-4: Measured temperature (in red) and wind speed (in blue) during the thermal test of the prototype of the DN30 package (Benchmark 2)
As the convection coefficients calculated with the formula in [SSG-26] para. 728.31 were too low in comparison with the test results, the convection coefficients were instead calculated by com-bining the values of the coefficient given in Table 6-16 and the values calculated with the formula given in [SSG-26] para. 728.30 0.036.
for a wind speed of 2.8 m/s.
The resulting cooling of the outer shell is still conservative for the numerical calculation compared to the fire tests (see sections 7.2 and 7.3).
The resulting convection coefficients are listed in Table 7-3.
47 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 7-3: Heat transfer by convection at the outer surface of the prototype of the DN30 package used for the benchmark analysis
Temperature Heat transfer coefficient [W/(m 2 K)]
[°C] Fire phase Cooling down
38.1 28.0 13.4 39 27.9 14.2 40 27.9 14.6 50 27.3 15.8 60 26.9 16.2 70 26.4 16.5 80 25.9 16.6 90 25.5 16.6 100 25.1 16.6 150 23.3 16.4 200 21.8 16.0 250 20.7 15.7 300 19.5 15.2 350 18.7 14.8 400 17.8 14.4 450 17.2 14.1 500 16.5 13.7 550 16.1 13.6 600 15.6 13.2 650 15.1 12.9 700 14.5 12.6 750 14.2 12.4 800 13.8 12.2 850 13.4 12.0 900 13.1 11.8 950 12.8 11.6 1000 12.5 11.4
7.1.7 Heat transfer in gaps For the heat transfer in gaps the same parameters as specified in section 6.9 are used.
48 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
7.2 Results of the analysis for Benchmark 1 The results of the analysis for Benchmark 1, conducted in November 2017, are shown in Fig-ure 7-5 to Figure 7-9 and Table 7-4.
Figure 7-5 shows an overview of the temperat ures measured during the experimental fire test Benchmark 1 and the calculated temperatures from the numerical simulation at the important temperature measurement positions like valve, plug and mantle of the 30B cylinder as well as the DN30 PSP outer shell. The measured data are shown with a dashed line and marked with the initial BM (for benchmark). The calculated values are shown in the same color, but with a solid line.
The measured temperature at the surface of the prototype is the average temperature of the four sensors placed at the surface of the prototype at the 0°, 90°, 180° and 270° position. The meas-ured temperatures drop faster to ambient temperatures after the fire phase than the calculated temperatures because the temperature drop is not instant but modelled over a period of 10 s to allow for an easier convergence. The difference during cooling down is caused by the sensor positioning; while the temperature sensors are placed 15 cm from the DN30 package, the calcu-lated temperature of the outer shell is derived directly from the relevant components.
The temperatures at the inner stainless-steel shell of the DN30 PSP were not measured in the thermal test. They are calculated and shown for information purposes.
1000 30B Valv e 30B Plug 800 30B Ma ntle
D N30 Inner Shell DN30 Outer Shel l 600 30B Valv e ( BM)
30B Plug (BM) 400 30B Ma ntle (BM)
DN30 Outer Shel l (BM)
200
0 0 5000 1000 0 1500 0 2000 0 Tim e [s ]
Figure 7-5: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 1 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down
Figure 7-6 shows the comparison of the measured and calculated temperatures at the mantle, the valve and the plug of the 30B cylinder. The measured curve of the temperatures of the 30B cylinder surface was used to determine the heat generation rates of the foam in the fire phase
49 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
and the early cooling phase as well as the he at transfer between DN30 overpack and 30B cylin-der. The best fit was reached for the values defined in subsections 6.4.4 and 6.6.2. While the calculated temperature curves for valve and plug are in good agreement with the measured tem-perature in the fire test, the difference between the temperatures at the mantle is significantly bigger. This is due to the different temperatures measured at the 30B mantle. While the bottom sensor at the 180° position reached its maximum of 120 °C after 13080 s, the sensor at the 270° position reached its maximum of 177 °C after 3300 s. For the comparison, the average temperature of the sensors in the 0°, 90°, 180°, and 270° position is used.
200
180
160
140
120 100 30B Valv e 30B Plug 80 30B Ma ntle D N30 Inner Shell 60 D N30 Outer Shel l 40 30B Valv e (BM) 30B Plug (BM) 20 30B Ma ntle ( B M)
D N30 Outer Shel l (BM) 0 0 5000 1000 0 1500 0 2000 0 Tim e [s ]
Figure 7-6: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 1 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down; detailed view
Table 7-4 shows the maximum temperatures at the important temperature measurement posi-tions as well as the time of their maximum in seconds after the start of the thermal test. The maximum temperatures and the times after which the respective maximum is reached are in very good agreement for the valve and plug of the 30B cylinder.
50 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 7-4: Comparison of measured and calculated maximum temperatures during the thermal test with the prototype of the DN30 package - Benchmark 1
Benchmark calculation Prototype test
Position Temperature Time Temperature Time
[°C] [s] [°C] [s]
Valve 137 11720 131 12680
Plug 148 9920 144 9450
Mantle 30B cylinder 150 9320 143 8505
Inner shell DN30 PSP 223 3630 - -
Outer shell DN30 PSP 982 1820 992 470
Figure 7-7 shows the calculated temperature distribution of the DN30 package at the end of the fire, Figure 7-8 at the time the burning of the foam stops, and Figure 7-9 shows the temperature distribution of the DN30 package around the time of maximum valve temperature.
Figure 7-7: Temperature distribution at the DN30 package for Benchmark 1 at fire end (t = 1810 s)
51 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-8: Temperature distribution at the DN30 package for Benchmark 1 at the end of the foam burning (t = 3630 s)
Figure 7-9: Temperature distribution at the DN30 package for Benchmark 1 around the time of maximal valve temperature (t = 11720 s)
The heat generation caused by the burning of the foam is shown in Figure 7-10 for the time 10 min after the start of the fire and in Figure 7-11 for the end of the pool fire. The highest heat generation of the burning of the foam is observed in the fi rst 10 min of the fire and gradually declines until the heat generation is stopped approximately half an hour after the end of the pool fire. The in-cineration front of burning foam advances from the outer shell towards the center of the package.
The cavities at the valve and plug are not empty in these two figures but show the convection link elements connecting the elements on the boundary of the cavity to the point mass in its center (see subsection 6.9.3).
52 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-10: Heat generation caused by the burning of the foam (t = 598 s)
Figure 7-11: Heat generation caused by the burning of the foam (t = 1810 s)
The remaining foam concentration after the fire test is shown in Figure 7-12 for RTS 120 and in Figure 7-13 for RTS 320. The remaining foam concentrations are in relative good agreement with the prototype after the experimental fire test shown in Figure 7-14 and Figure 7-15; the RTS 120 foam parts are burned to a higher degree than the RTS 320 foam parts, which are nearly un-burned in some areas.
53 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-12: Remaining foam concentration for RTS 120 calculated for the thermal test for Benchmark 1
Figure 7-13: Remaining foam concentration for RTS 320 calculated for the thermal test for Benchmark 1
54 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-14: Remaining foam for RTS 120 mantle side, top shell after the prototype fire test for Benchmark 1
Figure 7-15: Remaining foam for the valve side, bottom shell after the prototype fire test for Benchmark 1
55 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
7.3 Results of the analysis for Benchmark 2 The results of the analysis of Benchmark 2, conducted in September 2016, are shown in Fig-ure 7-16 to Figure 7-20 and Table 7-5. The main difference between both benchmark prototype tests is, that the prototype for Benchmark 2 is no t outfitted with the Microtherm insulation layer.
Figure 7-16 shows an overview of the measured and calculated temperatures at the important temperature measurement positions like valve, plug, mantle of the 30B cylinder and DN30 outer shell. The measured data are shown with a dashed line and marked with the initial BM (for bench-mark). The calculated values are shown in the same color, but with a solid line.
The measured temperature at the surface of the prototype is the average temperature of the four sensors placed at the surface of the prototype at the 0°, 90°, 180°, and 270° position. Compared to Benchmark 1, the average temperature of the outer shell is higher in the prototype fire test for Benchmark 2. The measured temperatures drop faster to ambient temperatures after the fire phase than the calculated temperatures because the temperature drop is not instant but modelled over a period of 10 s to allow for an easier convergence. The difference during cooling down is caused by the sensor positioning; while the temperature sensors are placed 15 cm from the DN30 package, the calculated temperature of the outer shell is derived directly from the relevant components.
The temperatures at the inner stainless-steel shell of the DN30 PSP were not measured in the thermal test. They are calculated and shown for information purposes.
1000 30B Valv e
30B Plug 800 30B Ma ntle D N30 Inner Shell DN30 Outer Shel l 600 30B Valv e ( BM) 30B Plug (BM) 30B Ma ntle (BM) 400 DN30 Outer Shel l (BM)
200
0 0 5000 1000 0 1500 0 2000 0 Tim e [s ]
Figure 7-16: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 2 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down
56 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-17 shows the comparison of the measured and calculated temperatures at the mantle, the valve, and the plug of the 30B cylinder. The measured curve of the temperatures of the 30B cylinder surface was used to determine the heat generation rates of the foam in the fire phase and the early cooling phase as well as the he at transfer between DN30 overpack and 30B cylin-der. The best fit was reached for the values defined in subsections 6.4.4 and 6.6.2. While the calculated temperature curves for valve and plug are in good agreement with the measured tem-perature in the fire test, the difference between the temperatures at the mantle is significantly bigger. This is due to the different temperatures measured at the 30B mantle. While the bottom sensor at the 180° position reached its maximum of 197 °C after 6240 s, the sensor at the 270° position reached its maximum of 283 °C after 3300 s. For the comparison, the average temperature of the sensors in the 0°, 90°, 180°, and 270° position is used.
250
200
150 30B Valv e 30B Plug 100 30B Ma ntle D N30 Inner Shell D N30 Outer Shel l 50 30B Valv e (BM) 30B Plug (BM) 30B Ma ntle ( B M)
D N30 Outer Shel l (BM) 0 0 5000 1000 0 1500 0 2000 0 Tim e [s ]
Figure 7-17: Measured and calculated temperatures at the prototype of the DN30 package for Benchmark 2 - fire phase and 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> cooling down; detailed view
Table 7-5 shows the maximum temperatures at the important temperature measurement posi-tions as well as the time of their maximum in seconds after start of the thermal test. The maximum temperatures and the time after the respective maximum is reached are in very good agreement for the valve and plug of the 30B cylinder.
57 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 7-5: Comparison of measured and calculated maximum temperatures during the thermal test with the prototype of the DN30 package - Benchmark 2
Benchmark calculation Prototype test
Position Temperature Time Temperature Time
[°C] [s] [°C] [s]
Valve 226 5720 229 7740
Plug 200 7520 206 7740
Mantle 30B cylinder 307 3980 238 4050
Inner shell DN30 PSP 334 2350 - -
Outer shell DN30 PSP 976 1920 1022 470
Figure 7-18 shows the calculated temperature distribution of the DN30 package at the end of the fire, Figure 7-19 at the time the burning of the foam stops, and Figure 7-20 shows the temperature distribution of the DN30 package around the time of maximum valve temperature.
Figure 7-18: Temperature distribution at the DN30 package for Benchmark 2 at fire end (t = 1930 s)
58 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 7-19: Temperature distribution at the DN30 package for Benchmark 2 at the end of the foam burning (t = 3740 s)
Figure 7-20: Temperature distribution at the DN30 package for Benchmark 2 around the time of maximal valve temperature (t = 5720 s)
7.4 Conclusion for the benchmark calculations The Benchmark models allow for a very good account of the experimental fire tests. The calcu-lated maximum temperatures in the critical valve and plug area as well as the times after which these maxima are reached, are in very good agreement with the measured temperature curves.
The calculated temperatures for the critical components are in good agreement with both experi-mental fire tests despite the considerable difference for the maximum temperatures for both tests (for example approximately 90 °C for the valve of the 30B cylinder).
The Benchmark models with the adjusted material properties of the foam and the parameters controlling the burning of the foam are therefore well suited for the thermal analyses of the DN30 package.
59 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
8 Calculation for RCT and NCT For the calculations for RCT and NCT, steady-state simulations are used for the calculations without solar insolation and with constant solar insolation. For the 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> with insolation / 12 h without insolation cycles, a transient simulation is used.
8.1 Results without solar insolation Due to the very low thermal power of the content, the temperatures at the DN30 package without solar insolation are only slightly higher than the ambient temperature. The values are listed in Table 8-1.
8.2 Results with solar insolation The results for the cycle of 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> with insolation / 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> without insolation are shown in Figure 8-1 and Table 8-1. The maximum temperature of 52 °C for the 30B cylinder and its com-ponents is reached after about 20 days. Hence, it can be assumed that the UF 6 remains solid under RCT and NCT conditions.
For the case of a constant insolation over 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> with 100 % of the insolation, the maximum temperatures reached are 58 °C for the 30B cylinder and its components.
The maximum temperature of 63 °C at the surface of the DN30 PSP complies with the initial conditions for the thermal test with a prototype documented in chapter 7.
The temperature distribution at the DN30 package for the steady-state calculation with constant solar insolation is shown in Figure 8-2.
Table 8-1: Temperatures at the DN30 package loaded with a filled 30B cylinder under RCT and NCT
Temperature [°C]
Position Without 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> insolation Constant insolation / no insolation insolation cycles
Valve 38 52 56
Plug 38 52 57
Mantle 30B cylinder 39 52 58
Volume 30B cylinder (UF 6) 39 52 58
Inner shell DN30 PSP 38 56 59
Outer shell DN30 PSP 38 63 61
60 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
65
60
55
50
30B Valve 45 30B Plug 30B Mantle 40 30B Volume (UF )
DN30 Inner S hel l DN30 Outer Shel l 35 0.00E +00 5.00E +05 1.00E +06 1.50E +06 2.00E +06 Tim e [s ]
Figure 8-1: Temperatures at the DN30 package loaded with a filled 30B cylinder, RCT and NCT, with 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> of insolation / no insolation
Figure 8-2: Temperature distribution for the steady-state analysis for constant solar inso-lation
61 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9 Calculations for ACT The following transient calculations for ACT are carried out using the calculation model for Bench-mark 1 with the microporous insulation layer Microtherm and the increased thermal conductivity for the RTS 120 foam at the plug side. Boundary conditions, material data, heat generation etc.
are as described in chapter 6, except for the listed deviations and differences.
For the following calculations for ACT in this ch apter, the ambient temperatures according to the ADR and IAEA regulations are used. These temperatures and the corresponding time steps listed in Table 9-1.
Table 9-1: Ambient temperatures and time steps for the ACT calculations
Time Step End time of time step [s] Temperature [°C] Phase
1 0 63 2 10 800 Fire phase 3 1810 800 4 1820 38 5 3620 38 Cooling down 6 25220 38
9.1 Empty 30B cylinder This calculation repeats the calculation for Benchmark 1 with the ambient temperatures defined in [ADR] and [SSR-6]. The deviations from the benchmark calculations are:
The fire temperature is set to 800 °C with a duration of 30 min.
During cooling down, the ambient temperature is 38 °C with solar insolation as defined in section 6.7.
Figure 9-1 shows an overview of the temperatures at the DN30 package loaded with an empty cylinder. The temperatures of the real thermal test are shown as dashed lines for comparison.
The following can be observed:
The temperatures at the surface of the DN30 package are significantly lower compared to the Benchmark 1 fire test.
The decrease of the temperatures at the surface of the DN30 package in the cooling phase is slower than the temperature curve measured for the prototype test. Below 400 °C the temperatures remain well above the measured temperature at the prototype.
62 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
1000 30B Valv e 30B Plug 800 30B Ma ntle D N30 Inner Shell D N30 Outer Shel l 600 30B Valv e ( BM)
30B Plug (BM) 400 30B Ma ntle (B M)
D N30 Outer Shel l (B M)
Temp er atur e li mit 18 3 °C
200
0 0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-1: Temperatures at the DN30 package loaded with an empty 30B cylinder - all temperatures for fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down
Figure 9-2 shows the temperatures of the valve, the plug, and of the mantle of the 30B cylinder.
Table 9-2 lists the temperatures at relevant positions of the DN30 package. The following can be observed:
The maximum temperatures at the valve and plug are with 122 °C resp. 120 °C well below the temperature limit of 183 °C.
The maximum temperatures of the 30B cylinder are considerably lower than the temper-atures measured at the prototype as the regulatory fire temperature according to the ADR/IAEA regulations is considerably lower than the temperatures measured during the experimental fire tests.
The maximum temperatures at the 30B cylinder and its components are delayed in com-parison with the temperatures measured in the thermal test.
63 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
200
180
160
140
120
100 30B Valv e
80 30B Plug 60 30B Ma ntle 30B Valv e ( BM) 40 30B Plug (BM) 20 30B Ma ntle ( BM)
Temp eratur e li mit 18 3 °C 0
0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-2: Temperatures at the DN30 package loaded with an empty 30B cylinder - man-tle of the 30B cylinder as well as valve and plug thread for the fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down
Table 9-2: Temperatures at the DN30 package loaded with an empty 30B cylinder
DN30 package loaded with an empty 30B cylinder Prototype test for Benchmark 1 Position Temperature Time Temperature Time
[°C] [s] [°C] [s]
Valve 122 12710 131 12680
Plug 120 10910 1441) 9450
Mantle 30B cylinder 124 10310 1431) 8505
Inner shell DN30 PSP 188 3650 - -
Outer shell DN30 PSP 785 1810 992 470
- 1) Average temperature during fire phase considerably higher at the plug side and longer duration of the fire (34 min), see section 7.2.
Figure 9-3 to Figure 9-5 show the temperature distribution in the DN30 package at different times during the thermal test. Figure 9-3 shows the temperatures at the end of the fire, Figure 9-4 at the time when the self-sustaining burning of the foam stops, and Figure 9-5 at the time of when the maximum temperature of the valve is reached.
64 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 9-3: Temperature distribution at the DN30 package loaded with an empty 30B cyl-inder at the end of the fire (t = 1810 s)
Figure 9-4: Temperature distribution at the DN30 package loaded with an empty 30B cyl-inder at the end of the burning of the foam (t = 3620 s)
Figure 9-5: Temperature distribution at the DN30 package loaded with an empty 30B cyl-inder around the time of the maximum valve temperature (t = 12710 s)
65 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.2 Filled 30B cylinder In this calculation, a full 30B cylinder is analyzed. The cylinder contains 2277 kg of UF 6 homoge-neously distributed in the cavity of 0.736 m 3 volume. The density is given in subsection 6.4.1 (3093.8 kg/m3). All other parameters are identical to the parameters used for the analysis of the DN30 package loaded with an empty 30B cylinder documented in section 9.1.
Figure 9-6 shows an overview of the temperatur es at the DN30 package loaded with a filled cyl-inder. The temperatures of the experimental fire test with an empty 30B cylinder are shown as dashed lines in comparison. The following can be observed:
The temperatures at the surface of the DN30 package and for the 30B cylinder are signif-icantly lower compared to the Benchmark 1 fire test, as the regulatory fire temperature according to the ADR/IAEA regulations is considerably lower than the temperatures meas-ured during the experimental fire tests.
The decrease of the temperatures at the surface of the DN30 package in the cooling phase is slower than the temperature curve measured for the prototype test. Below 400 °C the temperatures remain well above the measured temperature at the prototype.
1000 30B Valv e 30B Plug 800 30B Ma ntle D N30 Inner Shell D N30 Outer Shel l 600 30B Valv e ( BM)
30B Plug (BM) 400 30B Ma ntle (B M)
D N30 Outer Shel l (B M)
Temp er atur e li mit 18 3 °C
200
0 0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-6: Temperatures at the DN30 package loaded with a filled 30B cylinder - all tem-peratures; for the fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down
Figure 9-7 shows the temperatures of valve and plug and of the mantle of the 30B cylinder. Ta-ble 9-3 lists the temperatures at relevant positions of the DN30 package. The following can be observed:
The maximum temperatures at the valve and plug are well below the temperature limit of 183 °C with 110 °C and 107 °C.
66 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
The maximum temperatures are well below the temperatures measured at the prototype and the temperatures calculated for the DN30 package loaded with an empty 30B cylin-der.
200
180
160
140
120
100 80 30B Valv e 30B Plug 60 30B Ma ntle 30B Valv e ( BM) 40 30B Plug (BM)
20 30B Ma ntle ( BM)
Temp eratur e li mit 18 3 °C 0
0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-7: Temperatures at the DN30 package loaded with a filled 30B cylinder - mantle of the 30B cylinder, valve and plug thread, for the fire phase and 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> of cooling down
Table 9-3: Temperatures at the DN30 package loaded with a filled 30B cylinder
DN30 package loaded with a filled 30B cylinder Prototype test for Benchmark 1 Position Temperature Time Temperature Time
[°C] [s] [°C] [s]
Valve 110 10910 131 12680
Plug 107 9710 1441) 9450
Mantle 30B cylinder 118 9110 143 1) 8505
Inner shell DN30 PSP 188 3650 - -
Outer shell DN30 PSP 785 1810 992 470
- 1) Average temperature during fire phase considerably higher at the plug side and longer duration of the fire (34 min), see section 7.2.
67 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 9-8 to Figure 9-10 show the temperature distribution at the DN30 package at different times during the thermal test. Figure 9-8 shows the temperature distribution at the end of the fire, Figure 9-9 at the end of the burning of the foam, and Figure 9-10 at the time the maximum tem-perature of the valve is reached.
Figure 9-8: Temperature distribution at the DN30 package loaded with a filled 30B cylin-der at the end of the fire (t = 1810 s)
Figure 9-9: Temperature distribution at the DN30 package loaded with a filled 30B cylin-der at the end of the burning of the foam (t = 3620 s)
68 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 9-10: Temperature distribution at the DN30 package loaded with a filled 30B cylin-der around the time of the maximum valve temperature (t = 10910 s)
Figure 9-11 shows the temperature distribution for the UF 6 at the maximum temperature of the UF6. The maximum temperature is approx. 20 °C below the admissible temperature of 131 °C (see Table 3-1) and the temperature at the center of the UF 6 is still below 64 °C, hence there is no melting of the whole UF 6.
Figure 9-11: Temperature distribution of the volume (UF 6) inside the 30B cylinder at the maximum temperature (t = 10910 s)
69 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.3 Partially filled 30B cylinder For the partially filled cylinder with a filling ratio of 50 %, the material data of UF 6 (see section 6.4.1) is used with the same values for specific heat and thermal conductivity. For the density, the base value of UF6 is reduced to 50 % for the partially filled cylinder ( 1546.9 ). Con-servatively, the maximum heat generation of UF 6 of 4.1 W/m³ is used for the partially filled cylinder as well.
Figure 9-12 and Table 9-4 show the maximum temperatures at the relevant positions of the DN30 package for a filling ratio of the 30B cylinder of 50 % compared to the temperatures from the benchmark calculation.
200
180
160
140
120
100 80 30B Valv e 30B Plug 60 30B Ma ntle 30B Valv e ( BM) 40 30B Plug (BM)
20 30B Ma ntle ( BM)
Temp eratur e li mit 18 3 °C 0
0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-12: Maximum temperatures at the DN30 package loaded with a partially filled 30B cylinder
70 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 9-4: Maximum temperatures at the DN30 package loaded with a partially filled 30B cylinder
DN30 package loaded with a partially filled 30B cylinder Prototype test for Benchmark 1 Position Temperature Time Temperature Time
[°C] [s] [°C] [s]
Valve 113 11510 131 12680
Plug 111 9710 1441) 9450
Mantle 30B cylinder 119 9110 143 1) 8505
Inner shell DN30 PSP 188 3650 - -
Outer shell DN30 PSP 785 1810 992 470
- 1) Average temperature during fire phase considerably higher at the plug side and longer duration of the fire (34 min), see section 7.2
Figure 9-13 shows a comparison of ACT calculations for the valve thread and the volume inside the 30B cavity for the different filling ratios empty, partially filled (50 %) and filled (100 %). The maximum temperature at the valve is significantly higher for an empty 30B cylinder.
130
120
110
100
90 30B V alv e (Empty)
80 30B C avi ty (Empty) 30B V alv e (Par tiall y Fil led) 30B C avi ty (Par tiall y Fil led) 70 30B V alv e (Fill ed)
30B C avi ty (Fill ed) 60 0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-13: Comparison of valve and 30B cylinder cavity temperatures for different fill-ing ratios
71 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.4 Sensitivity analysis of reduced thickness of the Microtherm layer In these calculations, the influence of a reduced thickness of the Microtherm insulation layer is investigated. The first case is a reduction of the thickness from the nominal 10 mm to 9.5 mm, the lowest thickness within manufacturing tole rances. Thermal loads and boundary conditions are identical to the IAEA/ADR requirements for an analysis of the DN30 package loaded with an empty 30B cylinder as documented in section 9.1.
Figure 9-14 shows the calculated temperatures of valve, plug and mantle of the 30B cylinder for the model with reduced thickness compared to the calculation with the standard model with the nominal thickness. The corresponding maximum temperature are listed in Table 9-5. The follow-ing can be observed:
The maximum temperatures at the valve and plug are well below the temperature limit of 183 °C with 123 °C and 122 °C.
The temperature distribution is very similar to the calculation for the model with the nomi-nal thickness of the Microtherm insulation layer.
The difference in the maximum temperatures caused by a reduction of the thickness of the Microtherm insulation layer from 10 to 9.5 mm is very small with approximately 1 °C.
200
180
160
140
120
100 30B Valve (9.5 mm)
80 30B Plug (9.5 mm) 60 30B Mantle (9.5 mm) 30B Valve (Standard) 40 30B Plug (Standard) 20 30B Mantle (Standard)
Temperature limit 183 °C 0
0 5000 10000 15000 20000 25000 Time [s]
Figure 9-14: Maximum temperatures at the DN30 package loaded with an empty cylinder, thickness of Microtherm reduced to 9.5 mm
72 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 9-5: Temperatures at the DN30 package loaded with an empty 30B cylinder, thick-ness of Microtherm reduced to 9.5 mm
DN30 package (Microtherm DN30 package (no reduction of thickness reduced to 9.5 mm) Microtherm thickness)
Position Temperature Time Temperature Time
[°C] [s] [°C] [s]
Valve 123 12710 122 12710
Plug 122 10910 120 10910
Mantle 30B cylinder 125 10310 124 10310
Inner shell DN30 PSP 188 3650 188 3650
Outer shell DN30 PSP 785 1810 785 1810
The influence of a reduced thickness of the Microtherm insulation layer due to damages sustained from the free drop tests for ACT is calculated for a total remaining thickness of 7.5 mm, 5.0 mm, and 2.5 mm.
Figure 9-15 shows the temperature over time for the valve and plug of the 30B cylinder. The following can be observed:
The maximum temperatures are higher for a thinner Microtherm insulation layer.
The maximum temperatures are reached in a shorter time for a thinner insulation layer.
The temperatures are higher at the plug for a thinner Microtherm insulation layer com-pared to the valve due to the increased thermal conductivity of the foam at the plug side.
Even for a reduction of the thickness of the Microtherm insulation layer to 2.5 mm for the whole DN30 package, the maximum temperatures at the valve and plug are well below the temperature limit of 183 °C with a maximum of 136 °C for the valve and 151 °C for the plug.
73 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
160
140
120
30B Valv e ( 9.5 m m) 100 30B Plug (9.5 mm)
30B Valv e ( 7.5 m m) 80 30B Plug (7.5 mm) 30B Valv e ( 5.0 m m) 60 30B Plug (5.0 mm) 30B Valv e ( 2.5 m m) 30B Plug (2.5 mm) 40 0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-15: Maximum temperatures at the valve and plug of the 30B cylinder for different thicknesses of the Microtherm insulation layer
The influence of a partial loss of the Microtherm insulation layer due to damages caused by the free drop test on a bar for ACT is investigated, too. Therefore, a part of the Microtherm insulation layer is deleted, causing a gap in the Microtherm around the valve and the plug. Figure 9-16 shows the gap in the Microtherm insulation layer.
The width of the gap is 11.5 mm, so that the total area covered by the gap is equivalent to the cross-sectional area of a drop test bar with Ø 150 mm. Figure 9-17 shows a comparison of the cross-section of the DN30 overpack. The cross-sectional area of a drop test bar with Ø 150 mm is shown in red and the area of the gap with an area equal to the drop test bar is shown in blue.
Additional models with a width of the gap of 50 mm and 100 mm, equal to four resp. eight times the cross-sectional area of a drop test bar, are included as well. The position of the gap for these three modifications is listed in Table 9-6 with the inner radius r i and the width of the gap d Gap as indicated in Figure 9-16.
74 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 9-16: Gap in the Microtherm insulation layer (pictured is the gap for four times the area of drop test bar)
Area of dr op tes t bar Pos ition of valve w ith Ø 150 m m Area of gap in Mic rotherm equal to 1 x the ar ea of dr op tes t bar
D N30 pac kage Pos ition of plug
Area of dr op tes t bar Pos ition of valve w ith Ø 150 m m Area of gap in Mic rotherm equal to 4 x the ar ea of dr op tes t bar
D N30 pac kage Pos ition of plug
Figure 9-17: Comparison of the area of a drop test bar with Ø 150 mm and the gap with an equal area resp. four times the area
75 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 9-6: Position of the gap for the damaged Microtherm insulation layer
Equivalent area of a drop test bar with Ø 150 mm ri [mm] dGap [mm]
1 x AØ150 238.48 11.52
4 x AØ150 200.00 50.00
8 x AØ150 175.00 100.00
Figure 9-18 and Figure 9-19 show the temperature over time for the valve and the plug of the 30B cylinder. The following can be observed:
The maximum temperature at the valve and plug are higher for a wider gap in the Mi-crotherm insulation layer.
The maximum temperatures are reached in a shorter time.
The influence of the gap and its size is significantly higher for the plug side due to the increased thermal conductivity of the foam at the plug side.
Even if an area eight times the size of a drop test bar is lost, the maximum temperatures at the valve and plug are below the temperature limit of 183 °C with a maximum of 124 °C for the valve side and 176 °C for the plug side.
200
180
160
140
120
100
80
60 30B Valv e (com plete M icroth erm) 40 30B Valv e (area = bar Ø 150 mm )
30B Valv e (area = 4 x bar Ø 15 0 mm) 20 30B Valv e (area = 8 x bar Ø 15 0 mm)
Temp erature li mit 18 3 °C 0
0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-18: Maximum temperatures at the valve of the 30B cylinder for different gaps in the Microtherm insulation layer
76 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
200
180
160
140
120
100
80
60 30B Plug (c ompl ete Mi cr other m) 40 30B Plug (g ap area = 1 x area of a bar) 30B Plug (g ap area = 4 x area of a bar) 20 30B Plug (g ap area = 8 x area of a bar) 0 Temp er atur e li mit 18 3 ° C
0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-19: Maximum temperatures at the plug of the 30B cylinder for different gaps in the Microtherm insulation layer
77 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.5 Sensitivity analysis of extrapolated material properties of the foam A sensitivity analysis was performed to evaluate the influence of extrapolated thermal conductivity and specific heat of the foams. Therefore, the thermal conductivity and specific heat for RTS 120 and RTS 320 are varied in the range of +/-10 % for the thermal conductivity and specific heat above 250 °C. For the third set of thermal properties in this analysis, no extrapolation of thermal properties above 250 °C is considered. Thermal loads and boundary conditions are identical to the IAEA/ADR requirements for an analysis of the DN30 package loaded with an empty 30B cyl-inder as documented in section 9.1.
The modified material data for RTS 120 and RTS 320 is listed in Table 9-7 and Table 9-8.
Table 9-7: Thermal properties of RTS 120 foam / stainless-steel mixture used in the sensi-tivity analysis
Temperature
[°C] Standard Standard +10 % Standard -10 % No extrapolation Thermal conductivity [W/(m K)]
0 0.266 0.266 0.266 0.266 100 0.294 0.294 0.294 0.294 200 0.330 0.330 0.330 0.330 250 - - - 0.349 300 0.458 0.503 0.412 -
400 0.792 0.872 0.713 -
500 0.873 0.961 0.786 -
600 0.967 1.063 0.870 -
700 1.062 1.168 0.956 -
800 1.157 1.272 1.041 -
Specific heat [J/(kg K)]
23 1323 1323 1323 1323 50 1466 1466 1466 1466 75 1633 1633 1633 1633 150 1954 1954 1954 1954 200 2219 2219 2219 2219 250 2571 2571 2571 2571 300 2789 3068 2700 -
400 3318 3650 2986 -
500 3847 4232 3463 -
600 4377 4814 3939 -
700 4906 5396 4415 -
800 5435 5979 4892 -
78 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Table 9-8: Thermal properties of RTS 320 used in the sensitivity analysis
Temperature
[°C] Standard Standard +10 % Standard -10 % No extrapolation Thermal conductivity [W/(m K)]
0 0.197 0.197 0.197 0.197 100 0.256 0.256 0.256 0.256 200 0.320 0.320 0.320 0.320 250 - - - 0.426 300 0.451 0.496 0.406 -
400 0.465 0.511 0.418 -
500 0.487 0.536 0.438 -
600 0.558 0.614 0.502 -
700 0.629 0.692 0.567 -
800 0.701 0.771 0.631 -
Specific heat [J/(kg K)]
23 1284 1284 1284 1323 50 1415 1415 1415 1466 75 1542 1542 1542 1633 150 1912 1912 1912 1954 200 2082 2082 2082 2219 250 2772 2772 2772 2571 300 2876 3163 2588 -
400 3470 3816 3123 -
500 4063 4470 3657 -
600 4657 5123 4191 -
700 5251 5776 4726 -
800 5845 6429 5260 -
Figure 9-20 shows the temperature over time fo r the valve and plug of the 30B cylinder compared to the standard model described in section 9.1. The following can be observed:
The temperatures are on a similar level for the variation of the thermal properties in the range of +/-10 %.
The temperatures are considerably lower if no extrapolation of the thermal properties is considered.
79 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
130
120
110
100
90 30B Valv e (Standard) 30B Plug (Sta ndar d) 80 30B Valv e (+1 0 %)
30B Plug (+ 10 %)
70 30B Valv e (- 10 %)
30B Plug (- 1 0 %)
60 30B Valv e (No Ex trapola tion) 30B Plug (N o Ext ra polati on) 50 0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-20: Maximum temperatures at the valve and plug of the 30B cylinder for a varia-tion of thermal properties of the foam
80 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.6 Sensitivity analysis of parameters controlling the burning of the foam A sensitivity analysis was performed to evaluate th e influence of variations for the parameters of the modified Arrhenius equation controlling the burning of the foam in the range of +/-10 % to maximize resp. minimize the heat generation of the foam. Thermal loads and boundary conditions are identical to the IAEA/ADR requirements for an analysis of the DN30 package loaded with an empty 30B cylinder as documented in section 9.1.
The different parameters are listed in Table 9-9.
Table 9-9: Factors used for the calculation of the heat generation of the foam
Factor Standard Standard +10 % Standard -10 %
A 1.010-3 1.110-3 0.910-3 B 300 270 330 ENT 2.0107 2.2107 1.8107
Figure 9-21 shows the temperature over time fo r the valve and plug of the 30B cylinder compared to the standard model discussed in section 9.1. The following can be observed:
The temperatures are on a similar level for the variation of the parameters in the range of
+/-10 %.
130
120
110
100
90
30B Valv e (Standar d) 80 30B Plug (Sta ndard)
70 30B Valv e (+1 0 %)
30B Plug (+ 10 %)
60 30B Valv e (- 10 %)
30B Plug (- 1 0 %)
50 0 5000 1000 0 1500 0 2000 0 2500 0 Tim e [s ]
Figure 9-21: Maximum temperatures at the valve and plug of the 30B cylinder for a varia-tion of the parameters controlling the burning of the foam
81 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
9.7 Pressure build-up in the 30B cylinder For temperatures above the triple point of UF 6 (64 °C) a pressure build-up is investigated for melted UF6 contents. The following conservative assumptions are used for the calculations:
The maximum temperature of the UF6 is assumed to be the average temperature of all UF6 in the 30B cylinder. Calculations show that only the outmost region of the UF 6 in contact with the 30B cylinder shell will reach that maximum temperature while the core of the UF6 is still in its solid state; the average temperature calculated for all UF 6 is well below that maximum temperature. For example, for the HAC calculation for a full cylinder (100 %
filling), the maximum temperature of more than 50 % of all UF 6 in the 30B cylinder is still below the triple point of 64 °C.
For the calculation of the pressure build-up, the pressure of the empty 30B cylinder is assumed to be 5 psi instead of 3 psi (see [USEC 651] sections 5.3.5 and 8.3).
The calculations for the pressure build-up are calculated with the safe fill limit of 95 %
instead of the minimum volume (see [USEC 651]).
The admissible pressure for the 30B cylinder, the valve and the plug for HAC is conservatively set to the test pressure of 2.76 MPa/400 psig as specified in [ANSI N14.1] for these components.
The pressure build-up is investigated for filling ratios of 50 % and 100 % for a temperature of 131 °C/267.8 °F.
The vapor pressure data of UF 6 is extracted from [DeWITT] and listed in Table 9-10 below.
Table 9-10: Vapor pressure of UF6 extracted from [DeWITT]
Temperature [°C] Vapor pressure [MPa]
63.88 0.15061) 64.20 0.1527 91.88 0.3395 99.94 0.4174 108.07 0.5091 116.03 0.6129 124.17 0.7349 133.19 0.8892 141.44 1.0517 149.50 1.2300 180.57 2.1313 207.32 3.2436 230.20 4.6103 230.20 4.6103
- 1) vapor pressure of the solid, for all other temperatures: vapor pressure for the liquid
82 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
The calculation of the total pressure is based on [USEC 651]. The input data for a 30B cylinder is according to [ANSI N14.1]:
26 ft
, 5020 lb 68 °F 20 °C 267.8 °F 131 °C
, 317.8 lb ft (solid, at 20 °C/68 °F)
, 198.6 lb ft (liquid, at 131 °C/267.8 °F)
The pressure inside an empty 30B cylinder is 5 psi or less according to [USEC 651].
, 5 psi
The volume of UF6 and air inside the 30B cylinder for a filling ratio of 50 % are then:
, 0.5,, 0.5 5020 lb317.8 lb ft 7.90 ft
V, V V, 26 ft 7.90 ft 18.10 ft The air pressure at 20 °C/68 °F is:
p, p, VV, 5 psi 26 ft18.10 ft 7.18 psi
The vapor pressure of liquid UF 6 at 131 °C/267.8 °F is interpolated linearly from Table 9-10:
, 123.53 psi The volume and pressure of the air inside the 30B cylinder at 131 °C/267.8 °F are:
, 0.5,, 26 ft 0.5 5020 lb198.6 lb ft 13.36 ft
,,,, 7.18 psi 18.10 ft 404.15 K13.36 ft 293.15 K 13.42 psi
The total pressure at 131 °C/267.8 °F is the summation of the partial pressures:
,,, 123.53 psi 13.42 psi 136.95 psi 0.94 MPa The admissible pressure for HAC is the test pressure specified in [ANSI N14.1]. The test pressure of 400 psi/2.76 MPa is applicable for the cylinder itself and for the valves. This testing pressure is then corrected for the elevated temperature of 131 °C/267.8 °F (above the design limit of 121 °C/250 °F) using the strength values for ASME SA516 Grade 55 according to [ASME BPVC]:
200 °F 300 °F 55.0 ksi 250 °F 27.0 ksi 300 °F 26.5 ksi The yield strength at 131 °C/267.8 °F is interpolated linearly:
267.8 °F 267.8 °F 250 °F300 °F 250 °F 26.5 ksi 27.0 ksi 27.0 ksi 26.82 ksi
83 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
The corrected limit for an elevated temperature of 131 °C/267.8 °F is then:
,. ° 400 psi 26.82 ksi27.0 ksi 397.36 psi 2.74 MPa
The maximum pressure at 131 °C/267.8 °F is below the corrected admissible pressure of 2.74 MPa/397.36 psi with a safety margin of 2.90 for a filling ratio of 50 %.
The volume of UF6 and of the air inside the 30B cylinder for a filling ratio of 100 % are:
,,, 5020 lb317.8 lb ft 15.80 ft
V, V V, 26 ft 15.80 ft 10.20 ft The air pressure at 68 °F is:
p, p, VV, 5 psi 26 ft10.20 ft 12.74 psi
The vapor pressure of liquid UF 6 at 131 °C/267.8 °F is interpolated linearly from Table 9-10:
, 123.53 psi
The volume and pressure of the air inside the 30B cylinder at 131 °C/267.8 °F are:
,,, 26 ft 5020 lb198.6 lb ft 0.72 ft
,,,, 12.74 psi 10.20 ft 404.15 K0.72 ft 293.15 K 249.40 psi
The total pressure at 131 °C/267.8 °F is the summation of the partial pressures:
,,, 123.53 psi 249.40 psi 372.93 psi 2.57 MPa The maximum pressure at 131 °C/267.8 °F is below the corrected admissible pressure of 2.74 MPa/397.36 psi with a safety margin of 1.07 for a filling ratio of 100 %. As this safety is calculated with the testing pressure specified in [ANSI N14.1], this safety also includes an addi-tional safety margin because the testing pressure only utilizes a maximum of 95 % of the yield strength for primary stresses (see [ASME BPVC]).
Additionally, the safety margin for the 30B cylinder is investigated according to ASME Code Sec-tion VIII - Division 1 for the maximum internal pressure of 2.57 MPa.
The dimensions of the 30B cylinder are according to [ANSI N14.1]:
Outer diameter: 30 in 762 mm Nominal shell thickness: 1 2 in 12.7 mm
Minimum shell thickness: 5 16 in 7.94 mm Joint efficiency: 0.85 (spot RT as a minimum [ANSI N14.1])
Yield Strength at 131 °C, ° 26.8 ksi 184.93 MPa (interpolated linearly from [ASME BPVC])
84 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
At first, the cylinder itself is investigated. The outside radius, the inside diameter and the inside radius are calculated to:
2 762 mm2 381 mm
2 762 mm 2 7.94 mm 746.1 mm
2 746.1 mm2 373.1 mm
The required wall thickness is then calculated according to the internal pressure design calcula-tion of cylinders [ASME BPVC]. The formula and results are listed in Table 9-11:
Table 9-11: Required wall thickness for the cylinder for internal pressure
Paragraph/
Appendix Formula Required Wall Thickness treq Safety
I App. 1-1(1) 0.4 6.19 mm 1.28
II UG-27(1) 0.6 6.16 mm 1.29
III App. 1-2(1) 1 exp 6.18 mm 1.28
IV App. 1-2(1) exp 1 6.15 mm 1.29
Next, the elliptical heads of the cylinder are investigated. The inside diameter, the height of the head, and the K-Factor are calculated to:
2 762 mm 2 7.94 mm 746.1 mm
h D4 746.1 mm4 186.5 mm
K 16 2 D2 h 16 2 746.1 mm2 186.5 mm 1.00
85 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
The required wall thickness is calculated according to [ASME BPVC]. The results are listed in Table 9-12:
Table 9-12: Required wall thickness for the elliptical heads for internal pressure
Paragraph/
Appendix Formula Required Wall Thickness treq Safety
I App. 1-4(1) 2 2 0.1 6.14 mm 1.29
II App. 1-4(1) 2 0.2 6.11 mm 1.30
The minimum safety factors calculated according to [ASME BPVC] are 1.28 for the cylinder and 1.29 for the elliptical heads for a pressure of 2.57 MPa at a temperature of 131 °C/267.8 °F.
Additional calculations are done for the solder connection between the valve and the cylinder.
The ultimate tensile strength of the tin-lead-solder at room temperature is 42 MPa/6090 psi. With the conservative assumption, that the tensile strength is 0 MPa at the solidus temperature of 183 °C, the tensile strength at 131 °C is interpolated linearly:
131 °C 131 °C 20 °C183 °C 20 °C 42 MPa 0 MPa 0 MPa 28.6 MPa
The safety factor for the maximum internal pressure of 2.57 MPa is therefore 11.1 for the solder itself. Next, the solder connection is investigated:
The radius for the solder connection is calculated as the middle radius at the beginning of the thread (dimensions according to [ANSI N14.1]). The thread roots shall be filled approx. half full, therefore the height is the height of the solder at that radius:
0.5 0.5 1 964 14.5
0.5 0.5 1 14 15.9
2 14.5 15.9 2 15.2
15.2 14.5 0.70 With at least 7 inserted threads (see [ANSI N14.1]), the thread area is calculated to:
2 7 2 15.2 7 0.70 463.7 The area the internal pressure is acting on is calculated with the outer (, ) and inner diameter
(,) of the valve the pressure is acting on:
4,, 4 1 14 78 4 31.8 22.2 403.8 The stress for the solder connection is then:
2.57 403.8 463.7 2.24
86 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
The strength of the solder connection is conserva tively assumed to be 20 MPa (see [ROLOFF]).
The safety factor is then calculated to:
20 2.24 8.94
The admissible temperatures for the contents of the 30B cylinder is therefore set to 131 °C/267.8 °F (see Table 3-1). For this admissible temperature there is no danger of rupture because of a possible pressure build-up caused by elevated temperatures and melted UF 6 con-tents.
87 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
10 Proof of the package DN30 to meet the requirements of ADR and IAEA regula-tions
10.1 Ambient temperatures and pressures For the thermal analysis, an ambient temperature of 38 °C is taken into account (see sec-tion 6.8.1). Pressures that are likely to be encountered during RCT and NCT have no effect on the results of the thermal analysis. The requirements of [ADR], No 6.4.2.10 and [SSR-6] para.
616 are met.
10.2 Rupture of the containment system In section 9.7 a maximum pressure of 0.75 MPa at a temperature of 125 °C is evaluated. Accord-ing to [ISO 7195] and [ANSI N14.1], the 30B cylinder is designed for an internal pressure of 1.38 MPa. The requirements of [ADR] No. 6.4.6.2 c) and [SSR-6] para. 632 c) are met.
10.3 Temperature of the accessible surface In section 8.1 the maximum temperature at the surface is calculated with 38 °C. This is well below the admissible temperature of 50 °C. The requir ements of [ADR], No 6.4.8.3 and [SSR-6] para.
654 are met.
With this proof, it is also verified that the requirements of [ADR], No 6.4.8.4 and [SSR-6] para.
655 are met.
10.4 Influence of the thermal test on the shielding analysis The thermal analysis for ACT in chapter 9 shows that the stainless-steel shells of the DN30 PSP as well as the carbon steel envelope of the 30B cylinder are not affected by the thermal test in such a way that their shielding properties are reduced. It is verified in subsection 2.2.4 of the main part of the SAR that considering this result of the thermal analysis, the radiation level 1 m from the surface of the package does not exceed 10 mSv/h. In the shielding analysis, a complete loss of the foam is considered.
The requirements of [ADR] No. 6.4.8.8 and [SSR-6] para. 659 with respect to the thermal test are met.
10.5 Influence of the thermal test on the containment analysis The thermal analysis for ACT in chapter 9 shows that the temperature of the valve and plug thread does not exceed 183 °C, i.e., the solidus temperature of the tin-lead solder of the valve and plug threads is not reached. Hence, it can be assumed that the leakage rate of the DN30 package is not affected by the thermal test. The proof of containment in subsection 2.2.3 of the main part of the SAR considers the measured leakage rate for the prototype.
The requirements of [ADR] No. 6.4.8.8 and [SSR-6] para. 659 with respect to the thermal test are met.
10.6 Influence of the thermal test on the criticality safety analysis The thermal analysis for ACT in chapter 9 shows that the stainless-steel shells of the DN30 PSP as well as the carbon steel envelope of the 30B cylinder are not affected by the thermal test in such a way that their thickness and density is reduced. It is verified in subsection 2.2.5 of the main part of the SAR that considering this result of the thermal analysis, criticality safety is ensured. In the criticality safety analysis, a comple te loss of the foam is considered.
The requirements of [ADR] No. 6.4.11.10 and 6.4.11.13 as well as [SSR-6] para. 682 and 685 with respect to the thermal test are met.
88 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
10.7 Component temperatures of the DN30 package The maximal component temperatures during RCT and NCT are at their maximum 63 °C for the DN30 PSP, and 58 °C for the 30B cylinder and its contents (see section 8.2), and therefore lower than the admissible values specified in Table 3-1.
The maximum component temperatures of the DN30 PSP during ACT are close to the admissible temperatures defined in Table 3-1. However, tests with a prototype of the DN30 package showed that these temperatures have no negative effect on the function of the inner and outer shell of the DN30 PSP with respect to shielding and criticality safety.
The maximum temperatures of the 30B cylinder are well below the admissible temperature de-fined in Table 3-1. The temperatures of the valve and plug thread do not exceed the admissible temperature specified in Table 3-1.
89 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
11 Thermal Analysis without the valve housing In this chapter, the design change of the removal of the housing around the 30B cylinder valve for the series design of the DN30 PSP in comparison to the DN30 PSP prototypes used for the two experimental thermal tests is discussed. Other design changes like the improvement of the feet and the rotation device (see section 2.1.4.2 of the main part of the SAR) have no relevance regarding the thermal behavior of the package.
The intumescent material and the housing of the valve protecting device were added to the design of the DN30 PSP for the second fire test (Benchma rk 2) to improve the thermal protection of the 30B cylinder by preventing the stream of hot decomposition gases into the cavities between the DN30 PSP and the 30B cylinder. The housing affixed with the intumescent material was intended as an additional protection for the 30B cylinder valve to further reduce the stream of hot gases into the valve area by expansion of the intumescent material. The main function of the housing is to provide a casing for the intumescent material, so that the intumescent material can be placed in the vicinity of the valve and in case the temperature rises to 150 °C or above, tightly encloses the valve by expansion of the intumescent material. A small influx of hot gases is possible for ACT as the DN30 PSP is not air-tight and has no containment function (only the 30B cylinder). The drop tests for ACT (see Appendix 2.2.1.3 of the main part of the SAR) demonstrate, that no large openings or gaps towards the cavity occur at the package and that the closure devices keep both halves of the DN30 PSP closed together.
However, adding the microporous insulation layer Microtherm to the design of the DN30 PSP for the third fire test (Benchmark 1) was so efficient regarding the thermal protection of the 30B cyl-inder that the housing is no longer needed to meet the admissible temperatures, as the safety margin was at least 39 °C for the tinning of the cylinder valve and plug threads (see Table 3-1, maximum admissible temperature: 183 °C). Furthermore, the microporous insulation is so effi-cient that there is no visible expansion of the intumescent material inside the housing, demon-strating that the housing is not fulfilling its design safety function anymore due to the effectiveness of the microporous insulation layer.
The thermal analysis shows that the maximum temperature at the valve is reached during the cooldown phase, when the heat transfer is mainly determined by conduction through the 30B cyl-inder mantle, so that the removal of the housing only has a small impact on the maximum tem-perature of the valve and the thermal safety of the packaging.
The following sections evaluate the impact of the removal of the housing on the thermal safety of the packaging. They are focused on the maximum temperatures of the valve since the impact of the removal of the housing on other parts of the 30B cylinder will be even lower. More specifically, no significant influence is expected for the UF 6 contents, as the surface of the cylinder covered by the housing compared to the total 30B cylinder surface is very small. Accordingly, there is no risk of exceeding the UF 6 content temperature limit of 131 °C (see Table 3-1) due to the removal of the housing. This admissible temperature of the UF 6 contents of 131 °C is itself a very con-servative limit, as is shown in the thermal analysis for a filled 30B cylinder in section 9.2, that shows that the core of the UF 6 content remains solid for HAC, greatly reducing the possible pres-sure build-up due to melted UF 6 contents compared to the pressure calculated for the admissible temperature of 131 °C that conservatively assumes completely melted UF 6 contents.
The impact for RCT and NCT is briefly discussed in section 11.1. For ACT, two approaches are used that both use conservative assumptions and focus on the valve temperatures under ACT:
one using the results of the third and last experimental fire test, the Benchmark 1 fire test (sec-tion 11.2) and one using numerical analyses (section 11.3). Both approaches clearly demonstrate that the housing has only a small impact on the valve temperature and that large safety margins still remain for the valve temperature if the housing is removed.
90 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
11.1 Evaluation for RCT and NCT As the housing is designed to limit the influx of hot gases by expansion of the intumescent material for temperatures above 150 °C, there are no effects of the removal of the housing on the thermal analysis for RCT and NCT, as there are no hot gases present during RCT and NCT, and the temperature range encountered during RCT and NCT is less than 65 °C (see section 8.2). This is significantly below the expansion temperature of the intumescent material, which would be the only way for the presence or absence of the housing to theoretically influence the thermal analy-sis.
11.2 Evaluation for ACT based on the results of the experimental fire tests The impact of the removal of the housing on the thermal analysis of the DN30 package can be assessed directly using the results of the experimental fire tests.
11.2.1 Evaluation based on thermocouples near the valve not covered by the housing The assumption, that the removal of the housing has only a very minor influence on the maximum temperatures is confirmed by the evaluation of the data recorded by the temperature sensors at the valve side during the Benchmark 1 fire test, see thermal test report 3.2/817026 (Appendix 2.2.2.2 of the main part of the SAR). The positions of these temperature sensors are shown in Figure 11-1.
Figure 11-1: Positions of temperature sensors during the Benchmark 1 fire test
The temperature-time-curve during the cooldown phase of Benchmark 1, when the maximum temperatures are reached, is shown in Figure 11-2 for the thermocouples A-1 (on top of the 30B cylinder skirt at the valve side), A-2 (middle of the 30B cylinder head at the valve side next to the nameplate), B-1 (tip of the valve stem), and B-3 (valve body).
91 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
200 °C
180 °C A 30B Cylinder skirt
160 °C A 30B Cylinder head B Tip of Valve Stem 140 °C B Valve body Temperature limit 183 °C
120 °C
100 °C Tmax, A-1 = 134 °C Tmax, A-2 = 130 °C 80 °C Tmax, B-1 = 131 °C Tmax, B-3 = 131 °C 60 °C 0:00 2:24 4:48 7:12 9:36 12:00 Time [h:mm]
Figure 11-2: Maximum temperatures at the 30B cylinder during the cooldown phase of the Benchmark 1 fire test
The comparison of the temperatures recorded for these thermocouples shows that the tempera-ture at the valve sensors covered by the housing is nearly identical to the temperature at the sensor A-2 at the cylinder head and lower than the temperature at the sensor A-1 on top of the skirt of the cylinder, especially during cooling down when the maximum temperatures are reached. It should be noted that sensor A-2 is not covered by the housing, so the similarity of its temperature to the valve sensors shows, that the housing has no significant effect on thermal safety in the valve region, as the heat transfer during the cooldown phase is mainly determined by conduction through the 30B cylinder mantle.
An evaluation based on the sensors A-1 and A-2 provides a conservative estimation of the max-imum temperatures of the valve in case the housing is removed because:
Sensor A-1 is mounted on top of the skirt and is closer than the valve to the corner area of the cavity, and therefore closer to an increased heat input that may occur during the fire phase due to the steel reinforcements and the damages of the preliminary drop tests in the corner area.
Sensor A-2 is not covered by the housing and directly mounted to the 30B cylinder head and is therefore a good indication for the heat transfer caused by an influx of hot gases through the gap between the top and bottom half of the DN30 PSP during the fire phase.
Both sensors A-1 and A-2 are directly mounted to the 30B cylinder and are therefore a good indication of the heat influx to the valve during the cooldown phase, when the maxi-mum temperatures are reached, as the heat transfer during this time is dominated by conduction through the 30B cylinder mantle.
Assuming the maximum temperature recorded at the valve side for a sensor position without housing (thermocouple A-1: 134 °C) for the valve in case the housing is removed leads to a con-servative estimation of an increase in the maximum valve temperature of approx. 3 °C, with a large safety margin of 49 °C to the admissible temperature of the valve of 183 °C still remaining.
92 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
11.2.2 Evaluation based on the behavior of the intumescent material An additional confirmation for the assumption that the removal of the housing only has only a minor influence on the maximum temperatures and that the housing does not fulfill its design safety function anymore is the comparison of the condition of the intumescent material in the housing for the Benchmark 1 and Benchmark 2 fire tests as shown in Figure 11-3.
Figure 11-3: Housing with intumescent material after the Benchmark 1 and 2 fire tests
The comparison of the intumescent material leads to the following conclusions:
During the Benchmark 2 fire test, the maximum temperature at the valve was measured to be 234 °C. The temperature at the housing was significantly above 150 °C, clearly caus-ing an expansion of the intumescent material that is attached to the inside of the housing.
During the Benchmark 1 fire test, the maximum temperature at the valve was measured to be 131 °C. The temperature at the housing did not reach the expansion temperature of the intumescent material, and therefore did not significantly exceed 150 °C, as no expan-sion of the intumescent material inside the housing was observed.
Since no expansion of the intumescent material inside the housing was observed during the Benchmark 1 fire test, the intumescent material does not fulfil its intended function anymore due to the lower overall temperatures at the 30B cylinder that are caused by the addition of the very efficient microporous insulation layer. Additionally, the maximum temperature at the outer surface of the housing cannot be significantly higher than the onset temperature for the expansion of the intumescent material of approx. 150 °C (there cannot be any significant temperature difference across the 1 mm thick steel casing of the housing).
Accordingly, the temperature at the valve for a DN30 package without the housing cannot be significantly higher than 150 °C as well, confirming a large safety margin to the admissible tem-perature for the tinning of the valve thread of 183 °C (see Table 3-1).
93 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
11.2.3 Evaluation based on a comparison between valve and plug side Another conservative evaluation of the maximum temperature at the valve without the housing is to apply the maximum temperature recorded at the plug side during the Benchmark 1 fire test of 144 °C for the valve side as well instead of the actually recorded temperature of 131 °C. For the following reasons, deriving the valve temperature from the plug temperature is conservative for the analysis of the DN30 package without the housing:
The heat transfer mechanisms on the plug side are similar to those on the valve side with the housing removed, as there is no additional barrier between the intumescent material or the inner shell of the DN30 PSP and the valve or plug, respectively. The thicknesses of the steel sheets, foam components, and thermal insulation are, with exception of the axial steel reinforcements, the same for the valve and the plug side.
The distance between the DN30 PSP and the dished heads of the 30B cylinder is shorter on the plug side because of the shorter skirts of the 30B cylinder.
While the conditions during the Benchmark 1 fire test in general exceeded the regulatory requirements, the fire test conditions recorded for the plug side exceeded those on the valve side (see the discussion in section 7.2 as well):
o The average fire temperature on the plug side was 996 °C compared to 920 °C on the valve side.
o The fire duration on the plug side was longer than on the valve side (34 min com-pared to 30 min).
The damaging of the DN30 PSP was consider ably higher on the plug side, where cracks occurred both at the inner and the outer shell after the performed drop test sequences used for pre-damaging the DN30 PSP (see Figure 11-4 and Figure 11-5).
Figure 11-4: Crack around the plug protecting device at the inner shell after the pre-dam-aging drop tests for the Benchmark 1 fire test
94 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
Figure 11-5: Damages to the outer shell after the pre-damaging drop tests for the Bench-mark 1 fire test
Even with the conservative assumption that the maximum temperature at the valve side for reg-ulatory ACT would be equal to the maximum temperature at the plug of 144 °C recorded for the Benchmark 1 fire test, the safety margin would still be large as the maximum temperature is 39 °C below the limit of 183 °C for the tinning of the valve thread (see Table 3-1). Conservatively, this safety margin is evaluated for an empty 30B cylinder and based on the higher fire temperature and longer fire duration for the Benchmark 1 fi re test exceeding the regulatory requirements.
11.3 Evaluation for ACT using numerical analyses In addition to the evaluation based on the experimental fire tests in the previous section, the removal of the housing is investigated using t he benchmarked numerical model as well. Like the valve protecting device, the housing is not explicitly modelled in the numerical model for the ther-mal analysis of the DN30 package because of the constraints of an axisymmetric two-dimensional model; instead, the housing is taken into account in the benchmarked heat transfer mechanisms that are explained in section 5.1.
For a conservative simulation of the DN30 package without the housing, the heat transfer mech-anisms on the valve side are increased to match or exceed the heat transfer mechanisms on the plug side. This approach can be used because without the housing, the situation on the valve side is very similar to that on the plug side with regard to the thermal behavior. The thermal be-havior on the plug side can be directly deduced from the Benchmark fire tests, as no additional protection devices like the housing were present here during these tests.
The heat transfer mechanisms in the gaps between the DN30 PSP and the 30B cylinder are:
The conduction for the gap on the valve side is increased from 30 W/(m² °C) to 45 W/(m²°C), the corresponding value used for the gap on the plug side (see section 6.9.2).
The radiation coefficients are already equal for the gap on the valve and the plug side with an emissivity of 1.0 under ACT.
The convection coefficients for the gap at the valve side are increased by a factor of 5 compared to the benchmarked values (see section 6.9.3) during the fire phase and the first 30 min of the cooldown phase until the burning of the foam stops compared to the plug side and the standard conditions as listed in section 9.1. For the rest of the cooldown
95 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
phase, the convection coefficients for the gap at the valve side are equal to the plug side to keep the heat in the package.
All other thermal loads and boundary conditions are identical to the analysis of the DN30 package loaded with an empty 30B cylinder that consi ders the IAEA/ADR requirements as documented in section 9.1. The only difference is that the simulation without housing is run as two separate tran-sient analyses (one for the fire and the first 30 min and one for the rest of the cooldown phase) to allow for the implementation of different convection coefficients. This is only due to the way con-vection is modelled and has no influence on the results of the thermal analysis.
Figure 11-6 shows the maximum temperatures over time for the valve and the plug of the 30B cyl-inder obtained with the numerical model without the housing in comparison to the results obtained with the standard numerical model with housing used for the analysis in section 9.1. The following is observed:
As expected, the temperature curves at the plug side in the simulations with and without the housing are basically indistinguishable from each other.
The removal of the housing leads to only a very small increase for the maximum temper-ature on the valve side of approx. 1 °C.
At the time when the burning of the foam stops (30 min after the end of the fire), the tem-perature at the valve is approx. 2 °C higher for the model without housing compared to the standard model due to the higher convection and conduction in the gap. Further into the cooldown phase this difference decreases as other heat transfer mechanisms, espe-cially the conduction through the 30B cylinder mantle, dominate the thermal behavior of the package during cooldown.
200
180
160
140
120
100 30B Valve (standard model) 80 30B Plug (standard model) 30B Valve (without housing) 60 30B Plug (without housing)
Temperature limit 183 °C 40 0 5000 10000 15000 20000 25000 Time [s]
Figure 11-6: Maximum temperatures at the valve and the plug with and without the hous-ing of the valve protecting device
96 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
There are two main reasons for this very small increase for the maximum temperature at the valve:
The heat transfer in the gaps on the valve and plug side are generally dominated by radi-ation, and the radiation coefficients for the standard numerical model and the model with-out the housing both assume the maximum possible emissivity of 1.0.
The heat transfer to the 30B cylinder during the cooldown phase when the maximum tem-peratures are reached is dominated by conduction through the 30B cylinder mantle.
Therefore, rather large changes to the heat transfer mechanisms for the gaps on the valve and plug side result in only small changes in the maximum temperatures of the 30B cylin-der and its components.
Without the housing, the maximum temperature at the valve for ACT is less than 123 °C. This demonstrates a very large remaining safety margin of 60 °C in comparison to the admissible tem-perature for the 30B cylinder of 183 °C (see Table 3-1). Both the calculation with the standard model and the calculation with the model without housing use an empty 30B cylinder to achieve the maximum temperatures for the DN30 package.
Further sensitivity analyses show, that further changes to the conductivity of the air gap of 45 W/(m²°C) with a factor of 0.5 (new,1 = 22.5 W/(m²°C)) and 2 (new,2 = 90 W/(m²°C)) only lead to small changes in the maximum temperature at the valve of less than - 0.5 °C for a factor of 0.5 and less than + 0.5 °C for a factor of 2. This confirms that the conduction for the air gap only has a very small influence on the maximum temperature of the valve, as radiation dominated the heat transfer through the air gap and the heat transfer in general is mainly determined by the conduc-tion through the 30B cylinder mantle during the cooldown phase, when the maximum tempera-tures are reached.
In a further analysis to investigate the influence of the reduced amount of intumescent material in the valve area due to the removal of the housing, a conservative model is created, where the complete intumescent material at the inner shell in radial direction is removed for the valve area.
All other modifications for the aforementioned model without the housing are included as well.
The maximum temperatures over time for the valve and the plug of the 30B cylinder are shown in Figure 11-7 for the modified model without housing and without intumescent material at the valve side compared to the standard model from section 9.1. The following is observed:
As expected, the temperature curves at the plug side in the simulations with and without the housing are basically indistinguishable from each other.
The temperature at the valve rises faster compared to the standard model, however the maximum temperatures are less than 2 °C higher for the modified model compared to the standard model.
At the time when the burning of the foam stops (30 min after the end of the fire), the tem-perature at the valve is approx. 6 °C higher for the modified model without housing com-pared to the standard model due to the higher convection and conduction in the gap and the removed intumescent material. Further into the cooldown phase this difference de-creases as other heat transfer mechanisms, especially the conduction through the 30B cylinder mantle, dominate the thermal behavior of the package.
97 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
200
180
160
140
120
100 30B Valve (standard model) 80 30B Plug (standard model) 30B Valve (without housing) 60 30B Plug (without housing)
Temperature limit 183 °C 40 0 5000 10000 15000 20000 25000 Time [s]
Figure 11-7: Maximum temperatures at the valve and the plug for the modified model compared to the standard model
Without the housing and with the conservative removal of all the intumescent material from the radial part of the inner shell at the valve side, the maximum temperature at the valve is less than 124 °C. A large safety margin of 59 °C compared to the admissible temperature of 183 °C for the valve still remains.
11.4 Conclusion Both the evaluations based on the results of the Benchmark fire tests for ACT (section 11.2) and the evaluations using numerical simulations (section 11.3) show that the removal of the housing only has a very small effect on the thermal safety of DN30 package, and that the safety margin to the admissible temperature of the valve remains large. It should be noted that these evaluations are based on several conservative assumptions, providing additional weight to the conclusion that the thermal safety is still ensured when the housing is removed.
Furthermore, the removal of the housing has no impact on the proof of the DN30 package to meet the requirements according to the IAEA and ADR regulations provided in chapter 10:
The temperatures of the outer DN30 PSP shell under RCT and NCT are not affected at all by this design change so that the temperatures of accessible surfaces are identical.
Changes in the temperatures of the UF 6 contents under RCT, NCT, and ACT must be negligible, because the area of the cylinder shell that was covered by the housing in com-parison to its total area is negligible, and the heat transfer to the cylinder during cooling down, when the maximum temperatures are reached, is dominated by conduction be-tween the inner DN30 PSP shell and the mantle of the 30B cylinder. Accordingly, there is no risk of exceeding the admissible temperature of the UF 6 contents of 131 °C (see Ta-ble 3-1) due to the removal of the housing.
98 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
12 Summary In this report, a calculation model for the DN30 package is developed based on the results of two experimental fire tests with prototypes of the DN30 package. The important parameters in this model are:
For the foam between the inner and outer shell of the DN30 PSP, incineration is modelled to achieve good agreement with the temperatures measured during the two experimental fire tests that are used for the benchmarking of the calculation model.
The thermal conductivity is increased by a factor of 4 for the RTS 320 foam and by a factor of 5 for the RTS 120 foam at the plug side to achieve good agreement between the tem-perature curves of the fire tests and the benchmark analyses.
The temperatures calculated with the benchmark models and the temperature curves over time are in good agreement with the results of the fire tests as shown in sections 7.2 and 7.3.
With this calibrated numerical model, the temperatures at the DN30 package are calculated with the ambient conditions specified in [ADR] and [SSR-6]. The maximum temperatures at the rele-vant positions of the DN30 package loaded with an empty 30B cylinder for ambient conditions according to IAEA/ADR guidelines are below the maximum temperatures measured at the proto-types during the experimental fire tests. For the DN30 package loaded with a filled or partially filled cylinder, the temperatures are below the temperatures calculated for the empty 30B cylinder.
Sensitivity analyses show that even for a great reduction of the thickness of the Microtherm insu-lation layer from 10 mm to 2.5 mm, the maximum temperatures of the critical components are well below the temperature limit. If gaps up to double the size of a bar used for the free drop test for ACT are present in the Microtherm insulation layer, the maximum temperatures are below the temperature limit, too. Further sensitivity analyses show that variations in the range of +/-10 % in the thermal properties of the foam above 250 °C or in the range of +/-10 % for the parameters controlling the burning of the foam show only a minor impact on the resulting temperatures.
In chapter 10, it is demonstrated that the DN30 package fulfils the requirements of [ADR] and
[SSR-6] towards the thermal conditions under RCT, NCT and ACT loaded with a 30B cylinder containing any mass of UF 6 between empty and full (0 kg and 2277 kg).
The evaluation of the design change to remove the housing in chapter 11 shows that the removal leads to only a minor increase in the maximum temperature at the valve with a large safety margin still remaining. Due to the very efficient microporous insulation layer Microtherm, the housing does not fulfil its design safety function anymore is no longer needed.
99 / 100 0023-BSH-2016-002-Appendix-2.2.2.3-Rev5
References
[ADR] European Agreement for International Transports of Dangerous Goods by Road (ADR) of 30 September 1957 with Annexes A and B as applicable from 1 January 2023
[ANSI N14.1] Uranium Hexafluoride - Packaging for Transport, ANSI N14.1-2023
[ANSYS] ANSYS, Release 19.0, Help System, ANSYS, Inc
[ASME BPVC] ASME Boiler and Pressure Vessel Code, 2017 Edition
[DeWITT] R. DeWitt, Uranium Hexafluoride: A Survey of the Physico-Chemical Properties, Goodyear Atomic Corporation, Portsmouth, Ohio, GAT-280 Chemistry General, 1960
[DIN EN 10028-3] DIN EN 10028-3:2017-10: Flat products made of steels for pressure pur-poses - Part 3: Weldable fine grain steels, normalized
[FRANSSEN] Jean-Marc Franssen and Paulo Vila Real: Fire Design of Steel Structures, Annex A: Thermal data for carbon steel and stainless steel sections, Euro-pean Convention for Constructional Steelwork, 2012
[HARRIS SN50] Technical Information Sheet, 50/50 Tin Lead Solder, The Harris Products Group, 2017
[ISO 7195] ISO 7195:2020-11: Nuclear energy - Packagings for the transport of ura-nium hexafluoride (UF6)
[ROLOFF] Wittel, H., Jannasch, D., Voiek, J., Spura, C.; Roloff/Matek Maschinene-lemente: Normung, Berechnung, Gestaltung; 23. Auflage; Springer Vieweg Verlag, 2017
[SSG-26] Advisory Material for the IAEA Regulations for the Safe Transport of Radi-oactive Material (2018 Edition), Specific Safety Guide No. SSG-26 Rev. 1, IAEA, Vienna
[SSR-6] Regulations for the Safe Transport of Radioactive Material (2018 Edition),
SSR-6 Rev. 1, IAEA, Vienna
[USEC 651] The UF6 Manual - Good Handling Practices for Uranium Hexafluoride, Rev. 9, USEC, 2006
[VDI Heat] VDI Heat Atlas, 10th Revision, Springer Verlag, 2006
100 / 100