ML092290631
| ML092290631 | |
| Person / Time | |
|---|---|
| Site: | Washington State University |
| Issue date: | 04/20/2009 |
| From: | Wall D Washington State Univ |
| To: | Janice Nguyen Division of Policy and Rulemaking |
| Wertz G, NRR/DPR/PRTA, 434-326-1086 | |
| References | |
| TAC ME1589 | |
| Download: ML092290631 (35) | |
Text
STARTUP REPORT FOR THE WASHINGTON STATE UNIVERSITY NUCLEAR RADIATION CENTER TRIGA REACTOR APRIL 20, 2009 LICENSE NO. R-076 DOCKET NO.50-027 REDACTED VERSION*
SECURITY-RELATED INFORMATION REMOVED
- REDACTED TEXT AND FIGURES BLACKED OUR OR DENOTED BY BRACKETS
WASHINGTON STATE UNIVERSITY John Nguyen U.S. Nuclear Regulatory Commission Mailstop 012 DOO 11555 Rockville Pike Rockville, MD 20852-2738 Washington State University recently completed converting the WSU TRIGA reactor from operating on a mixed BEU/LEU core to a core fueled entirely by low enriched uranium.
WSU is required to submit a Reactor Startup Report within six months of the return of the reactor to normal operation.
The reactor was returned to steady-state operation in October 2008.
Since that time the reactor has been operated at steady-state in order to build up 14QSmto a relatively constant level before conducting pulse testing of the reactor.
The pulse testing was performed in March, 2009 and the results were used to calculate pulsing limits. After pulse testing was completed it was determined that the reactor could be returned to normal operation, i.e. both steady-state and pulsing operations, as of April 13,2009.
Donald Wall, Ph.D.
Director Nuclear Radiation Center Washington State University P.O. Box641300, Pullman, WA 99164-1300 509-335-8641
- Fax: 509-335-4433
- www.w5u.edu/nrc
Nuclear Radiation Center Washington State University Pullman, WA 99164
2.1 Overview 4
2.2 Core Loading 4
3.1 Overview 8
3.2 Results 8
4.1 Overview 9
4.2 Results 9
5.1 Overview 13 5.2 Results 13 6.1 Overview 14 6.2 Pre-Operational Power Calibrations: 25%, 75%, and 100% Power 16 6.3 Initial 100% Power Calibration for Operational Core 35A 18 6.4 Post-Conversion Calibrations 21 7.
THERMAL FLUX DISTRIBUTIONS 22 8.
REACTOR PHYSICS MEASUREMENTS 23 9.
PRIMARY COOLANT MEASUREMENTS 24 10.
PULSE MEASUREMENTS 25 10.1 Overview 25 10.2 Calculation of Pulsing Limit..
25 10.3 Conclusions 29
The Washington State University Nuclear Radiation Center (WSUNRC) converted its 1 MW TRIGA open pool research reactor from an HEUILEU mixed core to a low enriched uranium (LEU) core.
As part of the Global Threat Reduction Intuitive (GTRI), the HEU fuel that partially comprised the WSUNRC 1 MW reactor Core 34A was replaced with 30120 LEU fuel. The new core is designated as Core 35A, an all LEU core composed of both new 30/20 fuel and the 8.5/20 fuel that comprised the LEU portion of the previous 34-A mixed core.
Core 34A was shutdown for the last time on Friday, September 19,2008.
All fuel was removed from the grid plate.
On September 29, 2008, the U.S. Nuclear Regulatory Commission issued the order to convert, and the fIrst 30/20 LEU cluster went on the grid plate at 1553 hours0.018 days <br />0.431 hours <br />0.00257 weeks <br />5.909165e-4 months <br />, thereby completing the U.S. Department of Energy conversion milestone for the WSU Reactor (WSUR).
Subsequent loading of LEU Core 35A commenced shortly thereafter, reaching criticality on October 7, 2008 at 1526 hrs. The LEU core was declared steady-state operational on October 17, 2008 after extensive testing and measurements prescribed in the Conversion Safety Analysis Report (2007 CSAR).
Pulse testing on Core 35A was completed on March 6, 2009 utilizing the procedures outlined the 2007 CSAR.
A time line ofthe events is provided in Table 1.1.
ummary rme me 0 conversIOn re a e even s.
Date Event 9/19/08 HEUILEU Core 34A shutdown 9/23/08 Core 34A fuel and reflector unload complete Graphite reflectors loaded into core 9/29/08 Neutron source installed USDOE conversion milestone complete (1553 hrs) 10/7/08 LEU Core 35A goes critical (1526 hrs) 10/8/08 LEU Core 35A fuel loading complete 10/10/08 Graphite/fuel optimization complete 10/10/08 Final SDM and core excess complete 10/15/08 125% power scram test 10/17/08 Final power calibrations complete Core 35A Steady-State Operational 3/6/09 Pulse testing complete 4/15/09 Core 35A Pulse Operational
Comparison table for Cores 34A and 35A with measured and calculated values.
Core 35A 35A 34A 34A Number (BOL)
(calculated) mOL)
(calculated)
=
30/20; 8.5/20; 30/20; 8.5/20; 8.5/70; 8.5/70; Q
Core Type 8.5/20; mixed 8.5/20; mixed mixed LEU mixed LEU
~-
HEUILEU HEUILEU
'"' =
Q e U '"'
Critical oS Mass
=
(2 U-235)
($0.91)
($0.98)
($1.76)
($1.70)
Core
$7.44
$6.37
$6.42
$6.65
~
Excess
'"'~-~
Blade 1
$1.44
$1.34
$1.60
$1.32 e='"'
Blade 2
$3.71
$2.99
$3.50
$2.89
=
~.e-Rod 3
$3.20
$3.19
$3.08
$3.22
.*.*~.*.*-
Blade 4
$3.97
$3.02
$3.81
$2.86 CJ=~=:
Blade 5
$0.16
$0.43
$0.16
$0.40 Ap
$2.38
$2.52
$2.20
$1.45 (1 MW)
~.J:l D8 N/Aa 4.95 3.22 4.92 (5.18) bIJ~~
= =M
'"'- e D9 N/Aa 1.44 N/Aa 1.45 (2.10)
~
~
CJ
~----
-< = =
_ e~
E8 N/Aa 4.35 3.33 4.35 (4.39)
= '"' Q
.*.. ~-
~.=:~
-<E--'-"
E9 N/Aa 1.29 N/Aa 1.30 (1.75) f}eff 0.0075 0.0075 0.0070 0.0076
'"' ~
void coeff.
Q CJ (per 1%
N/Aa
-0.135% L\\k/k N/Aa
-0.080% L\\k/k CJ
~
= >.
H2O)
~.=:
=:~
PNTC 0.6 x 10-4to 0.54 x 10-4to
(-L\\klk-oC, N/Aa 1.27 x 10-4 N/Aa 1.51 x 10-4 23-1000°C)
(a) No data are available at this time.
(b) Values given for the axial average thermal neutron flux calculations are from MCNP5 and in parentheses, calculated from our WSUR in house Exterminator II code.
2.1 Overview The initial core loading process was carried out in accordance with the 2007 CSAR and our conversion SOP (CSOP).
The procedure comprised a stepwise plan to refuel the WSUR with LEU fuel in core positions (CPs) where the FLIP fuel was located in Core 34A.
2.2 Core Loading New graphite reflector elements were first installed in the same positions from which the old graphite reflector elements were removed CP F4 and F5, where used graphite reflectors were placed.
The IFE cluster was installed into CP C4 and the IFE was inserted into C4NW and connected to the temperature indication units in the console.
The pulse rod fuel assembly was installed into CP D5, followed by the new pulse rod in D5NW.
The pulse rod system was completed, tested, and declared operational.
Calculations showed that criticality would occur with 11 partially burned 8.5/20 LEU clusters (44 fuel elements) and 10 new 30/20 LEU clusters (39 fuel elements) for a total of21 LEU clusters (83 fuel elements) as detailed in the MCNP diagram, Figure 2.1 and Figure 2.2.
0.50 0.45 0.40 0.35 0.30 0.25
~.,..
0.20 0.15 0.10 0.05 0.00 30 40 50 60 70 80 9
-0.05
"# Fllel Elements
u cntlca mu tlpllcatlOn ata or ore Elements Count rate 11M Predicted total elements added (cpm) needed for criticality 51 14.8 1
55 34.8 0.4253 58 59 57.6 0.2569 65.1 67 267.2 0.0554 69.2 71 613.2 0.0241 74.2 75 4377.2 0.0034 75.5
0.0 50
3.1 Overview The critical mass is calculated from the number of elements needed to achieve criticality. In the case ofWSUR Core 35A, this number is 79 elements, composed of partially burned 8.5/20 and fresh 30/20 fuel. The amount ofuranium-235 comprising those specific elements is delineated as the critical mass. The amount of U-235 remaining in the 8.5/20 section of the core (44 elements total) was computed from the last SNM core inventory completed for the HEUILEU core, for the period ending September 30,2008.
3.2 Results The inventory in the critical assembly for the 8.5/20 section of the core is grams ofU-235.
The rest ofthe core is fresh 30/20 fuel, and the amount contained in this critical assembly is g ofU-235.
Therefore, the total critical mass for WSUNRC Core 35A is g ofU-235.
The calculations in the 2007 CSAR suggested that 83 elements were required to achieve criticality, accounting for g U-235. This is a difference of only one 4-rod 30/20 fuel cluster.
4.1 Overview This section outlines the characteristics ofthe core excess as calculated by MCNP5 and actual measured values for Cores 34A and 35A. In addition, the calculated optimal configuration for Core 35A was found to exceed the maximum allowable excess reactivity value of 5.6% ~klk once assembled, because the new ~effvalueof 0.0075 reduced the limit from $8.00 for Core 34A to $7.46 for Core 35A.
To resolve the situation, seven standard LEU fuel elements with a higher burn-up and lower U-235 content were moved into higher flux positions to reduce the overall reactivity of the core. Seven new reflectors were replaced by old reflectors, in postulating that they would be less efficient, and standard fuel was moved such that less reactive clusters were in higher power factor core positions.
Once a satisfactory core excess was obtained for the fueled core itself, four experimental rotator tubes, and experimental facilities were installed. Core 35A was then operated to accumulate enough burn up to build in samarium thereby allowing the flux distribution and poison levels to reach steady-state concentration. The amount oftime required is 371 MWH on the new fuel.
4.2 Results Core 34A was originally calculated to have an excess reactivity of $6.65 by MCNP5. When this value was measured just prior to disassembly for the conversion, it was found to be $6.31. This was an acceptable value, and was within the previous limit of
$8.00.
Core 35A was calculated by MCNP5 to have a core excess of$6.35.
This value was experimentally determined to be $7.69 after Core 35A achieved criticality. The large discrepancy between the actual and calculated values is due to the discrepancy in the MCNP5 calculated and experimentally measured values for the control element reactivity worths.
Immediately following the core reload and once criticality was achieved with a completely fueled core on October 8, 2008, a shutdown margin was performed and the core excess was found to be $7.69. Calibrations were then performed on all four control blades and one control rod to ensure a precise core excess was calculated. The following day on October 9,2008 a shutdown margin was performed with the freshly calibrated control elements and the core excess was calculated to be $7.65, still in excess of the acceptable value.
The following day, October 10,2008 a combination of solutions was determined by the Facility Director, Reactor Supervisor, and the GA consultant assigned to the project. Old reflectors, which were thought to be less efficient, were placed into core positions B4, B5, B7, B8, G5, and G6 and a shutdown margin was performed. The core excess value was found to be higher at $7.75, and the new reflectors were replaced in these positions.
Standard elements S21 were moved from B1 to C2, and S03 from C2 to B1. The core excess was determined to be $7.60. Standard elements S26 were moved from Fl to E2, and S06 from E2 to FI. The core excess was determined to be $7.54. Standard elements S22 were moved from F3 to E2, and S27 from E2 to F3, S26 from E2 to F2.
The core excess was determined to be $7.44 and which is below the upper limit in the technical specification of $7.46.
With the core excess within acceptable limits, a power calibration at 750 kW, 1 MW, and two initial approaches to 1 MW were performed.
The fission products produced by these actions further reduced the core excess to $7.35 as determined on October 20, 2008.
With the core now designated operational the four experimental rotator tubes were installed in positions E8, D8, E9, D9, with the cadmium lined tube in position E9. The final core excess of the operable core 35A was determined to be $7.12 on October 20, 2008.
Core 35A was then operated at one megawatt on a regular basis to build up samarium in the core, which was estimated to require 371 megawatt hours to reach steady-state within 1% of the final value.
Figure 4.1 traces the core excess from October 10,2008 to February 17,2009 and shows the build-in of fission product poisons, mainly 149Sm, as the WSUNRC procedures for SDM measurements to be with a cold-clean
- core, free of xenon poisoning.
The 149Sm buildup accounts for approximately
$7.44 - $6.86 =
$0.58 (kex(BOL)- kex(current>>)
of negative reactivity contribution from neutron poisons in the core.
It is important to note that the core as of February 17, 2009 has 367 MWH of bum up, and as such the equilibrium between samarium production and removal has not been
- achieved, seen in Figure 4.1.
Core 35A Startup Report
$7.50
$7.40
$7.30
.- $7.20 G'7'-"~
,;/
$7.10
$7.00 *
$6.90
$6.80 0
$1.20 G'7
~
~
$1.10 ~
$0.80 500 Core 35A burnup (MW-Hours)
Core excess C., - --)
and shutdown margin C+,-)
calculations based on experimental data taken after Core 35A was loaded and brought to an acceptable core excess value of $7.44.
Core35ABOL Core34AEOL Measured Calculated Measured Calculated Blade 1
$1.44
$1.34
$1.60
$1.32 Blade 2
$3.71
$2.99
$3.50
$2.89 Rod 3
$3.20
$3.19
$3.08
$3.22 Blade 4
$3.97
$3.02
$3.81
$2.86 Blade 5
$0.16
$0.43
$0.16
$0.40 Total
$12.48
$10.97
$12.15
$10.69
The Washington State University TRIGA reactor is calibrated using a calorimetric process.
Per the WSUNRC SOPs, the reactor was isolated on the west side of the pool by placing a large divider door over the pool divider opening.
A stable, homogenous pool temperature was then obtained utilizing the primary coolant loop and a pool mixer.
The coolant temperature was monitored and plotted in five minute intervals until stable 0.1 °C/h).
The reactor was then brought quickly to one megawatt and maintained at this power level by control element manipulation for forty minutes, while recording the primary coolant temperature at five minute intervals.
After forty minutes, the condition of the reactor was recorded in the reactor log, and the reactor was shut down by manual SCRAM forty five minutes after achieving power.
The coolant temperature continued to be monitored and plotted in five minute intervals.
After a time period sufficient to provide indication of relatively constant rate of pool temperature decrease, the reactor power level was calculated using the plotted data.
The initial and final pool temperatures were extrapolated from the calculated mid-time on the graph of coolant temperature vs. time.
Change in pool temperature was determined and the heating time (startup to SCRAM) was recorded.
From this data the temperature rise per hour was calculated by dividing the average rise in temperature by the heating time.
To obtain the actual power level of the reactor, the temperature rise per hour was divided by the tank constant for the WSUR (5.90 °C/Hr/MW).
The positions ofthe in-
Core 35A in its pre-operational arrangement, without experimental or irradiation facilities installed.
Core 35A Startup Report Operational Core 35A arrangement containing experimental and irradiation facilities installed into or around the grid plate.
26 25 U
°'-'
S" 24 Qi.*..
~
CI 23
=.*..
"C Qi.~- 22
=a QZ 21 20 0
/---
/
/
/
/
/
/
40 60 80 Time (min)
Figure 6.3 Reactor power calibrations for pre-operational Core 35A at 25%
(calculated 24.9%, - - -), 75% (calculated 70.6%, -
-)
and 100%
(calculated 102%, --)
power. Initial temperatures and times are offset to the lowest temperature recorded and the first rise in temperature, respectively.
a e.
ower ca 1 ra Ion a 00 ower on nuc ear InSrumen a Ion.
Control element positions Power channel indications Fuel (in)
(% IMW Temp.
No.1 No.2 No.3 No.4 No.5 crc mc Log-N (OC)
Start:
7.17 6.54 14.99 6.62 13.06 25 24 25 144 Finish:
6.72 6.76 14.99 6.75 12.20 25 25 26 149
Power calibration at 75% of full Control element positions in No.1 No.2 No.3 Start:
8.05 8.05 15.11 Finish:
7.80 8.10 15.11 Calculated Power: 0.706 MW No.4 8.04 8.10 No.5 11.12 9.80 ower on nuclear instrumentation.
Power channel indications
%1 CIC VIC 76 74 74 70 Fuel Temp.
(OC) 273 268 Power calibration at 100% of full Control element positions in No.1 No.2 No.3 Start:
8.26 8.25 15.11 Finish:
8.26 8.25 15.11 Calculated Power: 1.02 MW No.4 8.27 8.25 No.5 11.05 11.49 ower on nuclear instrumentation.
Power channel indications Fuel
% 1 Temp.
CIC UlC (OC) 100
~
3~
100 94 301
Power calibration at 100% of full power on nuclear instrumentation.
Calculated power from measured data shows actual core power at 90.2%
power.
Control element positions Power channel indications Fuel (in)
(% IMW)
Temp.
No.1 No.2 No.3 No.4 No.5 CIC VIC Log-N (OC)
Start:
8.62 8.60 15.11 8.57 11.09 100 98 100 303 Finish:
8.43 8.61 15.11 8.57 9.80 100 98 100 302 Power calibration at 90% of full power for verification on nuclear ft d
d*
InstrumentatIOn a er etector a IJustment.
Control element positions Power channel indications Fuel (in)
(% 1 M'V)
Temp.
No.1 No.2 No.3 No.4 No.5 CIC DIC Log-N (OC)
Start:
8.54 8.54 15.13 8.52 11.69 92 92 90 301/271 Finish:
8.52 8.49 15.12 8.55 9.97 92 92 90 304/276
26 25 U
0'-"e 24
~-
~
= 23 eo:-
"'0~.~- 22 eo:
6**
QZ 21 20 0
40 60 80 Time (min)
Figure 6.4 Reactor power calibrations for pre-operational Core 35A at 100%
(calculated 90.2%, --)
and then again at 90% power (calculated 90.1%,
-)
after detector movement.
Initial temperatures and times are offset to the lowest temperature recorded and the first rise in temperature, respectively.
Control element positions in No.1 No.2 No.3 Start:
8.53 8.52 15.12 Finish:
8.48 8.48 15.12 Calculated Power: 1.002 MW No.4 8.55 8.49 No.5 11.06 12.60 Power channel indications
%1 CIC mc 100 100 100 100 Fuel Temp.
(OC) 314/283 314/284
Preliminary measurements ofthermal neutron flux in irradiation position D8 have been made.
This was done by neutron activation Of63CU, 58Fe, 59CO, and
]97Au. The value was found to be 7.3 x 1012 n/cm2/s.
This is higher than a previously determined value of 4 x 1012 n/cm2/s in the same irradiation position.
- However, it should be pointed out that the uncertainty associated with the Core 34A value is not known.
Flux values for the other irradiation positions in Core 34A are unknown.
Reactor physics measurements have not been made at this time.
The meaSurementS will be completed shortly.
Table 1.2 provides the results for the calculated and measured values that we do have for BOL Core 35A and EOL Core 34A.
The primary coolant bulk pool water was analyzed by gamma spectrometry. A 500 mL water sample was taken after at least four hours of operation at 1 MW. It was then counted on a GeLi gamma detector for 80,000 seconds.
The primary coolant pool water was analyzed on October 29, 2008 and November 24,2008 for sealed source radioactivity. In addition, an analysis was performed on the same samples to look for evidence of fission product release that might occur at levels not picked up by the continuous air monitor (CAM). There was nOindication of a release.
10.1 Overview The variable heat capacity, along with the rod power factors, radial, and axial peaking factors can be used to predict peak temperatures for fuel rods.
Previously, the WSU mixed core utilized standard TRIGA and FLIP fuels that had the same uranium loading (8.5% by weight), and thus both fuel types had very similar heat capacities.
At present, the WSU reactor has fuels with significantly different uranium loadings (8.5%
and 30%) which leads to fuels with heat capacities that discernibly different.
- However, for the purpose of calculating temperature peaking in the hot rod position, and in the central core region, which is populated only by 30/20 fuel, it is sufficient to use the heat capacity value for 30/20 fuel. This will establish temperature peaking values for two reasons: first, power peaking takes place in the 30/20 region, and second, the 30/20 fuel has a smaller heat capacity, and thus will reach higher peak temperatures than the standard TRIGA fuel for the same energy and peaking factors.
10.2 Calculation of Pulsing Limit The heat capacity for 30/20 TRIGA fuel may be determined by the same means used to calculate the heat capacity for 8.5/20 standard TRIGA fuel.
The expression used to determine the heat content for values greater than 25 DC for <5-phaseZrHx for various stoichiometries (i.e. values of x < 1.65) is:
The enthalpy for uranium metal that has been previously used to calculate standard TRIGA heat capacity is given by Using a density of5.610 g/mL for ZrHl6and 19.05 g/mL for the density of uranium metal, the density of30%
U/ZrHl.6 is calculated as follows.
Therefore, the total volume of 1 gram of 30/20 TRIGA fuel is 0.1405 mL, giving a density of7.115 g/mL.
The volumetric heat content is given by the sum of the heat contents for the ZrH1.6 and uranium metal contributions.
For 1 mL of fuel the number of moles of ZrH1.6 is given by 0.05365 mollmL(0.03488T2
+ 33.706T
- 8.64.44 J/mol) =
1.871 x 1O-3T2+ 1.808T - 46.377 J/mL (2.1345 g U/mL fuel) x (6.483 x 10-5r + 0.1087T - 2.758 Jig)
= 1.384 x 10~2 + 0.2320T - 5.887 J/mL fuel
Adding the volumetric heat content for ZrH1.6 and uranium metal gives the volumetric heat content for the fuel:
~16.0 6
~;
14.0
~-~
'"'ti 12.0
'"'~=
~
10.0 1.15 1.35 1.55 Pulse size ($)
Figure 10.1 Energy release versus pulse size for Core 35A.
6.0 0.75
D4NE (25°C)
D4NE (280°C)
C4NW (280°C)
C4NW (280°C)
RPF 2.56 2.47 1.56 1.56 Radial 1.27 1.19 Not available Not available Axial 1.35 1.29 1.454 1.454 RPFxRxA 4.39 3.78 2.70 3
3.51 b (a) The product ofRPF x R x A was determined by using the radial peaking factor for the hot rod in D4NE, i.e. 1.19.
(b) The product ofRPF x R x A was determined by using the radial peaking factor for the core average, i.e. 1.55.
(2.04 J/mL °C x 765 0c) + (4.018 x 10-3 J/mL °C2 x (765 oC)2)/2 =
= 1560.6 +
1175.7 = 2736 J/mL
Using total fuel volume of 343.42 mLirod x 119 rods = 40,867 mL, the peak allowable energy density of 2736 J/mL and a peaking factor of 4.39, one can calculate the peak allowable energy release:
2736J 1mL x 40,867mLjuel = 25.47MJ lxl06 J 1MJx4.39 25.47MJ+7.17MJ
=$2.31 14.14MJ 1$
1.25 1.5 1.75 Pulse size ($)
Figure 10.2*
Energy release versus pulse size comparisons for Core 34A (Ll) and 35A (.).
25
~6 "0
20
~
fI}
=~-~
'"'~
15
'"'~=
~
10 5
0.75 1
The WSUR is the only mixed core comprising 8.5/20 and 30120 fuel in existence.
- Moreover, the calculations done for this conversion were unique because the 8.5/20 fuel was partially burned, while the 30120 fuel was fresh.
Error is to be expected here due to fuel bum up calculations in preparation for predicting BOL Core 35A; however, all ofthe calculations were within reason when compared to the measured values for the new mixed core.
The codes used were benchmarked by comparisons with the previous mixed FLIP/STD Core 34A measured and calculated values.
In all aspects, the conversion went relatively smoothly, with no major setbacks.
Problems did arise with a higher than predicted (and allowed) core excess, which was fixed by rearranging graphite and fuel into different orientations within the same core arrangement.
That is, no 8.5/20 fuel was switched for 30120 fuel or graphite, and vice versa.
With the conversion is now complete, LEU Core 35A is completely steady-state and pulse operational.