ML091470391

University of Wisconsin, Madison, Response to Request for Additional Information Regarding Heu/Leu Conversion for License Amendment No. 17

ML091470391
Person / Time
Site:	University of Wisconsin
Issue date:	04/10/2009
From:	Agasie R Univ of Wisconsin - Madison
To:	Document Control Desk, Office of Nuclear Reactor Regulation
References
RSC 1004, TAC MD9592
Download: ML091470391 (147)

Text

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO REQUEST FOR ADDITIONAL INFORMATION REGARDING HEU/LEU CONVERSION REDACTED VERSION SECURITY-RELATED INFORMATION REMOVED

.,,REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS:,.

Nuclear Reactor Laboratory UWNR University of Wisconsin-Madison 1513 University Avenue, Room 1215 ME, Madison, WI 53706-1687, Tel: (608) 262-3392, FAX: (608) 262-8590 email: reactor@engr.wisc.edu, http://reactor.engr.wisc.edu April 10, 2009 RSC 1004 United States Nuclear Regulatory Commission ATTN: Document Control Desk Washington, D.C. 20555

Subject:

Docket 5Q-156, License R-74 Response to Request for Additional Information for Amendment No. 17 to Facility License No. R-74 University Of Wisconsin Nuclear Reactor.

TAC No. MD9592

Dear Sirs:

By letter, dated February 26, 2009, the Commission has requested additional information in order to complete the review for the University of Wisconsin Nuclear Reactor's (UWNR) request to amend.facility license number R-74 and technical specifications to facilitate the conversion of the reactor from high enriched uranium (HEU)t1,o low enriched uranium (LEU) in accordance with 10 CFR 50.64(b) (2) (ii).

Enclosed are the responses to the request for additional information. The responses are provided in the same order as the Commission's requests. The format of the enclosure is to restate the request followed by the response. The original request is counter shaded to aid in the separation between request and response.

I certify under penalty of perjury that the foregoing is true and correct.

Executed on:

Robert J Ag se Reactor Director Enclosure 00 pJr

UWNR LEU Conversion Responses to Request for Additional Information Responses to LEU Conversion Request for Additional Information 1.1.,ri

.1.,~~~d...iieon .1 r..3,ouaplcatnsaestt yn.Se*,tio4.2 2* yostatethat.thetra.slpntod guhe tube lrl ar'. s,the r anset.r Idguide"*"

"tube beinr epe.

• as* a "part ofthe conveision? "so,f lease

ýrov justifieation.

Licensee's Response:

Yes. The transient rod guide tube is being replaced as'part of the conversion, This will minimize handling of the previously irradiated guide tube in order to keep staff doses ALARA. The new guide tube is constructed in accordance with GA drawing T4S2!0C152, as is the existing guide tube, and therefore is a direct like-for-liker replacement. See attachment 1.

Sections 1.11,.,4.2.2,ard *4.2 IiSetion 11 andi.:ypu rpp.iiat-io.* ttet i. at th ~uel'anhb '46ie 1stOrag will.b". gchanged. as::parto.tecn&i6owirn nly ectioni4'2.3 yo statefour additiona reflectors will be inst'I"ed Are., additionalrefl;ectors,.being added as part of the.,convýsion?.-I-f so lepase jrvifdjustification. -

Licensee's Response:

Yes. Four additional reflectors are necessary as part of the conversion. As described+in section 4.5,2 (pages 37-38), a reduction in the number of fuel bundles, from 23 to 21, is necessary to ensure shutdown margin. One of the current approved HEU operational cores uses 12 reflectors (123-R12), soq a proposed LEU core was analyzed with 12 reflectors (J21-R12). However, as shown in figure RAI-2-1 below, a much reduced core lifetime would result when compared to the HEU core. Calculations show that the core lifetime of 150 MW-days (with subsequent recovery out to 1100 MW-days) using 12 reflectors can be increased to 1800 MW-days using 2 additional reflectors for a total of 14 reflectors.

Page 1 of 95

UWNR LEU Conversion Responses to Request for Additional Information UWNR CORE LIFETIME 4.00

-'---HEUj FLIP 1234M1

-- Q"LEU 30120 J23-0 10 3.50 -*-LEU 30/20 J'21A0 2

-- *LEU 30/20 J21-IR14 100*

2.50 2.00 UJ 1.00 0,50 o no 000 0 250 500 750 1,000 1.250 1,500 1,750 2,000 2.250 2,500 2.750 3.000 MWd Figure RAIk2-1, UWNR Core Lifetime

~r Licensee's Response:

Yes. The new graphite refledtors:are constructed in accordance with Idaho National Laboratory drawing 600855. The new reflector design is consistent with the existing reflector design GE drawing 612D489. The GE drawing was confirmed via measurement to match existing reflectors. See attachments 2 and 3.

Licensee's Response:,

Yes, As a MTR conversion type TRIGA reactor, the bottom adapter and top handle cluster hardware (to include locking plates and bolts) used to create the .4-element cluster are being replaced, This will minimize handling of previously irradiated hardware in order to keep staff doses ALARA. The existing and new bottom adapters are constructed in accordance with GA drawing T4S210D104. The existing and new top handles are constructed in accordance with GA drawings T4S210C101 (4 element) and T4S210D111 (3 element), See attachments 4, 5, and 6.

Page 2.of 95

UWNR LEU Conversion Responses.to Request for Additional Information Licensee's Response:

Yes,. Enrichment should be stated in units of weight percent.

O pn~a ]t e~r. e 9t e 'os ,D~ Is j~ t on a Ii!,u ,n 0- t hn at e ra d ia tio h ba s t' s ie't"i l~ ,. ,' . .

Licensee's Response:

There are currently two approved operational HEU cores, 123-R10 and 123-R12. The difference between these cores is the replacement of the irradiation baskets in D3 and D7 with graphite reflectors. Because the 123-R12 operational HEU core already-allows removing the irradiation baskets, this is not a change to an experimental facility related to the LEU conversion. However, this application does not seek to approve the proposed J21-R14 LEU core with irradiation. baskets in D3 and D7. Any future core modifications would be performed under 10 CFR 50.59 and existing, procedures.

BO .E ~Please.discuss .the difference, between ,these ,twov, l~ies "

Licensee's Response:

The value of 1.334 +/- 0.0453 %Ak/k (page 27) was calculated in MCNP by modeling the rod drop methodology of measuring reactivity. The value of 1.467 : 0.105 %Ak/k (page 32) was calculated using MCNP and curve fitting by modeling the rising period rod bump methodology of measuring reactivity. The difference between these two methodologies accounts for the importance of the flux shape on the measurement of the worth of a control element., These values are identical within their respective.

uncertainties and agree well with the measured value of 1.374 %Ak/k (page 32).

Page 3 of 95

UWNR LEU Conversion Responses to Request for Additional Information ection '47*7Are, there any other measurements that have been perormed on the

-- '- o wicould4

r be usedto help benchmark'the MCNP'modei Withthe prse Licensee's Response:

The only data available from the first all-FLIP HEU core is in the core loading report

.from January 1980 consisting of differential and integral control element worth curves and axial plots of detector response in a number of fuel bundle locations. While comparisons to this data are possible, they do not result in a quantitative indication of computational bias. Such computational bias is routinely established by comparing simulated results of known critical configurations. The discrepancy between the simulated eigenvalue and the experimental eigenvalue (kexp = 1 by definition) is used to establish the bias. Comparison of simulated axial distributiorns, whether control element worth or detector response, does not provide the same quantitative basis, Discrepancies between such results can vary axially and do not :indicate a specific bias in the eigenvalue.. Furthermore, the control element worth curves are based on a manual fit to a small number of measurements and While a record exists of the curve generated by that manual fit,' the measurements themselves are not recorded.

Finally, the axial detector response data is in a form that does not give a clear indication of how it was measured and what simulation technique would be most appropriate for comparison.

o, 4..4,5..5, 45. 64.5.7 and 4.5,8,..the units fordifferential.

[rthriar%,hi ShoulLdthekunits be"r ,/k/inf 0

Licensee's Response:

Yes. The units should be [%Ak/k / in]. Note that within section 4.5.1, these units appear in Figures 4.5.4, 4.5.5, 4.5.6, 4.5.7, 4.5.8, and Table 4.5,3 (pages 30-32).

9 Sectiron4.54../452:.Iln Tables 4.5.6and,4.5.13, the.units, forthe.void; cQefficients are

... .[givenaast[Ak/ki! %void].Is that correct o'rshouldtfthenit"be. [ý Ak/k ivoidY Licensee's Response:

The reported units of [Ak/k I %void] are correct.

0. -Section .,-1 42. In TYablesi4i..6 nd 45.1, void othecoefficient are st v o erature cocent are, state. .,a s,:ipd-vp.

Licensee's Response:

The coolant temperature coefficients are positive as reported on. pages 35 and 45.

Although they are calculated as positive with MCNP, the values are small and comparable to values reported in the HEU SAR. Previous attempts to experimentally measure the value have been difficult because raising the coolant temperature will also raise the fuel temperature, and the negative fuel temperature coefficient is much larger than the calculated coolant temperature coefficient.

Page 4 of 95

UWNR LEU Conversion Responses to Request for Additional Information

~~tatii~~f~/YJ3 tci'6I bee- rtheitI 1'.,k Licensee's Response:

The reported units of [Ak/k / K] are correct.

Licensee's Response:

Yes. The units should be [%Ak~k / K]..

Licensee's Response:

The shutdown margin for the.LEU core was calculated to be 0.294 %Ak/k with the maximum allowable experiment installed, control blade 3 and the regulating blade fully withdrawn. This satisfies the requirements of Technical Specification 3.1. This calculation is based on the shutdown margin of 0.994 %Ak/k and blade as reported 3 and on page the regulating 38 fully blade of thewithdrawn.

analysis report The with no experiments Technical Specification shutdown margin was calculated by subtracting 0.7 %Ak/k, which is the maximum allowable reactivity of a non-secured experiment, to arrive at 0.294 %Ak/ak. e Page 5 of 95

UWNR LEU Conversion Responses to Request for Additional Information Se~tior~i,1.7 1, error 3.had a fundamental ona!Lborory d in the pointksI 1

inde used the alculation es the versiony f the c6du useduins the scond hng Netio,411-the RELA5 Licnse' MOD33cdusddenohav'ertefxsipemner~h Repose Nonl the seLon dcoesnoth frgondmneNatioal kine rwtic otquatio.

an Lrroratoin 208 the h decide INL Ws kia sode, ithefist In hodrectoIa.rrnos de I ahoNX ion oRAtory h reported that errors had been found and crrected in the point kinetics routine of ReLAP5-3D, which uses thersame point kinetics routine as RELAP2/MoD3r3. Two changes in the point kinetics routine were made. The first change was related to an index used in the calculation of the delayed neutron precursors. The second change related to logic that determined when to apply a quasi-steady form Of the point reactor kinetics equation. In 2008 INL decided that the first correction was erroneous: the index was correctly calculated in the original code. Now INL recommends applying only the second change.

Argonne National Laboratory has the source code for the UNIX version of RELAP5-3D/

version2.3.7t and has compiled the original version of the code with no point kinetics correction, as well as versions with the 2007 corrections and the 2008 corrections. To investigate the impact of these corrections to the code, calculation Inswere made for a

$2.0 step reactivity insertion transient in the University of Wisconsin reactor using all three code versions, Table RAI-14-1 shows powers calculated by these three versions at times near the power peak. The uncorrected version and the latest version give results that are identical to 6 significant figures. The 2007 version gives results that differ from the others by about 1%. Thus, for pulse calculations of interest for the University of Wisconsin reactor the differences between the uncorrected results and the results from the latest version are non-existent or negligible. Even the differences between the 2007 version and the latest version are minor.

Table RAI-14-1, Calculated Powers for a $2.0 Step Reactivity Insertion in the Universt fWisconsin Reactor T~ransient time, s 0.035 0.040 Reactor power, GW, for 1.00128 2.57338 the 200u code corrections Reactor power, GW, for 0,990375 2.59909 the 2007 coide corrections ________

Reactor power, GW, for 1.00128 2.57338--

the uncorrected code, _________ ________

Page 6 of 95

UWNR LEU Conversion Responses to Request for Additional Information

,hot nedq~ im rna Licensee's Response:

Yes, the maximum allowable bulk coolant temperature is administratively controlled by an automatic scram from the reactor protection system if the temperature reaches 130 0 F. However, adding a Technical Specification for pool water temperature is consistent with Table 1 of Technical Specification 3.3.3. The proposed change is included in response to question 56.

• rid outl~t rie~ssielIoss. oefficients*.

Licensee's Response:

A sensitivity study on the impact of the maximum fuel temperature, exit bulk temperature, flow rate,,and power to CHF and MDNBR as a function of the inlet and outlet pressure loss coefficients at 1.5 MW for HEU-BOL has been provided' in the tables below. The inlet and outlet pressure loss coefficients were independently altered by +/-20%.

Table RAI-16-1, Altering the Lower Pressure Loss Coefficient Only Adjustment -20% Nominal 201%

Lower Pressure Loss .0 .424 1.616 2,020 2.424 Coefficient Max Temperature ('C) 642.03 642.03 642.03 Max Clad Temlp (T) 140.90 140.90 14090 Exit Bulk Temp (9C) 99.66 100.44 101.12 Mass Flow Rate (kgis) 0.13904 0.13665 0.13462

% difference 1.75% 0.00% -1.49%

Power to CHF using 52.770 U.376" 52.036 Groenveld 2006 (kW) 5 . .036 Power to CHF 34.299 33.993 33,730 Bernath (kW)

MDNBR Groenveld 2006 1.998 1 1.983 1.971

% difference 0.75% 0.00% -0.65%

MDNBR Bernath

.% difference 1.299 1.287 1.277 090 0,90% .. 0.00% -0.77%

Page 7 of 95

UWNR LEV Conversion Responses to Request for Additional Information Table RA16-2, Altering the Upper Pressure Loss Coefficient Only Adjustment I-20% Nominal 20%

Upper Pressure Loss 1.104 1.380 1656 Coefficient _,0_____,5 Max Temperature ('C) 642.02 642.03 642.03, Max Clad Temp ('C) 140.90. 140 90 140.91 Exit Bulk Temp (9C) 99.90 100-44 100.94 Mass Flow Rate (kg/s) .0,13829 . 0.13665 0.13517

% difference -1.20% 0.00% 1.08%

Power to CHF usinf

• 52,647 52.376 52.129 Groenveld 2006 (kW)

Power Pernto to CHF CHF 34.204 33.993 33.802 Bernath(W)_____

MDNBR Groenveld 2006 1,994 1.983 .974

% difference . 0.52% -0,00% -0.47%

MDNBR Bernath . 1.295 .1.287 1.280

% difference 0,62% 0.00% -0.56%

Table RAI-16-3, Altering the Both the Upper and Lower Pressure Loss Coefficients Adjustment -20% Nominal 20%

Upper Pressure Loss 1.104 1.380 1.656 Coefficient Lower Pressure Loss 1.616, 2.020 2.424 Coefficient Max Temperature (9C) 642.02 642.03 642.03 Max Clad Temp (C) 140.89 140.90 140.91 Exit Bulk Temp ("C) 99.04 100.44 101.56 Mass Flow Rate (ka/s) 0.14099 I. 0.13665 0.13336

% difference -3.18% 0.00% 2.41%

Pow0,,er to CHF using . 5 5 Groenveld 2006 (kW) 53.086 52.376 51.822 Power to CHF 34.545 33.993 33.565 Bernath (kW) 34.545 33.93_ 3.56 MDNBR Groenveld 2006 2,010 1,983 1,962

% difference 1.35% 0.00% 06%

MDNBR Bernath 1.308 1.287 1.271

% difference 1.63% 0.00% -1.26%

As seen in these tables, altering the lower, or upper pressure loss coefficient does not have a significant impact on either the temperature, mass flow rate, or the MDNBR.

This is consistent with previous analyses by General Atomics, "TRIGA Reactor Thermal-Hydraulics Study STAT-RELAP5 Comparison," TRD 070.01006.04 (April 2008).

Page 8 of 95

UWNR LEU Conversion Responses to Request for Additional' Information Licensee's Response:

During startup testing with the all TRIGA-FLIP HEU core, the fuel temperatures in one quarter of the core were measured as reported in the HEU 2000 license renewal SAR, page 4-45, These measurements were done in support of the reload and startup testing. Once startup testing was complete, the operational core was chosen to have IFEs located in D4 SW and E3 NE. In the analysis report it was chosen to compare the model with a pin having a similar peaking factor which was not the case for E3 NE.

Therefore, E4 SE was chosen because both the model and the measured peaking factors were 1.10 as described on page 73.

However, data for E3 NE is available and the measured vs. predicted curve for the instrumented fuel. element at E3, NE is shown in the figure below. The bottom thermocouple measurement was not reported in the reload report due to the thermocouple burning out. duringstartup testing, E3NE Peak Radial Temperature Distribution with 0.1O'mil Gap Radial Distance from Fuel Centerline [in]

0.0 0,1 0.2 0.3 0.4 0.S 0.6 0,7 350. ..............................

66*2 612 300 562 U.

x 250 U a' 462 U

CL 20 412 5.

E ai I Terriperature Distribution U

I-150 . .Centr 212 X Top 100 ....................

0 0,2 0.A Q6 08 1 1.2 1.4 1.6 1.8 Radial Distance from Fuel Centerline,[cm]

Figure RAI-17-!, E3 NE Peak Radial Temperature Distribution Page 9 of 95

UWNR LEU Conversion Responses to Request for Additional Information Licensee's Response:

There are measured instrumented fuel element temperatures at D4 SW as seen in the figure below, D4SW Peak Radial Temperature Distribution with 0.10 miu Gap Radial Distance from Fuel Centerline (in]

().0 0.1A 0.2 0.3 0.4 0.M 0,6 0.7 J, 912 450 .........

400 ,. 82 712

.300 .... ......................

o Radial Temperature Distribution .

~~~~~--.. -. .. ..... .-...

a Center i 412 200 A Bottom X Top 100~ ..--................ ........... 212 0 0.2 0:4 0.6 0.8 1 1.2 14 1.6 1,8 RadialIDistance from Fuel Centerline [cm]

Figure RAI-18-1, D4 SW Peak Radial Temperature Distribution Licensee's Response:

Figure 4.7.11. on page 77 shows the radial temperature profile where the highest axial power peaking factor is, 5.5 inches (13.97 cm) from the bottom of the active fuel.

Page 10 of 95

UWNR L,*U Conversion Responses to Request for Additional Information Licensee's Response:

An error was discovered with the LEU-BOL axial power shape from MCNP5 that was input into the RELAP5 LEU-BOL models. The following figure shows

'Y1.3}','-*" - - - - ----- ---- LE ~ the comparison between the original t axial E0 power shape and the revised MCNP5 axial power shape.

f' - y fl -rneý!.2 0 rig naI Axial Power Pr~ofile Comparison' (LEU-BOL)

Lc"e' ReviRed

• 0 6i 1.4 ~ ~ ~ ~ ~ ~. . ......... ......

0.4............................................ ,'

o 5 10 15 20 25 0 40 Axial Height from Bottom of Active Fuel [cml Fiur RA-1AilPwrSae oparison Between Original and Revised (LEU-BOL)

As is evident by Figure RAi-201, the peak axial power changed from 1.368 in the origina analysis to 1.4032 in the new MCNP5 calculations for LEUsBOL. The pin power peaking factor did not change. This changed the maximum fuel temperature from 662.83°C to 673.86°C at 1.5 MW which is higher than the LEU-MOL maximum fuel temperature at 1 .5 MW of 665.06°C. Therefore LEU-BOL has the highest maximum fuel temperature with the highest rod power.

The new LEU-BOL steady state analysis is presented in the table below:

Page, 11.of 95

UWNR LEU Conversion Responses to Request for Additional Information Table RAI-20-1, T/H Compahison between Original and Revised LEU-BOL results

. . ... .. . . .... .. .C ore . ... Wt e i e . .. . .. . .

Pa eower LEU Conversion With Revised Parameter Power SAR R[MW PowerI Radial Axial Shape  % Difference Rod Power in DSSW 1.5 ý29.041 29,041 0.00%

1.3 25.169 25.169 0.00%

[kW] 1.0 19.361 19.361 0,00%

1.5 0,114878 0.14861 1,64%0/

Mass Flow Rate [kg/s] 1.3 0.13143 0.13105 1.61%

1.0 0.10535 0.10503 1_,54%

Maximum Fuel 1.5 662.83 673.86 0.30%

Centerline 1.3 594.40 604,10 0.28%

Temperature [0C] 1.0 490.15 497.81! 0.26%

.. .. m... . 1,5ý 141.60 142.02 0.30%

Maximum Outside 1,3 139.60 139,99 0.28%

Clad Temperature [°C] 1,0 136.30 136.66 0.26%

Exit O . Clad 1.5 127.47 127.78 0.24%

Exit Outer Clad 1,3 127.14 127.06 -0.06%

Temperature [°C] 1.0 125.09 125.02 !0,06%

Exit Bulk Coolant 1.5 101.32 100.95 40.37%

E1.3 100.04 100.17 0,13%

Temperature [*C] 1.0 98.23 98.37 0.14%

1.5 53.465 . 53.112 -0.66%

Critical Rod Power 1.3 .52.733 51.453 -2.49%

Groeneveld 2006 1,0 51.884 49.891 -3,99%

1.5 35.716 35.631 -0.24%

Critical Rod Power 1.3 33.488 33.403 0.25%

Bernath 1.0 29.437 .29.599 0.55%

Power to Reach CHF .Groeneveld 52.786 53.112 0.61 %

2006 at .631...1...  %

flowlastrate non-oscillatory Bernath 35,164 35.63t 1.31%

1.5 1.818 1.829 0.59%

MDNBR - Groeneveld 1.3 2.095 2.044 -2.48%

2006 1.0 2.680 2.577 -4,00%

1,5 1.211 1..227 1.30%

MDNBR - Bernath 1.3 .1.331 1.327 -0429%

1'0 1.520 1.529 0,68%

Where % Difference is defined as:

%Difference revised - origindl 100%

revised As can be seen in the table above, the changes in the axial flux profile lead to insignificant differences.

Page 12 of 95

UWNR LEU Conversion Responses to Request for Additional Information In regards to the MDNBRs shown in Tables 4.7.13 (page 114) and 4.7.16 (page 121),)

the overall MDNBR trends are identical. As the power increases, the MDNBR decreases. For the old LEU-BOL analysis, the Groeneveld 2006 correlation has a calculated MDNBR of 2.680 at 1.0 MW and a MDNBR of 1.818 at 1.5 MW. For LEU-MOL, the Groeneveld 2006 correlation has a calculated MDNBR of 2.678 at 1.0 MW and a MDNBR of 1.809 at 1.5 MW. A similar trend for the Bernath MDNBR also results as can be seen from the table, However, it can also be seen that the MDNBR for LEU-MOL is more limiting in 2 out of 6 cases in the table below, Table RAI-20-2, MDNBR Comparison between LEU-BOL and LEU-MOL JO

~~~%LEU-BOL6AU MDNBR Power LEU-BOL LEU-BOL Revised Correlation [MW] Conversion Revised LEU-MOL higher SRLEU-MOL than 1.5 1,818 1.829 1.829* 0.00%

Groeneveld 2oned 1,3 2.095 2.044 1.982 3.03%

1.0 2.680 2,577 2.678 -3.92%

1.5 1.211 1.227 1,240* -1.06%

Bernath 1.3 1.331 1.327 1.339 -0.90%

1.0 1,520 1.529 1.527 0.13%

• Pseudo-transient calculated stable flow rate at 1.5 MW, thus the MDNBR no longer is being calculated using the critical rod power from a lower power level that calculated a stable flow.

The percentage of LEU-BOL higher than LEU-MOL column shown is .ealculated as:

(LEU-BOL - LEU-MOL) / LEU-BOL

• 100%

In all cases, LEU-MOL has the MDNBR located at 19.05 cm above the bottom of the active fuel. Whereas, LEU-BOL the Groeneveld 2006 MDNBR is located at 16.51 cm above the active fuel and the Bernath MDNBR is located at 21.59 cm above the active fuel. Since the Bernath case at 1.0 MW giVes essentially the same result, the only evident discrepancy appears for Groenveld 2006 at 1.3 MW.

The important thermal hydraulic parameters used at 16.51 cm, 19.05 cm and 21.59 cm can be seen in the table below. Note that the critical heat flux ratio (CHFR) is defined as:

QHFR = q"(CHF) / q'(Iocal)

Page 13 of 95

UWNR LEU Conversion Responses to Request for Additional Information Table RAI -2M-3, Important Thermal Hydraulic Parameters for LEU-BOL vs. LEU-MOL F- Percernt-LEU-BOL higher LEU-BOL than Parameter (at 1.3 MW) Revised LEU-MOL LEU-MOL Mass Flow Rate [kg/s] 0.13105 0.13049 0.43%

Local heat fluxat 16.51 cm [W] 8..816.957 793.381 2.89%

Local heat.flux at 19.05 cm [W] ___ 777.822 782.666 -0.62%

Local heat flux at 21.59 cm [W] 733.290 726.828 0.88%

Local quality at 16.51 cm -0.06136 -0.06230 -1.53%

Local quality at 19.05 cm- -0.05372 -0,05458 -1.60%

Local quality at 21.59 cm ý0.04652 -0.04742 -1,93%

Local fluid velocity at 16.51 cm [mis] 0.283613 0,282283 0.47%

0.284573 0.283249 0.47%

Local fluid velocity at 19.05 cm [m/s]

Local fluid velocity at 21.59 cm [m/s] 0,285479 0.284147 0.47%

Predicted Groeneveld 2006 CHF at 16.51 cm .- 034%

[kW/m] 2079.638 2086.678 Predicted Groeneveld 2006 CHF at 19.05 cm 1989.248 1995.506 -0.31%

[kW/m 2 ]

Predicted Groeneveld'2006 CHF at 21.59cm 1911.717 191 -0.32%

[kW/m 2 ] 191 7.814 Predicted Bernath CHF at 16.51 cm [kW/mz] 1370.584 1380.707 -0.74%

Predicted Bernath CCHF at 19.05 cm [kW/m'] 1285.578 1294.833 -0.72%

Predicted Bernath CHF at 21.59 cm [kW/mz] 1205.403 1215,050 -0.80%

CHFR with Groeneveld 2006 at 16.51 cm 2.546 2.630 -3.30%

CHFR with Groeneveld 2006 at 19.05 .cm 2.557 2.550 0.27%

CHFR with Groeneveld 2006 at 21.59 cm 2.607 2.639 -1.23%

CHFR with Bernath at 16.51 cm 1.678 1.740 -3.69%

CHFR with.Bernath at 1.9.05 cm 1.653 1'.654 -0.06%

CHFR with Bernath at 21.59 cm 1.644 1.672 -1.70%

Critical Rod Power with Groeneveld 2006 51.453 49,625 3.55%

Critical Rod Power with Bernath 33.403 33.517 -0.34%

From this table, it is evident that LEU-MOL has a more limiting local heat flux at 19.05 cm than LEU-BOL, thus giving a lower MDNBR at this axial location. However, the minimum CHFR shows a less noticeable difference (0.27%) between LEU-MOL and LEU-BOL than the MDNBR does. The reason the MDNBR results are not satisfactory is due to the method of calculating the necessary rod power for the Groeneveld 2006 to reach a MDNBR of 1.0while keeping the flow rate constant. This is shown in the table above where the critical rod power at 1.3 MW with the Groeneveld 2006 correlation is 3.55% higher for LEU-BOL than LEU-MOL. This is a very large discrepancy and is due to code convergence problems with the K2 and K4 terms with switching from negative to positive quality.

Therefore, the differences between LEU-BOL and LEU-MOL are small and generally within the errors.of the correlations themselves. .Because the Groeneveld 2006 and Bernath correlations were not developed for use in TRIGA analysis, the more limiting Bernath correlation was used. However, Anderson, et al from the University of Wisconsin has proposed to ANL to precisely determine CHF for the three TRIGA fuel Page 14 of 95

UWNR LEU Conversion Responses to Request for Additional Information assembly types (hexagonal, circular and rectangular). The results displayed in Tables 4.7.13.and 4,7,16 show that the reactor will not reach CHF even at 1.5 MW for LEU-BOL and LEU-MOL.

1. Figtres4.7.1* 6 (p. 8I4) 4.7. 15 115),

1.. .7 5,1(p.-I!,' n 7 rah~deichotrod" powerto rec-- s ~nton of p1 OL,:.LEU-MOL, and LEU-EQL. ,.Plea provide the Vnumercritial rod valis.QAthe ower as'determined by the Groneveld 2006 and'the 1Bernathco6irreaiai'6n for co6:e ps

-f1: MW,.1.3 MW and 1.5 MW at, HEU-BOL, LEU-BOL, LEU-MOL, and LEU-EOL, Licensee's Response:

The values are provided in the table below, where both the original reported and revised LEU-BQLL numbers are provided. The revised numbers account for the revised axial and radial power distributions for LEU-BOL.

Table RAI-21-1, Critical Rod Powers LEU-SOL CHF Power HEU-BOL Conversion LEU-BOL LEU-MOL LEU-EOL Correlation [MW] [kW/rod] SAR Revised [kW/rod] [kWlrod]

[kW/rod] [kW/rod]

Groneveld 1.5 52.376 53.465 53.112 52.832 54,553 2006 1.3 49.573 52.733 51.453 49.625 54.651 1.0 47.579 51.884 49,891 51,579 51.314 1.5 33.993 35.716 35.631 35.820 35.492, Bernath 1.3 31.849 33.488 33.403 33.517 33.153 1.0 28.206 29.437 29.599 29.406 29.124

'ed .

• 2ý o 4,7',..,.4A :7 . Y,; ,W'.,,, qs,

2. ~ectin 47.4,4.77, 47.8-, and 4.7. 9. -When y6u ,calculatýd:coplanrit~flo4i.t'~ ~~yrv dt rod. power forvarious core configurtions, RE.LAP. 6 1ted fWosillation osme power. Above this hot rod power, you providegraphs showing projections ofof

oolant flowrate. Is the extrapolation of flow calculated by RELAP5 above the last

redicted'stable flow realistic? Please discuss.

Licensee's Response;.

The normal steady state method of solving the RELAP5/MOD3.3 transient failed to get a stable solution for core powers around 1.5 MW. However, another method of solving the steady state cases using a 'pseudo transient' was used that was able to extend the RELAP5/MOD3.3 region of applicability before having to use the extrapolated region.

By setting the initial mass flow rate to nearly zero and setting the power to nearly zero and then ramping up the power until RELAP can calculate a steady state solution produced the following graphs for LEUýBOL, MOL and EOL. The revised axial and radial power distributions for LEU-BOL were used.

Page 15 .of'95

UWNR LEU Conversion Responses to Request for Additional Information Coolant Flow Rate of Hot Rod vs. PoWer (LEU BOL) 0.0 MW 1.3 MW 1.5 MW,

--- *- RELAP Calculated

. I.... . . . .. Flow 0.20.4 Rate

.... ...... -I.. .. .... . ....... ......... ..... ........

-- - Extrapolated Region 001 "

0.0 ............

07 u 005 ............ ... . *" .... * ... 00,1 ..W 0 -,. .. - . . * - * " . . * * * * , 0 0 s 10 15 20 25 30 35 40 45 Power in Hot Rod [kW]

Figure RAI 1, Coolant Flow Rate vs. Power LEU-BOL Flow Rate vs. Power of Hot Rod (LEU BOL) 1.0 MW 1.3 MW 1.5 MW I.- Greeed20 70 W, 10 -' . . ...... .......

50 .....

. on 0 1 ---.-----. -Extrapolated Region 0.04 0.06 0.08 0.1 0.12 0.14 0.1,6 0 .18 02 012 2 FlowRate [kg/s]

Figure RAI-22-2, Power vs. Coolant Flow Rate LEU--0OL Page 16 of 95

UWNR LEU Conversion Responses to Request for Additional Information Coolant Flow Rate of Hot Rod vs. Power (LEU MOL) 1.0MW 1.3 MW 1.5 MW 00,5 Rate .*

Extrapolated Region 0.4 4.~

I0 0.05 0 .1 0 0.0 0 5 10 15 20 25 30 35 40 45 Power in Hot Rod [kW]

Figure RAI-22-3, Coolant Flow Rate vs. Power LEU-MOL Flow Rate vs. Power of Hot Rod (LEU MOL) 1.0 MW 1.3 MW 1.5 MW

70. . . . . . . . . . . . . . .

60 .

50 0

40 0

30 0 - -Groeneveld 2006 20 1

-- ~Bernat 10

-- Extrapolated Region 0

0,04 0.06 0.08 0.1 012 0,14 016 018 0.2 0,22 Flow Rate [I(g/s]

Figure RAI-22-4 Power vs, Coolant Flow Rate LEU-MOL Page 17 of 95

UWNR LEU Conversion Responses to Request for Additional Information Coolant Flow Rate of Hot Rod vs. Power (LEU EOL) 1.0 MW 1.3MW 1.5 MW 0.25 -~

o RELAP CalculatedFlow Os,5 Rate ,

- - Extrapolated Region -. 4

~0 cc*~ *--- 0.3 oo ~......... .. o,4 0 5 10 15 20 25 30 35 40 45 Power in Hot Rod [kW]

Figure RAI-22-5, Coolant Flow Rz vEU L Flow Rate vs. Power of Hot Rod (LEU EOL) 1.0 MW 1.3 MW 1.5 MW 70* . . .. .

60 ........- ....... . ...... .... . .. .... . . . .

40 30 ,........ ..........

1W I Groeneveld 2006 0 ........ . . . . . .

- .- Extrapolated Region 0,04 0.09 0.14 0.19 0,24 Flow Rate [kg/s]

Figure RAI-22-6, Power vs. Coolant Flow Rate LEU-EOL Page 18 of 95

UWNR LEU Conversion. Responses to Request for Additional Information As can be seen from these graphs, RELAP5/MOD3.3 predicts a stable flow regime at 1.5 MW for all LEU statepoints. In addition if the inlet coolant temperature lowers from 54.440 C to 439C or 30"C than the stable flow regime would be predicted until 37 kW/rod and 43 kW/rod respectively. Therefore, no flow oscillations are predicted at 1.5 MW or below during steady state operation.

The direct answer to the question 'Is the extrapolation of flow calculated by RELAP5/MQD3.3 above the last predicted stable flow realistic?" is that it is simply not known. It may be possible that stable subcooled nucleate boiling flow occurs considerably above the powers. that RELAP5/MOD3.3 predicted, but without applicable experimental data it would be difficult to say with certainty. Anderson, et al from the University of Wisconsin has proposed to ANL to precisely determ ine CHF for the three TRIGA fuel assembly types (hexagonal, circular and rectangular). However, it is important to emphasize that a reactor power of 1.5 MW with an inlet temperature of 54.440 C is far beyond not only normal operating conditions at 1.0 MW with about a 30 0 C inlet coolant temperature, but also the anticipated faulted conditions where the power tr.ip is no higher than 1.3 MW.

3. Section 4.7.4, 417.7, 4.7,8, and 4.7.9. RELA,5 calculated flow oscillations at a power 6f round 28 kW/rod. Demonstrate that DNB is a more conservative criterion than flow instability in determininQ the thermal limit of the UWNRk.

Licensee's Response:

It may be possible that DNB is not a more conservative limit than flow instability, but without applicable experimental data it would be difficult to say with certainty. Also, the inability of RELAP5/MOD3.3 to predict a stable flow above a specific power may not be indicative of anything more than a limitation of the code. It is important only that the power levels at which either of the two undesirable phenomena occurs be far above where the reactor is operated and they are. For LEU BOL, for example, at the normal reactor power level of 1.0 MW, RELAP5/MOD3.3 predicts stable values of flow up to 1.5 MW. These calculations were performed for a coolant inlet temperature of 54.440 C, which is the maximum administrative limit and far above the normal value of about 30 0 C.

For an inlet coolant temperature of 30'C, the corresponding maximum power level for which RELAP5/MOD3.3 predicts a stable flow is 2.2 MW. It is worth noting that decreasing the inlet temperature from 54.44 0 C to 309C will also increase the power levels at which the Bernath and Groeneveld correlations predict DNB to occur. Thus, decreasing the coolant inlet temperature improves both the maximum predicted power for which stable flow is obtained and the powers at which DNB are predicted to occur.

As seen in the new figures in Question 22 for the LEU flow rate using the pseudo transient, the flow is stable at 1.5 MW and thus flow oscillations are not predicted in steady state operation of the LEU core. Since the reactor is limited to operating below 1.5 MW of steady state operation at all times, core power is the determining power of the thermal limit of the UWNR.

Furthermore, Anderson, et al from the University of Wisconsin has proposed to ANL to precisely determine CHF for the three TRIGA fuel assembly types (hexagonal, circular and rectangular) which could provide the necessary experimental data.

Page 19 of 95

UWNR LEU Conversiln Responses to Request for Additional Information

.4. Sectio.4.75. a weighted or averaged fuel temperature used in the of and 4.7.10. Was alculationý-ý_ýa the reactivity iv feedback Licensee's Response:

When implementing the 2 channel model for pulsing analysis, the radial nodilization was constructed so each radial zone in the fuel meat had equal radial volume.

RELAP5/MOD3.3 gives the choice of defining the reactivity feedback as a function of volume density or volumetric average fuel temperatures. ý Having equal radial volumes, the simple average of nodal temperatures is automatically volume-weighted.

In addition, RELAP5/MOD3.3 performs point reactor kinetics by computing the core average fuel temperature to use in the reactivity calculation, Since the height of each node is the same across the core, then a simple nodal averaging is used. Thus with nodal temperatures volume-weighted and the reactivity calculated with simple nodal averaging, the reactivity feedback used in RELAP5/MOD3,3 is volume-weighted averaging across the whole core.

5. -Seitio 4.7.5 and 4.7.10.,Was the effect of direct gAmma heating ofIhel C.

Z61a'incororated in the RELAP5 model Licensee's Response:

.Yes, the effect of gamma heating was incorporated into the pulsing models in sections 4.7.5, 4.7.10, and 13.2. In RELAP5/MOD3.3 oard 30000001, the default gamma was chosen to provide the gamma heating from the standard fission product decay calculations. RELAP5/MOD3.3 automatically calculated the amount of gamma heating in addition to the fission power to give'the total power produced, 6,. etin 4.4.5 and 4.7.10., Wasthe power distribution, in the core maintained constantJ

--.uringthe pulseahd was the assumption conservative&

Licensee's Response:

Initially, the pulsing analysis was performed using the same radial, axial, and pin peaking factors as those used in the steady state. analysis. To verify that this was a conservative assumption, an MCNP case was run for the most limiting case, LEU-MOL, with the transient rod full out and the blades at the cold critical bank height to determine what the radial, axial, and pin power peaking factors would be. The LEU?-MOL pin power peaking factor for D5SW was 1.797, axial peaking factor of 1.284, and radial peaking factor of 1.566 as seen in the table below. These new power profiles were put into the LEU-MOL RELAP5/MOD3.3 model to give a new prompt peak fuel temperature of 790.45°C in the hot rod. This maximum fuel temperature is 8.74% higher than the previous LEU-MOL results of 726.95°C. These results are more limiting, but they do not exceed the maximum fuel operating temperature of 8309C.

Page 20 of 95

UWNR LEU Conversion Responses to 'Request for Additional Information Table RAI,26-1, Key Paranmeters :Comaarina Power Distributions iP4rameter ILEU-MOL LEU.MOLý1Percn

- - Original I T-Rod out I Difference Pin Power Peakin. Factor I 1-598797 - 4S

.Axial Pea-kn t**or . . -359 1* 8.284 -5.519 1 Radial Peaking'Factor _____ i 1.438] i.6 1 8.901%

fa[uEi epeueJ 1 726.95! 797.45 Peak Pulse Powepr [PVV _.. 252 1 2.521 0 0.1 0%_/

Additionally, the RELAP5/MOD3.3 pulsing modei includes'the following limiting assumptions:

• Instantaneous firing of the transient rod P D5 SW channel includes heated perimeter of the transient rod making power

• No reactivity feedback in the hot rod

• No moderator temperature/void reactivity feedback

• Power profile remains constant through transient Therefore this model is conservative and still demonstrates that the fuel temperature will not exceed 830,C during a 1 /4%Ak/k pulse for the most limiting state-point,

'Licensee's Response:

The power profile and the maximum temperature from pulse initiation until reactor scram (15 seconds after pulse initiation) for the four figures in question are given below. The revised LEU-BOL axial and radial power distributions were used.

Power and Temperature Profile after 1.4%Ak/k Pulse (HEU BOL) 900

• ' - 700 e.. !

I-,Max Temp 5Soo

! -- -- opera#ting L mit

{~~~..... ,..r"" ',0 *! n_ _ .......... ' 0 02 4 68 10 12 1 Time [s]

Figure RAI-27-1, Power and Temperature Profiles after Pulse HEU.-BOL Page 21 of 95

UWNR LEU Conversion Responses to Request for Additional Informration Power and Total Energy of 1.4% Ak/k Pulse (LEU BOL)

.oo

..- - - - - - .000 I.E+09 - --.. .. 8 00

• .* .700

~.-Rower.

ao - MaxTemp

- ,- Operating Limit t

.+0. . . . . .200 0 2 4 6 8 1to 12 14, Time Ws]

Figure RAI-27-2, Power and Temperature Profiles after Pulse LEU-BOL Power and Temperature of 1.4% Ak/k Pulse (LEU MOL) 1E+10- -- r1000

,_ ~70Q

- Max Tern p i CL - -* Operating LimitI Soo:

".... . 1200 0 2 4 6 8 10 12 14 F :Time [s]

Figure RAI-27-3, Power and Temperature Profiles after Pulse LEU-MOL Page 22 of 95

UWNR LEU Conversion Responses to Request for Additional Information Power and Temperature of 1.4% Mk/k Pulse (LEU EOL)

. . .. . . . . . . . . . ... ..... . ... . . . . . . .. . . 1000 900

-- - - -- - -- --- - - - - - - -0~

"- -OPwer Lii 700

• i,,.. -- Max Temp O "' -- Operating Limit

4. 00

- --- --.~---j~600

! 300 L'E+O 64 ... . . ... ........ ... ..... . .. . . . .... .... .. .. .......... 20 0 0 2 4 6 8 10 12 14 Time [s]

Figure RAI-27-4, Power and Temperature Profiles after Pulse LEU-EOL These graphs were produced with the following assumptions of the 2-channel model.

• D5 SW channel flow.area adjusted to be a typical cell so that the heated perimeter does not include the transient rod making power. This was the same change made in the LOCA analysis.

o Instantaneous firing of the transient rod changed to 0.1 seconds to fire the transient rod o Incorporating Doppler reactivity feedback in the hot rod

• Incorporating Moderator Temperature feedback from Tables 4.5.6 and 4,5.13 a Incorporating Moderator Density feedback from Tables 4.5.6 and 4.5.13 The major change to this model was the incorporation of the additional feedback mechanisms, specifically the moderator density feedback. Due to void production during the pulse, this adds an additional negative feedback in addition to the Doppler feedback, As can be seen from these graphs at .no time does the maximum temperature exceed 8301C. The maximum predicted temperatures over the 15 seconds are. 738.,35QC.

• 756.45"C, 826.150C, and 810.25°C for HEU-BOL, LEU-BOL. LEU-MOL, and LEU-EOL respectively.

Characteristically for pulses in TRIGA reactors, the fuel temperature peaks near the outer surface of the fuel meat near the peak power and shortly thereafter. As time passes, heat is redistributed toward the center of the fuel meat and the temperature rises, becoming larger at the, fuel center than the fuel outer surface temperature, as shown in the figure below.

Page 23 of 95

UWNR LEU Conversion Responses to Request for Additional Information Power and Temperature of 1.4% Ak/k Pulse (LEU MOL) r900

• --,-'=Max Centerline I 700Q 1.Ei l.-- cae0- of u

---

--- o'- Li tf perating 400~80 o a Max Outside Fue0 Edge I .E+ 0 6 . ..."............".. . ... ..* , .. ..........

.. 200 1.E+07--------------------------.

FiueRI2-,Pwr* n eprtr fe Powlý!er UM_____"

0 2 4 6 8 10 12 14 Time [s]

Figure RAt-27-5, Power and Temperature after Pulse LEU-MOL From the figure above, the prompt peak temperature on the outside edge of the fuel is 6130 C, and is substantially less than 830eC However, a as the heat is redistributed, the maximum temperature of the rod occurs at the fuel centerline at 1Ssec and in the limiting case of LEU-MOL, the fuel temperature approaches the operating limit of 830ot with a maximum of 82600.

The0 consequence of the fuel centerline temperature approaching the operational limit of 830 C is not significant when- considering the basis for this limit. It is known that after extensive steady-state operations at 1MW, the hydrogen in the ZrHx matrix will redistribute due to migration from the central high temperature regions of the fuel to the cooler outer regions, thereby increasing the ZrHx ratio from the nominal value of 1-6.

When the fuel is pulsed,* the instantaneous prompt temperature distribution, is such that the highest values- occur at the surface -of the element and the lowest values Qorcur at the center. The higher prompt temperatures in the outer regions occur in fuel with hydrogen to zirconium ratios that have now substantially increased above the nominal value. This produces hydrogen gas pressures considerably in excess of that expected for ZrH, 6. If the pulse insertion is such that the temperature of the fuel exceeds 8740 C, then the pressure will be sufficient to cause expansion of microscopic holes in the fuel that grow with each pulse (General Atomics, "Pulsing Temperature Limit for TRIGA LEU Fuel," GAwC26017, December 2007). However, at the center of the rod, the ZrH, ratio has decreased below the nominal value. Thus, as the center of the fuel rod approaches 8261C, in the limiting case of LEU-MOL, the fuel rod will not produce hydrogen gas pressures in excess of the expected ZrH, 6 .

Page* 24 of 95

UWNR LEU Conversion Responses to Request for Additional Information Finally, the analysis is still conservative in that lit neglects cross-flow between channels, Which is anticipated to be significant during a pulse transient. Also, additional margin exists because operationally the *reactorscrams within 5 seconds of the pulse rather than 15 seconds, Licensee's Response:

Yes, the axial power profiles were derived from MCNP using critical blade heights.

The critical bank height moves out to compensate for core burnup. The revised LEU-BOL curve is shown in the following figure, to replace Figure 4.7.34 on page 102.

Axial Power Distribution for LEU Core 1.6 BQL Rod's at 11.73.in 1.4 MOL Rods at 12.05in 1.2 A A EOL Rods at 13.14in A

A. A A A I,- 0.8 0 4 0.6 0.4 0.2 0.0 0 5 10 15 20 25 3Q 35 40 Axial Distance from Bottom of Fuel (cm]

Figure RAI-28-1, Revised LEU Axial Power Distribution Page 25 of 95

UWNR LEU Conversion Responses to Request for Additional Information S9-*section 4.7.6. Are there core locations other than D5 SWthat willFhave pin powerl teaking factors greater than 1.61 if a fresh LEU fuel pin is inserted in those locations atl oror l. .. . ..

Licensee's Response:

It is possible that there will be core locations other than D5 SW that will have pin power peaking factors greater than 1.61 if a fresh LEU fuel pin is inserted in those locations at MOL or EOL. However, D5 SW will always have the greatest, pin power peaking factors throughout core life if replaced with fresh fuel. This is because D5 SW will have the greatest burnup of any pin, and therefore the reactivity insertion caused by replacing any other pin location with a fresh fuel pin will be less than replacing the fuel pin at D5 SW. As stated on page 103 of the analysis report, inserting a fresh fuel pin in D5 SW at EOL will produce the largest pin power peaking factor of 1,74. However, there will be no other core location that will produce a pin power peaking factor greater than 1.74 at EOL.

• 7,-' ection 4.7.6. Your application states that if the hot rod at core location 05 SW needed o be replaced that the CHF limit would not be exceeded. What acceptance criteria is

)Tsed following replacement of the fuel? Would a 10 CFR 50.59 review be performed asi art of the fuel rod replacement..

Licensee's Response:

The acceptance criteria for replacement of fuel at core location D5 SW would be to ensure that the pin power peaking factor would not exceed 1.61. This would ensure that the design basis analysis of sections 4.7.6 - 4.7.9 would remain valid, It should be noted that replacement of fuel at core location D5 SW could not be with fresh fuel and still meet the acceptance criteria.

In addition, while the analysis of sections 4.7.6 - 4.7.9 have demonstrated that the hot rod at core location D5 SW is the limiting rod with respect to CHF with a pin power peaking factor of 1.61, inserting a fresh fuel pin next to a control blade shroud could be more limiting due to the increased wetted perimeter which decreases the margin to CHF due to reduced flow. In order to prevent the fresh fuel from decreasing the margin to CHF when placed in a location next to a control blade shroud, the pin peaking factor must be less than 1.47. This ensures that the margin to CHF is no less than what was analyzed for the hot rod (D5 SW) in sections 4.7.6 - 4.7.9.

A 10 CFR 50.59 review would be performed as part of a fuel rod replacement and core rearrangement, The analysis would need to show that the acceptance criteria for loading new fuel would be met; specifically, the fresh pin power peaking factor be <

1.47 when placed next to a control blade shroud and s 1.61 in all other locations.

Page 26 of 95

UWNR LEU Conversion Responses to Request for Additional Information L~i~eofr h777 -(fr 10) AO-what coolant temi55rature-4n po lvl ire'theLU calculations performed? If temperatures and levelsUsed are not licensed limits lease Licensee's Response:.

The LEU calculations assume a coolant temperature of 130OF and a pool level of 19 feet above the core. The pool level of 19 feet is at the limit given in Technical Specification 3.3,3(d). The, coolant temperature is at the administrative limit according to UWNR 100. In response to RAI question 15, the coolant temperature limit is being added to the Technical Specifications. See question 56.

7. (. 11i1), 4.7.15 (p. 119), and 4 7.17 (p 124-.-Why is the maxim Tabl-e 4.7.12 (pmfuel

' emerature at-L-EU"EOL lower than that at LEU-MOL.

Licensee's Response:

The key parameters for the differences in the maximum fuel temperature between LEU-MOL and LEU-EOL are summarized in the table below. The percentage of LEU-MOL higher than LEU-EOL column shown is calculated as:

(LEU-MOL - LEU-EOL),/ LEU-MOL

• 100%

Table RAI-32-1, Key Parameters LEU-MOL vs, LEU-EOL for Steady State Analysis P.er-ent LEU-MOL higher Parameter (at 1.5 MW) LEU-MOL LEU-EOL than LEU-EOL Pin Power Peaking Factor 1.598 1.567 1,940%

Axial Power Peaking Factor 1.359 1.304 4.047%

Outside Radial Power Peaking Factor 1.438 1.358 5.563%

interior Radial Power Peaking Factor 0.784 0.817 -4.209%

Centerline Temperature [(C] 665.06 641.91 - 3.481%

Qutside Cladding Temperature [°Q] 141,43 140,48 0,672%

Temperature Difference [°C] 523.63 J 501.43 L 4.240%

The pin power peaking factor, the axial peaking factor, and the outside radial power peaking factor for LEU-MOL are 1.940%, 4.047%, and 5,563% higher respectively than LEU-EOL. However, the interior radial power peaking factor for LEU-MOL is 4.209% lower than LEU-MOL. In the hottest LEU-MOL fuel rod, more of the power is deposited closer to the surface for the fuel rod and less is deposited closer to the center of.the fuel rod than in the hottest LEU-EOL rod. Since on average the power has further to travel in the LEU-EOL rod than in the LEU-MOL, the radial temperature rise in the LEU-EOL is a little bit closer to the LEU-MOL radial temperature rise than it would otherwise be.

This can be seen best by looking at the temperature difference between the clad and the centerline temperature. For LEU-MOL the temperature difference is 523,63QC and for EOL the temperature difference is 501.43 0 C. Thus, the LEU-MOL temperature difference is 4.240% larger than LEU-EOL, Page 27 of 95

UWNR LEU.Conversion Responses to Request for Additional Information It is also interesting to note that the pulse analysis shows the maximum power of the pulse is higher for LEU-EOL than LEU-MOL, but the temperature of LEU-MOL. is higher by 3.46TC than LEU-EOL. The key parameters for the pulse analysis can be seen in Table RAI-32-2, Table RAI-32-2, Key Parameters LEU-MOL'vs. LEU-EOL for Pulse Analysis LEU-MOL higher Parameter (at 1,5 MW) LEU-MOL LEU-EOL than LNV-EOL Pin PowerPeaking Factor . .1.59 1,567 1.940%

Axial Powe-rPeaking Factor 1.359 1.304 4.047%

Outside Radial Power Peaking Factor 1.438 1.358 5.5&63%o-Interior Radial Power Peaking Factor 0.784 0.817 ..... -4,200%

Maximum Pulse Power [GW] 2.52 3.06 -21.429%

Maximum Temperature [°C] 726.95 723.49 0.476%

Total Negative Temperature -9.30979 -8.604 7.571%

Coefficient entered into RELAP [$/K]

The maximum pulse power for LEU-EOL is higher than LEU-MOL since the total negative temperature coefficient entered into RELAP is 7.571% lower for LEU-EOL, While the maximum pulse poweris 21 .429% higher for LEU-EOL, the pin power, axial power, and outside radial power peaking factor are all higher for LEU-MOL. In a pulse, more power is being produced at the outer edge of the fuel and thus the maximum temperature occurs in this region. Therefore, with a higher hot rod power, axial peaking and outside radial power peaking factors, LEU-MOL has a higher fuel temperature despite having a lower maximum pulse power.. The difference between the maximum fuel temperatures is only 0.476% or 3.46=C and is not a significant difference.

-3. *Tables 4.7.P1(p.

ower than 111) and 4.7,15 (p. 119). The hot rod power shown at LEU-WMýi that at LEU-BOL however the maximum fuel temperature FeUMVOL.

I Please disc~us~s.1 is higher ati Licensee's Response:

This response is taken in its entirety from Question 20.

An error was discovered with the LEU-BOL axial power shape from MCNP5-that was input into the RELAP5 LEU*BOL models. The following figure shows the comparison between the original axial power shape and the revised MCNP5 axial power shape, Page 28 of 95

UWNR LEU Conversion Responses to Request for Additional Information Axial Power Profile Comparison (LEU-BOL)

--~Original

.- Revised 0,6~~.. .... .....

04 0 510 15 20 25 30 35 40 Axial Height from Bottom of Active Fuel [cm]

Figure RAI-33-1, Axial Power Shapes Comparison Between Original and Revised (LEU-BOL)

As is evident by Figure RAI-33-1, the peak axial power changed from 1 368 in the original analysis to 1.4032 in the new MCNP5 calculations for LEU-BOL. The pin power peaking factor did not change. This changed the maximum fuel temperature from 662.83°C to 673.86°C at 1.5 MW which is higher than the LEU-MOL maximum fuel temperature at 1.5 MW of 665.06°C. Therefore LEU-BOL has the highest maximum fuel temperature with the highest rod power.

The new, LEU-BOL steady state analysis is presented, in the table below:

Page 29 of 95

UWNR LEU Conversion Responses to Request for Additional Information Table RAI-33-1, T/H Comparison between Original and Revised LEU-BOL results

....... Core LU.e.o With Revised Parameter Power LEU Conversion Axial I Radial  % Difference WSAR Power Shape Rod Power in D5SW 1,5 29.041 29,041 0.00%

1.3 25.169 25,169 0,00%

[kVV] 1.0 19.361 19.361 0.00%

1.5 0.14878 0.14861 1.64%

Mass Flow Rate [kg/s] 1.3 0.13143 0.13105 1.61%

_______  ! .0 0. 10535 0.10503 1.54%

Maximum Fuel 1.5 662.83 673.86 0.30%

Centerline 1.3 594.40 604.10 0,28%

Temperature [°C] 1.0 490.15 497.81 0.26%

Maximum Outside 1.5 141,60 142.02 0.30%.

MaximumrOute 1.3 139.60 139.99 0.28%

Clad Temperature [°C] 1.0 136.30 136.66 0.26%

Ex.t.Out.r.Clad1.5 127.47 127.78 0.24%

Exit Outer Clad 1.3 127.14 127.06 -0.06%

Temperature ['0] 1,0 125.09 125,02 40,06%

1.5 101.32 100,95 -0.37%

Exit Bulk Coolant 1.3 100.04 100.17 0.13%

Temperature [°C] 1.0 98.23 98.37 0,14%

1.5 53.465 53.112 4-066%

Critical Rod Power 1.3 52.733 51,453 -2.49%

Groeneveld 2006 1.0 51.884 49.891 r3.99%

........ ..... 1.5 35.716 35.631 -0.24%

Critical Rod Power 1.3 33.488 33.403 0.25%

Bernath 1.0 29.437 29.599 0 Power to Reach . . CHF

. . . evoid 2006 52786 _ _ _ 53.112

_ _ _ _ _ _ _0.61%

at last flow rate non-oscillatory Bernath3.63 35.!64 35,631 1.31%

1.31%

MDNBR - Groeneveld 1,5 1.818  !.829 0.59%

1.3 *2.095 2.044 -246%

2006 1.0 2,680 2.577 4.00%

1.5 1.211 1.227 1.30%

MDNBR - Bernath 1.3 1.331 1.327 -0.29%

1.0 1.520 1.529 0.68%

Where % Difference is defined as:

%Diff revised4* original 100%

revised As can be seen in the table above, the changes in the axial flux profile lead to insignificant differences.

Page 30 of 95

UWNR LEU Conversion Responses to Request for Additional Information

-4.7 Pleas refer ta auesontios 56 and - 8 when respondlna to thle followinci aUes-'fib0K able 4.7.14.to The W appear' calculated be above thermocouple the.,.SSS temperatures limit of 40000C. at 1 MWWhat Please expliin, for location' are the Q4]

herrmocouple temperatures at 1.3 MW and 1.5MW?..

Licensee's Response:

The original basis for the LSSS of 400'C was based on the 1973 SAR estimate of peak fuel temperatures at the UWNR from the Torrey Pines TRIGA Mark Ill reactor analysis, despite the fact that these two reactors are geometrically dissimilar.

During the refueling of the UWNR to the TRIGA core, measured temperatures for D4 SW were reported to exceed 400°C at 1MW, as reported in the startup program and included in the HEU 2000 license renewal SAR (page 4-45). Therefore, the IFE connected to the fuel temperature safety channel was placed in a location that fuelthat would not exceed 400°C at 1MW, specifically E3 NE. It is fully expected temperatures in the interior of the core will be greater than 400 0C. This is why this application proposes a change to technical specification 2.2 to provide greater flexibility in placing an IFE in the central region of the core, if desired, that could be connected to the fuel temperature safety channel. See proposed change to

'technical specification 2.2 in the response to RAI question 56.

The calculated thermocouple temperatures at 1.0 MW, 1.3 MW and 1.5 MW using the same methodology that created Table 4.7.14 are shown in the three tables below.

The IFE temperatures for D4 SW were updated in order to incorporate the revised axial LEU-BOL shape for the hot rod, Table RAI-34-1. IFE Temperatures at 1.0MW IFE Summary Table at 1,0 MW IFE Locoiion 0.1 mil gap 0.05 mil gap 0.15 mil gap

.C OF QC OF QC OF Bottom 444.37 831,86 397.20 746.95 488.17 910.71 D4 SW Center 429.50 805.10 384.22 723.60 471.64. 800.95 Top 412.55 774.58 .369A47 . 697.04 452,75 846.95 Bottom 299.27 570.69 270.90 519.62 326.27 619.29 E3 NE Center 291.37 556.47 264.14 507.44 317.34 603.20 Top 279.87 535.77 254.30 489.74 30430 579.74 Table RAI-34-2,/lFE Temperatures at 1.3MW IFE Summary Table at 1.3 MW .... ___ .....

.FE L.ocation 0.1 mil gap 0.05 mil gap 0.1S5mil gap I iC OF QC F OC 0 F

Bottom 535.23' 995.41 476.65 889.96 589.16 1092.48 D4 SW Center 516.39 961.49 460.09 860.15 568.30 1054.93 Top 494.89 922.80 441.24 826.22 544.45 1012.01 Bottom 348.95 660.11 313.48 596.26 382.38 720.28 E3 NE Center 338.97 642.15 304.89 580.79 *371.16 700.08 Top 324,47 616.05 292.42 558,36 354.81 670.66 Page 31 of 95

UWNR LEU Conversion Responses to Request for Additional Information Table RA1-34-3, IFE Temperatures at 1.5MW IFE Summary Table at 1.5 MW

!FE Location 0.1 mil gap 0.05 mil gap 0.15 mi! gap

.C "F 9C OF 9C 9F Bottom 594.79 1102.62 528.99 984.18 654.94 1210,89 D4 SW Center 573.32 1063.98 510.07 950.13 631.29 1168.32 Top 548.82 1019.87 488.53 911.35 604.26 1119.67 Bottom 38.41 720.34, 342,28 648.10 420.00 788,00 E3 NE Center 371.06 699.90 332.46 630.43 407.27 765.09 Top 354.55 670.18 318,23 604.81 388.73 731.71 Licensee's Response:

The assumed masses of cadmium were, 1.04kg for the pneumatic tube and 1.57kg for the whale tube.

36.. Section 12.6, Are there any quality assurance teststhat University of Wisconrsin WiHl

__PppIy upon receipt of the fuel? If yes,_please briefly describef Licensee's Response:

Yes. The fuel rods are inspected by the CERCA Quality Inspectors at the fabrication facility, Following the CERCA inspection, the Idaho National, Laboratory performs an on-site Source Inspection of all of the CERCA QA inspection/verification records. The INL then performs a visual inspection of all fuel elements, records all imperfectionsi and verifies the imperfections are within the established design criteria. The INL also performs a verification of all accessible dimensions on a statistical sampling of fuel elements. If TRIGA fuel fabrication is underway at the facility during the inspection visit, the INL QA inspector will observe the fuel element assembly process to ensure that the process is being carried out as expected.

The UWNR will perform a Receipt Inspection of each fuel element while the INL Quality Assurance Inspector is at the UWNR facility. The Receipt Inspection ensures that fuel elements were not damaged during packaging or transport. The Receipt Inspection is a visual inspection, and observed imperfections can be compared to the Source Inspection records with the INL QA Inspector. The UWNR will also review all of the CERCA and INL QA inspection/verification records to ensure each fuel element complies-with design specifications.

Following the receipt inspection, fuel elements will be measured in accordance with UWNR 142, Procedure for Measuring Fuel Element Bow and Growth, to establish the baseline for future annual measurements, Page 32 of 95

UWNR LeU Conversion Responses to Request for Additional Information Licensee's Response:

The graphite reflectors will be inserted after loading to the proposed 21 bundle core, They will be inserted only after measuring their individual reactivity worth and verifying that adequate-shutdown margin will be maintained. After loading to the full J21-R14 core the shutdown margin will be measured and verified to meet Technical Specification limits.

Wil~theFE b tet nd calibrate? If so, when?,

Licensee's Response:

Each IFE will be tested prior to loading into the core. Upon receipt, the reslstance values will be verified to be consistent with manufacturer reported values to rule out a possible short or open circuit in the thermocouple. Than the signal from the IFE will be read with a calibrated process meter and the IFE reading will be verified to be consistent with a known reference temperature.

,oprnepmethO'd: Wi~lll~ye 'determne:-"excess,.reac*t. ?ilfso please diseu's, Licensee's Response:

Yes. Excess reactivity will be determined upon reaching criticality and the operational core using the rising period rod bump method. Excess reactivity will not be determined while loading from initial criticality to the operational core; however shutdown margin will be determined using the rod drop method after each fuel bundle addition to ensure compliance with Technical Specification 3.1.

0. $,ctipCt.27., How man power increment~stes ,w'l11beutihiedand how large are thej 0norem'eerents Licensee's Response:

10 increments in power level will be used, from low power (less than 1kw) up to 1MW full power in 100kW steps, Ie. ased, on'esri!dt orilased leri cmputer :models*

Licensee's Response:

Power and fuel temperature coefficients of reactivity will be calculated based on measured data during startup*testing.

Page 33 of 95

UWNR LEU Conversion Responses to Request for Additional Information

• Section 13. From a review of your accident analyses, it appears that some ot theI scenarios (e.g. maximum hypothetical accident (MHA), loss-of-coolant accident (L!C 3tc.) may have a potential radiological impact outside the reactor facility. From a

'eview of your emergency plan, dated 5/14/04, it is not clear how response is hand e n any potentially impacted areas outside the goerations bounda in the en ineering D-iding. Please discussf Licensee's Response:

In accordance with UWNR 150 emergency procedure, "Reactor Accident, Fission Product Release, or Major Spill of Radioactive Materials," the site boundary (as defined in the emergency plan revision 4) is evacuated. It is recognized that an inconsistency exists between the emergency plan definition of the emergency planning zone (EPZ) and the evacuation zone defined in UWNR 150. Therefore, it is proposed that the emergency plan be revised such that the emergency planning-zone is defined by the site boundary, rather than the operations boundary. The revision 6 of the emergency plan is attached for approval*. See attachment 7.

The emergency plan was also revised to account for updated dose calculations for, four accident events in Table 2, as well as to update the emergency action levels due to revised released inventories as a result of the conversion to LEU. After correcting for the LEU-BOL power distribution, the BOL case was found to be more limiting than the MOL case which was reported in the LEU conversion SAR, therefore the changes to revision 6 of the emergency plan use the revised LEU-BOL power distribution. The dose calculations reflect the revised LEU 30/20 core design as well as current '_..

methodologies of calculation in the analysis report as reported in sections 13,1.5.2, 13,1.6.3, 13.1.7.3, and 13,1.8.1. However, the results as reported in the analysis report were updated to use more appropriate fission product release fractions. The analysis report calculated the release fraction based on the maximum centerline temperature, but a more accurate approach is to use an effective release fraction calculated by volume integrating the release fraction equation across the fuel temperature distribution, both axial and radial, This is appropriate since the release fraction measurements were made on small isothermal fuel samples (General Atomics, "The U-ZrH, Alloy: Its Properties and Use in TRIGA Fuels." GA E-1 17-833, February 1980, page 5-5). The revised release fractions are:

approximately 10% of the valueslisted-in chapter 13 of the analysis report. This results in changes to the emergency action levels and potential 1

exposure as detailed below, The emergency action levels were changed in revision 5' of the emergency plan as a result of modifications to the ventilation system. The original calculations for ventilation system operable assumed that the release was instantaneous, and that it was vented at a constant rate for the amount of time it would take to exchange one confinement volume, where the confinement volume was assumed to be 2000m3 The revision 4 emergency action level, was derived by assuming the insoluble beta emitter activity of was released. This activity was released to the confinement volume and vented at a constant rate thereby producing a concentration of:

2000m .

This was rounded down to for revision 4 of the Emergency Plan.

Page 34 of 95

UWNR LEU Conversion Responses to Request for Additional Information Revision 5 of the emergency plan was performed in accordance with a 10 CFR 50.54(q) analysis which included changes resulting from the new ventilation system. The new ventilation system is designed to sweep air into the reactor laboratory from public access space through theauxiliary support spaces surrounding the reactor confinement.

This limits the potential spread of airborne contamination. To accomplish this, exhaust is taken from the reactor confinement and the auxiliary spaces and combined in a common plenum, prior to release from the stack. The new ventilation system has a nominal exhaust flow rate from confinement of 2700 scfm and an exhaust flow rate of 9600 scfm (4.531.i 3/s) in the mixing plenum where the stack sampie is taken. The additional dilution would decrease the release concentration and therefore the emergency action level was revised.

As calculated in the 2000 license renewal SAR Rev 2, Appendix A, page A-4, the time to vent confinement is i 569s. The total volume of air exhausted in I 569s is 4.531m 3/s*1569s z 7109m3 . Therefore the revision 5 action level is:

7109m3 3 This was rounded down to iE-4pCi/ml for revision 5 of the emergency plan.

For revision 6, only the activity of was revised. By using the revised B-OL power distribution and the revised release fractions, the activities of the insoluble beta emitters are BOL, MOL, and EOL. The insoluble beta emitters are Kr-85m, Kr-85, Kr-87, Kr-88, Kr-89, Xe-I 33, Xe-1 35, Xe-137, and Xe-I 38.

If the more limiting value of at EOL is used, then the revision 6 action level is:

71 09m3 This is rounded down to for revision 6 of the emergency plan, when the vent Idtion system is operable. When the ventilation system is inoperable, the revision 1

6 action level is:

2000in This is rounded down to for revision 6 of the emergency plan, Page 35 ef 95

UWNR LEU Conversion Responses to Request for Additional Information The following 4 accident events are therefore revised in revision 6:

1. Event: Severe fuel clad leak-approaching MHA size with pool level normal and ventilation system operative, Potential Release:

Action Level; whole body, whole. body

'revised to thyroid revised to thyroid I

2. Event:

Potential Release:

Action Level:

Severe fuel clad leak-approaching MHA size with pool near empty and ventilation system normal.

whole body, whole body revised to thyroid revised to thyroid I

3. Event: Severe fuel clad leak-approaching MHA size with pool level normal and ventilation system inoperative.

Potential Release: whole body, , thyroid revised to whole body. thyroid Action Level: revised to

4. Event: Severe fuel clad leak-approaching MHA size with pool near empty and ventilation system inoperative.

Potential Release:

Action Level:

whole body, whole body, revised to thyroid revised to thyroid I

' Note: Action levels for ventilation system normal are reported as revision 4 of the emergency plan. These values were recently changed, as a result of a new ventilation system, to in I in revision 5, which was submitted under separate cover following a 10 CFR 50.54(q) analysis.

Changes to the emergency plan as a result of the LEU conversion are I

submitted here as revision 6. For convenience, revisions 4, 5 and 6 are included in attachment 7, both with and without strikeouts indicating changes from the previous version.

Page 36 of 95

UWNR LEU Conversion Responses to Request for Additional Information

• '3.ý Section 13.1. It is stated that certain isotopes (eig., 1-130m, 1-136,- KrF89, Xe-137) werer hot included in the estimates for whole-body and thyroi dose .because of their "short

,kalf-lives (less than10 minutes).", Given the short exposure time of five mi*n*utes lhese isotopes willmake a contribution to the doses. Please iustifyltheir exclusion, ,or tgubmit revised doses.

Licensee's Response:

The primary reason as stated in the analysis report for neglecting these isotopes is the lack of any published dose coefficients. However, using the methodology in reference 24 of the analysis report, the whole-body effective dose coefficients were manually computed for the short-lived isotopes and revised results are shown below (Table numbering represents the original numbering in the LEU Conversion Analysis SAR). The revised LEU-BOL power distribution was used. Also, the revised results use the more realistic temperature distribution integrated release fraction as described in the response to question 42. The thyroid dose contributions were not revised. The source for the thyroid dose coefficients, "Federal Guidance Report No.

11: Limiting Values of Radionuclide Intake and Air Concentration and Dose.

Conversion Factors for Inhalation, Submersion, and Ingestion," states on page 25 that the biological half-time to transport iodine from the blood to the thyroid is 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.

Any short-lived isotopes would therefore have a negligible cohtribution-to the thyroid dose.

Revised Table 13.1.4 MHA OccupationalExternal Dose by Isoto ge Isotope Effective Revised HEU LEU BOL LEU MOL LEU EOL Dose Coef. External External External External

• Worker Dose Worker Dose Worker Dose Worker Dose (rem-m 3/Ci-s) (rem) (rem) (rem) (rem)

Br-82 4.810E-01 8.029E-07 6.870E-07 5.344E-06 1.029E-05 Br-83 1.413E-03 1.052E-06 3.753E-06 3.555E-06 2.769E-06 Br-84 3.482E-01 4.872E-04 1.735E-03 1.635E-03 1.265E-03 Br-85 7.898E-03 1.374E-05 4.884E-05 4.601E-05 3.554E-05 Br-87 2.680E-01 8.022E-04 2.853E-03 2.678E-03' 2.059E-03 1-130m 1.637E-02 5.174E-08 3.119E-08 3.591E-07 7.408E-07 1-131 6.734E-02 2.702E-04 9.491E-04 9.421E-04 7.643E-04 1-132 4.144E-01 2.480E-03 8.866E-03 8.609E-03 6.954E-03 1-133 1.088E-01 1.018E-03 3.621E-03 3.498E-03 2.796E-03 1-134 4.810E-01 5;084E-03 1.809E-02 1.742E-02 1.387E-02 1-135 2.953E-01 2.573E-03 9.156E-03 8.840E-03 7.064E-03 1-136 3.931E-01 1.668E-03 5.967E-03 5.717E-03 4.531E-03 Kr-83m 5.550E-06 4.132E-09 1.472E-08 1.396E-08 1'.087E-08 Kr-85m 2.768E-02 4.868E-05 1.730E-04 1.631 E-04 1.260E-04 Kr-85 4.403E-04 1.142E-08 5.143E-09 5.828E-08 9.676E-08 Kr-87 1.524E-01 5.421E-04 1.926E-03 1.809E-03 1.392E-03 Kr-88 3.774E-01 1.896E-03 6.736E-03 6.327E-03 4.865E-03 Kr-89 1.411E-01 8.994E-04 3.196E-03 2.994E-03 2.296E-03 Xe-131m 1.439E-03 5.716E-08 1.963E-07 2.017E-07 1.592E-07 Xe-133m 5.069E-03 1.387E-06 4.455E-06 4.800E-06 3.868E-06 Xe-133 5.772E-03 5.402E-05 1.821 E-04 1.856E-04 1.485E-04 Xe-135m 7.548E-02 1.195E-04 4.242E-04 4.165E-04 3.394E-04 Xe-1 35 4.403E-02 2.550E-04 9.873E-04 9.367E-04 7.222E-04 Xe-137 2.604E-02 2.166E-04 7.703E-04 7.421E-04 5.919E-04 Xe-138 2.135E-01 1.848E-03 6.571E-03 6.284E-03 4.959E-03 Page 37 of 95

UWNR LEU Conversion Responses to Request for Additional Information Rpvis(~d TaHr 11 L 6 Mi-IA Total fleounational flns~ during 5 minutr ~vaeuatmnn External Dose Thyroid Dose TEDE (mrem) (mrem) (mrem)

Previous HEU SAR 10 N/A N/A Revised HEU Analysis 20.3 2,110 83.7 LEU BOL Analysis 36.1 3,730 148 LEU MOL Analysis 34.6 3,670 145 LEU.EOL Analysis 27.4 2,960 116 Revised Table 13.1.8 MHA Building Occupant Doses for GroundReleasee External Dose Thyroid Dose TEDE (mrem) (mrem) (mrem)

Revised HEU Analysis 1.26 131 5.18 LEU BOL Analysis 2.24 231 9.16 LEU MOL Analysis 2.14 227 8.95 LEU EOL Analysis 1.70 183 7.20 ReviseaI Table 13.1.11 Near MHA with Pool Intact OccupationalDose during 5 minm'te evacuation External Dose Thyroid Dose TEDE (mrem) (mrem) (mrem)

Previous HEU SAR N/A 18,900 N/A Revised HEU Analysis 7.32 211 13.7 LEU BOL Analysis 13.1 373 24.2 LEU MOL Analysis 12.4 367 23.4 LEU EOL Analysis 9.69 296 18.6 Revised Table 13.1.12 Near MHA with Pool Intact Building Occupant Dosesfor GroundRelease External Dose Thyroid Dose TEDE (mrem) (mrem) (mrem)

Revised HEU Analysis 0.453 13.1 0.845 LEU BOL Analysis 0.808 23.1 1.50 LEU MOL Analysis 0.768 22.7 1.45 LEU EOL Analysis 0.600 18.3 1.15 By including the short-lived isotopes, the TEDE numbers are higher than previously reported by at least 4%, but no more than 11%, and are all still within limits.

However, a further reduction by a factor of approximately 10 is achieved using the more realistic release fraction. This is still a conservative calculation since no credit is taken for radioactive decay of the isotopes.

[S1 ection 13.1.2. In equation 13.1.1, the exponent is given as exp(-1.34x10"/T). Should he exponent be eXp(-1.34xl 04 /T)

Licensee's Response:

Yes, the exponent should be exp(-1.34x10 4 /T).

Page 38 of 95

UWNR LEU Conversion Responses to Request for Additional Information hpefifdA $1 d e -cfrem Licensee's Response:

Individuals on the fifth floor are evacuated with the rest of the building. The volume of the fifth floor was neglected in dose calculations for conservatism and because the open air atrium in the central wing of the .building would allow.

for readily mixing of building air between the first through fourth floors, while' the fifth floor is isolated from the atrium. The reported doses to building occupants apply to all individuals in the building, including those on the fifth floor.

• ur.n~g ~t~.Wine andsummermonths tiv e  ? What assumptionis.mosD.

onse*rative.

Licensee's Response:

The momentum rise, as calculated using Equation 13.1.4 on page 187, is 11 .3m regardless of the time of year.

The methodology for calculating the buoyancy rise is taken from "Workbook of

'Atmospheric Dispersion Estimates" by Turner, 1994 (page 3-2), which is reference 27 of the LEU conversion analysis report. First, the intermediate variable of buoyancy flux is calculated using the equation below:

F 2 9T/(4T gvd 8)

Where:F = buoyancy flux (mI/s3) g acceleration of gravity (9.8 ms,2 )

v . stack gas exit velocity (m/s) d = top inside stack diameter (m)

AT = stack gas temperature minus ambient air temperature (K)

T1 z stack gas temperature (K)

The monthly average temperatures as reported in the HEU 2000 license renewal SAR are assumed. The coldest month is January with a temperature of 16.80 F (264,7K) and the warmest month is July with a temperature of 71.4 0 F (295.0K).

The stack outlet temperature is assumed to be 72 0 F (295.4K) year-round. The stack gas exit velocity is 17.272m/s and the top inside stack diameter is 0.7747m as reported in the LEU conversion analysis report, The buoyancy flux is therefore calculated to be 2.6m 4/S3 in the winter, and 0.2m 4/sl in the summer.

The buoyancy rise is given by the following equation:

4H = 21.425F.14/u if F is < 55 Where, u wind speed at top, of stack (m/s)

The buoyancy rise is calculated to be 12.4m in the winter and 1.8m in the summer.

Page 39 of 95

UWNR LEU Conversion Responses to Request for Additional Information A lower buoyancy rise is more conservative because it will result in a lower effective stack height which in turn will cause higher ground-level concentrations.

Because the buoyancy rise is not steady year-round, It was assumed to be Om in the analysis report, which is more conservative.

7, Section !3. i7, 3.8-and 13.1 9. For those scenarios where the Ventilation system is in-peration, what is the dose to persons in the Mechanical Engineering Building from shine om the volumetric source term in the confinement building until it is ventilated to the I rnvironment.

Licensee's Response:

The dose was calculated using an MCNP model of the confinement structure. The revised LEU-BOL power distribution was used, as was the revised temperature distribution integrated release fraction as discussed in the response to question 42. If the total released inventory is assumed to be uniformly dispersed in the confinement volume, then the dose rate in the nearest unrestricted area of the building is calculated to be If this dose rate existed during the time required to exhaust the entire confinement, 26 minutes, it would result in a dose of approximately

8. Section 13.1.7. Can a person in the'unrestricted environment receive a doisefrom shine rom the plume passing overhead greater than-the immersion dose when theplume]

)ieaches the ground.

Licensee's Response; No. The shine dose from an overhead plume was calculated using an MCNP5 model of the plume as a solid cone source. The cone was sub-divided along its length into 10m segments (frustums) and the plume was modeled as a puff release. The source term was defined as the entire released inventory, which was inserted into a single 1om segment of the plume. After correcting for the LEU-BOL power distribution, the BOL case was more limiting than the MOL case as reported in the LEU Conversion SAR, therefore the revised BOL case was analyzed. Also, the revised temperature distrib4tion integrated release fraction was assumed as discussed in the response to question 42, The dose rates from this source term were calculated at various fixed receptor locations, and then the source term was moved into the next 1Om segment of the plume to simulate the puff cloud moving down-wind. For each calculation, the calculated dose rate, in mrem/hr, was multiplied by the time required for the puff cloud to travel the 10m distance assuming the minimum monthly average wind speed of 3.54m/s. This time of travel for the puff cloud is 2.8s, or 7.8E-4hr for each IOim length of down-wind travel, In this manner, the dose contributions from the passing puff cloud were calculated at each receptor location individuallyfor each 10m distance and then summed together. The farthest receptor location was at 150m, because this corresponds to the distance'of highest ground-level concentration reported on page 190, therefore any exposure beyond this point is a result of immersion in the plume rather than shine from overhead, The puff was modeled out to a distance of 250m, because beyond this distance the dose contribution to the receptor located at 150m was negligible, The puff distance of 250m corresponds toa total exposure time of 71s (assuming 3.54i/s). The radius of the cone at each down-wind distance was determined by the class A vertical dispersion coefficient used in the Gaussian plume model (Equation 13.1.7 on page 189). Class A was chosen because it is the most limiting, since it results in the most rapid expansion of the cone radius bringing the edge of the cloud closer to the ground, therefore Page 40 of 95

UWNR LEU Conversion Responses to Request for Additional Information 0 decreasing the distance from the plume to the receptor and increasing dose. Within each segment of the cone source, the concentration of the plume was uniform. The geometry of the problem is shown in the figure below.

Figure RAI-48-1, Plume Shine Model Calculations of the external dose from the overhead plume do not exceed 0.01 mrem.

This .is well below the maximum dose due to immersion of 0.324 mrem at 148m as reported in section 13.1.7. Even when the shine dose is added to the immersion dose, the maximum combined dose is 0.329 mrem compared to 0.324 mrem reported in section 13.1.7. The combined dose does not exceed the previously reported 0.324 mrem until the down-wind distance approaches 150m, which is approximately the point of maximum ground-level concentration. Therefore, the maximum dose due to immersion, as calculated in the conversion analysis, is more limiting than the shine dose from the plume passing overhead. The total shine dose is given in the table below as a function of receptor distance.

Table RAI-48-1, Plume Shine Doses Receptor Distance (m) Dose (irem) 26 2.9 30 3.1 40 3.4 50 3.6 60 3.8 70 4.0 80 4.1 90 4.3 100 4.4 110 4.5 S 120 130 4.6 4.8 Page 41 of 95

UWNR LEU Conversion Responses to Request for Additional Information 150 _ 5.2 150%

~ c v 0re~jo &_: att Licensee's Response:

No, 8.86 %kIk ýof reactivity is not available to scram the reactor. The shutdown

.margin with no experiments and all control elements fully inserted- was determined to be 5.677 %Ak/k, where the critical bank height was 10. 13in. The power defect from low power to 1.3MW, with a bank height of 11 .73in, was determined to be 1.411 %Ak/k at LEU BOL. Therefore, the shutdown margin at 1,3MW would be 7.088 %Ak/k.

Using the revised value for shutdown margin inserted 2 seconds. after the transient, the revised results of section 13.2 are given below. As can be seen in the LEU BOL plot$

the change is negligible and this trend is similar for HEU, LEU-MOL, and LEU-EOL plots.

Core Power after Blades SCRAM at Zs following 1.4% Ak/k pulse at 1.3 MW (LEU BOL)

-- 7.088% delta k/k reaetivity inserted 2SI  !-ý,---86%delta k/k reactivity inserte d I.E+04 0 to 20 30 40 50 60 70 80 90 100 Time [s]

Figure RAI.49-1, Power After SCRAM LEU-BOL Page 42 Of 95

UWNR LEU Conversion Responses to Request for Additional Information Maximum Hot Rod TO 'mperature after Blades SCRAM at 2s following 1.4% ýk/k Pulse at 1.3 MW (LEU BOL) 12 0 0 " . '*"... . .... . .... ......... .

-- . - .- -- " - -- " - "- -2032 "f-I ll~~tl IL- I. I. ... 2037*.

l0ogo . . . ... . .. ... ......

inserted 800 S-- Safety Analysis Limit 1150%C 600

-8.86 %delta k/k reactivity 103 2 E 400 II lký 200

....... 32 0 10 20 30 40 50 60 70 89 90 1QQ Time [s]

Figure RAI-49-2, Temperature After SCRAM. LEU-BO3L.

Licensee's Response:

Typically, for TRIGA reactors, the rapid addition of reactivity accident is often analyzed as a pulse from full power. However, pulsing from full power requires willful violation of procedure and failure of the pulse mode interlocks and is therefore not considered credible. Therefore, the rapid addition of reactivity accident analysis assumes a failure of an experiment with a total worth of 1.4 %Ak/k. Therefore the total worth of the shutdown margin is available, including the transient rod, at 2 seconds, because the transient rod must comply with TS 3.3.1 when not'in pulse mode, Page 43 of 95

UWNR LEU Conversion Responses to Request for Additional Information Licensee's Response:

From Table, 13.3. 1, the calculated dose rate to the 3d floor non-restricted classroom is significant, but in the event of a loss of coolant accident the building evacuation alarm would alert people to evacuate these classrooms before the core was completely uncovered. In order to estimate the integrated dose received by a

-member of the public during the evacuation, the MCNP5 model. of the unshielded core was modified to include partial water shielding at several time steps. The core gamma source term was also modified to simulate an appropriate level of decay from full power. The integrated dose to the 3"d floor classroom was calculated at various times during the pool water loss and is shown in the figure below.

Figure RAI-51-1, 3 r Floor Integrated Dose During LOCA The pool water would drain to approximately 7.4 ft above the core in at which point it would trip the bridge area radiation monitor, which in turn would automatically initiate the building evacuation alarm. A 5 minute evacuation time from the sounding of the evacuation alarm is assumed. Therefore, the hypothetical member of the public that remains in the 3 'd floor classroom for 5 minutes following the automatic initiation of the building evacuation alarm after start of the LOCA) would receive a dose of during the first and an integrated dose of before evacuating at Realistic doses Page 44 of 95

UWNR LEU Conversion Responses to Request for Additional Information would be far less than this, because the preceding analysis does not take into account time spent in hallways and stairwells (where the dose rate is much lower) during the evacuation. Because the time spent in the high dose rate field in the 3" floor classroom would be far less than 5 minutes, the integrated dose would be substantially lower due to the majority of the dose being received in the final minute as shown in the figure above.

Using the same model, the maximum dose rate at the site boundary, which is the I

area evacuated, was calculated to be immediately after the core is uncovered( after the start of the LOCA). This dose rate would remain for no more than'24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> for a total dose of by which time the emergency procedures would refill the pool with water.

Licensee's Response:

The drain time of represents the time 1.

uel eam.i leel u5,avr~ed rhalfevrdiwae-Licensee's Response:

The calculated dose rates in Table 13.3.1 assume the core is completely uncovered.,

However, in order to determine the impact of the competing effects of increased shielding and increased reflected scatter from the water in the event the core was partially covered, a modified case with the water level at core mid-plane was analyzed and found to haverno statistical difference from the uncovered case.

Licensee's Response:

The wrong initial starting temperature was reported for LEU-BOL. Using the correct axial power shape for LEU-BOL, the starting temperature is 506.1 W9C, end of water transient temperature ,is 75.29PC, and the maximum temperature in hot rod is 652.591C 8,350 seconds after start of transient. The new LEU-BOL complete LOCA curve can be seen below:

Page 45 of 96

UWNR LEU Conversion Responses to Request for Additional Information Maximum Hot Rod Temperature during LOCA Transient (LEU BOL)

Steady State - Water, Draining out of Pool Transient

- Complete Air Cooled Transient - -- Air Cooled Safety Analysis Limit 1 900V 800 ' -... ............ ... .

700 6.) .5.. ......

300 . . . .. ............. 1,132 4032 0831

.500Q 5000 15000. 25000 35000 .45000. 55000 65000 75000 85000

• t..............

. im ~ ]

........... Figui' RA!.54-' Ma-mu~m Hot.od Temperatr during LCA(E BQ). .

P2ge 463f2 Page 46, of 95

UWNR LEU Conversion Responses to Request for Additional Information S5*....Section 13.3.3. Y*ur application states afialysis has been pefi-ieormdthat demehtras complete LOCA is more limiting than a Partial LOCA. Please discuss or provide eferences 33 and 34.17 Licensee's Respontie:

An excerpt from Reference 33 is provided below detailing the LOCA. The mqdel is detailed in sectiens 2.1.3 and 2.2.

2.1.3 LOCA Model The RELAP5 LOCA model uses the same 2-channel model used in the pulsing analysis, with a few exceptions. First, since the power is input manually, the point reactor kinetic equations were not used. Secondly, itwas necessary to split the problem into three' parts:

1. 2-channel steady state model at 1.02 MW, 5.7912m (19 feet) of water above the core, and inlet water temperature of 54.44"C (130°F)

2. 2-channe! transient model where the loss of water is modeled by losing water pressure until 101.3 kPa achieved at to simulate 1 losing 5.7912m (19 feet) of water.,

3. 2-channel transient model where the water coolant has been replaced with air.. It is assumed the inlet air temperature is 25 0C (77-F).

The entire LOCA transient was run for a total of 86,400 seconds or 1 day- in order to ensure the peak fuel temperature was captured during the analysis. In addition, since the decay heat corresponds with the steady state operational history, the pin power, axial, and radial peaking factors are identical to the steady state analysis. The delayed neutron power after the rods drop into the core is not substantial enough to change the decay heat power profile from the steady state power profile.

• .2 EES EES1 was used to analyze both the complete and partial LOCA for the UWNR.

EES allows the user to enter in equations and use the embedded fluid libraries "to calculate specific parameters at a particular set of conditions. This is very helpful so that the user does not need to-constantly interpolate tables to find the specific property for his/her particular problem. By entering in the governing equations for the gravitational pressure gain and the frictional pressure loss, a natural convection loop could be calculated in order to determine the maximum fuel temperature during a LOCA.

2.2.1 Complete LOCA Model In order to perform the complete LOCA analysis by hand, it is necessary to understand the governing equations used in the analysis. The main governing equation is the gravitational buoyancy pressure gain set equal to the frictional pressure loss. The gravitational buoyancy pressure gain is:

Page 47 of 95

UWNR LEU Conversion Responses to Request for Additional Information L~1 APOMP = ?*d,, fl, -,(T,- r.')) +' -.L,,,,_t=,,Pis '9oS. (TI, - .)

Where:

g 9.81 m/s2 node 0.0254 m (15 axial nodes) p11 4 density of the air at the local air temperature fb= local volume expansion coeffilonet of air at the local air temperature Ti." local air temperature T. 298 K Lnon.fue= 0.1905 m (length of fuel above / below active fuel region)

The frictional pressure loss is determined by the following equation:

apr,.

f.L,,,_f,*p,,T+ f, .,.r,,.,, --- + node "fT "

2 ,2 Where:

f 1%!RR Assuming Re < 2000 DH 0,0154703 m (hydraulic diameter of the entire core) v !ocal air velocity "in"I corresponds to inlet core conditions before reaching heated fuel "out" corresponds to outlet core conditions after being heated by fuel K= 2.02 (inlet pressure loss coefficient)

Kout K 1 38 (outlet pressure loss coefficient)

The local Reynolds number is calculated by, where p is the local viscosity of the air:

In order to determine the properties of air, it is necessary to determine the energy. released per node, air temperature per node, and the mass flow rate of the air by the following equations:

Page 48 of 95

UWNR LEU Conversion Responses to Request for Additional Information Where:

q; =. local energy released .per node across the core total mass flow rate of air in the core Cpi = local specific heat at the local air temperature AT, = local change in temperature across the node ppf, local axial power peaking factor S= total core power as a function of time previous local air temperature. When i-1, To 298 K A 0.046598868 m2 is the flow area (with rods) or 0. 138943 to-do(without rods)

After determining these conditions, it is possible to find the mass flow rate-of air and the temperature of the air. In addition to this, it is necessary to calculate the maximum fuel temperature of the rods. This is done by using the Dittus-Boelter equation:

Nu,*= O.023ReP,"8 P 4 Wh ere:

Pri local Prandtl number at the local air temperature h = local heat transfer coefficient kj =, local air thermal conductivity Having determined the heat transfer coefficient, the temperature of the fuel rod can be determined via the following equations:

-qi-

+ -

Where:

q = local heat flux out of the rod N rod 91 (number of rods in the HEU core)

Drod 0.0353894 m (diameter of the rod)

ToutercladI = local temperature of the outer clad Page 49. of 95

UWNR LEU Conversion Responses to Request for Additional Information While there should not be a significant difference between the outer clad temperature and the maximum fuel centerline temperature, additional calculations were performed in order to determine' the maximum fuel centerline temperature. This is necessary because the RELAP5 results for the LOCA used the maximum fuel temperature and not the inner cladding temperature in which fuel cladding interactions would become an issue. Thus, the following equations were constructed:

r,..,fla,, = OV~4 +Zir,,4ao,,q',..... ., .......

Where:

Tgap~iad j local fuel temperature of the gap/clad region TfvelgaI pji local fuel temperature of the fuel/gap region Tfuwemax, local fuel temperature of the fuel centerline regio rr= 0.0179197 .m (radius of the clad) r9 ep = 0.01740154 m (radius of fuel + thickness of gap) tgaqp = 2.54E-06 m (assumed thickness of gap) rfuql = 0.017399 m (radius of fuel) ktuei =18 W/m-K (thermal conductivity of the fuel)

The thermal conductivity of the cladding is determined by the following equation in W/m-K:

kela =OO1466Tvina, -+ 10.84697 The thermal conductivity of the gap is determined by the following equation in W/mýK; k=ap'i 8.58773E - 15 ,( ,.4 3.06727B -1i ,

(r,,w.d.J) + 5.83945.0 -

.(Tqaciad,)2- 9..7506.E - 6. (7'g+peO= - 1.0,1597B - 2 The equations for the thermal conductivity of the stainless steel cladding and the gap were determined using a best fit line to the known data points provided by ANL, The known data points were incorporated into the RELAP5 script as seen in the appendix. Interestingly, the difference in temperature between the maximum fuel centerline and the outer clad is approximately 2+C, Page 50 of 95

UWNR LEU Conversion Responses to Request for Additional Information After entering these equations into EES, it was possible to find the maximum temperature of the fuel rods as a function of time if the initial core power was given as a function of time. For the HEU core, the air-cooled portion of the LOCA power transient can be given by the following power function:

= 2.279'92 B + 05 - (t-fve)-°6s981 To determine the maximum temperature of the hot rod using the hand calculation model, the power function was multiplied by the pin power peaking factor of the hot rod. This gave a first approximation of what the maximum fuel temperature in the hot rod would be.

2.2.2 Partial LOCA Models in conjunction with the total LOCA hand calculation, analysis was also performed looking at the case in which the water does.not completely drain from the core but partially covers the core, Kevin Austin calculated the water would cover the bottom 4.5 inches (11.43 cm) of active fuel and the remaining 10.5 inches (26.67 cm) of active fuel would be air cooled. For simplicity with the 15 axial nodes employed in the model, this analysis assumed the water would cover the bottom 5 inches. It is not anticipated this difference would drastically change the results.

At first, an air-only model was constructed looking at the case in which axial conduction was ignored and air cooling was only supplied. Since the beam ports do not make direct contact with the fuel to provide fresh air, only air coming down the empty slots of the grid box could be used to create a driving flow of air.

The most limiting case was looked at where the area of interest was rods next to blade 3 and the regulating blade. Air flow would come down the two empty grid boxes and then. go across 5 rods before going up. If the mass flow rate were assumed to be the same at the top of each fuel channel, a pressure loss scenario similar to the total LOCA could be created. Since it would have been very cumbersome to create a model based on 10 axial nodes for'10 rods (5 x 2),

it was assumed the rods were heated uniformly axially. The maximum predicted temperature from'this model was well over ary safety limit, but this calculation produced a mass flow rate of air of 2.272xi0" 3 kg/s for 10 rods. This number would be used in subsequent analysis.

After constructing this model, it was then suggested to look at the mass flux of water vapor being provided by the portion of the rods submerged in water. This model, called the air water vapor model, is identical to the complete LOCA model with a few exceptions. First, it was assumed there were no inlet frictional losset since the mass flow rate of steam is not interacting with rods atfthe top of the core. Second, the mass flux of steam was determined by the following equation:

Where:

sub= total power produced by the lower portion of the rods hfg = heat of vaporization of water at 54A44 0C Page 51 of 95

UWNR LEU Conversion Responses to Request for Additional Information Then, the total mass flow rate was determined to be the sum of the mass flow rate of air calculated in the previous' calculation averaged over the core plus the mass flow rate of steam. The humidity of the air was then determined assuming the initial air had a relative humidity of 50% at 298 K. Then the humidity ratio of the air was determined via the following equations 2 :

PO _ Y, pr

'9P P 0 =Yt.t P, T4 P, ., -

Ma Where:

-* relative humidity (50%)

ID= vapor pressure Pg = saturated vapor pressure at the temperature of the air y, mole fraction of water yv=sat saturated mole fraction of water P.~ 101.3 kPa W humidity ratio m, =initial mass flow rate of vapor in the air m,a initial core average mass flow rate of air msteam = calculated core average mass flow rate of steam The remaining difference is that the fluid properties are based off of an air-water mixture and not air-only as calculated in the total LOCA model.

While the air-water vapor model did reduce the maximum temperature calculated, the core average temperature was close to 900C, It was then decided to create a third model, called the axial conduction model, based off of axial conduction as seen in Figure RAI-55-1 with a uniform heat generation along the entire length of the fuel rod, where.L = 0.127 m (5 in).

x-O Figure RA!-55-1, Axial conduction model used for partialLOCA Page 52 of 95

UWNR LEU Conversion Responses to Request for Additional Information The governing equations for this model are derived from El Wakil's nuclear heat transport for a fin.- The general equation for the water portion is:

WT =T-T mw &Aea The particular solution to the second order differential equ'ation is:

Thus the solution is, where A and B are constants to be solved:

Ow =:Ae-"' + 5ec'- +. w, The corresponding solution for the air cooled section of the rod have had all 'WI subscripts replaced with 'a'. Additionally, the constants C and D are to be solved as well:

- c"~+ +

In order to solve the four unknown constants, the following four boundary conditions were used, C

-AdTw] -kA dx1x]

-kA]

Page 53 of 95

UWNR LEU Conversion Responses to Request forAdditional Information Thus four equations and four boundary conditions can be used to solve the four unknowns:

A+ B+.T,, + k v.=C. + D*+ Ta+ ,,,

' mwA+rnwB- *.nC+tmoD .

-m,, t m,,,B = - o,,],=.,

-k (-m,,,wAedk +

+/- ( n 6

-kM(-m mMDeA(2Lmw))

+e2~)1 4191MAxL The constants are defined as:

k= 18 W/m-K L.= 0.127 m Radips- 0.0179197 m Perimeter 2rRadius Area * *.Radiusz The remaining terms, Tw, Ta, hw, and h. are the bulk fluid temperature and heat transfer coefficient in air or water. The properties of the air are determined via an interpolating spheme with the previous model.. In order to model conduction, the slope of the air.temperature line at the air-water inter phase (x=0) is used to determine the heat conducted to the water by the following equation:

• , ~dT~l q-kA -kA(mD - m*C)

This conduction heat is then added to the air-water vapor model as additional heat generation in the water-submerged portion of the rods, The heat lost in the air portion of the rods is subtracted out of the rod power emitted to the air. The air-water vapor model determines the fluid properties, of the air while the axial conduction model determines the maximum fuel temperature of.the rod and -the heat conducted down the rod. Then an interpolation can be performed between the two models to determine the correct fluid properties, heat conduction, and maximum fuel temperature. The only parameters that have not been explicitly incorporated into either model are the temperature of the water and the heat transfer coefficient of the water. By doing a sensitivity study, a tolerance band can be created to see how the maximum temperature is dependent upon these parameters.

With the partial LOCA models created,, it is then possible to see which accident is more limiting, the complete LOCA or the partial LOCA, by cQmparing the maximum fuel temperature of the core average position.

Section 6 from reference 33 is excerpted below, which details the LOCA analysls.

Page 54 of 95

UWNR LEU Conversion Responses to Request for Additional Information 6,0 Loss of Coolant Accident After analyzing a pulse at full power, the other accident considered is a loss of coolant accident (LOCA). A LOCA occurs due to a sheared and open beam port, a very unlikely event. The time in which it took the fuel to be uncovered from the start of the LOCA transient was determined to be by Kevin Austin in the HEU to LEU conversion analysis. Since the beam ports are in the mid-plane of the core, water will only flow out of the core to the bottom of the beam. port leaving 11.43 cm (4.5 in) of active fuel still covered by water.

GA had previously calculated a LOCA using complete air cooling and did not present any analysis for reactors, such as the. UWNR, to have a case in which water is still cooling the bottom third of the active fuel. The reason to be concerned about having the bottom third with water is that a natural circulation loop of air is much harder to form when the water would act as an insulating layer against air trying to go down into the core and up along the fuel rods to cool the fuel rods. . However, the presence of water will allow for axial conduction from the fuel rod to the water and the generation of water vapor that would carry away the heat to portions of the fuel that are air cooled. Thus, the LOCA analysis presented here will look at two different cases. One in which there is a complete LOCA with the other being a partial LOCA where water is still in contact with the bottom third of the active fuel.

Another issue that has been brought to our attention by ANL is the fuel temperature limit. When the rods are water cooled, the Safety Analysis Limit (SAL) fuel temperature limit is 11150C (21000 F). This number is a function of the gap size, hydrogen content in the fuel, and the cladding being at a much lower temperature than the centerline temperature in the fuel. However in air, the cladding temperature is at roughly the same temperature as the fuel centerline, and thus the fuel temperature SAL was determined to be 50QC (1740 0 F) in NUREG-1282. In addition to the analysis presented by TRIGA International,, the methodology for calculating the cladding strength as a function of temperature was performed in GA-9064 4 and the fuel temperature SAL in air of 950*C (1740"F) was determined in the TAMU 1979 SAR.

In order to determine the fuel temperature during both LOCA calculations, it is necessary to make a few appropriate assumptions. It is assumed the reactor is operating at 1.02 MW (1 MW nominal power + 2% uncertainty) for 50 days of continuous.operation. Whi le previous accident analysis, such as the pulse at full power was performed at 1.3 MW, it is unreasonable to believe the reactor operators would operate the reactor beyond the scram set point for 50 days straight, The UWNR has never operated continuously since the UWNR is a research reactor and not a power reactor, and thus continuous operation at 1.02 MW is still a very conservative assumption.

In.addition, the hot rod channel thermal hydraulic parameters were changed from previous analysis to be more physically representative of the actual hot rod channel. Previously, the hot rod channel assumed the channel had a limiting flow area due to the presence of the transient rod and still assumed the transient rod was producing power. In order to make the assumption more accurate, the flow area was changed from 4.7429 cm 2 (0.73516 in2) to 5.0144 cm2 (0.77723 in2) and the hydraulic diameter was changed~from 1.66318 cm (0.65479 in) to Page 55 of 95

UWNR LEU Conversion Responses to Request for Additional Information 1,78143 cm (0.70135 in). This analysis still assumed the four quarter rod segments were powered by a rod with the same pin power peaking factor of D5 SW. The assumptions made for all previous analysis are still.valid so that the predicted maximum fuel temperatures and CHF values are bounding. For the LOCA, the change in assumption was made so that the actual' conditions in the core would be modeled and then the accident scenario would occur.

In the event of massive water loss, the reactor would be shut down after receiving the pool high/low alarm. Kevin Austin has previously calculated it would take at least to uncover the fuel due to the' LOCA. During the first the fuel would still be water cooled and the power of the core would be decreasing from the power of delayed neutron fission and decay heat. At and beyond, it is assumed that all water cooling is lost, and the fuel only has air cooling.

To determine the maximum fuel temperature during the LOCA transient, It is assumed that blades 1, 2, 3, and the transient rod SCRAM into the core 2 seconds after the pool level alarm is activated. The regulating blade is not inserted during a SCRAM and is assumed to not be manually inserted into the core, This causes a prompt drop in power, and the dominant source of power for the transient is delayed'neutron fission power and decay heat. Using an 80 second delayed neutron period, the power from fission after the blades have dropped in can be calculated. At ,the power in the core is -

determined predominately from the decay heat. , Since the decay heat is

]

determined from the previous steady state operation, the axial'and radial power distributions are identical to those used in steady state analysis. In addition, the core power peaking factors are also identical to those used in the steady state analysis.

Since no benchmark can be performed with measured results for the LOCA, a hand calculation or analytic solution was created in order to determine the fuel temperature from first principals. By comparing the RELAP5 results to the analytic solution performed with the assistance of EES, confidence in the model dan be gained.

6.1 Decay Heat during LOCA transient To determine the overall power transient during the LOC.A, the ORIGEN2 data used in the conversion report to determine the radiation levels in an unshielded core was also used to determine decay heat. Since the ORIGEN2 data was only constructed for the hot pin, the decay heat for the hot pin was multiplied by the number of rods in the core and divided by the hot rod power peaking factor at the respective time of core life.

In addition to the decay heat, there is additional fission power that needs to be added for the first part of the transient due to the influence of delayed neutrons.

The prompt negative jump due to the effect of control' blades failing in is:

p Page 56 of 95

UWNR LEU Conversion Responses to Request for Additional Information For HEU BOL these parameters are where Pshutdown is computed by adding the shutdown margin and the worth of blade 3 together:

Po 1.02 MW 0-0,73%

PNiociwn "3,691%4 k/K Prbt~sd0rop~ 172830.7831 W Following the prompt jump, the fission power decreases as a function of time as follows:

For HEU BOL the parameters are:

Paqor 4ýacieý drQp = 172830,7831 W At hutdown 8 is the time since SCRAM in seconds T B8s The summation of the fission power and the decay heat gives the final total core power curve entered into the 2-channel RELAP5/MOD3.3 model, The LOCA power curves did not incorporate the positive reactivity effects of the rods cooling down. While this would increase the fission power, it is not anticipated to effect the total power curves computed for the air cooled transient. The total power curve for HEU BOL is shown The entire total core power transient is shown in Figure RAI-55-2, and the air cooled portion of the transient is shown in Figure RAI-55-3. Further analysis will follow for LEU BOL, MOL, and EOL cases. The core exposure for LEU was 50 MWd for BOL, 800 MWd for MOL, and 1800 MWd for EOL. The input conditions for all stages of core life are shown in Table RAI.-55-1.

Rge 57 of 95

UWNR LEU Conversion Responses to Request for Additional Information Total CorePower During LOCA Transient (HEU BOL)

I.CQF-f07 .......

- Steady State

- Water Cooled Portion of Transient CE-lOG - Air Cooled Portion of Transient 1.00 0 .................. .

.s000 5000 1o000 25000 3SO00 45000 55000 65000 - 75000 85OO0 Time [sl Figure RAI-55-2, Total core power used in LOCA starting at 1.02 MW (HEU.SOL)

Total Core Power during Air cooled portion of LOCA Trans ient (HEU BOLW 200o - ........ .. ..

18000 - .- .- --.--.----

16000, 120001 ----...

--- ~ .... ..... ....... ........... .... ...

1000Q0 .-. . .------- . ....

80000 . ...... .....

12000 6-000 4000ooo I _ _.,. -................ .......

400 0 10000 20000 30000 40000 50000 60000 7000 80000 900M0 Time Is]

Figure RAI-55-3, Total core power used in LOCA during airtransient (HEU 6OLl Page 58 of 95

UWNR LEU Conversion Responses to Request for Additional Information Table RAI-55-1, Input conditions to determine the power profile for the LOCA transient Steady State Power 1.02 MW 1.02 MW 1.02 MW 1.02 MW Infinite Operation Time 50 days 50 days 784.3 days 1764.7 days 130.753  % 0.782% 0,774% 0,7389%

Pshutdown -3.691 %Ak/k -3.593%Ak/k -3,902%Ak/k -*5. 114%Aklk Pafter blades drop 172,830.78 W 182,317.71 W 168,853.09 W 128,769.58 W T 80s .80s 80 805 Power at the start of the air 18,369.86W 18,883.59"W 19,833.55"W 19,719.38W cooled transient, t=836s 18,3_9_8_ W _1_,883.5_ W _!_,833_55 W _____1_.38_

Using the results shown in Table RAI-55-1, the power profile for the LEU core at BOL, MOL, and EOL can also be constructed. These power profiles are then put into the RELAP5 input decks with the same methodology employed for the HEU BOL case. The maximum fuel temperature can then be calculated depending upon whether a total LOCA or partial LOCA, occurs.

6.2 Complete water drainage from core during LOCA transient After inputting this power profile into each component of the transient, the maximum temperature during the transient can be calculated. During the water cooled portion of the transient, the temperature falls from the steady state temperature of 483.00°C (901.406F) to 73.46 0 C (164.23"F) after for the HEU BOL case, At this point in time the water level would reach the top of the fuel, and the remaining water is assumed to vacate the core.

In order to.run the RELAP5 case with air cooling, the initial mass flow rate of air is set to nearly zero (0.0001 kg/s) to simulate the buildup of the natural circulation of air. With this condition, and the decay heat curve as shown by the red line in Figure RAI-55-2, RELAP5 would predict a maximum fuel temperature "to jump by over 800 degrees centigrade in the first iteration. Due to this nrn-physical effect, it was necessary to alter the first few seconds of the power transient to allow RELAP5 the ability to converge on the correct mass flow rate and heat transfer coefficients. By starting at a low power level and increasing the power to the correct power level after 4 seconds, RELAP5 was able to converge on the mass flow rate and give realistic results. For HEU BOL, 18,369.86 W is the calculated decay heat power at the start of the air cooled transient. This power perturbation is not expected to significantly alter the maximum fuel temperature calculated as seen in Figure RAI-55A4.

Page59 of9

UWNR LEU Conversion Responses to Request for Addition! Information Maximum Hot Rod Temperature during LOCA Transient (HEU BOL) 1000 1832 goo~ - ...........

1632 800 80 ------- Steady State U~

600 --- ~~-_

-C omplete Air Cooled Transient C96.89 1232 ,

I1

-- Air Cooled Safety Analysis Limit 83 C, 4004 E- . .........

2 300 .

'00 ...................

---

.. 43 2

0 2

-5000 5000 15000 25000 35C000 45000 55000 65000 75000 85000 Time [s]

Figure RAI-55-4, Temperature profile during LOCA transient(HEU 80Q)

The maximum temperature for HEU BOL during the air cooled portion of the transient was calculated to be 596.890 C (I 106.400 F). This temperature occurs 7,750 seconds from the initiation of the accident. While clearly a LOCA is a significant accident, no damage to the fuel is predicted since the fuel temperature does not exceed the SAL.

While RELAP5 is predicting a particular temperature, it is unclear at first whether the predicted maximum temperature during a LOCA is accurate without having

'another calculation to compare it with. Thus, a hand calculation was performed with the assistance of EES. This analytic solution was constructed using first principals of balancing the .buoyancy -pressure gains due. to the heated air with the frictional pressure drop over the entire core. By determining the mass flow rate, heat transfer coefficients, and air temperature, it was possible to determine the maximum fuel temperature in the core.

The analytic model was designed for the entire core, and thus it is. necessary to compare with the 2nd channel of the. 2-channel model. In addition, since the hand calculation model assumes that the power level is at steady state, the first portion of the air cooled transient where the fuel rod is heating up is not modeled, Thus it was assumed that steady state results would start at around 5000 seconds when the total core power is about 10,000 W. The maximum core channel result comparison between the analytic solution and the RELAP5 2"-'

channel is shown in Figure RAI-55-5.

Page 60 of 95

UWNR LEU Conversion Responses to Request for Additional Information Maximum Averag 'e Core Temp during LOCA Transient (HEU BOL) 1000 . .. .1832 900 .632 800

- Complete Air Cooled Transient I

700 -.. . .. ... ... .

1232 E 600 ...........................

- Hand Caic w/ Dittus 'oeite 1032 500 1 8312 400 4-632 300 432 200

~ 232 10 0 ........ .

-5000 5000 15000 25000 35000 45000 55000 65000 75000 85000 Time [s]]-

Figure RAI-55-5, Core Channeltemperature comparison during LOCA (HEU BOL)

Figure RAI-55-5 shows that the analytic solution and the RELAP5 solution for the core averaged channel give very similar results from about 25,000 seponds onwards, The analytic and RELAP5 solutions diverge during the peak portion of the transient since this is still in the transient region. The maximum difference between these two lines is approximately 45'C around 12,500 seconds. In order to compare with the hot rod temperature, the total core power was artificially increased by the hot rod power peaking factor, 1.6. This produced a maximum temperature profile as shown in Figure RAI-55-6.

Page 61 of 95 L

UWNR LEU Conversion Responses to Request for Additional Information Maximum Hot Ro d Temperature during LOCA Transient (HEU BOL) r1832

-Steac ly State I

900 163?

-- Complete Air Cooled Transient 700

-Hand Calc w/ Dittus Boelter 1232 500 500 8132 300 iI i.-*..**632 qi 200 f---

100 0

-5000 V 5000 15000

-~ .----..--.

25000 , 35000

~

45000 55000 65000

~. . ...

75000

-432 85000 232..

32 Time [si Figure RAI-55-6, Hot channel temperature comparison during LOCA (HEU.80L)

While the core channel analytic solution compared very well with the RELAP5 results, the hot channel analytic solution does not compare as well with the RELAP5 results as seen in Figure RAI-55-6. The maximum difference between these hot channel results is approximately 90°C at 12,000 seconds, or about double the difference of the core channel results. To look at why the results are different, the mass flow rate, heat transfer coefficient, exit air temperature, and exit air velocity for the analytic solution and the RELAP5

,results are shown for the core channel in Figures RAI-55-7 through RAI-55-1 0, Page 62 of 95

UWNR LIEU Conversion Responses to Request for Additional Information Mass Flow Rate of LOCA (HEEU BOL) 0.c 0,

05 .......... ...... .... ...... ........

O 0.

-- "RELAP:

-,Analytic Solution 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 Time [s]

Figure RAI-55. 7, Mass flow rate comparison during LOCA (HEU BOL)

Heat Transfer Coefficient 6.5" frombottom of fuel (HEU BOL) 14 1

~1J 2

II

- i "*Analytic Soluti on 0 1,0o0o 20000 30000 40000 50000 60000 70000 80000 90000 Time.[sW J

Figure RAI-55-8, Heat transfercoefficient comparison during LOCA (HEU BOL)

Page 63 of 95

UWNR LEU Conversion Responses to Request for Additional Information Air Temperature at Core Exit (HEU BOL) 200 I.J QI 1O5Q ..........

E di I.-.

-RMLAP 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 Time Is]

Figure RAI-55-9. Air temperature at core exit during LOCA (HEU BOL)

Air Velocity at Core Exit (HEUBOL) 140;8 ............. . ............

1A - - --

- RELAP

-Analytlc Solution 0 100 oo20000 30000 40000 00o0o. 6000 90000 Time (s]

Figure RAI-55-10, Air velocity comparison at core exit during LOCA (HEU BOL)

What is apparent from these comparison figures is that while the temperature profile of the hot channel matches up very well between the two mode s, they do not compare nearly as well when looking at the fundamental constituents.

There are significant differences between each of the constituent terms, Page 64 of 95

UWNRLEU Conversion: Responses tol Request for Additional Information especially the mass flow rate and the air velocity at the core exit The mass flow rate RELAP5 calculates is nearly half the mass flow rate of the analytic solution, and the air temperature RELAP5 predicts at the top of the core is about 100C higher than the analytic solution. Interestingly, RELAP5 is calculating a heat transfer coefficient ,higher than the analytic solution, making up for the lower mass flow rate and higher air temperature. in addition, a comparison between the axial temperature profile of the analytic solution and RELAP5 is shown in Figure RAI-55-1 1.

Core Channel Axial Temperature Profile Comparison at 3S,000 sec (HEU BOL) 30 0~ .....

' 4 .e..~

4 0 4 a 4 9 9 200.

150 -.-- p I-

• RELA.P

=Analytic Solution S 10 15 20 25 30 35 40 Axial Distance from Bottom of Fuel [cm]

Figure RAI-55-1 1, Core channel axial temperaturecomparison at 35, 000 seconds (HEU BOL)

While the maximum temperature is approximately the same, the axial location shifts between RELAP5 and the analytic solution by 2 nodes (2 inches). This difference is probably coming fromi the difference in the air temperature in the two. models and the latent heat of the fuel rod accounted for in the RELAP5 transient model and not modeled in the analytic solution. Notwithstanding these differences, the full LOCA RELAP5 model has been compared with an analytic

'model, and the RELAP5 model gives reasonable results. Therefore, only the RELAP5 results will be' presented. As shown in Figure RAI-55-4, the maximum temperature calculated by RELAP5 for the HEU BOL during the LOCA transient is 596.89°C (1 106.40°F). Using the same methodology used for the HEU core, the LEU core; LOCA results can be calculated: The most'limiting stage of core life for the. LOCA transient is LEU MOL. The maximum hot rod temperature during the LOCA transient is shown in Figure RAI-55-12.

Page 65 of 95

UWNR LEU Conversion Responses to Request for Additional Information Maximum Hot Rod Temperature during LOCA Transient (LEU MOL) 1900 1637

-Steady State 00 J, 143, 0Complete Air Cooled Transient

-- Air Cooled Safety Analysis Limit 1032 500 S400 1'0 . . . .- .

-0 . .....

. .. ........ .. . . ... . 2

.... ............. 32 00 -----

0 -........... ý .

-5000 .000 15000 25000 35000 450oo 55000 65000 75000 8S000 Time Is]

Figure RAI-55-12, Temperature profile during LOCA transient (LE.U MOL)

The maximum temperature predicted by RELAPS is 695.10°C (1283,18°F).

This temperature is below the SAL by 254.9°C. Many of the very conservative and limiting assumptions could be dropped to recover even more margin. For example, it could be assumed thatit takes to drain the pool as opposed to as currently calculated if the water level was at the true operating level. In addition, a more realistic operational schedule could be used to determine the total power level. Also, if such an accident were to occur, the

,operators also have the option of using the emergency city water pump to slow the rate of water draining from the pool, Furthermore, the most likely scenario for a beam port. rupturing is if a large heavy object were dropped on the beam port from the pool top. Handling of such objects is not anticipated until several hours after reactor shutdown.

Page 66 of 95

UWNR LEU Conversion Responses to Request for Additional Information Toble RA/-5-5-2, Summary of LOCA temperature resufts Startingftep during . 483.000C 498.75C 498.759C 4821 PC steady st atemoperation (901.40F) . (929.75CF) (929.75-F) .(900788F)

Temperature at end of 73.460C 74.87FC 75.730C 74(.81PC water cooled transient (164,23"F) (166.77-F) (168.31"F) (166,6-F)

-Change in temperature 409.54 0 C 423.889C 423.02 0 C 407.900C.

I71 from steady state to (737.17-F) (762.88-F) (761.44-F) (734.220F)

Core Power at 18,369.86 W 18,883.59 W 19,833.55 W 19,719.38 Maximum temperature 596.890 C 648.37 *C 695.100C 879.1 SC in hot rod during LOCA (1106.40-F) (1199.07 'F) (1283,18-F) (1254,57-F)

Time of maximum 7,750 sec 7,775 sec 8,450 sec 9,300 sec temperature in hot rod In summary, all of the LOCA temperature results are shown in Table RAk55-2.

As can be seen, at no time is the predicted temperature greater than fuel temperature safety analysis limit in air, 6,3 , Partial water drainage from core during LOCA transient In conjunction with the complete LOCA, an analysis was conducted on the partial LOCA. As stated in section 2.2.2, three different models were created in order to analyze the partial LOCA. The mass flow rate of air in the air-only model was computed to be 2,272x1 0"3 kg/s for 10 rods or 0.0206752 kg/s for the entire 91 elements in the HEU core, Then, using this mass flow rate, th6 fluid properties of the air were calculated in the air-water vapor model. Water Vapor

,was generated due to the heat of the rodsrsubmerged in water and axial conduction calculated in the axial conduction model. By interpolating between the air water vapor model and the axial conduction model, the maximum fuel temperature can be calculated using the axial conduction model, Since the fluid properties of the water were never modeled explicitly, a sensitivity study was done when performing the analysis. .The first assumption was to assume the water temperature was 100"C, and the heat transfer coefficient of the water was 1,000 W/m2-K. This produced the axial heat generation curve at 5,000;seconds seen in Figure RAI-55-13. This produced a maximum fuel temperature in the average core position to be 417.9 0 C (784.22°F), Further analysis on having the water temperature be either 60 0 C or 100"C and the water heat transfer coefficient being either 500 W/m2-K or 1,000 W/m2-K is shown in Table RAI-55-3. The'temperature of the air was taken to be the highest air temperature calculated in the air-water vapor model.

Page 67 of 95

UWNR LEU Conversion Responses to Request for Additional Information Temperature Profile of Average Rod with uniform heat flux at t 5000s 45 0 ....

350

.. 300 250 ..........

S100 ....... . ....

so .............

0:1s .-0.1 .0.05 0 0.05 0: 0.15 2 0:25 (0 Distance from Water / Air poundarV [m)

Figure RAI-55-13, Axial ternPperatureprofile during partialLOCA (HEU BOL)

Table RA1.55-3, Summary of partialLOCA results without weighted overage h, 4,585 166.6 60 5.105 500 406.3 4,698 1639 60 5.109 1000 397.6.

4,324 172.7 100 5.096 500 1 426.1 4,431' 170.2 100 5.100 1000 1 417.9 When determining the heat transfer coefficient of air with the air-water vapor model, the average was taken for every node except the node directly above the

• water. This node had a very large heat transfer. coefficient and thus was thrown Qut for conservatism. However, if this node was added, and a weighted average on the heat transfer coefficients was performed, the following results Would result as seen in Table RAI-55-4.

/ S Table RAI-5&-4, Summary of partial L OCA results with weighted average ha 4,181 176.15 60 8.592 i 500 351.7 345.4 4,317 172.9 60 8575 I 1,000 3,846 184.3 1 100 8,631 . 500 367.5.

3,972 181. 1_5 100. 8,617 1,000 361.5

-'..'.t: .. ":;.

Page 68 of 95

UWNR IEU Conversion Responses to Request for Additional Information As can be seen in these tables, incorporating the first axial node above the water line makes a large difference in the heat transfer coefficient and also the maximum fuel temperature calculated. For conservatism, no benefit will be assumed for the first axial node. Further refinement of the model could be done to look at the effect of how the power shape would impact the analysis.

However, this would either take incorporating a power shape in the volumetric heat generation term, complicating the solution, or solving an equation for each axial node..

Additionally, it is importantto determine whether the partial or complete LOCA is more, limiting. The maximum fuel temperature of the average rod in the complete LOCA has been calculated to be 431.90C (809.42°F) at 5,000 seconds.

Even with the highest.water temperature and the.lowest heat transfer coefficient, the fuel temperature in the partial LOCA never exceeds this temperature. Thus, it is concluded the complete LOCA is more limiting than the partial LOCA.

In addition to the analysis just at 5,000 seconds,, the partial ,LOCA analysis can.

be performed at other time, intervals to create a transient ýcurve, This; required, iterating at each time step, so fewer time steps were taken than the 1;,000 points created automatically for complete LOCA analysis. The analysis for the partial LOCA used a water temperature of 1000C, and the heat transfer coefficient of the water was 1,000 W/m2 -K throughout the iteration process. The two transient solutions are.presented in Figure RAI-55-14.

Maximum Temperature of core calculated for each LOCA 450 .

-"*Partial LOCA 400 ~

• ,Total LOCA

,* 350 -*-'.-. *.----'-.- . .. . ",

2 SO 0 10000 20000 30000 40000 50000 60000 70000 80000 90000 Time [s]

Figure RA!-55-14, Temperature comparison between total and partial LOCA (HEU 8QL)

This plot shows that the maximum temperature of the core average rod during the partial LOCA is lower than the total LOCA until their intersection point around' 17,500 seconds. After this point the partial LOCA has a higher calculated temperature. By that time step, the peak fuel temperature has already been Page 69 of 95

UWNR LEU Conversion Responses to Request for Additional Information reached as seen in the previous total LOCA analysis. Therefore, the total LOCA is still -expected to be more limiting that the partial LOCA.

References:

1. Klein, S.A, "Engineering Equation Solver Version 7.968", Madison, WI, September 2007.

2. Moran, Michael J and Shapiro, Howard N., Fundamentalsof Engineering Thermodynamics, 51h Edition. John Wiley and Sons: 2004.

3. EI-Wakil, M. M. Nuclear Heat Transport. The American Nuclear Society: La Grange Park, IL, 1993.

4. GA-9064, "Safety Analysis Report for the Torrey Pines TRIGA Mark II Rector." General Atomics, Inc., January, 5, 1970.

In addition to the preceding analysis taken from reference 33 of the analysis report, an independent analysis was conducted by ANL which confirms the conclusion that the complete LOCA is more limiting than the partial LOCA. Applicable excerpts from reference 34 of the analysis report are included below.

5.2 Partial LOCA The centerlines of the four UWNR beam -tubes are aligned with the core mid-plane, which is located at 7.5 inches above the bottom of the fuel, Since the beam tubes are 6 inches in diameter, the lowest initial water level for the partial LOCA analysis is 4.5 inches above the bottom of the fuel. For a fuel rod power of 19.7 kW and a drain time of , the analysis in Appendix B predicted a peak fuel temperature of 5780 C for the partial LOCA, compared to 585 0 C for the complete LOCA. Since this temperature is 3720 C below a maximum fuel temperature in air of 950"C, a LOCA initiated by a failure of one of the UWNR beam ports will not result in failure of the hottest fuel rod.

B.2 COMPUTATIONAL FLUID DYNAMICS (CFD) MODELING A computational fluid dynamics (CFD) model was developed with the STAR-CD code' and used to do the computations. The geometry was kept simple so that the problem would run quickly and be easy to understand while enabling the concept to be tested and easily understood. An axisymmetric wedge of a single rod was analyzed. The coolant channel geometry was assumed to be annular rather than the shape of a cusp between four adjacent rods on a square pitch.

The -channel flow area associated with a single rod in the UWNR was preserved.

The assumed axial power shape in the rod is shown in Figure Bt. The decay heat curves for infinite operation and 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br /> per Week of operation are shown in Figure B2 and are based on Reference 2. The one for infinite operation is based on 10'3 seconds of continuous operation and includes the Gma,(t) factor, as given in Table 13 of the Reference 2, to account for neutron capture in the fission products. The curve for 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br /> of operation does not include this factor and is based on 40 years of operation.

Page 70 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure 83 shows the CFD model geometry, including 16

• the mesh. used in the CFD analysis. The computational 14 volume is a 3Y axisymmetric wedge of a cylindrical region.

In the figure, the wedge is viewed from a skewed angle that makes the fuel rod appear very short and very large in diameter. The total fuel rod length is 30 inches, Some of the key dimensions are provided in Table B1 for 4

the UWNR. The. zirconium rod is 0.25 inches in 2 diameter. The fuel and the upper and lower reflector 0 outer diameters all are

,assumed to be 8.80 mils 0.5 0.6 0.7 0.8 0,9 1 1.1 1.2 1.3 (0.0088 inches) less than Relative Power 1t371 in'ches. 'The clad thickness is 0.020 inches Figure 81. Fuel Rod Axial Power Shape and the fuel rod outer diameter is .1.411 inches, Since it was not practical to represent the very small typical gap thickness between the fuel and the clad of 0,1 mils in the CFD mesh, a larger radial gap of 0.01117 cm (4.40 mils) was used. In order to compensate for the thicker gap in the model, the gas thermal conductivity was increased to 0.699 W/m-K. This resulted in a gap conductance of 6260 W/m 2-K, which is a representative value for a 0.1--mil gap in a TRIGA fuel rod.

0,06

,, 0.05 0

CL 0.04

- 0.03

0 o 0.02 L. 0.01, 0.00 -'

0,001 0.01 01 1 10 100 Time After Scram, hours Figure 82. Decay Heat Power Levels for Infinite Operation and 120 hr/wk Operation Page 71 of 95

UWNR LEU Conversion Responses to Request for Additional Information Model Layout Model Layout with CFD Mesh End Fixture t0

--, .125" Gap Gap

• Water Level 4.inches

' Clad End Fixture Figure 53. CFD Model Geometry (3Y wedge viewed from a skewed angle)

The fuel rod upper and lower end fittings are represented as solid stainless steel cylinders. The turquoise-colored region in Figure 63 represents the flow channel for the steam. It starts at the water.level, which in Figure B3 is 4 inches above the bottom of the fuel, and extends along the exposed lateral fuel rod surface to the top of the rod. This dimension was adjusted as needed, but was fixed within any given, transient solution, Hence, several QFD mesh models similar to the one shown in Figure B3 were used so that the appropriate water level was used in each, analysis. The gas in the 1 /8,th inch gap above the upper reflector was assumed to be, a,.mixture of 86%: xenon and 14% krypton by mole fraction. This gas mixture determines its thermal conductivity. The theý'mal conductivity of this gap is expected to have very little influence on the peak temperatures predicted, Therefore, a very limiting assumption was used. The top surface of the rod portion of the model is assumed to be insulated. The steam enters the channel at 100°C. The outer lateral cylindrical boundary of the steam .is modeled as a "symmetric boundary" in that it is thermally insulated and provides no viscous shear forces to the flowing steam. The channel flow area and channel hydraulic diameter of the UWNR were preserved in this axisymmetric representation. The material properties used in the CFD analysis Page 72 of 95

UWNR LEU Conversion Responses to Request for Additional Information for the fuel, the stainless steel clad and end fittings, the zirconium rod, and the graphite reflectors are listed in Table B2.

Table B1. Model Parameters Maximum Licensed Reactor Power, MW  !.01 Peak Rod Power, kW 1.. .

Rod OD, in. (cm) 1,411 (3.58394)

Rod Arrangement in Limiting Channel Square Pith, i 1K.80 Flow Area per rod, in2 0.77723 Channel OD for OFD model, in. (cm) 1,7204 (4.3851)

Clad Thickness, in, 0.020 Radial Gap Conductance, W/m. -K 6260.

Radial Cap Thickness in CFD Model, cm 0.01117 Fuel Pellet OD in CFD model, cm. 3.71396 Total Rod Length In CFD Model, in. (em) 30.0 (76,2)

Length of Upper End Fitting, in. (erm) 4.387 (11.1430)

Thickness of Upper Gap, in. (cm) 0.125 (0.3175)

Length of Upper Reflector, in. (cm) .3.45 (8.763)

Length of Fuel, in. (cm) 15.0 (38,1)

Length of Lower Reflector, in. (cm) 3.45 (8.763)

Length of Lower End Fitting, in. (cm) 3.588 (9.1135)

Beam Tube ID Outside of Pool, in. 6.0W Beam Tube ID thru Wall, in, 6.0 Beam Tube Centerline below Core Centerline, in,. 0.0 Initial Water Level (0=bottom of fuel), in. 4.5 Time after Scram to reach water level, min.

Effective Pool Surfa. e Area, t&

1 The analysis uses a power level of 1.02 MW to account for a 2% uncertainty in power level meas*urement. ;Peak red power including 2% uncertainty (19.3 x 1.02 19.7 kW).

The exterior surface of the fuel rod that is immersed in the water is assumed to have a 11 0C constant temperature boundary condition. The I1109 C temperature is based on the assumption that the surface temperature is 10"C above the 100 0C water saturation temperature. If the surface were adjacent to a flowing subcooled liquid, then the McAdams, Jens and Lottes, and Thom et al.

correlations would indicate that 10 C above the water saturation temperature would be a reasonably conservative estimate of the rod surface temperature.

Since it is the agitation of the liquid caused by nucleate boiling, rather than. the flowing of the liquid, that keeps the temperature rise relatively small, the 1100 C value is judged to be a reasonable upper bound, although the liquid is essentially stagnant.

Page 73 of95

UWNR LEU Conversion Responses to Request for Additional Information Table B2. Material Properties of Solid Regions in CFD Solutions

' k - 18 W/mnK p= 7150 k9/0~

cp : (132.67 + 0:565 T) J/kg-K, where T [k]

Cl.jl and gnd Pittinao (stainless steel) k = (9.038 it 2.182 x0;2 T - 8.040 x 10! T2 + 2.491 1T0"T3) W/m-k, where T; [k]

P; 8000 kg/m3 cp= (308.3 + 0.7890 T - 8.245 x 10' T2 + 3,345 1O* T) J/kg-k, where T (k]

Zirconium Rod4 k = 20 W/m4-K P = 6500 kg/mr

% = 300 Jfkg-K Refloector (wiaphite) k -= 46 WtrnK p 2000 kq/mý 70 J/kg*K

= 700 The pool water temperature is assumed to be 25"C, typical of the value that would exist during normal operation. Water at this temperature can be thought of as being supplied to a coolant channel that extends the length of the fuel rod and is filled with liquid below the water level and'vapor above, Seventy-five calories are required to raise 1 gram of 25 0 C water 75°C to the 100"C boiling point. An additional 539 calories per gram (2257 kJ/kg) are required to convert the saturated water to saturated steam. Thus, 539 + 75, or 614, calories are required to convert 1 gram of 25 0 C water to saturated steam., An equivalent perspective is that 614 calories/gram (2571 kJ/kg) is needed to increase the specific enthalpy of 25 0C liquid water to that of saturated steam, In the model of a single rod and its associate coolant channel, this heat is supplied by the submerged end of the fuel rod, At each instance in time the heat flux integrated over the I 10VC surface of the submerged end of the rod provides the power that converts liquid to steam. It is assumed that for every 614 calories of energy provided, 1 gram of saturated steam is sent up the rod coolant channel. In the model the coolant channel is assumed. to originate at the surface of the water. Ideally, the STAR-CD code would calculate the power being delivered to the water at each time step and would determine the steam flow rate to be used for the succeeding time step.

However, it appears that the required heat fluxes are available only during post processing after all of the time steps have been solved. Therefore, an iterative approach was used. A guess was made of the steam production as a function of time. Then the results were post-processed to deduce the required steam production function. The deduced function was assumed and the CFD solution Page 74 of 95

UWNR LEU Conversion Responses to Request for Additional Information was repeated. After about two or three Iterations, a converged solution was obtained. The analysis considered only steam flow and includes no air recirculation. The analysis assumes that the top of the tank is completely open so that the superheat steam can freely flow into the large room where the reactor is located.

The initial condition for the rod is that the temperature is I00"C everywhere except on the submerged boundary, which ist 1100C. This is slightly conservative because when the fuel rod is completely submerged all of the temperatures- will be essentially that of the of the pool water, which is assumed to be 250C. Perhaps, a less pessimistic approach would be to assume the pool water temperature for the Initial condition. However, this may be too optimistic since the fuel heats up as the level drops from the top of the fuel to the bottom of.

the beam port, Also, 30'C is a more typical pool water temperature. The 5°1 higher pool temperature will increase the steam production by about a factor of 614/609, since each degree corresponds to I calorie per gram, This represents only a 0.8% increase in steam production and will have only a very small effect on the peak fuel temperature.

The fuel meat heats up very slowly due to its relatively low (decay heat) power and its very large heat capacity (2.46x1 0' J/m3 -K at 100"C), For the volume of the fuel meat in one rod (3.51x 10"4 M3 ), the heat capacity of one rod (evaluated at 100°C) is 863 J/K. Some of the transients are assumed to start at after 1 the scram. Integration of the (infinite) decay heat curve from i.e., the-firsi of the transient, indicates that 9.36 full power seconds of energy are generated in the fuel. For a 20 kW rod, this corresponds to 20,000 W x 9,36 s, or 1.87 x105 J. This energy divided by the fuel meat heat capacity, 863 JIK yields a temperature rise of 2170C.' This implies that,if all of the power generated in fuel meat during, the firstý. of the 1 transient stayed in the fuel meat rather than be transferred away, the fuel meat J temperature on average would rise only 2170C, or less than an average of a half of a degree per second.

8.3 .CFD RESULTS AND DISCUSSION OF CFD RESULTS

.3.1 University of Wisconsin The maximum licensed power for the UWNR is 1.0 MW. The uncertainty in the measured power level is 2%. It is reasonable to assume that the reactor will not be operated for an extended period of time above 1.0 MW (1.02 MW, including the uncertainty). Thus, the shutdown decay heat in this analysis is based on operation at 1.02 MW. The calculated 3 peak rod power of 19.3 kW at 1.0 MW was increased by 2% to 19.7 kW for the UWNR partial LOCA analysis.

The centerlines of the beam tubes are aligned with the core mid-plane, which is located at 7.5 inches above the bottom of the fuel. Since the beam tubes are 6 inches in diameter, the lowest initial water Ievel for the partial LOCA analysis is 4.5 inches above the bottom of the fuel. It is assumed that the water drains down to its initial level after the reactor is scrammed due to low water level, Page 75 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure B4 shows the maximum temperature of each region within the 19.7 kW fuel rod as a function of time. The peak fuel temperature, 558°C, occurs at 4.07 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> after the start of the transient (4.32 hours3.703704e-4 days <br />0.00889 hours <br />5.291005e-5 weeks <br />1.2176e-5 months <br /> after the scram).

The top (dark blue) curve in Figure B5 shows the total decay power history in watts for the 19.7 kW UWNR fuel rod. The solid pink curve indicates how may of those watts at each instance are coming out of the submerged portion of the rod. This was obtained by post processing the CFD results to find the integral of the heat flux over the submerged portion.of the rod. Similarly, the yellow curve represents the watts coming out of the exposed portion of the rod and superheating the steam that is flowing in the channel. The turquoise curve was obtained by subtracting the total the power transferred from the rod, which is the sum of the pink and yellow curves. from the power generated, which is the dark blue curve. This difference must be the power that is stored in the fuel pin. A positive value indicates that the rod is heating up. A negative value indicates that the rod is cooling off. The zero value between the heat-up and the cool-down, which is reached at 4.07 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> into the transient, is the time of the peak temperature.

Uw, 4,", 19.7 kW, 600 O500 4 400 t

• E300 200 0 2 4 6 8 10 12 14 16 Time in Transient (hours)

Figure B4. UWNR Maximum Temperature by Region (4 5-inch water level; 19.7 kW/rod: drain)

The red dashed curve, which coincides with the solid pink curve, is the guessed value (that was used in the CFD analysis) of the amount of heat that is used to generate steam, This should closely match the heat that is transferred from the submerged surface of the rod to the water, the solid pink curve. It took a few iterations in which the pink curve from one iteration became the red dashed curve of'the next before the pink and red curves became essentially coincident, Page 76 of 95

UWNR LEU Conversion Responses to Request for Additional information The ratio of the pink curve to the dark blue curve is shown in Figure B6. Thus, at very early times much of the heat is stored in the rod as the red is heating up, leaving only a relatively small amount (<50%) for steam production. At an hour, more than 60% is going to steam production and after 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> more than 70% is going to steam production. At the time of the peak temperature, 4.07 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />, about 72% of the power is going to steam production. These relatively large percentages are to be expected because 4.5 inches, or 30% of the total length of the fuel meat, is submerged. If only about 60% of the power produced by the exposed170% of the fuel meat length is conducted down the length of the rod to the water, then the portion going to the water would be. about 72% of the total decay power.

4:00 UW, 4.56, 19.7 kW, 3 350-2 300 250t pi

• 200 150 100.

50

-500 0 2 4 6 8 10 12 14 16 Time in Transient (hours)

Figure 85. UWNR Distribution of Decay Power

.(4.5cinch water level; 19.7 kW/rod; ) "3 Figure B7 shows the spatial distribution of temperature throughout the model at 14,640 a (4.07 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />) into the transient (4.22 hours2.546296e-4 days <br />0.00611 hours <br />3.637566e-5 weeks <br />8.371e-6 months <br /> after the scram), which is the time that the peak fuel temperature of 557.5'C is reached. The peak steam temperature of 555,8"C is within 29C of this value. This small difference is to be expected. There is very little radial temperature variation, in the fuel rod and the temperature of the steam that is in contact with the rod should be the same as the rod. The horizontal multicolored stripes in the figure clearly demonstrated that the temperature gradient is predominantly in the axial direction, as is the heat flow.

The portion of the fuel rod. that is below the water level (where the steam inlet in the model cross section occurs) appears in Figure B7 as a solid blue color because the 1100 C boundary condition tends to keepsý this region at the boundary temperature, The minimum temperature shown in the legend, 103.4°C, is less than this value because the steam enters at 100°CC Page 77 of 95

UWNR LEU Conversion Responses to Request for Additional Information The peak temperature in the fuel rod occurs very close to the top of the fuel column, The temperature distribution in the figure along the axial length of the fuel column approximates that of steady-state one-dimensional heat transfer in the axial direction in a solid that has a uniform heat generation rate. Therefore, the rod temperature essentially increases with the square of the distance from the peak rod temperature. This is shown more clearly in Figure B8, where the temperatures along a vertical line at the inner edge of the fuel (outer edge of the zirconium rod) are plotted.for the same time as in Figure B7. Both figures show the increase in the magnitude of the temperature gradient with distance downward from the peak temperature, as is expected in heat-generating solids.

Percent heat into water (UW, 4.5", 19,7 kW, 100%

90%

80% j 70% A 60%

r_4 4"0%

" 40%

30%

20%

.10%

0%.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Time in Transient (hours)

Figure 16. UWNR % of Decay Power to Water (4.5-inph water level; 19.7 kW/rod, 1 The horizontal colored stripes bend upward and look like halves of parabolas in

.the fuel region. This is because the steam; temperature decreag'es from the clad surface to the outer fluid boundary. This decrease in temperature

.becomes obvious' if one follows a thin imaginary horizontal line from the clad to the outer edge of the fluid region. Both the rod and the steam temperatures increase with axial distance from the water level and are nearly the same at the channel exit.

Figure B6, which was developed for the channel surrounding the highest power rod, is assumed to be representative of the entire core. The reduction in water level with time is obtained by integrating the decay power going into the liquid with respect to time. The integration starts at the start of the transient, which is after the scram.. Thus, for any time during the transient, the total energy that was used to convert 25 0 C water to 100 0 C saturated steam can be determined. Since for every 614 calories of energy,. 1 gram of steam is produced. it is a simple matter to determine the mass and the volume of water Page 78 of 95

UWNR LEU Conversion Responses to Request for Additional Information that has been boiled from the pool. The reduction in level is simply this volumo divided by the surface area of the pool in the core region, The surface area of the (empty) UWNR pool is given as 89.13 ft2. The fuel rods and the reflector and other structures in the core region will occupy some of this surface area, Thus, the surface area in the core region is estimated to be 85 ft2. Figure 839 shows the reduction in water level in inches and the decay power history in percent, Thus, the water level goes down 0.25 inches in 4,04 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />, 0,50 inches in 7,98 hours0.00113 days <br />0.0272 hours <br />1.62037e-4 weeks <br />3.7289e-5 months <br />, 0.75 inches in about 14 hours1.62037e-4 days <br />0.00389 hours <br />2.314815e-5 weeks <br />5.327e-6 months <br />, 1,00 inches in 21,95 hours0.0011 days <br />0.0264 hours <br />1.570767e-4 weeks <br />3.61475e-5 months <br />, and 2.00 inches in about 50 hours5.787037e-4 days <br />0.0139 hours <br />8.267196e-5 weeks <br />1.9025e-5 months <br />. The, corresponding decay power percentages can be read from the figureo End Fitting Reflector 557.6 525.2 492.7 460,3 427.8 395.4 Fuel 362.9 330.5 298.1 265.6 233,2

.200.7 168.3 135.8 103.4 Y

Figure ý7. UWNR Temperature Distribution (C) at 4.07 Hours (4.5-inch water level; 19.7 kW/rod, 1)

When the transient analysis is performed with a fixed water level of 4.5 inches, the peak fuel temperature, as shown in Figure B4, is predicted to be 554.7"C and to occur at 4.07 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br />. As Figure B9 indicates, the water level drops by 0.25 inches by about the time the peak is reached. Therefore, the peak fuel temperature could be significantly higher than the predicted 554.7"C. Since at this time the behavior of the system is quasi steady state, a steady-state solution was obtained with the decay power level corresponding to 4.04 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (4.29 hours3.356481e-4 days <br />0.00806 hours <br />4.794974e-5 weeks <br />1.10345e-5 months <br /> after the scram) and with the water level reduced to 4.25 inches. This steady-state solution produced a peak fuel temperature of 577.5 0 C. Additional steady-state analyses corresponding to points later in time along Figure B9 were also considered, As the results shown in Table B3 indicate, later times with lower water levels and decay powers produce lower temperatures.

Page 79 of 95

UWNR LEU Conversion Responses to Request for Additional Information Some may question the notion that for times near or beyond the peak temperature, a steady-state solution can be used in place of a transient one. It was a simple matter to demonstrate the degree of validity of this theory with the aid of additional steady-state solutions. Therefore, in Table B4 the transient solution results at 4.04 seconds and an initial water level of 4.5 inches are compared with its steady-state solution counterpart. The differences between the two sets of results are extremely small.

600 UW, 4.5", 19,7 kW. 4. 07 houre I

500 U 400 2300' E

• -200 1007-0 5 10 15 20 25 30 Axial L.ocation, inches Figure B8. UWNR Axial Temperature Distribution (C) at 4.07 Hours along Vertical Line at Inner Edge of Fuel (4.5-inch water level; 19.7 kW/rod, I

Page. 80 of 95

UWNR LEU Conversion Responses to Request for Additional Information 2

UW, 4.5", 19.7 kW I

1.75 1.5 06

, 1.25

• 0

, 0.75 a) 0,5 0.25 0

0 4 8 12 16 20 24 28 32 36 40 44 48 52' Time in Transient, hours Figure B9. UWNR Decay Power and Level Reduction (4.5minch water level; 19.7 kW/rod, Table B3. Maximum Fuel Temperature for the University of Wisconsin Nuclear Reactor uV-11 UW-2 4.u4 7.98 1 4./oD 400 "to°u 152.7 ,(u VUu 560.4 70.90 01,.o UW-3 21 95 3.50 115.7 72.59 528.4

• Time = 0 is when the water level has just reached the bottom of the beam port, which is assumed to be after the scram. All of these results are based on steady-state analyses.

Table B4. Comparison of Transient and Steady-State Results for the University of Wisconsin Nuclear Reactor UW-4 j Transient 4.04 178,6 I 72.05 1 557,5 I

..UW-5 Steady State 4.04 4.50 1 71.27 1 554.7 I

• Time = 0 is when the water level has just reached the bottom of the beam. port, which is assumed to be after the: scram. I Page 81 of 95

UWNR LEU Conversion Responses to Request for Additional Information References; I. STAR-CDO, CD-adapco, Plymouth, MI, USA.

2. American National Standard Decay Heat Power in Light Water Reactors, ANSI/ANS-5.1-2005, American Nuclear Society, La Grange Park, Illinois, USA, 2005.

3. Safety Analysis Report for the Conversion of the University of Wisconsin

- Madison TRIGA Reactor from HEU to LEU Fuel, University of Wisconsin - Madison, August 2008.

The results of the analyses performed in references 33 and 34 are summarized in the table below. It is evident that in both analyses, the complete LOCA is more limiting than the partial LOCA.

Table RAI-55-5, Summary of Comp!te lts. Partial LOCA Results Complete LOCA Partial LOCA Max.Temp (OC) Max Tempr(n.)

Reference 33 432* 418*

-Reference 34 585 578

• Reference 33 compares the maximum temperature in a core-averaged location instead of the hot-rod.

1. 6. -8eection 14. Please provide replacement TS pages with the proposed haniges-to the]

Licensee's Response; Replacement Technical Specification pages with the proposed changes are attached.

See attachment 8.

. Se*tion 14Fo~Fer each proipsed change tO the TS, provide a justification.I Licensee's Response:

Justifications for each proposed change to the Technical Specifications are attached.

See attachment 8.

Page 82 of 95

UWNR LEU Conversion* Responses to Request for Additional Information 8.pas rfetouetions -30,ian-d:88when res b~ondinto the foll oWiin4uetion]

,*tion .l,2.Y2Y.What.,aqre.thel prqd *,t, pet ues'of the. IFEin th*rexi§ig; o=re o0Sitionswhen the limitinci safetviyssem setting(LS$S): for ,powr iswreached.

Licensee's Response:-

The predicted IFE temperatures in D4 SW and E3 NE at 1.3MW are:

Table RAI-58-1, IFE Ternperatures at 1.3MW IFE Summary Table at 1.3 MW IFE Location 0 0.1 mil gap F° 0.05 mi! gap F=C° 0.15 mil gap

______F__cc __ OF____ OF Bottom 535,23 995.41 476.65 889,96 589.16 1092.48 D4 SW Center 516.39 961.49 460;09 860.15 568.30 1054.93 Top 494.89 922.80 441.24 826.22 544.45 1012.01 Bottom 348.95 660.11 313.48

• 596.26 382.38 720128 E3 NE Center 338.97 642.15 304.89 580.79 371.16 700.08 Top 324.47 616.05 292.42 558.36 354.81 670.66 Note that the revised LEU-BOL axial power shape-was used for D4 SW.

... Seciron,*1,4.2.:?. Tedi*` `si*sn of a~is~forthel fo, t.mp.ertuIrefers

,hee o6a6Z50C margih.,,,t6:thfj11t*ierrsafet lin~if, Provide the anaTyis for t.e

• .Iev6p!pmeht of ihis margin to the sfet AlM Licensee's Response:

The basis in section 14.2.2 is incorrect. The 25 9 C margin is applied to the LSSS and not the safety limit. Section 4.7.6 of the.conversion analysis does not predict that Ifthe IFE thermocouple reaches 400°C then the maximum fuel temperature would be no greater than 1125°C.

The basis for the 25 0C margin, as applied to the LSSS, and not the safety limit, was derived by noting the average maximum difference between two thermocouples within an IFE, as shown in Table 4.7.14, page 116, to be 20.625 0C. Additionally, the uncertainty of the measurement from a thermocouple embedded in an IFE was determined to be :t 3.720C'-2 and the uncertainty of the calibrated fuel temperature safety channel was determined to be +/- 0.3'C 3, for an overall measurement uncertainty of +/- 3.739C. The uncertainty in the measurement was applied to the thermocouple ran9e, to derive 24.3550C using error propagation (square root of sum of squares).

Therefore, if the hottest thermocouple was reading 4250C, then the coldest thermocouple would be reading no less than the LSSS of 4000C.

The analysis of section 4.7.6 applied this margin of 25OC to the LSSS, to determine that any pin with a peaking factor of at least 0.866 will have a maximum thermocouple temperature of 4250C, and from the foregoing analysis, the coldest thermocouple reading of no less than 4000C.

Page 83 of 95

UWNR LEU Conversion Responses to Request for Additional Information The relationship between the LSSS and the safety limit is analyzed in section 4.7.6.

The analysis shows that when D5 SW measures 6789C, a pin with a peaking factor of at least 0,866 will measure 4250C under the following conservative assumptions. First, because the Groeneveld 2006 and Bernath correlations were not developed for use in TRIGA analysis, the more limiting Bernath correlation was used. However, Anderson, et a14 from the University of Wisconsin has proposed to ANL to precisely determine CHF for the three TRIGA fuel assembly types (hexagonal, circular and rectangular). Second, the flow rate is assumed to be constant from the point in which RELAP5/MOD3,3 predicts flow oscillations to occur. Under these assumptions, CHF is predicted to occur in D5 SW at a rod power of 35.6kW BOL, 35.9kW MOL, and 35.5kW EeL. However, assuming the flow rate continues to increase linearly in the extrapolated region of Figure 4.7.43, page 113, a more realistic rod power would be 41 kW to achieve CHF in D5 SW with the Bernath correlation. This provides a 15% margin to DNB, This assumption is consistent with the results predicted in the response to question 23. Finally, the axial power shape of the thermocouple rod used the axial power shape of the cold rod, The cold rod at B3 NE has a smaller axial peaking factor than the hot rod located at D5 SW.

The lower axial peaking factor translates into less power being generated near the mid-plane of the element and therefore lower thermocouple temperatures would be calculated: Any element with a pin power peaking factor of at least 0.87 would have a higher axial peaking factor than the cold rod. A lower predicted thermocouple temperature means the necessary pin power peaking factor must be higher to get to the temperature trip set-point. With these assumptions, the margin to the fuel temperature limit is calculated as 4720C and not 25TC.

Finally, it is important to note that to achieve a rod power of 35.6kW or 41kW in D5 SW would require a core power of 1.8MW and 2.1MW, respectively, which is in excess of the power level safety limit of 1.5MW.

References:

1. Sandia Report SAND2004-1023, April 2004, Uncertainty Analysis of thermocouple measurements used in normal and abnormal thermal environments: experiments at Sandia's Radiant Heat Facility and Lurance Canyon Burn Site. By James T. Nakos.

2. Manual on the Use of Thermocouples in Temperature Measurement, Fourth Edition, ASTM Manual Series: MNL12, Revision of Special Technical Publications (STP) 470B, 1993.

3. Operators Manual, "DP81/DP82 Digital Process Indicators," Omega.

4. Critical Heat Flux in TRIGA Research Reactors, M.Anderson proposal to Argonne National Laboratory.

Page 84 of 95

UWNR LEU Conversion Responses to Request for Additional Information

,tection 1 a SSif l b 00for ttie f 1.6 and owever, s iss n stcti6 .6 of tej AR.,Please address.

Licensee's Response:

The original basis for the LSSS of 400"C was based on the 1973 SAR estimate of peak fuel temperatures at the UWNR from the Torrey Pines TRIGA Mark Ill reactor analysis, despite the fact that these two reactors are geometrically dissimilar.

During the refueling of the UWNR to the TRIGA core, measured temperatures for D4 SW were reported to exceed 4000C at 1MW, as reported in the startup program and included in the HEU 2000 license renewal SAR (page 4-45). Therefore, -the IFE connected to the fuel temperature safety channel was placed in a location that would not exceed 400°C at 1MW, specifically E3 NE. It is fully expected that fuel temperatures in the interior of the core will be greater than 4000C. Since the analysis for the proposed LEU core shows that the central region of the core would exceed 4000C at 1.0MW, the proposed alternate LSSS of 5000C for the central region of the core allows greater flexibility if it is desired to place the IFE closer to the hot rod. The calculation of the relationship between the peaking factors and the LSSS is detailed below.

Currently the technical specification for the IFE allows the IFE to be placed anywhere in the core with FLIP fuel, The trip set point is when a thermocouple in the IFE hits the LSSS of 400 0C providing margin for the fuel temperature limit of 11 500C. However, depending upon which portion of the core the IFE is operating at, it would be necessary for the reactor to operate in excess of the 1.25MW trip set point to even approach 4000C. Ifthe IFE were placed in the coldest location of the reactor (B3 NE) the hot rod location would experience CHF before the IFE would' experience a fuel temperature greater than 4009 C. Therefore itwas necessary to determine a range of core locations that would still protect the core from possible fuel damage.

The power when CHF would occur in the hot rod is calculated with the Bernath correlation to be 35.6307 kW/rod, 35.8807 kW/rod, and 35.4920 kW/rod for LEU-BOL, LEU-MOL, and LEU-EOL respectively using the pseudo-transient results to extend the RELAP5 predicted non-oscillatory flow rate for rod powers up to 29 kW/rod. This correlates to a predicted core power to reach CHF of 1.837 MWj 1.863 MW, and 1.879 MW for LELJ-BOL, LEU-MOL, and LEU-EOL respectively with their respective pin power peaking factors. This number makes the limiting assumption the mass flow rate of water will not increase'as core power increases once the RELAP5 model predicts the flow will oscillate. Further calculations have shown if one assumes thelflow rate continues on the same linear trend as shown by the dashed line in Figure 4.7.38, the power to reach CHF would be approximately 41 kW/rod with the Bernath correlation or a total core power of 2.113 MW as seen in Figure 4.7.40. This gives a limiting assumption of the power to reach CHF by 15%. --All analysis used the predicted core power to reach CHF using the last known RELAP5 calculated flow rate in order to determine the most limiting thermocouple temperature.

Furthermore, Instead of using the hot rod axial and radial power profiles; the cold rod axial and radial power profile was used to calculate the temperatures of the thermocouples. The thermocouples are located 0.3 in (0.762 cm) from the fuel centerline, 6.5, 7.5, and 8.5 in (16.51, 19.05, 21.59 cm) above the bottom of the active Page 85 of 95

UWNR LEU Conversion Responses to Request for Additional Information fuel. The maximum axial peaking factor for B3NE is 1.2943, 1.2428 and 1.2466 for LEU-BOL, LýU-MOL, and LEU-EOL respectively. The maximum radial peaking factors for B3 NE is 1.7154, 1.6414 and 1.602 for LEUMBOL, LEU-MOL, and LEU-EOL respectively. Thus, it is expected that by using the cold rod axial peaking factors for all rods, the maximum predicted thermocouple is bounding, All analysis conducted used a hot gap size of 0.1 mils as previously assumed for the hot rod. Then, using the pin power peaking factors of the core, the maximum thermal couple reading for each rod in the core was calculated. It was noticed during the course of the analysis the predicted maximum thermal couple reading had a very linear shape as a function of the pin power peaking factor. Thus a least squares regression line was found to see what the maximum thermocouple temperature is as a function of the pin power peaking factor. For all cases-of core life this function had an R2 value very close to 1, and thus was deemed to be acceptable for predicting the thermocouple temperature without having to run 83 different cases.

While 400°C is the LSSS for the thermocouple, acceptable thermocouples may not read exactly 400"C at the predicted lower bound of power to reach CHF. ,That does not leave adequate margin, based on analysis detailed in question 59, and it was decided to give. a 25 0 C of further margin to the LSSS of 4000C. With the predicted temperatures, and the acceptance criteria for the thermocouple locations, maps of acceptable IFE locations for all times of core life can be seen in Figures RAI-60-1 to RAI-60-6. The red (M)signifies the predicted temperature is less than 400 (500 for inner core positions), a yellow exclamation (I) signifies the temperature is between 400 and 425 (500 - 525 for inner core positions) and a green check mark (/) signifies the temperature is greater than 425 (525 for inner core positions) at the predicted core power to reach CHF. It is recognized that with burnup the peaking factors in the core will change and therefore acceptable IFE locations will also change. ýTo account for this, Figures RAI-60-7 and RAI60-8 show acceptable IFE. locations for all phases of core burnup.

Page K6 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure RAI-60-1, Locations for IFE where IFE would hit at least 425TC (LEU-BOI1)

Page 87 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure RAI-60-2, Locations for IFE where lFE would hit at least 5250C (LEU-BOL)

Page 88 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure RAI-60-3, Locations for'IFE where IFE would hit at least 425T (LEU-MOL)

Page $9 Qf 95

UWNR LEU Conversion Responses to Request for Additional Information Figure RAI-60-4, Locations for IFE where IFE would hit at least 525"C (LEU.MOL)

Page 90 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure RAI-60-5, Locations for IFE where IFE would hit at least 425"C (LEU-EOL)

,Page 91 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure RAIk60-6, Locations for IFE where IFE would hit at least 525"C (LEU-EOL)

After finding the locations where the IFE would hit at least 425 or 525 for all times of core life, 2 summary figures were constructed to show where the IFE could be placed for all times of core life. If the IFE predicted at least 425 or 525 throughout all times of core life, then the IFE was deemed to be acceptable to be placed there throughout core life as seen in Figures RAI-60-7 and RAI-60-8. The IFE locations will be D4 SW and E3 NE for the LEU core which are in the same location as the HEU core, If an IFE were moved from these positions a 50.59 review would be performed, Page 92 of 95

UWNR LEU Conversion Responses to Request for Additional Information Figure tRAI-60-7, Where FE would hit at least 425°C for all times of LEU core life Figure RAI-60-8, Where IFE would hit at least 525°C for a11 times of LEU core life Page 93 of 95

UWNR LEU Conversion Responses to Request for Additional Information Licensee's Response:

Existing Technical Specification 3.2 limits reactivity insertions to 1A %Ak/k. Sections 4.7.5 and 4.7.10 show that the prompt peak fuel temperature after a 1.4 %Ak/K, pulse is 726.950C (1340.510 F) for LEU-MOL. The response to RAI question 27 shows that the maximum temperature within 15 seconds after the pulse is 8260C, although question 27 clarifies that the 8350C limit only applies to the prompt peak temperature, not the maximum temperature within 15 seconds after the pulse. Therefore, a 1.4 %Aklk reactivity insertion does prevent the reactor from reaching 8300C for all normal conditions of operation at all times in core life and no separate technical specification is needed to prevent exceeding 8300C.

Licensee's Response:

It is agreed that the proposed technical specification 14.3.3,4 is redundant and is not needed. This specification has been removed from the proposed Technical Specifications, See question 56.

Licensee's Response:

Note that the proposed LSSS of 50000 is for an IFE pin power peaking factor of at least ! 116. It is agreed that the function for the fuel element temperature channel is inconsistent with the proposed Technical Specification 2.2. A revision to specification 3.3.3 is included in the proposed Technical Specifications.,. See question 56.

Licensee's Response:

It is agreed that the bases for TS 5.6 are inconsistent with the conversion safety analysis report. Revised bases are included in the proposed Technical Specifications, See question 56, Page 94 of 95

UWNR LEU Conversion Responses to Request for Additional Information

':B.

rs at ethA~ a~ 16 F Pa76 'M1s1 ienn4 n~ ei poss

~~eeiV~~~~~~, ~neto oera on, R

~ ve* ssus i et, itný n e an Sh, h e,11 s Eari 011 Mii a e'al a m Irodue d b. eainertefei be ala -P_35ataq r t Oe n I Licensee'* Response:.

The possession limit of 15 kg U-235 is based on for . which is 14.55 kg, This number is rounded up to allow for some pins which may exceed as built. Furthermore, it is agreed to restate the license conditions as proposed to make them clearer to, understand, and to remove the exempt status table.

I Page 95 of 95

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 1 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 2 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 3 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 4 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 5 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 6 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

UNIVERSITY OF WISCONSIN NUCLEAR REACTOR LICENSE NO. R-74 DOCKET NO. 50-156 RESPONSE TO RAI REGARDING HEU/LEU CONVERSION ATTACHMENT 7 - REDACTED SECURITY-RELATED INFORMATION REMOVED REDACTED TEXT AND FIGURES BLACKED OUT OR DENOTED BY BRACKETS

LEU Conversion Changes to Technical Specifications: Item-By-Item Justification Note: Text from the Technical Specifications appearsin fixed-width Courier New font.

Table of Contents' The following entries are deleted:

1.18 Standard Core ........ 4 1.19 Mixed Core ........... 4 1.20 Flip Core ............ 4 Furthermore, the fiollowing entry is added:

1.10 LEU 30/20 Core ....... 4' Numbering of later entries is left unchanged. See changes to Technical Specifications Page 4 for further justification.

TABLE OF CONTENTS 1.0 DEFINITIONS 1,.1 Reactor Shutdown . . .. . . . . .... . . .. .

1.2 Reactor Secured . . . . .. . .... . 1.

1.3 Reactor Operation .... ............ ........ 1 1.4 Cold Critical . . 1 I..

1.5 Steady State Mode . ........... ... 1 1.6 Pulse and Square Wave Modes . .... ..... 1 1.7 Shutdown Margin ................. ............ 2 1.8 Abnormal Occurrence ..... 22...........

1.9 Experiment ............. .................. 2 1.10 Experimental Facilities . ..... ............ 3 1,11 Shim-Safety Blade . . ... ... ..... 3 1.12 Transient Rod ... . . . .......... ... 3 1.13 Regulating Blade ................... .. ..... 3 1.14 Fuel Element. ..................... ....... 3 1.15 Fuel Bundle .... ....... ...... .... . . ... 3 1.16 Core Lattice Position ...... ........... 3 1.17 Instrumented Element................... ... 4 1.18 LEU 30/20 Core ..... ................. . . 4 1.21 Operational Core .................. 4 1.22 Safety Limits. ......... .

.... .... 4 1.23 Limiting Safety System Settings- . .4....4 1.24 Operable.. .............. ....... ...... 5 1.25 Reactor Safety Systems ........ 5.......5 I.26 . Experiment Safety Systems ....... ....... . 5 1.27 Measured Value. 5 1.28 Measured Channel ..... ................ 5 1.29 Safety Channel................ 5 1.30 Channel Check . . ............... 5

,1.31 Channel Test . .5.....................5 1.32 Channel Calibration ............... 6 2.0 SAFETY LIMITS AND LIMITING SAFETY SYSTEM SETTINGS 2.1 Safety Limits . . . .... . . . ..... 7 2.2 Limiting Safety System Setting ..... ....... 8 3.0 LIMITING CONDITIONS FOR OPERATION 3.1 Reactivity Limitations ....... ............ i0 3.2 Pulse Mode Operation ......... .............. 10 3,3 Control and Safety System .!. ................ 11 3.3.1 Scram Time. . .i. ............................. 11 Amendment No. 17

Technical Specifications ,Pag 3 TS 1.14, which says:

A fuel element is a single TRIGA fuel rod of either standard or FLIP type Is changed as follows:

A fuel element is a single TRIGA fuel rod of.ethev r-- **r ndardl

• type To read:

A fuel element is a single TRIGA fuell rod of LEU 30/20 type Justification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore the definition of "Fuel Element" is revised to add LEU 30/20 fuel type and remove standard and FLIP types.

- TS 3 110 EX!PERIMEVTAL FACILITIES Experimental facilities shall mean beam ports, including extension tubes with shields, thermal columns with shields, vertical.tubes, through tubes, in-core irradiation baskets, irradiation cell, pneumatic transfer systems and in-pool irradiation-facilities.

REACTOR COMPONENTS

1. 11 SHIM-SAFETY BLADE A shim-safety blade is a control .blade having an electric motor drive and scram capabilities. It may have a fueled follower section.

1.12 TRaNSi-NT ROQ The tran'sient rod is a control rod with, scram capabilities that can be rapidly ejected from the reactor core to produce a pulse.

It may have a voided follower.

1.13 REGULATING BLADE The regulating blade is a low worth control blade that need not have scram capability and may have a fueled follower. Its position may be varied manually or by the servo-controller.

1.114. 'UEL_ ELEDAENT A fuel element is a single TRIGA fuel rod of LEU 30/20 type.

.15FUEL BUNDLE.

A fuel bundle is. a cluster of three or four fuel elements secured in a square array by a top handle and a bottom grid plate adaptor.

1.1.6 COBE LATTICE POSITO The core lattice position' is thatregion-in the core (approximately 3" by 3") over a grid plug hole. It may be.1 occupied by a fuel bundle, an experiment or experimental facility, or a reflector element.

Amendment No. 17

Technical Specifications Page 4 TS 1.18, 1.19, and 1.20, which say:

1,18 STANDARD CORE A standard core is an arrangement of standard TRIGA fuel in the reactor grid plate.

1.19 MIXED CORE A mixed core is an arrangement of standard TRIGA fuel elements with at least 35 TRIGA-FLIP fuel-elements located in a central region of the coreQ 1..20 FLIP CORE A FLIP core is an arrangement of TRIGA-FLIP fuel in the reactor grid plate.

Are deleted. Furthermore, the following definition is added:

1,18 LEU 30/20 CORE A LEU 30/20 core is an arrangement of TRIGA LEU 30/20 fuel in the reactor grid plate.

Justification; The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore, the

.. Mixed Core," and "Flip Core" are deleted because definitions of "Standard Core,"

following conversion to LEU 30/20 fuel there is only one valid operational core.

Similarly, a LEU. 30/20 core is defined to be an arrangement of TRIGA LEU 30/20 fuel in the reactor grid plate.

Technical Specifications Page 4, Continued TS 1.21, which says:

An operational core may be. a standard core, mixed core, or FI41P core for which the core parameters of shutdown margin, fuel temperature, power calibration, and maximum allowable reactivity insertion have been determined to satisfy the requirements of the Technical Specifications.

Is changed as follows; An operational core may be a st.ndar-d . r., 4 A r.. FP ee-e ** for which the core parameters of shutdown margin, fuel temperature, power calibration, and maximum allowable reactivity insertion have been determined to satisfy the requirements of the Technical Specifications.

To read:

An operational core is an LEU 30/20 core for which the core parameters of shutdown margin, fuel temperature, power calibration, and maximum allowable reactivity, insertion have been determined to satisfy the requirements of the Technical Specifications.

Justification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore, the definition of"Operational Core" is changed to identify only an LEU 30/20 core as being a valid operational core.

- TS 4 1,17 INSTRUMENTED ELEMENT An instrumented element is a special fuel element in which sheathed chromel-alumel or equivalent thermocouple is embedded in the fuel near the-horizontal center plane of the fuel element at a point approximately 0.3 inch from the center of the fuel body.

1.18 LEI 30/20 CORE A LEU 30/20 core is an arrangement of TRIGA LEU 30/20 fuel in the reactor grid plate.

1.21 OPERATIONL CORE An operational core is an LEU 30/20 core for which the core parameters of shutdown margin, fuel .temperature, power calibration, and maximum allowable reactivity insertion have been determined to satisfy the requirements of the Technical Specifications.

REACTOR INSTRUMENTATION 1.22- SAFETY LIMIT Safety limits are limits on important process variables which are found to be necessary to reasonably protect the integrity of certain of the physical barriers which guard against the uncontrolled release of radioactivity.

1.23 LIMITING SAFETY SYSTEM SETTINGS Limiting safety system settings are settings for automatic protective devices related to those variables having significant safety functions.

Amendment No. 17

Technical Specifications Page 7 TS 2.1, which says:

a. The temperature in a TRIGA-FLIP fuel element shall not exceed 11501C under any conditions of operation.

b. The temperature of a standard TRIGA fuel element shall not exceed 1000°C under any conditions of operation.

c. The reactor power level shall not exceed 1500 kW under any conditions of operation.

Is changed as follows:

a. The temperature in a TR!GA-4+1.-! fuel element shall not exceed 1150'C under any conditions of operation.

6. The teffPeoroturV of atn~r r-IC elemetr~ hl-ful Se~eeed 100 0 uno oycodti of oporatione.

ej. The e reactor power level shall not exceed 1500kW under any conditions of operation.

To read:

a. The temperature in a TRIGA LEU 30/20 fuel element shall not exceed 11501C under any conditions of operation, b, The steady-state reactor power level shall not exceed

.1500kW under any conditions of operation.

Justification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore, the fuel temperature safety limit for LEU 30/20 fuel is added, and the fuel temperature safety limits for FLIP and standard fuels are removed because after the c6nversion only LEU

.30/20 fuel is used. The outline numbering for the power level safety limit is updated, and clarified to be steady-state power since TIIGA LEU 30/20 fuel is designed for pulse operations.

- Ts 7 -

2.0 SAFETY LIMITS ND LIMITING SAFETY SySTeM SETTINGS 2.1 Saety Limits Apolicability This specification applies to fuel element temperature and steady-state reactor power level.

Objective The objective is to define the maximum fuel element temperature and reactor power level that can be permitted with confidence that nO fuel element cladding failure will result.

SRecifiction.

a. The temperatur6 in a TRIGA LEU 30/20 fuel element shall not exceed 1150'C under any conditions of operation.

b? The steady-state reactor power level shall not exceed 1500 kW under any conditions of operation.

Bas-es A loss of integrity of the fuel element cladding could arise from a buildup of excessive pressure between the fuel moderator and the cladding'if the fuel temperature exceeds the safety limit.

The pressure is caused by air, fission produce gases, and hydrogen from dissociation of the fuel moderator. The magnitude of this pressure is determined by the fuel moderator temperature and the ratio of hydrogen to zirconium in the alloy, Amendment No. 17

Technical Specifications Page 8 TS 2.1 bases, which say; It has been shown by experience that operation of TRIGA reactors at a power level of 1500 kw will not result in damage to the fuel. Several reactors of this type have operated successfully for several years at power levels up to 1,500 kW. It has been shown by analysis and by measurements on other TRIGA reactors that a power level of 1500 kW-corresponds. to a peak fuel temperature of approximately 500C, Thus a Safety Limit on power level of 1500 kW provides an ample margin-of safety for operation.

Are changed as fbllows:

It has been shown by experience that operation of TRIGA reactors at a power level of 1500 kW will not result in damage to the fuel. Several reactors of this type have operated successfully for several years at power levels up to 1500 kW. -he-*e-Hi ifflý, -e ether- TRICA feaetcr.

S, ..... that a power level of 1500 kW corresponds to a peak fuel temperature of approximately &9O-ý-M . Thus a Safety Limit on power level of 1500 kW provides an ample margin .of safety for operation.

To read:

It has been *hown by-experience that operation of TRIGA reactors at a power level of 1500 kW will not, result in damage to the fuel. Several reactors of this type have operated succes sfully for several years at power levels up to 1500 kW. The LEU Conversion SAR section 4.7.8 shows by analysis that a power level.

of 1500 kW corresponds to a peak fuel temperature of 6659C. Thus a Safety Limit on power level of 1500 kW provides an ample margin of safety for operation.

justification:

Calculations performed as part of the conversion analysis show a peak fuel temperature of 665'C at 1.5MW. These calculations are based on the proposed specific TRIGA LEU 30/20 core design at the University of Wisconsin and are not based on analyses of other TRIGA reactors. The basis for Technical Specification 2.1 is therefore updated to reference this calculation in the LEU Conversion SAR.

Technical Specifications Page 8. Continued TS 2.2(1), which says:

The limiting safety system setting for -fuel temperature shall be 4000C '(750 0 F) as measured in an instrumented fuel element. For a mixed core, the instrumented element shall be located in the region of the core containing FLIP type elements.

Is changed as follows:

The limiting safety system setting for fuel temperature shall be 4000C +47&2 as measured in an instrumented fuel element

-ftt-aallbe eeaed n he-r-~e-q-in of the eore cer-n-tý F13iP type ... tuer.t-e To read:

(The limiting safety system setting for fuel temperature shall be A4000C as measured in an instrumented fuel element with a pin power peaking factor between 0.87 and 1.16, or 500*C as measured in an instrumented fuel element with a pin power peaking factor of at least 1.16.

Furthermore, the outline headings for TS 2.2(1) and TS 2.2(2) are rewritten as 2,2(a) and 2.2(b) to be consistent with the rest of the Technical Specifications.

Justification:

Calculations perftrmed, as part of the conversion analysis show that it is possible that an IFE located in a core position with a pin power peaking factor of less than 0.87 will not protect the fuel temperature safety limit from being reached at the LSSS of 400'C. The technical specification is modified to impose this limit. Additionally the analysis for the proposed LEU core shows that the central region of the core would exceed 400"C at 1.0MW. The proposed altemrateLSSS of 500'C forthe central region of the core allows greater flexibility if it is desired to place the IFE closer to the hot rod. However, the pin power peaking factor at the core location of the IFE must be at least 1.1 6 to prevent the fuel temperature safety limit from being reached at the LSSS of 500'C.

Additionally, the technical specification references to mixed cores, which are no longer

.approved for use, are removed.

Technical Specifications Page 8, Continued (continuinB on page TS-9)

TS 2.2(1) bases, whirh say:

The limiting safety system setting is a temperature which, if exceeded, shall cause a reactor scram to be initiated preventing the safety limit from being exceeded. A setting of 4000C provides a safety margin of 7500C for FLIP type fuel elements and a margin of 600'C for standard TRIGA fuel elements. A part of the safety margin is used to account for the difference between the true and measured temperatures resulting from the actual location of the thermocouple. If the thermocouple element is located in the hottest position in the core, the difference between the true and Measured temperatures will be only a few degrees since the thermocouple junction is at the mid-plane of the element and close.to the anticipated hot spot. If the thermocouple element is located in a region of lower temperature, such as on the periphery of the core, the measured temperature will differ by a greater amount from that actually occurring at the core hot spot. Calculations indicate that, for this case, the true temperature at the hottest location in the core. will differ from the measured temperature by no more than a factor of two. Thus, when the temperature in the thermocouple elements reaches the trip setting of 4001C, the true temperature at the hottest location would be no greater than 8000C providing a margin to the safety limit of at least 200'C for standard fuel elements and 3500C for FLIP type.elements. These margins are ample to account for the remaining uncertainty in the accuracy of the fuel temperature measurement channel arid any overshoot in reactor power resulting from a reactor transient during steady state mode operation. For.a mixed core (i.e., one containing both standard and FLIP type elements), the requirement that the instrumented element'be. located in the FLIP region of the core provides an even greater margin of safety since the peak to average power ratio wi.thin that region will; be smaller than over an entire core composed of elements of the same type.

in the pulse mode of operation, the same limiting, safety system setting will apply. However, the temperature channel will have no effect on limiting the peak powers generated because of its relatively long time constant (seconds) as compared with the width of the pulse (milliseconds). In this mode, however. the.

temperature trip will'act to reduce the amount of energy generated in the entire pulse transient by cutting of the' "tail" of the energy transient in the event the pulse rod remains stuck in the fully withdrawn position.

I

Technical Specifications. Page 8, Continued (continuing on page TS=9)

Are replaced to read ( 2 nd paragraph does not change):

The limiting safety system setting is .a temperature which, it, exceeded, shall cause a reactor scram to be initiated preventing the safety limit from being exceeded. Analyses performed in section 4.7.6 of the LEU Conversion Analysis show that with-the IFE in a core location with a pin power peaking factor of at least 0.87, the maximum fuel temperature would be no greater than 678 00 C if the IFE thermocouple reaches 40000 providing a margin of 472 C to the safety limit. The same analyses also show that with the IFE inl a core location with a pin power peaking factor of at

least 1.16, the maximum fuel temperature would be no greater than 6780C if the IFE thermocouple reaches 500'C providing a margin of 4720C to the safety limit.

In the pulse mode of operation, the same limiting safety system setting will apply.. However, the temperature channel will have no effect on limiting the peak powers generated because of its relatively long time constant (seconds) as compared with the width of the pulse (milliseconds). In- this mode, however, the temperature trip will act to reduce the amount of energy generated in the entire pulse transient by cutting off the "tail" of the energy transient in the event the pulse rod remains stuck in the fully withdrawn position.

Furthenrmore, the TS 2.2(l) bases are rewritten as TS 2.2(a) bases to be consistent with the rest of the Technical Specifications.

Jumlification ,

Calculations performed as part of the conversion analysis show that it is possible that an IFE located in a core position with a pin power peaking factor of less than 0.87 will not.

protect the fuel temperature safety limit from being reached at the LSSS of 400'C. The technical specification is modified to impose this limit. Additionally the analysis for the proposed LEU core shows that the central region of the core would exceed 400'C at 1.0MW. The proposed alternate LSSS of 500'C for the central region of the core allows greater flexibility if it is desired to place the IFE closer to the hot rod. However, the pin power peaking factor at the core location of the IFE must be at least 1.16 to prevent the fuel temperature safety limit from being reached at the LSSS of 5001C, Additionally, the technical specification references to mixed cores, which are no longer approved for use, are removed.

Technical Stoecificationrs Pane 9 TS 2.2(2) bases, which say:

Calculations and measurements for similar TRIGA rea ctors indicate at 1.25MW, the peak fuel temperature in the core will be approximately 400 0 C so that the limiting power level setting provides an ample safety margin to accommodate errors in power level measurement and anticipated operational transients.

Are changed as follows:

m--s e-E--

-i, 4-Ji 1a r-T A re 9f.9 _4' -(,a

.*,the peak-fuel temperature in the core will be approximately 4-G-G M so that the limiting power level setting provides an ample safety margin to accommodate errors in power level measurement and anticipated operational transients.

To read:

Analysis in section 4.7 of the Conversion Analysis SARshows that at 1.3 MW, the peak fuel temperature in the core will be approximately 604 0 C so that the limiting power level setting provides an ample safety margin to accommodate errors in power level measurement and anticipated operational transients.

Furthermore, the TS 2.2(2) bases are rewritten as TS 2.2(b) to be consistent with the rest of the Teclhical Specifications, Justification:

Calculations performed as part of the conversion analysis show a peak. fuel temperature of 6041C at 1.3MW, providing margin to the departure of nucleate boiling and additional protection of the fuel temperature safety limit at the reactor power LSSS. These calculations are based on the proposed specific, TRIGA LEU 30/20 core design at the University of Wisconsin and are not based on calculations and measurements of similar TRIGA reactors. Therefore, the calculation of the maximum fuel temperature at the power level LSSS is updated,

- TS 8 -

The safety limit for the TRIGA-FLIP fuel element is based on data which indicate that the stress in the cladding due to hydrogen pressure from the dissociation of zirconium hydride will remain below the ultimate stress provided the temperature does not exceed 11500C and the fuel cladding is water cooled (pages 3-1 to 3-23 of GA-9064).

The safety limit for the standard TRIGA fuel is based on data including the large amount of experimental evidenceobtained during high performance reactor tests. of this fuel. These data indicate that the stress in the'cladding (due to hydrogen pressure from the dissociation of zirconium hydride) will remain below the. ultimate stress provided that the temperature of the fuel does not exceed 1000'C and the fuel cladding is water cooled (GA-9064, pages 3-1 tO 3-23),

It has been shown by experience that, operation of TRIGA reactors at a power level of 1500'kW will not resglt in damage to the fuel.. Several reactors of this type have operated successfully for several years at power levels up to 1500kW. The LEU Conversion SAR section 4.7,8 shows by analysis that a power level of 1500 kW corresponds'to a peak fuel temperature of 6650C. Thus a Safety Limit on power level of 1500 kW provides an ample margin of safety for operation.

2.2 LIMITING SAFETY SYSTEM SETTING Aoplicabilitv This specification applies to the scram setting which prevents the safety limit from being reached.

• biecti-Ve The objective is to prevent the safety limits from being reached.

Specifications

a. The limiting safety system setting for fuel temperature shall be 4Q00C as measured in an instrumented fuel element with a pin power peaking factor between 0.87. and 1.16, or 5000C as 'measured in an instrumented fuel element' with a pin, power peaking factor of at least 1.16.

b. The limiting safety-system setting for reactor power level shall be 1.25 MW.'

Amendment No. 17

- TS' 9 -

a. The limiting safety system setting is a temperature which, if exceeded, shall cause a reactor scram to be initiated preventing the safety limit from being exceeded. Analyses performed in section 4.7.6 of the LEU Conversion Analysis show that with the IFE in a core location with a pin power peaking factor of at least 0.87, the maximum fuel temperature would be no greater than 678"C if the. IFE thermocouple reaches 400 0 C providing a margin of 472 0 C to the safety limit. The same analyses also show that with the' IFE in a core location with a pin power peaking factor of at least 1.16, the maximum fuel temperature would be no greater than 678VC if the IFE thermocouple reaches 300"C providing a margin of 472'C to the safety limit.

In the pulse mode of operation, the same limiting safety system setting will apply. However, the temperature channel will have no effect on limiting the peak powers generated because of its. relatively long time constant (seconds) as compared with the width of the pulse (milliseconds). In this mode, however, the temperature trip will act to reduce the amount of energy generated in the entire pulse transient by cutting of the "tail" of the energy transient in the event the pulse rod remains stuck in the fully withdrawn position.

b. Analysis in section 4.7 of the Conversion Analysis SAR shows that at 1.3 MW, the peak fuel temperature in the core will be approximately 6040C so that the limiting 'power level setting provides an ample safety margin to accommodate errors in power level measurement and anticipated operational transients.

Amendment No. 17

Tcehnical Specifications Page I1 TS 3.2.bases, which say:

Measurements performed on the Puerto Rico Nuclear Center TRIGA-FLIP reactor indicated that a pulse- insertion of reactivity of 1.4 %Ak/k resulted in a maximum temperature rise of approximately 400 0 C, With an ambient water temperature of approximately 100 0 C, the maxiimum fuel temperature would be approximately .5000C resulting in a safety margin of 500 0 C for standard fuel and 650*C for FLIP type fuel. These margins allow amply for uncertainties due to the accuracy of measurement or location of the instrumented fuel element or due to the extrapolation of data from the PRNC reactor.

Are replaced in their entirety to read:

The LEU Conversion SAR section 4.7.10.shows by analysis that a 114. %Ak/k limitation on pulse reactivity will result in a maximum fuel temperature of 790 0 C. This leaves a margin to the 1150"C Safety Limit of 360'C, and a margin of 40'C to the 830 0 C operational limit'recommended by General Atomics, "Pulsing Temperature Limit for TRIGA LEU Fuel," GA-C26017 (December, 2007).

Justification:

Calculations performed as part.of the conversion analysis show a peak fuel temperature of 7271C following a 1.4 %Ak/k pulse reactivity insertion. These calculations are based on the proposed specific TRIGA LEU 30/20 core design at the University of Wisconsin and are not based on measurements of the Puerto Rico Nuclear Center reactor,

- TS 11 -

Spe cif ica-tion The reactivity to be inserted for pulse operation shall be determined and mechanically limited such that the reactivity insertion will not exceed 1.4% A k/k.

The LEU Conversion $AR section 4.7.10 shows by analysis that a 1,4% Ak/k limitation on pulse reactivity will result in a maximum fuel temperature of 79Q0C. This leaves a margin to the 1l50QC Safety Limit of 360'C, and a margin of 40°C to the 830"C operational limit recommended by General Atomics, "Pulsing Temperature Limit for TRIGALEU Fuel," GA-C26017 (December, 2007).

3.3 CONTRQLJ AND SAFETY SYSTEM 3.3.1 Scr ie This specification applies to the tinge required for the scrammable control elements to be fully inserted from the instant that a safety channel variable reaches the Safety System Setting.

Ob1e c tive The objective is to achieve prompt shutdown of. the reactor to prevent fuel damage.

The scram time measured from the instant a simulated signal reaches the value of the LSSS to the instant that the slowest scramnable control element reaches its fully inserted position shall not exceed 2 seconds.

This specification assures. that the reactor will be promptly shut down when a scram signal is initiated. Experience and analysis have ,indica'ted that for the range of transients anticipated for a TRIGA reactor, the .specified scram time is adequate to asssure the safety of the reactor.

Amendment No. 17

Technical Specifications Pag_133_

TS 3.33(a), which says (headings reproduced for clarity):

Safety System Or Minimum No. Function & operating Measuring Channel O2erable Mode in'Which Required

a. Fuel Element 1 Scram at 400 0 C. All modes.

Temperature Ischanged as follows:

Safety.System Or Minimum No. Function & Operating Measuring Channel Operable Mode in Which Req uired

a. Fuel Element 1 Scram at 4000c Temperature To read:

Safety $ystem Or Minimum No. Function & Operating Measuring Channel Operable Mode in Which Required 1 .Scram at 4009C for IFE

a. Fuel Element Temperature peaking-factors 0,87-1.16 or 500 0 C for IFE peaking factors >1.16, All modes.

Justification:

Calculations performed as part of the conversion analysis show that it is possible that an IFE located in a core position with a pin power peaking factor of less than 0.87 will not protect the fuel temperature safety limit from being reached at the LSSS. The technical specification is modified to impose this limit. Additionally the analysis for the proposed LEU core shows that the central region of the core would exceed 400'C at 1.0MW. The proposed alternate LSSS of 500°C for the central region of the core allows greater flexibility if it is desired to place the WFE closer to the hot rod. The pin power peaking factor requirements for the 1FE ensure that the reactor will scram before reaching the fuel temperature safety limit. This revision is in response to RAI question 63.

Technical Specifications Page 13, Continued The following is added to the end of Table I of TS 3.3.3 (headings reproduced for.

clarity):

Safety $ystem Or Minimum No. Function & Operating Measuring Channel Operable Mode in Which Reqguired j, Reactor Pool-water 1 Scram if water temperature is Temperature greater than 130 0 F; All modes.

Justification:

Calculations performed as part of the conversion analysis are baised on a maximum core inlet temperature of 130TF as originally assumed in the SAR, In order to remnair within the design basis of this analysis, a new technical specification 3.3.30) is added to limit pool water temperature. The limit already exists as an administrative limit, but it is now added to the Technical Specifications in response to RAI question 15.

- TS !3 -

Safety System Or Minimum No. Function & Operating Measurin. Channel. Opegable Mode i Which Reauired a, Fuel Element Scram at- 400*C for IFE peaking Temperature factors 0.87-1.16 or 5000C for IFE peaking factors 51.16. All modes.

b. Reactor Power Level Scram at 125% of full licensed power level; Square Wave &

Steady State Modes.

c. Manual Pushbutton Scram; All modes.

d. Reactor Pool-water Scram if water-level is less Level than 19 feet above top of core; All modes.

e. Log N Prevent firing transient rod when drive is not full in and power level is above 1 kW in all modes.

f. Log Count Rate Prevent control element with-drawal when neutron count rate is less than.2 per second; All modes,

g. Preset Timer Transient rod scram 15 seconds or less after pulse; Pulse mode.

h. Nigh Voltage Monitor Scram on loss of high voltage supply to neutron. and gamma ray power level instrumenta-tion detectors; All modes.

i. Pulse Mode Control, Prevents withdrawal of control Blade' Withdrawal blades while in pulse mode.

Interlock.

j. Reactor Pool-water Scram if water temperature is Temperature greater than 130OF; All modes.

Amendment No. 17

Techmical Speeifications Page 14 The following is added to the end of TS 3.3.3 bases:

The thermal-hydraulic ,analysis in the SAR assumes a pool water temperature of 130'F. If the temperature exceeds 130OF then the scram will prevent continued operation in an un-analyzed condition.

Justification:

Calculations performed as part of the conversion analysis are based on a maximum core inlet temperature of 1301F as originally assumed in the SAR. In Order to remain within the design basis of this analysis, a new technical specification 3.3.3j) is added to limit pool water temperature. The limit already exists as an administrative limit, bu't it is now added to the Technical Specifications in response to RAI question 15. The bases are added for the new technical specification 3.3.30) to explain that the specification will prevent operation in an un-analyzed condition, since the LEU Conversion SAR assumes a pool water temperature of 130'F.

- TS 14 -

The fuel temperature scram provides the protection to assure that if a condition occurs in which the limiting safety system setting is exceeded, an immediate shutdown will occur to keep the fuel temperature below the safety limit.

The reactor power level scrams are proVided in steady state and square wave modes as added protection. against abnormally high fuel

.temperatures and to assure that reactor operation stays within the licensed limits.

The manual scram allows the operator *ameans of rapid. shutdown in the event of unsafe or abnormal conditions.

The reactor pool water level scram assures shutdown of the reactor in the event of a serious leak in the primary system or pool.

The Log N interlock prevents firing of the transient rod at power levels above 1.0 kW if the transient rod drive is not in the full down position. This effectively prevents inadvertent pulses which might cause fuel temperature to exceed the safety limit on fuel temperature.

The Log N interlock does not allow: control element withdrawal unless the neutron count rate is high enough to assure proper instrument response during reactor startup.

The preset timer assures reduction of reactor power to a low level after a pulse.

The high voltage monitor prevents operation of the reactor with other systems inoperable due to failure of the detector high voltage supplies.

The pulse mode cbntrol blade withdrawal interlock prevents reactivity addition in pulse mode other than by firing the transient rod.

The thermal-hydraulic analysis in the SAR assumes a pool water temperature of 130'F. If the temperature exceeds 130"F then the scram will prevent continued operation in an un-analyzed condition.

Amendment No. 17

Technical Specifications Page 26 (continuing on page TS-27' TS 5.1, which says:

a. TRIGA-FLIP Fuel The individual. unirradiated FLIP fuel elements shall have the following characteristics:

(1) Uranium content: maximum of 9 Wt-% enriched to nominal 70% Uranium 235.

(2) Hydrogen-to-zirconium atom ratio (in the ZrHx):

nominal 1.6 H atoms to 1.0 Zr atoms.

(3) Natural erbium content (homogeneously distributed):

nominal 1.5 Wt-%.

(4) .. Cladding: 30.4 stainless steel, nominal 0.020 inch thick -k (5) Identification: Top pieces of FLIP elements Will have characteristic markings to allow visual identification of FLIP elements employed in mixed cores.

b. Standard TRIGA fuel The individual unirradiated standard TRIGA fuel elements shall have the following characteristics:

(1) Uranium content: maximum of 9.0 Wt-% enriched to a nominal 20% Uranium 235.

?) Hydrogen-to-zirconium atom ratio (in the ZrHx) nominal 1.7 H atoms to 1.0 Zr atoms.

(3) Cladding: 304 stainless steel, nominal 0.020 inch thick.

Is changed as follows:

a. T-ICA FIp Fu ,o+\I The individual unirradiated Frl-P f Uel elements shall have the following characteristics:

(1) Uranium content; maximum of 4 H Wt-'% enriched to

• nominal

• Urnu 235.

(2) Hydrogen-to-zirconium atom ratio (in the ZrH,):

nominal 1.6 H atoms to l.Q Zr atoms (3) Natural erbium content (homogeneously distributed);

nominal 4-.4 W Wt-%.

(4) Cladding: 304 stainless steel', nominal 0.02.0 inch thick.

earae--- ies'-&fe--ma* --t- -110w viuo. ido.ti.itotio.

(h) ei-d'iifl 1hi '.;lihr,,.,orootti 1 unir !odiotod or i-rTlCaeftu otooziar of-~imtnot oiumo . t nihdt e-~lTIP4te7

Technical Specifications Page 26 (continuing on page TS-27)

+2+-.-----HyEdrecjef te etreentuin atem fatic (i n thc rII+-e R{~-G~emdinaz: 304 T

-ef*a13 Mtidzzzol .2 rc To read:

a. TRIGA LEU 30/20 Fuel The individual unirradiated TRIGA LEU 30/20 fuel elements shall have the following characteristics:

(1) Uranium content: maximum of 30 Wt-% enriched to rnaximum of 19.95 Wt-% with nominal enrichment of 19.75 Wt-% Uranium 235.

'(2) Hydrogen-to-zirconium atom ratio (in the ZrH,):

nominal 1..6 H atoms to 1.0 Zr atoms with a maximum Hi to Zr ratio of 1.65.

(3) Natural erbium content (homogeneously distributed) nominal 0.9 Wt-%.

(4) Cladding; 304 stainless steel, nominal 0.020,inch thick.

Justification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore, the design features of Standard and FLIP fuel are removed and replaced with the design features of LEU 30/20 fuel since it is the only type of fuel used after the conversion, NUREG-1282 documents the LEU 30/20 fuel design features.

0

- TS 26 -

W 5.0 DEPI1 FEAURES 5.1 REACTOR FUEL Applicability This specification applies to the fuel elements used in the reactor core.

The objective is to assure that the fuel elements are of such a design ano -fabricated in such a manner as to permit their use with a high degree of reliability with respect to their physical and nuclear characteristics.

Soecifica'ions

a. TRIGA LEU 30/20 Fuel

• The individual unirradiated TRIGA LEU 30/20 fuel elements shall have the following characteristics:

(1) Uranium content: maximum of 30 Wt-% enriched to maximum of 19.95 Wt-% with nominal enrichment of 1.9.75 Wt-.% Uranium 235.

(2) Hydrogen-to-zirconium atom ratio, (in the ZrHF) : -nominal 1.6 H atoms to 1.0 Zr atoms with a maximum H to Zr ratio of 1,65.

(3) Natural erbium content (homogeneously distributed), nominal

'0.9 Wt-%.

(4) Cladding: 304 stainless steel, nominal 0.020 inch thick.

Amendment No. 17

Technical Specifications Pagg 27 TS 5,1 bases, which say:

a. A maximum uranium content of 9 Wt-% in a TRIGA-FLIP element is about 6% greater than the design Value of 8.5 Wt-%.

Such an increase in loading would result in an increase in power density of about 2%. Similarly, a minimum erbium content of 1.1% in an element is about 30% less than the design value. This variation would result in an increase in power density of only about 6%. An increase in local power density of 6% reduces the safety margin by at most ten percent. The maximum hydrogen-to-zirconium ration of 1.65 could result in a maximum stress under accident conditions in the fuel, element clad about a factor of two greater than the value resulting from a hydrogen-to-zirconium ration of 1.60. However, this increase in the clad stress during an accident would not exceed the rupture strength of the clad.

When standard and FLIP fuel elements are used in mixed cores, visual identification of types of elements is necessary to verify correct fuel loadings. The accidental rotation of fuel bundles containing standard and'FLIP elements can be detected by visual inspection. Should this occur, however,, studies of a single FLIP element accidentally rotated into a standard fuel'region indicate an insubstantial increase in power generation in the FLIP element.

b. A maximum uranium content of 9 Wt-% in a standard TRIGA element is about 6% greater than the design value of 8.5 Wt-%. Such an increase in loading would result in an increase in power density of less than 6%. An increase in local power density of 6% reduces the safety margin by at most 10%. The maximum hydrogen-to-zirconium ratio of 1.8 will produce a maximum pressure within the clad during an accident well below the rupture strength of the clad.

Are replaced in their entirety to read:

The fuel specification permits a maximum uranium enrichment of 19.95%. This is. about 1% greater than the design value for 19.75% enrichment. Such an increase in loading would result in an increase in power density of less than 1%. An increase in local power density of '% reduces the safety margin by less than 2% (Texas A&M LEU Conversion SAR, December 2005).

Technical Secifications Page 27. Continued The fuel specification for a single fuel element permits a minimum erbium content of about 5.6% less than the design value of 0.90 Wt-%. (However, the quantity of erbium in the full core must not deviate from the design value by more than -3.3%).> This variation for a single fuel element~would result in' an increase in fuel element power density of about 1-2%. Such a small increase in local power density would reduce the safety margin by less than 2% (Texas A&M LEU Conversion SAR, December 2005).

The maximum hydrogen-to-zirconium ratio of 1.65 could result in a maximum stress under accident conditions in the fuel element clad about a factor of two greater than the value resulting from a hydrogen-to-zorconium ratio of 1.60. However, this increase in the clad stress during an accident would not exceed the rupture strength of the clad (M.T. Simnad, "The U-ZrH, Alloy: its Properties and Use in TRIGA Fuel," General Atomics Report E-117-833, February, 1980).

Justification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore, the bases of TS 5.1 for the design features of the fuel are revised for the new LEU 30/20 fuel. The effects of the uranium and erbium design limits have already been estimated at Texas A&M using LEU 30/20 fuel. The effects of the hydrogen-t6-zirconium design limit has been reported by General Atomics in GA report E-1 17-833.

- TS 27 -

The fuel specification permits a maximum uranium enrichment of 19.95%.

This is about I% greater than the design value for 19,75% enrichment.

Such an increase in loading would result in an increase in power density of less than 1%. An increase in local power density of 1%

reduces the safety margin by less than 2% (Texas A&M LEU Conversion SAR, December 2005).

The fuel specification for a single fuel element permits a minimum erbium content of about 5.6% less than the design value of 0.90 Wt-%.

(However, the quantity of erbium in the full core must not deviate from the design value by more than -3.3%). This variation for a single fuel element would result in an increase in fuel element power density of about 1-2%. Such a small increase in local power density would reduce the safety margin by less than 2% (Texas A&M LEU Conversion SAR, December 2005).

The maximum hydrogen-to-zirconium ratio of 1.65 could, result in a maximum stress under accident conditions in the fuel element clad about a factor of two greater than the value resulting from a hydrogen-to-zirconium ratio of 1.60. However, this increase in the clad stress during an accident would not exceed the rupture strength of the clad (MT. Simnad, "The U-ZrH, Alloy: Its Properties and Use in TRIGA Fuel,"

General Atomics Report E-117-833, February, 1980).

.5.2 REACTOR CORE Acpolicability This specification applies to the configuration of fuel and in-core experiments.

Amendment No. 1'7

Technical Specifications Page 28 TS 5.2(a), which says;

a. The core shall be an arrangement of. TRIGA uranium-zirconium hydride fuel-moderator bundles positioned in the reactor grid plate, j

Is changed as follows:

a. The core shall be an arrangement of TRIGA M R uranium-zirconium hydride fuel-moderator bundles positioned in the reactor grid plate.

To read:

a. The core shall be an arrangement of TRIGA LEU 30/20 uranium-zirconium hydride fuel-moderator bundles positioned in the reactor grid plate.

Justification:

The only type of fuel approved for use is.TRIGA LEU 30/20 type. Therefore, the core arrangement is clarified to be exclusively LEU 30/20 fuel to preclude any operation with other TRIGA fuel.

Technicl Specifications Page 28, Contnnued TS 5.2(b), which says:

b. The Triga core assembly may be standard, FLIP, or a combination, thereof .(mixed core) provided that any FLIP fuel be comprised-of at least thirty-five (35) fuel elements, located in a contiguous, central region, And TS 5.2(b) bases, which say:

b. In mixed cores, it is necessary, to arrange FLIP elements in a contiguous, central region of the core to dontrol flux peaking and power generation peak values in individual elements.

Are deleted.

Justification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. Therefore, thedesign specification for mixed cores is removed because only cores using LEU 30/20 fuel are used after conversion.

Technical Specific.flons Page 28, Continued TS 5.2(a) bases, which say:

Standard TRIGA cores have been in use for years and their characteristics are well documented, The Puerto Rico Nuclear Center and the Gulf Mark ITI all-FLIP cores have operated and their characteristics are available. Gulf has also performed a series of experiments using .standard and Flip fuel in mixed cores and a mixed core has been used successfully in the Texas A&M.

University TRIGA reactor. In addition, studies performed'at Wisconsin for a variety of mixed core arrangements indicate that such cores with mixed loadings would safely satisfy all operational requirements (SAR Chapters 4 and 6).

Are changed as follows:

a TRIGA cores have been in use for. years and their characteristics are well documented The Purt.... Nu-ea Ge-te

.... r* Co t..orr nth-4e-Gu4Ma 4a].-F-~-FP-ii eer hve eeopora*td and their kL Es -1 characteristics are available. Gulf has alz-s perfored a ani-e-ef e-xp*rifrfints using .tandarfi andFlip4 fuel

"- min eei eio and a fflixed-eeEe--hasbe--ax-sý5eeee&9 4y ll-t-h T- a s, A -- e4-t-TRICA focotar. In addition, studies TR performed at Wisconsin 4-ej- a var...t. . . .. r indicate that safely satisfy all operational requirements *-'pr To read:

TRIGA cores have been in use for years and their characteristics are well documented. LEU cores including 30/20 fuel have also been operated at General Atomics and Texas A&M and their successful operational characteristics are available. In addition, the analysis performed at Wisconsin indicates that the LEU 30/20 core will safely satisfy all operational requirements.

See chapters 4 and 13 of the LEU Conversion Analysis SAR.

Jlustification:

The only type of fuel approved for use is TRIGA LEU 30/20 type. The bases are updated to reference other current facilities successfully operating with LEU 30/20 fuel, and to refer to detailed calculations in the LEU Conversion Analysis SAR.

Technical Specifications PaRq2& Cotinued TS 5.2(c and d), which say:

c. The reactor shall not be operated with a core lattice position vacant except for positions on the periphery of the core assembly.

d. The reflector, excluding experiments and experimental facilities, shall be water or a combination of graphite and water.

Are changed as follows:

ea. The reactor shall not be operated with a core lattice position vacant except for positions on the periphery of the core assembly.

d. The reflector, excluding experiments and experimental facilities, shall be water or a combination of graphite and water.

To read:

b. The reactor shall not be operated with a core lattice position vacant except for position. on the periphery of the core assembly.

c. The reflector, excluding experiments and. experimental facilities, shall be water or a combination of graphite and water.

Justification:

The outline numbering is revised because a previous entry referring to mixed cores, which are no longer approved, was deleted. No wording is changed.

Technical Specifications Page 28. Continued (continuing on page Ts-29 TS 5.2(c and d) bases, which say:

c. Vacant core lattice positions will contain experiments or an experimental facility to prevent accidental fuel additions to the reactor core. They will be permitted only on the periphery of the core to. prevent power perturbations in regions of high power density.

d. The core will be assembled in.the reactor grid plate which i.s located in a pool of light water. Water in combination with graphite reflectors can be used for neutron economy and the enhancement of experimental facility radiation requirements.

Are changed as follows:

ep. Vacant core lattice positions will contain experiments or an experimental facility to prevent accidental fuel additions to the reactor core. They will be permitted only on the periphery of the core to prevent power perturbati-ons in regions of high power density,

e. The core will be assembled in the reactor grid plate which is located in a pool of light water. Water in combination with graphite reflectors can be used for neutron economy and the enhancement of experimental facility radiation requirements.

To read:

b. Vacant core lattice positions will'contain experiments or an experimental facility to prevent accidental fuel additions to the reactor core. They will be permitted only on the periphery of the core to. prevent power perturbations in regions of high power density.

c, The core'will be assembled in the reactor grid plate which is located in a pool of light water. Water in comDination with graphite reflectors can be used for neutron economy and the ýenhancement of experimental facility radiation requirements.

Justification:

The outline numbering is revised because a previous entry referring to mixed cores, which are no longer approved, was deleted, No wording is changed.

- TS 28 -

0b~ ective The objective is to assure that provisions are made to restrict the arrangement of fuel elements and experiments so as to provide assurance that excessive power densities will not be produced.,

S0ecificatiqas

a. The core shall be an arrangement of TRIGA LEU 30/20 uranium-zirconium hydride fuel-moderator bundles positioned in the reactor grid plate.

b. The reactor shall not be operated with a core lattice position vacant except for positions on the periphery of the core assembly.

c. The reflector, excluding experiments and experimental facilities, shall be water or a combination of graphite and water.

Base*

a. TRIGA cores have been in use for years and their-characteristics are well documented. LEU cores including 30/20 fuel have also.

been operated at General Atomics and Texas A&M and their successful operational characteristics are available. In addition, the analysis performed at Wisconsin indicates that the LEU 30/20 core will safely satisfy all operational requirements.

See chapters 4 and 13 of'the LEU Conversion Analysis SAR.

b. Vacant core lattice positions will contain experiments or an experimental facility to prevent accidental fuel additions to the reactor core. They will be permitted only on the periphery of the core to prevent power perturbations, in regions of high power density.

Amenebnent No. 17

- TS 29 -

c. The core will be assembled in the reactor grid plate which is located in a pool of light water. Water in combination with graphite reflectors can be used for neutron economy and the enhancement of experimental facility radiation requirements.

5.3 Control Elements Applicability These specifications apply to the control blades and transient control rod.

Obj ective The objective is to assure that control elements are fabricated to reliably perform their intended control and safety function.

a. The safety blades shall be constructed of boral plate and shall have scram capability.

b. The regulating blade shall be constructed of stainiess steel.

c. The transient rod shall contain borated graphite or boron and its compounds in a solid,.form as a poison in an aluminum or stainless steel clad. The transient control rod shall have scram capability and may incorporate an aluminum or air follower.

The boral safety blades and stainless steel regulating blade used in the reactor have been shown to'p.rovide adequate reactivity worth, structural rigidity, and reliability to as-sure reliable operation and long life under operating. conditions. 'The transient control rod materials and fabrication techniques have been used in many TRIGA.

reactors and have demonstratedreliable operation and long life.

5.4 Radiation Monitoring Svstems Aplicicb ility These specifications describe the functional performance and essential components of the radiation monitoring systems.

The objective is to describe those systems which provide information on radiation levels and effluent radioactivity.

specificatiQns

a. The area radiation monitoring system shall provide gamma radiation level information at the control console for at least three locations in the Laboratory. It shall cause an alarm at the control console and initiation of an evacuation alarm if high radiation levels occur and prompt remedial action is not taken, Amendment No. 17

Technical Specifications Page 31 TS 5.6(b), which says-All air Or other gas exhausted from the reactor room and associated experimental facilities shall be released to the environment a minimum of 17 meters above ground level.

Is changed as follows:

All air or other gas exhausted from the reactor room and associated experimental facilities shall be released to the environment a minimum of -4 meters above ground level.

To read:

All air or other gas exhausted from the reactor room and associated experimental facilities shall be released to the environment a'minimum of 30.5 meters above ground level.

Justification:

Calculations performed in the conversion analysis are based on a minimum stack exhaust height of 30.5m above ground level. In order to remain within the design basis, the specifications for ventilation stack height are revised to reflect current stack design and to be consistent with the methodology of calculations in the LEU Conversion SAR.

Technical Specifications Page 31. Continued TS 5.6 bases, which say:

Calculations in Chapter 6 of the Safety Analysis Report show that, exposure of occupants of the Laboratory can be kept below 10 CFR part 20 limits for occupational exposure under accident conditions if the room volume is 2,000. m3. Calculations in Chapter 6 of the SAR based on release of radioactive effluent at ground level show that concentrations of radioactive materials are within limits of 10 CFR Part 20 for non-restricted areas during the accidenits considered. Further calculations based on release at the stack height show a further reduction by a factor of 10 due to operation of the ventilation system and release of effluent at a height.qf 17m.

Are changed as follows:

Calculations in Chapter -6 of the Safoty Annly - *,pcr l.. 10.. C R part

.. 20 limito... exposure unde-a 4ent zondiýit a . ,u

-hee+/- ~ vAa ae e.e '14~

if the ree volume i 2,000 m. aleulatiens in Chapter 6 of the SAR boased n releas -f- rordicootive ef-fluent-at groun- -,vl----ow that 20 icna ....... mati 4oo...5octivo afe witinii of-et -0fR t 2G-0er--non otitdoe-

,dur-ing the aeeidenitsoonis.Elr-d. Fufthe3r caletulatione boaee4 on aý e ur-, e-r- Y a cm-C'Erw

ý'4 i A -- 1 -. 9 er Tv mue ta epefa.--.. -e- A A A In 4-To read:

Calculations in Chapter 13 of the SAR demonstrate that the occupational doses in the event of the maximum hypothetical accident do not exceed limits if the lab volume is at least 2000m 3 . Furthermore, calculations in chapter 13 that assume operation of the ventilation system assume a stack height of 30.5m.

Justification; Calculations performed in the conversion analysis are based on a minimum stack exhaust height of 30.5m above ground level, In order to remain within the design basis, the specifications for ventilation stack height are revised to reflect current stack design and to be consistent with the methodology of calculations in the conversion analysis. The bases are modified to reference the methodology of calculations in the LEU Conversion SAR.

TS 31 5.6 Reactor BuIlding These specifications apply to the room housing the-reactor and the ventilation system controlling that room.

Qbh-tctive The objective is to provide restrictions on release of airborne radioactive materials to the environs, Specifications

a. The reactor shall be housed in a closed room designed to restrict leakage. the minimum free volume shall be 2,000 cubic meters.

b. All air or other gas exhausted from the reactor room and associated experimental facilities shall be released to the environment a minimum of 30.5 meters above ground level, Baýses Calculations in Chapter 13 of the SAR demonstrate that the occupational doses in the event of the maximum hypothetical accident do not exceed limits if the lab volume is at least 2000m3 . Furthermore, calculations in chapter 13 that assume operation of the ventiiation system assume a stack height of 30.5m.

57REACOR POOL WATER SYSTEM Applicability This specification applies to the pool containing the reactor and to the cooling of the core by the pool water.

01?j ejcttive The objective is to assure that coolant water shall be available to provide adequate cooling of the reactor core and adequate radiation shielding.

Amendment No. 1"7

Technical Specifications Pagc 32 The following is added to the end of TS 5.7;

f. A pool water temperature alarm shall indicate if water temperature reaches 130"F.

Furthermore, the following is added to the end of TS 5.7 bases:

f. The thermal-hydraulic analysis in the $AR assumes a pool water temperature of 130'F. If the temperature exceeds 130%F then the alarm will prevent continued operation in an un-analyzed condition.

Justification:

Calculations perfbrmed as part of the conversion analysis are based on a inaximum core inlet temperature of 130 0 F. In order to remain within the design basis, a new specification 5.7(1-) for pool water temperature is added in response to RAI question 15.

Technical Specifications Page 32. Continued TS 5.7 bases, which say:

a. This specification is based on thermal and hydraulic calculations which show that the TRIGA-FLIP core can operate in a safe manner at power levels up to 2,700 kW with natural convection flow of the coolant water. A comparison of operation of the TRIGA-FLIP and standard TRIGA Mark III has shown operation to be safe for the above power level. Thermal and hydraulic characteristics of mixed cores are essentially the same as that for TRIGA-FLIP and standard cores.

Are replaced in their entirety to read:

a. The LEU Conversion SAR section 4.7.8 shows by analysis that the natural convective cooling of the reactor core is sufficient to maintain the fuel in a safe condition up to at least a power level of 1500 kW (the power Safety Limit).

1Justification:

Calculations performed as part of the conversion analysis show a peak fuel, temperature of 665'C at 1.5MW under natural circulation conditions. These calculations are based on the proposed specific TRIGA LEU 30/20 core design at the University of Wisconsin and are not based on calculations and measurements of similar TRIGA reactors. Therefore, the bases for natural convection cooling are revised to refer to current calculations for the LEU core in the LEU Conversion SAR.

- TS 32 -

W Sjpediications

a. The reactor core shall be cooled by natural convective water flow.

b. The pool water inlet and outlet pipe to the demineralizer shall not extend more than 15 feet into the top of the reactor pool when fuel is in the core. The outlet pipe from the demineralizer shall be equipped with a check valve to prevent inadvertent draining of the ,pool.

c. Diffuser and other auxiliary systems pumps shall be located no more than 15 feet below the top of the reactor pool.

d. All other piping and pneumatic tube systems entering the pool shall have siphon breakers and valves or blind flanges which will prevent draining more than 15 feet of water from the pool.'

e. A pool level alarm shall indicate loss of coolant if the pool level drops approximately one foot below normal level.

f. A pool water temperature alarm shall indicate if water temperature reaches 130O9.

Bases

a. The LEU Conversion SAR section 4.7.8 shows by analysis that the natural convective cooling of the reactor core .is sufficient to maintain the fuel in a safe condition up to at least a power level of 1500 kW (the power Safety Limit).

b. The inlet pipe to the demineralizer is positioned so that a siphon action will drain less than 15 feet of water. The outlet pipe" from the demineralizer penetrates the pool below core level and a check valve prevents leakage from tpe pool by reverse flow from pipe ruptures or improper operation of the demineralizer Valve manifold.

c. In the event of pipe failure and siphoning of poo1 water, the pool water level will drop no more than 15 feet from the top of the pool.

d. Other pipes which enter the pool have siphon breakers'which prevent pool drainage. Valves are provided for pneum{atic tube system lines and primary cooling system pipe. Other piping installed in the pool has blind flanges permanently installed.

e. -Loss of coolant alarm, after one foot.of loss, requires corrective action. This alarm is observed in the reactor control room and outside the -reactor building.

f. The thermal-hydraulic analysis in the SAR assumes a pool water temperature of 1300F. If the temperature exceeds 130 0 F then the alarm will prevent continued operation in an un-analyzed condition.

Amendment No. 17