Sandia National Laboratories conducted experiments using a Dry Cask Simulator (DCS) to validate computational fluid dynamics (CFD) models for spent nuclear fuel storage cask thermal analyses. The DCS uses a scaled prototype boiling water reactor assembly inside a pressure vessel to simulate internal conditions of dry casks. Thermocouples measure temperature profiles under various decay heat and helium pressure conditions to model aboveground and belowground cask storage. Initial testing of the aboveground configuration showed good agreement between experimental temperature measurements and CFD simulations, within estimated uncertainties.

1. Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia
Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department
of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.
Dry Cask Simulator
Experiments for
CFD Validation
Sam Durbin, Eric Lindgren,
Abdelghani Zigh*, and
Jorge Solis*
* Nuclear Regulatory Commission
SAND2017-4330 C

2. Overview
 Purpose: Validate assumptions in CFD
calculations for spent fuel cask thermal design
analyses
 Used to determine steady-state cladding
temperatures in dry casks
 Needed to evaluate cladding integrity
throughout storage cycle
 Measure temperature profiles for a wide range
of decay power and helium cask pressures
 Mimic conditions for above and belowground
configurations of vertical, dry cask systems with
canisters
 Simplified geometry with well-controlled
boundary conditions
 Provide measure of mass flow rates and
convection heat transfer coefficients
 Use existing prototypic BWR Incoloy-clad test
assembly 2
Underground Storage
Source: ww.holtecinternational.com/productsandservices/
wasteandfuelmanagement/hi-storm/
Aboveground Storage
Source: www.nrc.gov/reading-rm/doc-
collections/fact-sheets/storage-spent-fuel-fs.html
(m)
(m)
Temp. (K)

3. Project Structure
 Boiling Water Reactor Dry Cask Simulator (DCS)
 Partnership between USNRC and DOE
 Equal cost sharing
 NRC staff leads technical review
 Mutual benefits
 Thermal-hydraulic data for validation exercises
 Complimentary data for High-Burnup Cask Demonstration Project
 Includes thermal lance comparisons to peak cladding temperature (PCT)
3

4. Past Validation Efforts
Full Scale
 Full scale, unconsolidated
 Castor-V/21 cast iron/graphite with polyethylene rod shielding
 1986: EPRI NP-4887, PNL-5917
 21 PWRs
 95 Thermocouples (TC’s) total
 Unventilated
 Sub-atmospheric (air and He) and vacuum
 REA 2023 prototype steel-lead-steel cask with glycol water shield
 1986: PNL-5777 Vol. 1
 52 BWRs
 70 TC’s total
 Unventilated
 Sub-atmospheric (air & He) and vacuum
 Full scale, consolidated
 VSC-17 ventilated concrete cask
 1992: EPRI TR-100305, PNL-7839
 17 consolidated PWRs
 98 Thermocouples (TC’s) total
 Ventilated
 Sub-atmospheric (air and He) and vacuum 4

5. Past Validation Efforts (cont.)
Unconsolidated Fuel
5
 Small scale, single assembly
 FTT (irradiated, vertical) and SAHTT (electric, vertical & horizontal)
 1986 PNL-5571
 Single 15x15 PWR
 Thermocouples (TC’s)
– FTT: 187 TC’s total
– SAHTT: 98 TC’s total
 BC: Controlled cask outer wall temperature
 Atmospheric (air & He) and vacuum
 Mitsubishi test assembly (electric, vertical & horizontal)
 1986 IAEA-SM-286/139P
 Single 15x15 PWR
 92 TC’s total, all distributed over 4 levels inside tube bundle
 BC: Controlled outer wall temperature of fuel tube
 Atmospheric (air & He) and vacuum
 Not appropriate for elevated helium pressures or
belowground configurations

6. Current Approach
 Focus on pressurized canister systems
 DCS capable of 24 bar internal pressure @ 400 ◦C
 Current commercial designs up to ~8 bar
 Ventilated designs
 Aboveground configuration (This presentation)
 Belowground configuration
 With crosswind conditions
 Thermocouple (TC) attachment allows better
peak cladding temperature measurement
 0.030” diameter sheath
 Tip in direct contact with cladding
 Provide validation quality data for CFD
 Complimentary to High-Burnup Cask Demo. Project
6

7. DCS Pressure Vessel Hardware
 Scaled components with instrumentation well
 Coated with ultra high temperature paint
7

8. Prototypic Assembly Hardware
 Most common 99 BWR in US
 Prototypic 99 BWR hardware
 Full length, prototypic 99 BWR
components
 Electric heater rods with Incoloy
cladding
 74 fuel rods
 8 of these are partial length
 Partial length rods 2/3 the
length of assembly
 2 water rods
 7 spacers
8
Nose piece and
debris catcher
BWR channel, water tubes
and spacers
Upper tie plate

9. Thermocouple Layout
 97 total TC’s internal to assembly
 10 TC’s mounted to channel box
 7 External wall
 24 in. spacing starting at 24 in. level
 3 Internal wall
 96, 119, and 144 in. levels
9
Radial Array
24” spacing
11 TC’s each level
66 TC’s total (details below)
Axial array A1
6” spacing
20 TCs
Axial array A2
12” spacing – 7 TC’s
Water rods inlet and exit – 4 TC’s
Total of 97 TCs
24”
48”
72”
96”
119”
144”
Internal Thermocouples
a b c d e f g h i
Q
R
S
T
U
V
X
Y
Z
24” & 96” levels 48” & 119” levels 72” & 144” levels
a b c d e f g h i
Q
R
S
T
U
V
X
Y
Z
a b c d e f g h i
Q
R
S
T
U
V
X
Y
Z

10. CYBL Test Facility
 Large stainless steel
containment
 Repurposed from earlier
CYLINDRICAL BOILING Testing
sponsored by DOE
 Excellent general-use
engineered barrier for isolation
of high-energy tests
 3/8 in. stainless steel
 17 ft diam. by 28 ft cylindrical
workspace
 Part of the Nuclear Energy
Work Complex (NEWC)
10

11. Aboveground Configuration
11
Pressure
Boundary
 BWR Dry Cask Simulator (DCS)
system capabilities
 Power: 0.1 – 15 kW
 Pressure vessel: 3E-3 – 24 bar
 Vessel temperatures up to 400 C
 ~200 thermocouples throughout
system
 Test conditions presented here
 Power: 0.5 – 5 kW
 Pressure: 3E-3 – 8 bar
 Air velocity measurements at inlets
 Calculate external mass flow rate

12. Internal Dimensional Analyses
 Internal flow and convection near
prototypic
 Prototypic geometry for fuel and basket
 Downcomer scaling insensitive to wide
range of decay heats
 External cooling flows matched using
elevated decay heat
 Downcomer dimensionless groups
12
Parameter
Aboveground
DCS
Low Power
DCS
High Power
Cask
Power (kW) 0.5 5.0 36.9
ReDown 170 190 250
RaH
* 3.1E+11 5.9E+11 4.6E+11
NuH 200 230 200
Downcomer
“Canister”Channel
Box
“Basket”

13. External Dimensional Analyses
13
External
cooling
flow path
Parameter
Aboveground
DCS
Low Power
DCS
High Power
Cask
Power (kW) 0.5 5.0 36.9
ReEx 3,700 7,100 5,700
RaDH
* 2.7E+08 2.7E+09 2.3E+08
(DH, Cooling / HPV) × RaDH
* 1.1E+07 1.1E+08 4.8E+06
NuDH 16 26 14
 External cooling flows evaluated
against prototypic
 External dimensionless groups
1 in.1 cm

14. Steady State Values vs. Decay Heat
14
 PCT and air flow  as
simulated decay heat 
 Significant increase in
PCT for P = 3E-3 bar
 Due to air in “canister”
instead of helium

15. Transient Data
15
 Power = 2.5 kW
 Internal pressure = 1.0 bar
 Steady state values
 PCT = 570 K
 Q = 673 slpm

16. CFD Modeling
16
 Computational fluid dynamics modeling
 ANSYS Fluent 16.1
 Discrete Ordinates (DO) for radiation heat transfer
 Semi-Implicit Method for Pressure-Linked Equations (SIMPLE)
 Link for momentum and continuity equations
 3-D mesh with symmetric mid-plane
 Fuel represented as porous media
 Internal laminar flow
 External Low-Re k-ε
 Modeling performed consistent with best practices and
best available data representing fuel properties
 NUREG-2152, “CFD Best Practice Guidelines for Dry Cask Applications”
 NUREG-2208, “Validation of CFD Methods Using Prototypic Light
Water Reactor Spent Fuel Assembly Thermal-Hydraulic Data”

17. Steady State Comparisons
17
Press.
(bar) Quantity
Power (kW)
0.5 1 2.5 5
Test CFD Diff.
Error
(%)
Test CFD Diff.
Error
(%)
Test CFD Diff.
Error
(%)
Test CFD Diff.
Error
(%)
1.0
PCT (K) 376 378 2 0.5 434 438 4 0.9 570 569 -1 -0.2 715 717 2 0.3
Q (slpm) 335 326 -9 -2.7 448 449 1 0.2 673 669 -4 -0.6 874 877 3 0.3
4.5
PCT (K) 367 368 1 0.3 426 423 -3 -0.7 545 549 4 0.7 689 698 9 1.3
Q (slpm) 306 293 -13 -4.2 415 396 -19 -4.6 603 601 -2 -0.3 830 826 -4 -0.5
8.0
PCT (K) 359 362 3 0.8 410 408 -2 -0.5 521 523 2 0.4 659 663 4 0.6
Q (slpm) 280 277 -3 -1.1 392 387 -5 -1.3 593 579 -14 -2.4 793 773 -20 -2.5
 Agreement within experimental uncertainty for
majority of results
 Only one instance of difference beyond estimated uncertainty
 Further analysis shown next

18. Graphical Steady State Comparisons
18
 PCT average difference of 2 K
across all conditions
 95% exp. uncertainty
 +/- 1% reading in Kelvin
 (UPCT, max = 7 K)
 Max. observed difference = 9 K
 (5 kW and 4.5 bar)
 Air flow rate average difference of
-8 slpm for all conditions
 95% exp. uncertainty of UQ = 35 slpm
 Max. observed difference = -20 slpm
 (5 kW and 8.0 bar)

19. Summary
19
 Dry cask simulator (DCS) testing complete for
aboveground configuration
 12 data sets available for pressurized canister
conditions
 3 data sets available for sub-atmospheric
 Comparisons with CFD simulations show
favorable agreement
 Within experimental uncertainty for nearly all cases
 Additional steady state comparisons for basket,
“canister”, and “overpack” also show good agreement

20. EXTRA SLIDES
20

21. Custom TC Lance
21
 Compliments the TC lance in
the Cask Demo Project
 Same fabricator (AREVA)
 “Same” materials and
fabrication process
– Closure method for SNL TC
lance significantly different
– Sealed using brazing method
with water-based flux
 TC elevations match BWR
assembly TCs
 Provides direct comparison
between lance TCs and clad TCs
TC Lance

22. Thermocouple (TC) Lance Anomalies
 “Glitches” observed in SNL TC
lance
 Sharp changes in dT/dt
 Coincidentally occurring near
~100 oC?
 Generally recovered by end of
Steady State
 Discussions with vendor
revealed unique closure for SNL
TC lance
 Hypothesis developed that TC
chamber contaminated with
water
 Closure formed by brazing
with water-based flux
22

23. Proposed Solution: Vent TC Lance
23
 Pierce lance collar below
brazed seal
 Introduce vent path for any
trapped water
 Breach created using rotary
tool with grinding wheel
 Performed May 2nd, 2017

24. Well?
 Test conditions repeated
for 2500 W, 1 bar He
 Significant difference in
response
 Success?
 Supports water
contamination
hypothesis
 Good news for Cask Demo
24

25. Belowground Configuration
 Modification to
aboveground ventilation
configuration
 Additional annular flow path
 Currently testing
 Inlet and outlet based on prototypic
configuration
 Scaling analysis completed
 Favorable comparisons
 Modified, channel Rayleigh
number (Ra*)
 Reynolds (Re) number
25

26. Cross-Wind Testing
26

27. Wind Machine Output
27
Velocity
(m/s)

28. CFD Cross-Wind Streamlines
28

29. Effect of Wind Speed on External Air Flow
29