ML20128N327
| ML20128N327 | |
| Person / Time | |
|---|---|
| Site: | River Bend  |
| Issue date: | 02/17/1993 |
| From: | John Miller, Jacqueline Thompson GULF STATES UTILITIES CO. |
| To: | |
| Shared Package | |
| ML20128N319 | List: |
| References | |
| EA-PT-91-0003-S, EA-PT-91-0003-S2-R00, EA-PT-91-3-S, EA-PT-91-3-S2-R, NUDOCS 9302230206 | |
| Download: ML20128N327 (19) | |
Text
-
.i
~
Repod Number - EA-PT-91-0003-S2__. .-i Revision 0 Page l of l9 !
Responses to NRC Qutstions on River Bend Station I Plant Transient Analysis Methodology, Supplement 1,
)
EA-PT-91-0003-SP
(
Febntary 17,1993-Thermal /IIydraulle Analysis Group.
Engineering Analysis Section Engineering Department River Bend Nuclear Group -
Gulf States Utilities Contributing Encineers T. W. Oliphant '
D. R. Swope .
.(
Reviewer: .
- m. 2 /7!9X J. I!. Thompson Da'te Supervisor-Thennal/IlydrauL Analysis Approval: '.. er d$M{er Z l 7!9L' JDiiller. d I D'te a
Director - Engineering Analysis
" ' -- 9302230206 930217 '
PDR AEDCK 05000458
_ P- ppg
t
-i. .,
Report Number . EA PT-91-0003-S2 _
Revision 0 -
Page 2 of 19 1.0 Summary {
q Gulf States Utilities (GSU) submitted topical report EA PT-91-0003-Mtti, entitled - ;
" River Bend Station Plant Transient Analysis Methodology," to the U. S. Nuclear j Regulatory Commission (NRC) in May,1991. This report describes GSU's computer ..
progmms, system models, and methods used for plant transient analysis a; River Bend- 'l Station (RBS). The report also provides comparisons to Peach Bottom 2 transients and- i selected River Bend transients. ;
A topical report supplement, EA PT-91-0003-SPt21, entitled " River Bend Station Plant i Transient Analysis Methodology, Supplement 1," was submitted to the NRC in l October,1991. The supplementary repoit describes.GSU's ACPR methodology, hot - !
channel model, and uncertainty. analysis methodology. -The supplement also presents--- >
applications of the RBS methodology to calculation of thennal limits; Additional' comparisons with RBS transients and revised Peach Bottom 2 transient comparisons ;
were provided. ;
In October,1992, the NRC requestedi)I additional .information on the- tmnsient methodology topical report EA-FT-91-0003 M. GSU's responses to the NRC .
questions ' arc documented in EA-PT-0003 Sild! which was submitted to the'NRC in .
December,1992.
An NRC request isl for additional information on the topical _ report supplement, EA PT-91-0003-SP, was received in December,1992. Section 2.0 of this.. report '
prwides responses to those questions. ;The fonnat for Section 2.0 consists of a
- restatement of the NRC~ question followed by the-GSU response. - References are :
provided in Section 3.0, I
t I
L t
t
~-
4
.. - - = - - . - , - - . - . - . - . . - ---- .- - ~ . , = - --
Report Number EA IT-91-0003-S2 Revision 0 Page 3 of 19 2.0 Responses to NRC Questions on River Bend Station Plant Transient Analysis Methodology, Supplement 1, EA-IT-91-0003-SP Question 1 Describe the Chapter 9 revision to the core physics procedures which required the regeneration of the Peach Bottom 2 kinetics data. Why does this change only have a signincant effect on the TTI predictions?
Response
The revision to the core physics procedures involved the use of ENDF/B-V kinetics tables (delayed neutron fractions, betas, and fast and thennal neutron velocities, V) instead of ENDF/B-III tables or modified versions. The cross sections remained ENDF/B-III. There were also minor changes in the SIMULATE-E nms due to corrected rod patterns in several depletion steps and other minor changes.
The change in procedun: was used for all three Peach Bottom tdp events and resulted in slight increases in calculated power over the calculations made befon: the previous revision. In addition, some changes to thennal/hydnmlic data were made for all three Peach Bottom trip events. The decrease in peak power seen in the TTl calculation was due to changes in thennat hydraulic data resulting from sensitivity studies with a criteria of improved calculation of measured parameters.
The most significant of these changes was a redistribution of separator inenia between inlet and outlet. TTl was most affected because it has the smallest total separator inenia and saw the largest percentage change at the separator inlet, which has the largest impact on the n:sults. The response to Question 7 provides a description of the separator inenia changes.
Question 2 I Provide the basis for the 95/95 uncenainty values used for the core leakage flow, total core pressure losses, jet pump M-ratio, TCV stroke time, initial thennal power, and
- maximum nmout time.
Response
To quantify the methodology uncenainty, ACPR sensitivity studies were perfonned which perturbed the components of uncertainty by an amount judged to be two standard deviations from their nominal value at a 95 percent con 6dence level.- Engineering
R: port Number EA-PT-91-0003-S2 Revision 0 Page 4 of 19 judgment was used to select and quantify these components of uncenainty, supplemented by infonnation available from technical references and test data repons.
In addition, License Topical Repons previously submitted by fuel vendorsle,71, other licenseest8'l, and national laboratoriesHol were reviewed. The standard deviations quoted by these prior uncertainty analyses were compamd to the values derimi for River Bend specific systems and conditions, and helped guide our selection process.
For many of the low significance uncenainty components, bounding values were used as the 2a uncertainty level. The bases for the sl cciGc i uncedainty components requested are provided in Table I below.
TAl&Ill BASES FOR UNCERTAINTY COh!PONENTS Component Uncertainty ACPR References (2a) Sensitivity for Bases Core Leakage Flow 10 % Low 8, and 9 Total Pressure losses 10 % Iow 6, 8, and 9 Jet Pump hi Ratio 5% hiedium i1 TCV Stroke Time 20 % hiedium Plant Data Initial Thermal Power 2% Insignificant 7 Feedwater hiaximum:
Runout Flow 5% Iow 8, and 9
. Ramp Rate 50% InsigniGeant Bounding
, Question 3 I
The Table-6.8 values of 6RCPR95t95 and resulting statistical adjustment factors have been detennined for specific transients having relatively small values of RCPR. Will the Table-6.8 values of 6RCPR,3,93 be applied in the RBS licensing analyses on a-percent basis (i.e. $RCPR/RCPR is assumed constant)? If not, provide justification for the method used to apply these 6RCPR95/95 uneenainty values.
Response
The RCPR9s/95 values will not be used in the GSU licensing analysis on a percent basis.
The RCPR93,95 value is used 10 determine the statistical adjustment factor (SAF) for
I
- Report Number _ EA-PT-914KKO SL__.
Revision 0 Page _5_uL19-each inmsient (e.g. load rejection without bypass, pressure regulator downscale failure). Specl0cally, SAPu = (RCPR95/95)u - (RCPRa,,i,,,,,,;,,,u )u where N is the transient type, in turn, the SAP for each transient is applied to detennine the required operating limit h1CPR (OLMCPR) for each transient as follows:
SiliCPR N " 1 - [(RCPR a,,;,, o,,;,,,u.)u + SAPu]
where SIAICPR is the safety limit MCPR.
Since transient specine SAFs are used to account for uncertainties in the calculation of transient specific OIAICPRs, RCPR ratios are not needed to account for uncertainty differences for various inmslents. Section 6 of Referrnce 2 contains a description of this process.
RCPR is denned as ACPR/ICPR as described in Section 5.0 of Reference 2. The RC PR,3,,3 values presented in Table 6.8 are not delta values, but represent the 95%
probability ACPR/ICPR value at a 95% conGdence level for a specific tmnslent.
ARCPRs, given in Tables 6.4 through 6.6, are used in the RBS methodology for calculating the standard deviations which in turn are used in calculating the RCPR95/95 values presented in Table 6.8.
The RCPR values for the three events are typical of values calculated for rapid transients tenninated by a reactor protection system (RPS) trip in a BWR/6. The BWR/6 design has a higher control rod insertion rate than the BWR/3, 4, and 5 designs. This feature nduces the severity of the pressurization events (e.g. load rejection, feedwater controller failure, and pressure regulator downscale failure).
Question 4 The scram and doppler reactivity contributions to the excessive reactivity are relatively.
small initially, but ultimately increase suf0ciently to dominate the excess reactivity and tenninate the transient. Ilow does the calculated reactivity compare with the reactivity l
inferred from the Peach Bottom 2 measurements, as a function of time, up to the point of minimum CPR7 Justify the use of the uncertainty in the peak excess reactivity for the scram and doppler reactivity uncertainties.
l l
l i
I
I
? '.
Report Number EA-PT-91-0MlbS2. _
+
Revision 0 Page 6pfl9
Response
Core power is strongly dependent upon excess reactivity, so changes in total reactivity are mirrored by changes in power. In References 1 and 2, GSU chose to compare the time dependent calculations to the directly measured parameter, core power, for each of the Peach Bottom 2 turbine trip tests. In Section 6.1.2 of Reference 2, the peak reactivity of the Peach Bottom turbine trip tests was inferred fnnn the point reactor kinetics equation, and is therefore proportional to (1/ Power)(dPower/dt). A comparison of the calculated and inferred mattivities for Til as a function of time is shown in Figure 1.
Reactivity as a Function of Time for TT1 i . _ ._ . _ _ _ _ -
o.8 ,,
o.6 -
0.4 -
- g o.2 .
on - . - - - ,. r.-:- -.
8 o.2 o 2o o.4o oro o.so 1.00 .,1.2 2 ~ " ~ - ~ - ~ ~ ~ ~ - ' ~ " * - ~ ~ " ~ ~ '.
CC -o.4 '
GSU Calculated .
.-o.6
- Inferred from Peach .
.o.B Bottom '.
3 Measuremente '.'.
1.2 - - - - - - - - - - - - '- - ----
Time (seconds) i Figure 1 Reactivity as a function of time for'IT1.
The point reactor kinetics equation was used to help quantify the uncertainty lii kinciks parameters in units of reactivity. GSU compared the inferred reactivity from each of the Peach Bottom tests to calculated excess reactivity. The standard deviation in peak excess reactivity was used as a convenient single measure of the adequacy of the neutronics data (i.e., two group cross sections, delayed neutron fractions, decay
Report Number EA-VI'-91-(KMB 52 Revision 0 Page 7 cf.19 fractions, and mean neutron generation times). Use of the peak excess tractivity uneenainty is justined for two reasons:
- 1) Being proportional to the derivative of the power spike rate, peak excess reactivity is a good indicator of the transient response including vold and Doppler feedbacks, just prior to rod insenion, and
- 2) Although scram worth is a strong feedback, its contribution to the transient uncertainty is small since once an axial ponion of the core is fully controlled, power generation in that por11on is dramatically reduced. The major uncenalnty associated with scram is the timing of the scam, not the rod worth. Scram time uncertainty is treated separately in the GSU methodology, Question S Compare the GSU and vendor methodologies for detennining the now dependent MCPRf limit, and justify any differences. How do the GSU and vendor predictions of the MCPR, limit compare?
Response
The vendor and GSU methods are essentially the same. Both use a steady state thennal-hydraulics program (FIBWR in the case of GSU). Both of these progmms are similar in nature, and use equivalent methods and correlations in the solution of flow distribution within the BWR core, as well as MCPR detennination.
The method to be employed, is to perfonn calculations at the 105%-of rated steam flow control rod line, at the maximum core How rate attainable with the recirculation flow control valves at the electronically limited valve position. The purpose of these calculations is to adjust the bundle relative power until the calculated bundle MCPR is equal to the safety limit MCPR. Keeping the relative bundle powers and axial power shapes intact, a series of calculations are perfonned at different core flows along the l 105 %-of-rated steam flow control line. The calculated MCPR at the given flow point l is the MCPRr. This is the same approach as outlined in the Technical SpecificationH21, BASES 3/4.2.3. The GSU and vendor calculated MCPRfcurves are similar.
l l
l Question 6 Discuss the difference between the contml rod insenion curves for the pressurization I and overpressurization events.
Repo 1 Number EA-FT-91-0003 S2 Revision 0 Page 8 of 19
Response
l The serrun function of the control ruds is accomplished by scram accumulators located J at the hydraulic control unit of each control rod. The accumulator is partially filled with water, and is maintained at a relatively high pressure by means of a nitrogen ,
blanket. Upon receipt of a scram signal, the scram valves open, admittirg the water in the accumulator to below the drive piston, which causes the control rod to rapidly insert against reactor pressure. Therefore, the rate of inseition is dependent upon the -
reactor pressure. ,
The overpressurization' event results in higher vessel pressures during the insertion of control rods than pressurization and non pressurization events. Therefore, the rate at l which the control rods are inserted is slower for the overpressurization transient. The insertion rates used by GSU are the same as those used by the fuel vendor. Figure 2 below shows the nominal scram rates used for evaluation of non pressurization, pressurization, and overpressurization events. Also shown is the . Technical Specification scram speed twiuirement for reactor dome pressure of 1040 psia. Note ,
that all scram curves used for non-pressurization, pressurization, and overpressurization -
events are more conse vative than the Technical Specification requirement. j Scram insertion Rate L
(Feet inserted Vs. Time) 12 T / / ',
/ ,/
10 ,/ ,/ - '__
./ ,/ -
/ '
8
,/,
6 f,- .--
j
" - haemsmateen 4 f
/
. / . ::= .
O d -+ , + ,
o 0.5 3.5 1 .1.5 2 2.5 3
. Time (sec)
Figure 2 Scram insertion rate used for various events.
I:
)
4
.--a ,,-v u , ,w--
i Report Number EA-PT-91-000332__.
Revision 0 i Page 9 of 19 j i
l l
Question 7 l Ilas the RBS model been adjusted to improve agreement with the Peach Bottom 2 measurements and, if so, what is the effect of this adjustment on the inferred peak excess reactivity?
Response
The only adjustment made, other than to match specine initial conditions for each of the three turbine trip tests, was the manner of allocating separator inedia to the separator inlet, standpipe inlet, and separator return junctions.
The total inertia was determined as a function of separator inlet quality from vendor data as described in Section 7.3.2 of Reference 1. The allocation to specific junctions, also described in Reference 1, has been revised to put a constant value at the separator return Junction, half of that part attributed to the standpipes at the standpipe inlet and the remainder of the total at the separator inlet. As separator inlet quality and total separator inenia vary,.only the separator inlet inedia changes. The separator outlet, which is almost all liquid, has constant inedia. This gave improved power response for Til without signincantly affecting the other Peach Bottom events.
The power curve, and therefore the peak excess reactivity, was found to be sensitive to the separator inenia distribution. A larger difference between calculated and measured powers will indicate a larger difference between calculated and inferred peak excess reactivity.
Question 8 How were the standard deviations of Table 6.8 detennined? Do the RCPRes values of Table 6.8 include a two standard deviation allowance as indicated in Figure 6.17
Response
Table 6.4 through 6.6 in Reference 2 present transient specific uncenainty values for each component of uncenalnty evaluated. Since the initial penurbations were two standard deviation (2a) values, the uncenainties are also 2a values. The square root of the sum of the squares (SRSS) of the ARCPR values was calculated to detennine an overall uncenainty in CPR units for the transient at the 2a level. The one standard deviation (a) value for the transient is then half of the SRSS of the ARCPR values.
Report Number EA#r-91-0003-S2 Revision 0 Page 10 of 19 The RCPR,333 values shown in Table 6.8 do not include a two standard deviation allowance. Instead, the RCPR 95 s3 value is the mean RCPR value plus 1.645 standard deviations.
Question 9 What is the difference between the ARCPR95 ss data of Table 6.8 and Tables 6.4 through 6.77
Response
As discussed in the response lo Question 3, the RCPR,333 values in Table 6.8 are not delta values, but represent the 95/95 ACPR/ICPR value for a specific tninslent. The RCPR,333 values are used to calculate the transient specific SAFs.
The ARCPR values shown in Tables 6.4 through 6.6 represent the transient specific uncertainty at the 2a level for a particular component of uncertainty (e.g. Initial thennal power). The ARCPR values for a transient are combined statistically and then used to calculate the RCPR standard deviation which is in turn used to calculate the transient specific RCPRy333 Table 6.7 does not contain ARCPR data.
Question 10 What is causing the ~20% GSU underprediction (relative to the vendor) of the ACPR for the pressure downscale failure (PDR) and feedwater controller failure (FWCF) events? Note that this difference is outside the expected ACPR95ss given in Table 6.8 for these events.
Response
i The ACPR agreement between the GSU and vendor methods is very good. Due to the very small ACPR for the pressure regulator failure and the feedwater controller failure events, the percentage difference is large. However, the magnitude of this difference is not significant. A difference of 0.01 in ACPR between the vendor and GSU calculations is to be expected given the potential for subtle modeling differences, and round-off.
The values in Table 6.8 are based on an uncertainty analysis for the GSU methodology which is used to calculate SAFs. The SAFs are used to adjust the OLMCPR for L
Report Number - EA PT-910003-S2_,
Revision - 0 Page ___ 11 of 19 uncertainties in the GSU computer codes and analytical models. Since the statistical:
values in Table 6.8 are based entirely on GSU calculated results, differences between the vendor and GSU calculated values do not have any relationship to the expected
~
uncertainty given in Table 6.8.
Question 11 q
. u What initial conditions (e.g., core flow, cycle exposure, power distribution, feedwater flow and temperature) will be assumed in the LRNH, PDF, FWCF and MSIV closure .
licensing analyses and are _these conditions conservative relative to expected operating :
~ ' '
conditions.
Response
Tables 7.1 and 7.2 of Reference 2 list initial- conditions for pressurization and 3 overpressurization events, respectively. Tables II and Ill list the values used for initial j conditions and parameters in addition to those listed in the table referenced above. j Pressurization events are initiated at rated conditions with uncertainty accounted for in" the statistical adjustment factor (SAF). The SAF provides a conservative estimate of MCPR. Overpressure events are initiated at conditions consistent with a conservative bounding initial condition of 102 % power, ,
a 1
4
(- ,
1 i
4 4
b
,- ...-.n-- 1 _ ..._ _ ,. _ ,___.. . _ . .
_ . , ,,_, ._......_.1,,. . . _ . _ _ , , c=c- - .c.-- , - . , -
'Repor1 Numher EA-IT-914003-S2 I Revision 0 Page ,,, 12 of 19 .
i TABLEIl j INITIAL CONDITIONS AND PARAMETERS FOR PRESSURIZATION EVENTS Parameter Value Justification !
Core Flow 84.5 Mlb/hr Rated Core Flow l Cycle Exposure End of Cycle Bounding, pruvides i I
conservative void and scram feedback.
Power Distribution:
System Model Italing This is the optimized end - t of cycle power distribution target for operation.; -
Ilot Channel 1.4 peaked chopped cosine Conservative.
Feedwater Flaw 12,45 Mlb/hr - Consistent with 100%
power heat balance.
Feedwater Temperature 420 'F - Consistent with 100% f
. power heat balance.
T
-r
+
k 9
B
+
c' ee-- * * * --,,eew-. -,-+-w . , , , ,s,-- .-v..we ,-, _ , , - . ., . . -,m-4 m,w.--, -. %n . .-.- , e,,,,.
Report Number EA-PT-91-0003-S2 l Revision 0 ,
Page 13 of 19 IMILR111 INITIAL CONDITIONS AND PARAhf BTERS FOR TIIB OVERPRESSURIZATION EVENT Farameter Value Justincation Core Flow 84.5 hilb/hr Rated Core Flow Cycle Exposure End of Cycle Bounding, provides conscivative void and scram feedback.
Fower Distribution:
System hiodel IIaling This is the optimized end of cycle power distribution target for operation.
Ilot Channel N/A Ilot Channel not used in AShfE overpressuri'ation analysis.
Feedwater Flow 12.74 hilb/hr Consistent with 102%
power heat balance.
Feedwater Tempemture 422 *F Consistent with 102 %
power heat balance.
Question 12 Will the bypass flow be modeled as a negative fill in the FWCF analysis to ensure there is no overshoot of the bypass flow capacity?
Response
While GSU believes that some overshoot of bypass Oow is realistic, our design basis methodology conservatively models the bypass How using a negative All in all cases where the bypass valves are predicted to open. The Peach Bottom turbine trip transients were best estimate calculations with a modi 0 cation consisting of piping with heat sink from the bypass valves to a time dependent condenser. The intention was to simulate those tnmslents as closely as possible, including any overshoot that may have occurred, llowever, the RBS design basis model, which models th: bypass as a negative fill for the FWCF, is intentionally conservative in this regard relative to best estimate models.
Report Number _EA 1"I'-91-oo03-52 Revision 0 l' age 14oL19 Question 13 IIow are calculational uncertainties accoun:cd for in the htSIV closure overpressurization analysis.
Response
A bounding appruach is used for the MSIV overpressurization analysis to account for calculation uncertainties. The initial power is assumed to be 102% of rated thennal power, rnther than 100% of rated thennal power which was assurned for the pn:ssurization events. Other initial conditions (dome pressure, steam flow, feedwater temperature, etc.) were detennined from a heat balance for the 102% power and 100%
core flow condition.
Question 14 Describe the modeling of the closing of the MSIVs and opening of the SRVs in the RBS overpressurization analysis. Does this treatment conservatively bound the perfonnance of these systems?
llesponse The MSIVs are assumed to close linearly over a three second period, which is iie l fastest speed the valves may close by design. This also is the fastest rate allowable by Technical Specifications n21, The pressure relief system at River Bend Station employs sixteen dual acting safety / relief valves (SRVs). The SRVs are actively opened in the relief mode by an air actuator when sensed pressure in the reactor vessel exceeds the relief set pressum. ' The SRVs are passively opened its safety mode when steam line pressure exceeds the SRV spring set pressure.
Following the present licensing basis, eight SRVs with the highest relief set pressures are ellowed to open in the relief mode. The remaining eight SRVs are those with the highest safety set pressures and are allowed to open in safety mode. The mode selected for each of the SRVs to opcate results in the highest peak pressure for the overpressurization event. System perfonnance is conservatively modeled using design logic delays, and valve opening characteristics. Mass How through the valve is reduced by decreasing the valve flow area from the best estimate value by 10%, which is consistent with the ASMB cc4Je requirements.
Report Number EA-PT-91-0003-S2
^
Revision 0 Page 15 of 19 The above measures ensure that the htSIVs close in a conservatively fast manner, and that the predicted safety relief flow is conservatively low prior to reaching the peak pressure.
Question 15 Provide the predicted peak pressure and margin to limits for each of the reactor coolant pressure boundary components.
Response
Peak pressure in the reactor vessel is lirnited to 110% of the design pressure of 1250 psig, or 1375 psig. Table IV below gives the pressures, limits, and margin for a variety of h> cations within the reactor coolant pressure boundary.
TABLE IV MARGIN TO PRESSURE LlhilTS FOR REACTOR PRESSURE BOUNDARY COMPONENTS Location Peak Limiting Margin Pressure Pressure gs;)
(psig) (psig)
Vessel Bottom IIcad 1247 1375 128 Steam Line 1218 1375 157 The peak pressure at the bottom of the vessel (peak pressure location) reported above is slightly greater than the value reported in Reference 2. Since the issuance of Reference 2, it has been discovered that the reactor tripped approx;mately 0.226 seconds early due to a minor input error. The difference in pressure is small, and is still well within the acceptance limit of 1375 psig, l
\.
l Report Number EA IT-91-0003-SA l Revision 0 Page 16 of 19 Question 16 In the hiSIV closure overpressurization analysis, how many SRVs are considered to be inoperable and is this consistent with the maximum allowed in the technical specincations?
Response
The overpressurization analysis repo:1ed in Reference 2 was performed with all SRVs operable (eight in safety mode and eight in relief mode as discussed in the response to Question 14). This is consistent with the analysis provi%I by the fuel vendor with each reload. As discussed in the response to Question 15, the GSU overpressurization analysis shows a peak pressure well below the acceptance limit of 1375, with somewhat greater margin than shown for 16 SRVs operable in the analysis presented in the RBS Updated Safety Analysis Report 031 (USAR). of Reference 13.
The RBS Technical Specifications require Gvc SRVs to be opemble in the safety mode, and an additional four SRVs to be operable in the relief mode. The requirement for nine operable SRVs is based on the USAR analysis. USAR Figure 5.2-4 shows the results of the analysis for eight through sixteen SRVs opemble. The analysis shows that nine operable SRVs are sufficient to prevent reactor vessel pressure from exceeding the acceptance limit, with larger margin for more valves operable, as expected.
The GSU calculations conGnn that the worst cycle conditions result in a wider margin to the pn:ssure limit with 16 SRVs operable than do the original USAR calculations. It is expected that a CSU calculation with only nine SRVs operable would also show a greater margin than does the original USAR analysis and therefore would bound the Technical SpeciGcations.
Question 17 Ilow is the uncertainty in the time-dependent hot-channel radial peaking factor accounted for in the licensing analyses?
Response
The uncertainty in the hot channel radial peaking factor is accounted for by choosing a value which boun(1s the highest value expected to occur during steady-state. One of the major assumptions of one-dimensional (axial) transient analysis is that the radial power distribution remains constant throughout the transient. hiost radial power fluctuations which occur as a result of a rapid core-wide power spike will tend to flatten the initial power distribution, hence the constant radial peaking factor assumption is conservative.
Report Number EAoPT-910MG-S2 Herision 0 >
Page 17 oil 9 The trend observed in the local power mnge monitor (LPRM) measumments taken during the Peach Bottom turbine trip tests indicate little change in the relative bundle powers during the tmnsient. These LPRM measurements funher substantiate the adequacy of the constant radial peaking factor assumption.
Question 18 What effect does the GSU core physics procedures revision have on the inferred (calculation-to-measurement) peak reactivity differencer and the determination of RCPR95/93 7
Response
The core physics procedure revision occurred prior to the completion of the sensitivity studies and uncenainty analyses, Since that time, the revised procedure has been used for all Peach Bottom core physics calculations. No RCPR95/95 calculation msults were submitted using the old procedure. As the requested analyses were never performed with the obsolete method, we are unable to generato them without a substantial effon.
o, -
, - ~ . --
RGport NEmber EA-PT-91-000pS2' Revision 0 Page~ 18 of 19 3.0 References
- 1. " River Bend Station Plant Transient . ? Analysis - : Methodology,"
EA-PT-91-0003-M, April 1991, transmitted to the NRC by RBG-34,939 dated May 2,1991.
- 2. " River Bend Station Plant Transient Analysis Methodology; Supplement 1 Delta CPR Methodology and Additional _ Benchmarks", EA-IT-91-0003-SP, October 1991, transmitted to t'te NRC by REG-35876 dated October 31,1991. =
- 3. Letter from D. V. Pickett (NRC) to J. C. Deddens'(GSU) dated October 30,_
- 1992, River Bend Station, Unit 1 - Request for: Additional Information Re Topical Report _ EA-Pr-91-0003-M, " River ' Bend -Station Plant Transient Analysis Methodology," (TAC No. M80315), RBC-42927, 4.
" Responses to NRC Questions on River Bend Station Plant Transient _ Analpsis_
Methodology - EA-PT-91-0003-M," EA-PT-91-0003-S1, December z 1992, tmasmitted to the NRC by RBG-37,930 dated December 18,1992.
- 5. Letter from E. T. Baker (NRC) to P. D. Graham (GSU) dated December 22, 1992, River Bend Station, Unit 1 - Request for . Additional Infonnationi Regarding Topical Report EA-PT-91-0003-M, Supplement 1,- " River Bend Station Plant : Transient Analysis Methodology" (TAC No. M80315),-
RBC-43190. '
t
- 6. " Safety Evaluation for the General Electric Topical Report Qualification' of the One-Dimensional Core - Transient Model for Boiling Water Reactors NEDO-24134 and NEDE-24154-P, Volumes I, II,' and III", USNRC, June 1980.
- 7. " Analytical Model for Loss-of-Coolant Analysis in Accordance With 10CFR50 Appendix K," NEDO-20566A, General Electric Company, September 1986.
TVA-TR-81-01, Tennessee Valley Authority, December 1981.
PECo-FMS-0004, September 1987. .,
- 10. "Saisty Evaluation for the General Electric Topical Report Qualification of the -
One-Dimensional Core Transient Model for Poiling-Water Reactors," USNRC, June 1980.
I1 " Testing of Improved Jet Pumps for the BWR/6 Nuclear System",
NEDO-10602, General Electric Company, June 1972.
d
Report Number EA-M'-91-0003-52 Revision 0 Page 19 of l9
- 12. River Bend Station Technical Specifications.
- 13. River Bend Station Updated Safety Analysis Repon (USAR).
l-l*
i