ML18046B082

Forwards Record of 810901-1023 Nrc/Util Telcons in Support of Review of Cycle 5 Reload Tech Spec Change Request. Re-evaluation of Steam Line Break Accident for Cycle 5 Encl

ML18046B082
Person / Time
Site:	Palisades
Issue date:	11/17/1981
From:	Johnson B CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.)
To:	Crutchfield D Office of Nuclear Reactor Regulation
References
NUDOCS 8111230517
Download: ML18046B082 (29)

Text

consumers Power company General Offlc89: 212 Wast Michigan Avenue, Jackson, Ml 49201 * (517) 788-0550 November 17, 1981 Director, Nuclear Reactor Regulation Att Mr Dennis M Crutchfield, Chief Operating Reactors Branch No 5 US Nuclear Regulatory Commission Washington, DC 20555 Consumers Power Company Technical Specification Change Request dated July 21, 1981 and entitled., Linear Heat Rate Limits, Radial Peaking Fa.ct ors and Excore Power Distribution has been under review by the NRC staff and has resulted in a number of questions. In an effort to reduce the amount of time required for this type of review, both the NRC staff and Consumers Power Company have performed the majority of question asking and answering through telephone conversations. Now that for the most part all *the questions hav.e been satisfactorly answered, an effort is underway to document the dialogue that has gone on.

Mr. P.S. Kapo of the NRC staff made the first attempt at this documentation by submitting a Record of Paraphrased Telephone Conversations for Consumers Power Company's review and approval. We have editorialized this documentation through additional telephone conversation with Mr. Kapo and are now submitting this Record of Telephone Conversations in its' final form as Attachments 1 and 2.

Also included, as Attachment 3, is Consumers Power Companys response to the question regarding the Reevaluation of the Steam Line Break Accident for Palisades Cycle 5.

Brian D Johnson Senior Licensing Engineer CC JGKeppler, USNRC NRC Resident Inspector-Palisades Attachments

... ATTACHMENT 1 .

RECORD O~ PHONE CONVERSATION TO OBTAIN INFO!UtATION PERTINENT TO THE PALISADES CYCLE 5 RELOAD REVIEW Telephone Conversation Dates (9-1-81 Through 10-23-81)

NOTE: This record is a paraphrase of the phone conversation, and not a verbatim transcript.

NOTE: All questions are from the NRC. All answers are from CPC.

NOTE: In this record, the following abbreviations are used._

AO - Axial Offset APL - Allowed Power Limit CAOC - Constant Axial Offset Control CPC - Consumers Power Company DBE - Design Basis Event DNB - Departure From Nucleate Boiling DNBR - DNB Ratio LCO - Limiting Condition for Operation I.HR - Linear Heat Rate LOCA - Loss of Coolant Accident LSSS ~ Limiting Safety System Setting - RPS Trip Set Points RPS - Reactor Protection System TS - Technical Specification or Technical Specifications T-H - Thermal-Hydraulic null81-0173a-43-45 1

Ql Do you have a startup physics testing program approved in a previous cycle that you intend to use in the cycle 5 startup?

Al The start-up physics testing program for cycle 4 was approved by the NRC.

For cycle 5, we intend to use the cycle 4 program, except that we will drop the Moderator Temperature Coefficient measurement at power and the Power Coefficient measurement at power. The reason we are dropping these two measur.ements is that they are rather inaccurate, and we can calculate these. parameters mo.re accurately than we can measure them.

We will still be performing the zero power Moderator Temperature Coefficient measurement, which .is an accurate measurement.

Q2 Will the effluents in cycle 5 increase from those in cycle 4 so that we would need an environmental impact analysis?

A2 No.

Q3 Will you have any modified plant equipment for cycle 5?

A3 Yes. A number of modifications in plant equipment are being made, most of them to conform to new regulations brought about by the Three Mile Island incident. These modifications are discussed in submittals separate from the cycle 5 reload submittal, and need not be considered in the cycle 5 reload review.

Q4 Did you use any modified analytical design methods in the analysis of cycle 5?

A4 Yes. Exxon has adopted a new method for analyzing the Ejected Rod Event and has analyzed our Ejected Rod Event using their new method. This method is described in topical report XN-NF-78-44, which has not yet been approved by the NRC.

QS Are there any changes in the external chemical design of the new fuel which could be a source of corrosion?

AS The new fuel added in cycle 5 has exactly the same external chemical composition as that added in cycle 4.

Q6 Are there any aspects of the fuel mechanical design of the cycle 5 fuel which were not present in the fuel in previous. cycles which would impact the T-H or LOCA analysis?

A6 The cycle 5 fuel has almost the same mechanical design as the .fuel used in previous cycles. The only changes in the cycle 5 fuel are the following:

(1) The outside diameter of the cycle 5 fuel cladding is 2 mils greater than the outside diameter of the cladding used in previous cycles.

nu1181-0173a-43-45 2

This small dimensional change will slightly increase the heat transfer surface area for the T-H and LOCA analysis and slightly increase the reflood rate in the LOCA analysis.

(2) the fuel pellet dimensions have changed by an amount too small to have any impact on the T-H or LOCA analysis.

(3) The number of gadolinium bearing assemblies and the gadolinium cont.ent of these assemblies has been progressively increased in the last three reloads. This will be explained in the following paragraphs:

All commercial power reacto.rs contain burnable poison fuel pins for the purpose of improving the power shape. In the recent past, most pressurized water reactors have used boron as the burnable poison.

Because of certain advantages of using gadolinium as a burnable poison, Exxon embarked on an experimental gadoliniuril program in cycles 3 and 4 of the Palisades reactor. In these two cycles, the gadolinium behaved as predicted, giving Exxon and Palisades the impetus to go to a cycle 5 core design which wbuld optimize the use of gadolinium as a burnable poison and provide a model for future reload cores.

The number of gadolinium bearing assemblies and their gadolinium content added in the last three cycles is as follows:

Fuel Number of Gd Number of Gd Pins Gd Content of Cycle ~ Assemblies per Assembly Pins (w/o) 3 G 8 4 1 4 H 4 8 4 5 , I 12 8 4 The NRC has received topical reports XN-NF-79-61 and XN-NF-79-56, Rev 1 on the use of gadolinium as a burnable poison. With the approval of these two reports and the approval of the cycle 5 reload design, the use of gadolinium in future reloads will be implicitly approved, bringing the commercial venture of the gadolinium experiment to a successful conclusion.

The effect of the enrichment and the gadolinium content of the cycle 5 fuel is included in the analysis of Section 7.1 of the reload safety report, XN*NF-81-34. Based on this analysis, new radial peaking factors have been included in the TS.

The reasons that gadolinium is preferable to boron as a burnable poison is that it is a better neutron absorber than boron and it burns out faster than boron. This gives gadolinium the following advantages over boron as a burnable poison:

(1) Gadolinium burns out somewhat faster than the fuel. Because of this the gadolinium bearing assemblies maintain the same reactivity, or nu1181-0173a-43-45 3

even increase slightly in reactivity during the first part of the cycle. This helps achieve a flatter core power shape throughout.the cycle than is possible using boron as the burnable poison.

(2) Since most of the gadolinium has burned out by the end of a cycle, a gadolinium bearing core has more reactivity than a boron bearing core toward the end of a cycle, which makes it possible to stretch the length of a cycle.

(3) Boron is a relatively weak neutron absorber, and whole fuel pins must be replaced by boron pins for the boron to be an effective poison. .By comparison gadolinium is a very strong neutron absorber, and a gadolinium bering pin which contains the usual amount of nuclear fuel plus 4~~ Gd o is an effective poison pin. Thus all the 2 3 pins in a gadolinium bearing assembly are active fuel pins. This increases the total U235 core loading which helps to extend the cycle.

(4) Because the gadolinium bearing assemblies contain all active fuel pins, as explained in (3) above, the LHR of the active fuel pins can be made smaller without decreasing the assembly power.

Q7 Does CPC have the capability of computing the LCO and LSSS for Palisades?

.If so, have you submitted a topical report?

A7 Exxon provides the limiting aU.owable core power for DNB protection as a function of flow, pressure and inlet temperature. This limiting core power is given in TS Figure 2.3. From this figure, CPC is able to compute the DNBR LSSS.

Exxon provides radial peaking factors and axial power distribution envelopes for steady-state DNBR margin. These data plus the already specified coolant condition limits are our DNBR LCO, Exxon provides the limiting LHR for LOCA protection which is the LHR LCO of TS Section 3.23.1.

CPC has not yet submitted a topical report on these computations.

Q8 In TS 3.11.2, which specifies the LCO for the AO and APL as measured by the excore detectors, the calibration of the AO and APL of the excore detectors to the incore detectors is discussed. How do you account for the error between the excore and incore AO and power? Do you have a topical repor.t which describes the calibration procedure?

A8 Errors between the incore and excore mea~ured AO and power are minimized by periodic calibration of the excore measured AO and power to the incore measured AO and power.

With the CAOC strategy the LCO on AO is +/- 5% of the target AO. The excore AO calibration is done exactly at the target AO. With only a 5%

nu1181-0173a-43-45 4

variation allowed in the AO from its calibration point, the error between the excore and incore measured AO is too small to be of importance.

With periodic excore power calibration and the CAOC operating stategy, the error inthe excore measured power is also kept small. With CAOC the core power shape remains essentially constant between excore-incore power calibrations. With the constant power shape, the ratio between the excore and incore detector signals remains essentially constant and the error between the excore measured power and the incore measured power will be quite small.

To date no topical report has been submitted which describes the excore-incore AO and power calibration procedure.

Q9 The list of DBEs analyzed in XN-NF-77-18 is smaller than the list of DBEs in the FSAR. Do you have NRC approval for considering only those DBEs in the XN-NF-77-18 list, rather than the FSAR list for all future reloads?

A9 The analyses in the Transient Analysis Report, XN-NF-77-18, were performed to justify operation of the Palisades reactor at 2530 MWT during cycle 2. Only those events for which the FSAR analysis was not bounding at 2530 MWT were reanalyzed. The reanalyses described in XN-NF-77-18 and the justification for not including the reanalyses of the other FSAR events were approved in the .NRC Safety Evaluation accompanying Amendment No 31 to our Operating License No DPR-20 [

Reference:

letter from A Schwencer (NRC) to David Bixel (CPC), November, 1977]. The reason that certain FSAR analyses were bounding for cycle 2 operation at 2530 MWT are listed in Table 4.0-3 of the NRC Safety Evaluation, which is reproduced on the following page. The FSAR analyses of these events remain bounding for cycle 5 operation for the same reasons.

nu1181-0173a-43-45 5

TABLE 4.0-3 TRANSIENTS AND ACCIDENTS NOT REANALYZED Incident Reason Not Reanalyzed Boron Dilution At start-up or refueling, the FSAR analysis is still bounding. At power, the incident is bounded by the Rod Withdrawal incident.

Steam Generator Tube The FSAR analysis, done at 2650 MWt, Rupture is bounding.

Turbine Generator Overspeed the FSAR analysis is still valid since it is* not affected by the power increase.

Fuel Handling Accident A bounding analysis was performed in connection with the spent fuel pool storage expansion approved by us in a license amendment issued on June 30, 1977.

Idle Loop Start-Up Start-up of the reactor* is not per-mitted with less than 4 pumps irioperation.

Malpositioning of Part-Length Operation of the reactor is permitted Control Rod Group only with the part-length control rods completely withdrawn from the core.

QlO On page 35 of the Reload Safety Report, XN-NF-81-34, it is stated that the DBEs listed in table 7.2 on page 39 were reanalyzed. However the title to table 7.2 is "Transient Events Considered iIJ. the Palisades Cycle 5 Plant Transient analysis." Were these DBEs, in fact, reanalyzed for cycle 5 or not?

AlO With the exception of main steam *line break, the DBEs in table 7.2 were not reanalyzed for cycle 5. *we compared the cycle 5 inputs to these DBEs with the inputs to the reference DBE analyses described in XN-NF-77*18 and in all cases, with the exception of the steam line break accident, found the XN*NF-77-18 analysis to be bounding.

A discussion of methods and results.of the main steam line break reanalysis for cycle 5 is provided in an attachment to this letter.

nul181-0173a-43-45 6

Qll Why does the list of DBEs .in table 7 .2 of the Reload Safety Report, XN-NF-81-34, not include Rod Ejection, which is included in the Transient Analysis Report, XN-NF-77-18?

All Rod Ejection is discussed later in the Reload Safety Report.

Q12 In TS 3.11;1 you require that only 50%.of the incore detectors be operable. Mostpiants require 75% of the incore detectors be operable.

How do you justify operating with 50~~ operable detectors without taking a penalty in peaking .factors, quadrant tile, or AO?

Al2 This is an old TS which has already been approved by the NRC. Also note that the standard Tech Specs refer to 75% of the incore detector locations (ie, strings). The Palisades requirement refers to individual detectors.

Ql3 In TS 3.11.2,b&c, why is no time period ~or calibration of the excore detectors specified?

Al3 The time period is specified in TS 4.18.2.1. It would probably make the TS seem mor.e coher.ent if this time period were specified in TS 3 .11.1, but we have specified it in TS 4.18.2.1 to make our TS format closer to the standard TS format. We are doing this in anticipation that some day we may adopt the standard TS format.

Ql4 You submitted XN-NF-80-47, which describes CAOC, in October 1980. Has Palisades been running with CAOC in cycle 4, or will this be a new mode of operation in cycle 5?

Al4 CAOC is the new mode of operation for cycle 5.

Ql5 What is Upburn?

Al5 Upburn is the increase in the power generated by burnable poison assemblies as a result of the burnable poison being depleted faster than the fuel in*the assembly. The maximum upburn in the fuel currently used in Palisades is about 2% between the monthly surveillance intervals, and about 20% in the course of a cycle.

null81-0173a-43-45 7

ATTACHMENT 2 RECORD OF PHONE CONVERSATION TO OBTAIN INFORMATION PERTINENT TO THE PALISADES CYCLE 5 RELOAD REVIEW Telephone Conversation Dates (9/15/81 throug.h 10/20/81)

NOTE: This record is a paraphrase of the phone*conversations, and not a verbatim transcript.

Note: All questions are from the NRC. All answers are from CPC.

NOTE: In this record the following abbreviations are used.

BOC = Beginning of Cycle CEA = Control Element Assembly = Control Rod CPC = Consumers Power Company DNB = Departure from Nucleate B6ili~g DNBR = DNB Ratio DTC = Doppler Temperature Coefficient EOC = End of Cycle HFP - Hot Full Power HHP Hot Half Power HZP = Hot Zero Power LCO = Limiting Condition for Operation LHR =.Linear Heat (generation) Rate (in a fuel pin)

MDNBR = Minimum DNBR MTC = Moderator Tempera*ture Coefficient PCP OW = Percent POWer PCM = Percent Mil = 10- 5 RCS = Reactor Coolant System RPS = Reactor Protection System RTD = Resi~tance Temperature Detector SG = Steam Generator TM/LP = Thermal Margin/Low Pressure (reactor trip)

TS = Technical Specification or Technical Specifications T-H = Thermal-Hydraulic Ql T-H ANALYSIS How do you perform your T-H analysis?

Al Our design code for T-H analysis is COBRA. Inputs to COBRA are RCS flow, RCS pressure, core inlet temperature, *an axial power profile, the one-eighth assembly pin by pin power distribution, and an assembly power.

COBRA assumes the assembly power distribution is one-eighth assembly symmetric, and that the assembly is immersed in an infinite sea of identical assemblies. In its computation COBRA takes account of cross flow, mixing, and two-phase flow. Using all of the above COBRA computes DNBR. A more detailed description of the procedure for calculating DNBR with COBRA can be found in XN-NF-77-22.

null81-0173h-43-46 1

. I I

Q2 PEAKING FACTORS AND TRANSIENT ANALYSES How does GPC demonstrate that the transient analyses in XN-NF-77-18 bound the transient results for the current cycle?

A2 In XN-NF-77-22, Section 2 it is shown that for the peaking factors and coolant conditions used in the transient analyses in XN-NF-77-18, the reactor would have a steady-state ~IDNBR = 1. 30 at a steady-state 115 PCPOW. This means that every transient analyzed in XN-NF-77-18 produces a DNBR degradation no greater than would be produced by running at 15% overpower.

For every cycle the peaking fact.ors (LCOs which appear in the TS) are chosen so that at 15~~ overpower and "Design" coolant conditions the MDNBR is 1.30. The numerical values for these peaking factors for cycle 5 are given on Page 38 of XN-NF-81-34. "Design" coolant conditions are the most adverse coolant co.nditions of RCS flow, RCS pressure, and RCS inlet temperature allowed by the TS during normal operation. The specific values for these design RCS parameters also appear on Page 38 of XN-NF-8i-34. The sections of the TS which delineate the allowed coolant conditions do not change from cycle to cycle.

With the peaking factors for cycle X chosen in this way, in cycle X there exists a steady-state margin in DNBR at HFP equivalent to 15% overpower.

Thus i f the degradation in DNBR is no greater than the degradation which appeared in the analysis of XN-NF-77-18, then for cycle X the transient would not result in a DNBR of less than 1. 3.

In most transients the core is represented by a point kinetics model.

For these cases the course of the transient does not depend on the detailed geometry of.the core, but only on the point kinetics reactivity parameters, the rod drop time, and the shutdown margin. If these parameters are no more adverse for cycle X than for the reference transient analysis, then the reference transient analysis bounds the transient analysis for cycle X. Transient analyses which must account, at least in some measure, for three dimensional effects are the Dropped Rod Event, the Ejected Rod Event, the Single Rod Withdrawal Event, and the Steam Line Break Event. For these events, the power peaking in the core must be computed, and thus for these events the parameters which affect power peaking, as well as the reactivity parameters, must be shown to be less adverse for cycle X than for the reference analysis for the reference analysis to be bounding.

Q3 T-H COMPUTATION OF RADIAL PEAKING FACTORS RELOAD SAFETY REPORT, XN*NF-81-34, PAGE 38 Exactly how are the T-H pinwise peaking factors computed?

A3 The FR (Assembly radial peaking factor) is taken from the TS. The maximum power assembly for the core in question is modeled in an infinite sea of similar assemblies and the pin by pin power distribution for that assembly is determined by a 2D PDQ. F£ is (maximum pin power)/(assembly average pin power). The total radial pin peaking factor is FR*F£. The null81-0173h-43-46 2

HFP axial power profile from Page 13 of XN-NF-78,.16 (peak at 60% of core height) is used.

Osing the design RCS conditions., the above assembly pin power distribu-tion and axial power profile,. and a powr of 115 %POW, COBRA is run to find MDNBR. Invariably MDNBR turns out to be greater than 1.30. The power in the hottest pin is raised until COBRA computes an MDNBR of 1.30.

The Fii. for this new assemblypin power distribution is computed and the new Fii. times the original FR is taken as the total radial pin peaking factor which becomes the new TS LCO limit.

Note there are several types of total radial pin peaking fac:,tors which are explained in the answer to question 30.

Q4 COMPUTATION OF RADIAL PEAKING FACTORS RELOAD SAFETY REPORT, XN"'.NF-81-34, PAGE 38 Three values for FR are used on this page: 1.43 and 1.46 for the T-H analysis, and 1.45 for the LOCA analysis. The TS value is 1.43. Please explain this apparerit anomaly.

A4 The anomaly on Page 38 arose because of a misunderstanding of definitions.

Originally this calculation was done with all *FRs equal to 1.45, which was the TS value.

The definition of FR is FR = (Hottest assembly power)/(Average assembly power)

For the T-H calculation, somehow Exxon misconstrued the.definition of FR to be FR= (Highest Assembly LHR)/(Core Average LHR), where (Assembly LHR) = (Total Assembly Power) and (Number of feet of active fuel pins in assembly)

(Core LHR) = (Total Core Power)

(Number of feet of active fuel pins in core)

We have compensated for this misunderstanding by making the TS FR = 1.43, the lowest FR which appears on Page 38.

This misunderstanding has been clarified to all parties involved, and will not recur in future reloads.

nu1181-0173h-43-46 3

Q5 CORE KINETICS PARAMETERS - DTC & MTC Give the DTC and MTC used in analyzing the various transients in the Transient Analysis Report, XN-NF-77-18, and the DTC and MTC which apply to the Reload Safety Report, XN-NF-81-34.

AS In the Reload Safety Report, XN-1'.lf-81-34, the values of DTC and MTC reported are best estimate computed values for cycle 5. In the Transient Analysis Report, XN-NF-77-18, the values of DTC and MTC may be best estimate computed values, or else values somewhat more conservative than the best estimate computed values. The more conservative values would be used in order to increase the probability that the inputs for the transient analysis being performed wi11 bound the inputs which occur in future cycles, thus obviating a reanalysis iri future cycles.

In the HZP analysis in XN-NF-77-18 the HFP values for DTC are used, which is a conservative approximation as the DTC is more negative at HZP than at HFP.

The best estimate computed values of DTC have an uncertainty of somewhat less than 10%. To compensate for this uncertainty and to bound the cycle to cycle variation in DTC, the values of DTC used in the transient analyses in XN-NF-77-18 are the nominal values times either 0.8 or 1.2, whichever makes the results of the transient more adverse.

The cycle 2 values of DTC (pcm/degf) used in XN-NF-77-18 are as follows:

BOC-HFP BOC-HZP EOC-HFP EOC-HZP 0.8*Nominal -0.87 -1.20 -1.10 -1.50 Nominal -1.09 -1.50 -1. 38 -1. 88

1. 2*Nominal -1. 31 -1.80 -1.66 -2.26 The best estimate computed cycle 5 values of DTC (pcm/degf) which apply to XN-NF-81-34 are as follows:

BOC-HFP BOC-HZP EOC-HFP EOC-HZP

-1.29 -1.55 -1.49 -1. 73 Bounding values of MTC are used in the transient analyses in XN-NF-77-18.

In XN-NF-77-18 the BOC HFP-MTC is +5.0 pcm/degf, which is the maximum value of MTC allowed by TS 3.12. TS 3.12 specifies that HZP MTC shall be measured at start-up, specifies what uncertainty will be applied to the measurement, and requires that the result not exceed +5.0 pcm/degf.

The values of MTC (pcm/degf) used in the Transient Analysis Report, XN-NF-77~18 are as follows:

BOC EOC

+5.0 -35.0 nu1181-0173h-43-46 4

The cycle 5 values of MTC (pcm/degf) from XN-NF-81-34 are:

BOC EOC HFP +2.0 (No Xe) -25.6 HZP +3.0 Not computed Although the EOC HZP ~ITC was not computed, the most negative possible value is about the same as the HFP value and well within the bounds of the transient analysis. The principal factor contributing to the change in MTC with power level is the attendant change in boron concentration, and the lowest possible concentration (zero) is assumed in the EOC HFP MTC.

Q6 CORE KINETICS PARAMETERS - MTC RELOAD SAFETY REPORT, XN*NF-81-34 In this report I see the following three different sets of HFP MTC (pcm/degf):

BOC EOC Page 4 2.0 -25.6 Page 24 -4.5 -25.6 Page 40 5.0 -35.0 Please explain this apparent anomaly ..

A6 XTG computes reactor conditions time step by time step throughout the cycle. The BOC MTC on Page 4 is what XTG has at the beginning of the first time step. XTG assumes the reactor goes instantly from HZP to HFP without building in any Xenon. In fact after reaching HFP Xenon builds in and the RCS boron must be diluted to keep the reactor at HFP. With the decreased boron the HFP MTC becomes -4.5 as is indicated on Page 24.

Th~ values on Page 40 are the values in the Transient Analysis Report, XN-NF-77-18 and are simply reproduced in the Reload Safety Report for comparison. The rationale for reiterating these parameters is given in the answer to Question 7.

Q7 CORE KINETICS PARAMETERS USED IN TRANSIENT ANALYSIS RELOAD SAFETY REPORT, XN-NF-81-34, PAGE 40 Why are these parameters labeled "Core Kinetics Parameters Used in the Palisades Cycle 5 Plant Transient Analysis" when they are actually the parameters used in the Transient Analysis Report, XN-NF-77-18, and have nothing to do with cycle 5?

A7 These parameters are listed for comparison with the cycle 5 parameters.

Each cycle 5 parameter is more conservative than the corresponding_

parameter listed on Page 40. Comparison of cycle 5 parameters to the parameters used in XN-NF,.77-18 in this manner permits us to decide whether or not a transient must be reanalyzed for cycle 5.

nul181-0173h-43-46 5

QB CORE KINETICS PARAMETERS RELOAD SAFETY REPORT, Xi'J".'NF-Bl-34, PAGE 40 Aren't the EOC and BOC column headings .reversed?

AB Yes, they are.

Q9 CORE KINETICS PARAMETERS - OTC - ROD DROP EVENT AND EXCESS LOAD EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-lB, PAGES 43-50 AND 7B-BO For these two events you have not stated what Doppler coefficient multiplier was used in the transient analysis. What Doppler coefficient multipliers did you use?

A9 BOC-HFP Doppler Multiplier = O.B EOC-HFP Doppler Multiplier = 1.2 QlO SHUTDOWN MARGINS USED IN TRANSIENT ANALYSES TRANSIENT ANALYSIS REPORT, XN-NF-77-lB There are no shutdown margins stated for any analyses in this report, except for the Steam Line Break from HFP and HZP, where the shutdown margin is given as 2.0%. What shutdown margins were assumed for the other transients?

AlO In all cases except Steam Line Break a 2.9% net rod worth is assumed.

The HFP shutdown margin depends on the OTC and MTC which vary from event to event and hence the HFP shutdown margin varies from event to event.

For any OTC and MTC assumed in the transient analyses, the 2.9% net rod worth corresponds to a HFP shutdown margin of <2.0% at BOC and <<2.0% at EOC. For HZP events a shutdoWI1 margin of 2.0% was assumed.

The asslimed 2.9% net rod worth is very conservative. A typical value for net rod worth would be about 4.5%.

For the Steam Line Break Event more shutdown margin is required, and for both the Hf.p and HZP cases the shutdown margin is assumed to be 2. 0%.

For this event a value is assumed for net rod worth which equals the HFP to HZP reactivity defect plus 2%. This assures that the reactor trip shuts the core down by 2.0%. After the trip the course of the transient is tracked using a point kinetics model. the usual constant values used for MTC and OTC in other transients would be inadequate for the Steam Line Break analysis because of the large temperature swing. To accommodate this large temperature swing this analysis uses the curves of Figure 3.Bl (reactivity vs moderator density) and Figure 3.B2 (reactivity vs power level) which essentially provide temperature and power dependent values for MTC and OTC.

The assumed shutdown margin of 2. O~~ has been in the TS since cycle 1.

Fortuitously, in the HFP Steam Line Break analysis in XN-NF-77-lB, the 2.0% shutdown margin gave an MDNBR of exactly 1.30. Since the Steam Line Break is an unlikely event (Class IV incident), an MDNBR of somewhat less than 1.30 would be permissable. Since the Steam Line Break.Event is the nu11Bl-0173h-43-46 6

• )

j limiting event for determining required shutdown margin, having a 2.0%

shutdown margin is a mandatory requirement for the core design, and this 2.0% margin is specified in TS 3.10.

Qll SHUTDOWN MARGIN COMPUTATION RELOAD SAFETY REPORT, XN-NF-81-34, PAGE 25 In the first column of this page, what are the (1), (2), and (1-2)?

All The row labeled (1-2) is row (1) minus row (2).

Note there is a numerical error in the table. Under Cycle 4, BOC, HZP the (1-2) number should be 2.22, not 2.24.

nul181-0173h-43-46 7

Ql2 TRANSIENT ANALYSIS TEMPERATURE & PRESSURE INPUTS TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 6.

Here you have T.in. = 537.5 +/- 5.0 DEGF and Press= 2060 +/- 50 PSIA. When you analyze a transient you must choose a single value for each input parameter. What values did you use in the transient analys{s?

Al2 Except for the Loss of Load Event, all events are DNB limited. For all the DNB limited events we used the most adverse values for these two parameters, namely T. = 542.5 DEGF and Press = 2010 PSIA.

in The Loss of Load Event can be DNB limited or pressure limited depending on what assumptions you make about the operability of the pressurizer relief valve and spray and the steam dump and turbine bypass valves. The results of the Loss of Load Event under different assumptions are shown on Page 95 *of XN-NF-77-18. Here it can be seen that when the event is DNB limited the initial pressure is 2010 PSIA and when the event is pressure limited the initial pressure is 2110 PSIA. *In all cases the inlet temperature used is 542.5 DEGF.

Q13 INLET TEMPERATURE USED IN TRANSIENT ANALYSES In both the Reload Safety Report, XN-NF-81-34, and the Transient Analysis Report, XN-NF-77-18, for 102% Power Tin= 537.5 +/- 5.0 DEGF. However, T.

in is not specified for 52% Power or 0% Power.

What values of T. were used in the transient analyses at 52% Power and in O~~ Power?

Is Tin programmed to these values?

Is Tin specified in the TS?

A13 In the transient analyses T. = 542.5 DEGF regardless of the power level.

in This is conservative because . T.in will always be less than that.

T. is not explicitly programmed, but the T program determines what T.

in av in will be at any given power level. At HZP T. = 532 DEGF and at HFP T. -

in in 536 DEGF.

T. is specified in TS 3.1.1.g. This TS calls for the limits on T. to in in vary with pressure and flow (ie, the number of SG tubes plugged).

T. is checked by the operator and logged once per shift, but the in frequency of monitoring is not specified in the TS .

. nuli81-0173c-43 8

• 9 Ql4 UNCONTROLLED ROD WITHDRAWAL EVENT In the Reload Safety Report, XN-NF-81-34, page 39 you have the following reactivity in~ertion rates:

102% Powex: 1.0E-5 S Ap/s S l.4E-4 52% Power: 6.0E-5 S Ap/s s 6.0E-4 In the Transient Analysis Report, XN-NF-77-18, pages 18-42, you have the following reactivity insertion rates:

102% Power: l.OE-5 S 6p/s S 3.. 0E-4 52% Power: 1.-0E-5 < Ap/s < 6.0E-4 Why are the limits different in the tw.o reports?

A14 The only analysis done was in XN-NF-77-18. The range of Ap/s indicated above appears to clearly bound the MDNBR for the inputs used in XN-NF-77-18.

the 6p/s limits stated above for XN-NF-81-34 very closely bound the 6p/s that correspond to the minimums of the MDNBR on the MDNBR vs Ap/s graphs in XN-NF-77-18 (except that for the 52% Power case the lower limit should be 4.0E-5).

Actually there is no reason to suppose that for the inputs in XN-NF-81-34 the MDNBR would lie within the same bounds of Ap/s as they did for the inputs in XN-NF-77-18, and thus it wo~ld have been more appropriate simply not to make any statement about the Ap/s range in XN-NF-81-34.

The important parameters in determirig MDNBR for Uncontrolled Rod Withdrawal are the DTC and MTC. In XN-NF-77-18 maxima and minima values were assumed for DTC and MTC, and these maxima and minima bound the maxima and minima OTC and MTC given for cycle 5 given in XN-NF-81-34.

Thus the XN-NF-77-18 analysis is bounding and it is not necessary to repeat the analysis for cycle 5.

Ql5 . ROD DROP EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 43.

In the middle of the page it states "The reactor regulating system was assumed to be in the manual mode, therefore conservatively inhibiting automatic rod insertion during the transient." It seems to me that the word "conservatively" should be scratched, and instead of reading "automatic rod insertion" it should read "automatic rod withdrawal." Is this correct?

.Al5 Yes.

null81-0173c-43 ' 9

• 1 e Q16 ROD DROP EVENT TRANSIENT ANALYSIS REPORT, x"'N-NF-77-18, PAGE 45.

There is a slight increase in power for .a' few seconds after.the rod drop, but then the power drops off after that. Why doesn't the MTC cause the power to return to its original level?

Also, why does the reactor trip a.fter 73 seconds?

Al6

• There is no return to power because this is a BOC case where the MTC is positive.

The reactor trip is caused by low pressurizer pressure.

Q17 ROD DROP EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 43.

Why is the automatic turbine runback feature for the Rod Drop Event disabled?

A17 The Turbine Runback feature was initially installed to equalize the primary and secondary power in a Dropped Rod Event as soon as the rod dropped. This feature offered two advantages:

(1) The reactor is less likely to trip i f the primary and secondary powers are equalized.

(2) A return to. power on the primary side is prevented by the Turbine Runback, thus reducing the probability that DNB will occur.

However, it was found that the Turbine Runback feature caused more problems than it solved. The two most prominent problems were:

(1) At BOC the MTC may be positive. If the turbine runs back the secondary power to less than the primary power, the primary heats up. With the positive MTC this causes the power in theprimary to increase, thus exacerbating the adverse effects of the transient.

(2) Due to the primary heatup if the turbine runs back the secondary power to less than the primary power, the Turbine Runback may cause an overpressurization event.

Because of these problems, it was decided to disconnect the Turbine Runback feature.

nu1181-0173c-43 10

.1 Ql8 ROD DROP EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 43.

Here it states that for the Control Rod Drop Event "The analysis was performed for the maximum and minimum .expected dropped rod wo.rths at both beginning and end of cycle conditions."

Rod Drop analysis is_ ordinarily done by first computing the (flp,FR) combinations corresponding to individually dropping each rod in the core.

Then the (llp ,FR) combinations which appear to be likely cand;i.dates for producing the lowest MDNBR in the transient are used in analyzing the transient. Lowering llp and raising FR make the transient more adverse.

From your description in XN-NF-77-18 it appears that you are using an incorrect procedure. Please explain this apparant discrepancy.

The (llp,FR) combinations for cycle 5 do not appear in XN-NF-81-34. What were the cycle 5 values of (llp,FR)?

Al8 The method we use to analyze this event is correct, but the description in XN-NF-77-18 is incorrect. We do first compute the (llp,FR) combinations corresponding to dropping each rod in the core, which is not explicitly stated in XN-NF-77-18. The quote from XN-NF-77-18 cited in the question above should read "The analysis was performed for the (llp,FR) combination containing the minimum llp and the (llp,FR) combination containing the maximum FR at both beginning and end of cycle conditions."

Since the analysis in XN-NF-77-18 was performed, we have observed the effects of the following simplifying assumption made in our analytical model: Our model assumes a constant FR in the core throughout the course of the transient, which is the FR -that applies after the rod is dropped.

Actually the core FR goes from its initial steady-state value to its maximum value in about the same time period that the reactor goes from its initial steady-state power to the power corresponding to the negative reactivity insertion of the dropped rod.

After running a number of Rod Drop Events with our analytical model, we have observed that the effect of this simplifying assumption is that in our analysis the thermal conditions of power, pressure, temperature, and flow are the most adverse at the beginning of the transient before the rod has had a chance to drop. From this we draw three important conclusions:

(1) Our analysis is always conservative.

nu1181-0173c-43 11

• J *

. i (2) The value of t:.p has no influence on our computed MDNBR, and thus it is only necessary to examine the (t:.p,FR) combinations with the largest value of FR.

(3) The MTC and OTC have no influence on our computed MDNBR.

For cycle 5 at BOC the (t:.p,FR) pair with the largest FR was (0.121,~, 1.505). For this pair the XN-NF-77-18 analysis is bouri.ding by a wide margin.

The cycle 5 operating power distribution is flatter at EOC than at BOC.

The FR after the rod drop is fairly well approximated by: (FR before rod drop)* (azimuthal tilt caused by dropped rod). We have no reason to suppose the tilt caused by the dropped rod will be significantly higher at EOC than at BOC. Therefore we expect the maximum dropped rod FR at BOC to be larger than at EOC. This was born out by our analysis in XN-NF-77-18. For this reason we expect the BOC Rod Drop Event to have more adverse consequences than the EOC Rod Drop Event, and for cycle 5 we computed the (t:.p,FR) pairs only for BOC conditions.

Also the table at the bottom of Page 43 should be deleted, since it misleads the reader into thinking that we only look at t:.p, and all the information in this table is presented in the table on the top of Page

44. It would be better i f the table on Page 44 were presented as follows:

Reactivity Case Feedbacks (t:.p,FR) Condition (t:.p,FR) Value MDNBR 1 BOC BOC Minimum t:.p (-0.04%, 1. 60) . 1.42 2 BOC BOC Maximum FR (-0.12%, 1.66) 1.35 3 EOC EOC Minimum t:.p (-0.04%, 1.60) 1.42 4 EOC EOC Maximum FR (-0.12%, 1.64) 1.40 Incidentally XN-NF-77-18 has mistakenly listed t:.p in units of numerical change. It should be in % change.

nul181-0173c-43 12

... .J ...

Q19 REDUCTION IN FEEDWATER ENTHALPY EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 66.

For the Reduction in Feedwater Enthalpy Event, why do you use BOC OTC &

MTC? (Most positive values)

For the Increased Feedwater Flow Event you use EOC DTC & MTC (Most negative values) which maximizes the power increase due to cooldown. It seems that EOC conditions would produce the most adverse results in the*

Reduction in Feedwater Enthalpy Event as well.

A19 In the FSAR the Reduction in Feedwater Enthalpy Event was analyzed for both BOC & EOC conditions. The BOC case was more limiting and therefore this was analyzed in XN-NF-77-18.

Later this was questioned by the NRC and we ran the EOC case as well.

This is reported in XN-NF-81-25, Rev 1. The results were as follows:

BOC Coefficients MDNBR = 1. 75 EOC Coefficients MDNBR = 1.65 Q20 REDUCTION IN FEEDWATER ENTHALPY EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 66.

For the Reduction in Feedwater Enthalpy Event at 102% Power you assume an enthalpy reduction of 52 BTU/LB. Where did this number come from?

A20 This is the enthalpy added by the high pressure heater in the feedwater train. The failure of any other heater would have a smaller effect.

Q21 INCREASED FEEDWATER FLOW EVENT TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 67.

From the wording on this page I would get the impression that X% Power corresponds to X% Feedwater Flow. Don't you actually have X% Power corresponds to less than X% Feedwater Flow, so that at 100% Power you still have some excess feedwater capacity?

A21 Yes, we do have excess capacity.

It would be more accurate to.state that in this event the feedwater controller increases the feedwater flow from a flow corresponding to 50%

Power to a flow corresponding to 100% Power.

Q22 EXCESS LOAD EVENT In the Transient Analysis Report, XN-NF-77-18, the Excess Load Event is initiated from HFP and HZP.

null81-0173c-43 13

In the Reload Safety Report, XN-NF-81-34, the Excess Load Event is initiated from HFP.and. 52'~ Power.

Why the difference?

A22 Editorial error.' It should be HFP and HZP in both reports.

Q23 EXCESS LOAD EVENT TRANSIENT ANALYSIS REPORT, XN*NF-77-18, PAGES 78-80.

For the HZP case (graph on page 86) there is a return to about 9% Power at 128 se~onds. After this the power drops very slowly (about 7% Power at 200 seconds and then the graph ru.ns out). Yet the core continues to cool down.

Where is all the heat going?

A23 In this analysis it is assumed that the steam dump and turbine bypass valves fail open and there is no operator intervention to close the valves. This is a cqnservative assumption because the opening of these valves causes the cooldown which, in turn, causes the return to power.

The same assumption i~ made in the FSAR. With the reactor tripped and the steam dump and turbine bypass valve controller in automatic, these valves would automatically open when the RCS temperature gets above 532 DEGF, thus preventing the core from overheating. If th*e controller were in manual and*the operator takes no action, the spring loaded SG safety code valves would open at a SG pressure of 1000 PSIA, so the core would still be protected from overheating.

In XN-NF-77-18 a net rod worth of 2.9% was assumed. Actually the net rod worth is about 4.5%, so that in the real reactor the return to power would probably not occur.

Q24 STEAM LINE BREAK

• TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGES 118-136.

For this event instead of using a MTC and a OTC, the moderator and power effects on reactivity are given by a graph of moderator density vs reactivity on page 122 and a graph of power vs reactivity on page 123.

How are these curves used in the computation?

A24 This question is best answered by summarizing the whole procedure for computing MDNBR.

PTS is our Plant Transient Simulation code. PTS performs a time step by time step computation of total core power and coolant conditions, modeling the core cooldown due to SG blowdown. For each time step PTS performs a point kinetics computation of total core power, and then performs an axially zoned T-H analysis to compute the point kinetics reactivity feedbacks to be used in the point kinetics computation in the next time step.

nu1181-0173c-43 14

. *II I

The computation of the reactivity feedbacks from the axially zoned model of the core are performed as follows.

The first step in computing the reactivity feedbacks in the axial slices is to determine the average power in each axial slice. This is done by combining the total core power from the point kinetics computation with an input axial power shape. Having an average power for each slice~ the reactivity feedback for that slice is picked off the graph on Page 123.

The axial slice reactivity feedbacks are averaged to get a Doppler reactivity feedback for the point kinetics .equation.

Having the average power in each slice, a T-H analysis is performed to, determine the average temperature, pressure, and flow in each slice. The average moderator density for each slice is computed as if these .

parameters were constant throughout the slice. The moderator reactivity feedback for each slice is picked off the graph on Page 122. The axial slice reactivity feedbacks are averaged to get a moderator reactivity feedback for the point kinetics equation. Note the PTS calculation does not model the voiding in the vicinity of the stuck rod, which adds conservatism to the calculation. The important outputs from PTS are the time step by time step values of total core power, inlet temperature, pressure, and flow.

The power peaking in the vicinity of the stuck rod is computed by 3D XTG.

XTG accounts for Doppler feedback and void reactivity feedback in the vicinity of the stuck rod. The void reactivity feedback uses a simple closed channel T-H model to compute the void. Inputs to XTG are inlet temperature, pressure., and flow from the PTS calculation.

COBRA then computes MDNBR at each time step using the total core power, inlet temperature, pressure, and flow computed by PTS combined with the radial peaking and axial power shape computed by XTG.

Q25 SG TUBE PLUGGING AND RUPTURED SG TUBE EVENT TRANSIENT. ANALYSIS REPORT, XN-NF-77-18, PAGE 8.

Here you have indicated that a number of SG tubes have required plugging.

In view of the problems you are having with SG tubes, how do you justify not performing a Ruptured SG Tube Event analysis in XN-NF-77-18?

A25 The Ruptured SG Tube Event was analyzed in the FSAR. The FSAR input parameters bounded the input parameters used in XN-NF-77-18, and hence the Ruptured SG Tube Event was not reanalyzed in XN-NF-77-18.

nu1181-0173c-43 15

Q26 SG TUBE.PLUGGING TRANSIENT ANALYSIS REPORT, XN-NF-77-18, PAGE 8.

Here you indicate that in 1977 you had the following number of active SG tubes:

SG #1 8519 tubes total 1929 tubes plugged 6590 active tubes SG #2 8519 tubes total 1744 tubes plugged 6775 active tubes How many active tubes do you have now? Do analyses in XN-NF-77-18 which depend on SG characteristics still apply in cycle 5?

A26 None of the analyses in XN-NF-77-18 require a knowledge of the number of active SG tubes. What these analyses do require is a knowledge of what RCS flow the SGs wi 11 support. According to TS 3. 1. 1. c, the measured total RCS flow rate at HZP must ~e at least 126.9 Mlbs/hr. Due to the change in density when going from HZP to HFP, this HZP flow rate corresponds to an HFP flow rate of 121. 7 Mlbs/hr.

• 121. 7 Mlbs/hr is the HFP flow rate assumed in XN-NF-77-18. The HZP RCS flow rate measurements are conducted once per cycle at start-up and after each tube plugging.

The last measured value of HZP RCS flow was 12.8. 9 Mlbs/hr. This assures that the SGs have the flow capacity assumed for all transients analyzed in XN-NF-77-18.

In order that it would be bounding for future cycles, the LOCA analysis in XN-NF-77-24 was performed assuming that an extra 502 SG tubes were plugged. These 502 tubes were divided between the two SGs, with most of the 502 plugged tubes being in SG #1. The number of plugged and active tu.bes assumed for each SG were as follows:

SG #1 8519 tubes total 2407 tube~ plugged 6112 active tubes SG #2 8519 tubes total 1768 tubes plugged 6751 active tubes In the Large Break LOCA analysis the break was conservatively assumed to

• be on Cold Leg #1. Since 1977, when this analysis was performed, we have had to plug about 80 additional tubes. Thus the XN-NF-77-24 analysis still has a lot of margin with respect to number of active SG tubes. The number of active SG tubes listed here was also used in the LOCA analysis described in XN-NF-78-16. TS 4.14 requires that. i f 64% of an SG tube wall has been worn away the tube must be plugged.

Q27 LOCA: RELOAD SAFETY REPORT, XN-NF-81-34.

The expression "Time of PC" appears on page 46. Shouldn't* this be "Time of PCT"?

A27 Yes.

nul181-0173c-43 16

I*

I I

Q28 RTD TIME RESPONSE Recently there has been a great deal of.attention paid to RTD time responses at a .number of reactor facilities. I see no mention of RTD time responses in either your TS or transient analyses. Why?

A28 Palisades was one of the earliest plants constructed, and at the time littl~ attention was being paid to RTD response time. Thus RTD response time was not in our original TS or Transient Analyses, and the issue has not been raised *ince that time.

Q29 TM/LP SETPOINT CIRCUITRY In the CE Setpoint Methodology Report, CENPD-199, it is stated that in the RPS for "early systems" the Til/LP circuitry computes Q = Max(Neutron Power, ~T Power)

PVAR = a

• PF(Q)

• Q+ 6

• TCAL + Y

~PF In your TS you state PT rip

. =A

• TH - B

• TC - C.

Which of these equations actually represents your TM/LP circuitry?

A29 Our TS is correct. The TM/LP circuitry at Palisades is an even earlier design than what is called an "early system" in CENPD-199. Our TM/LP circuitry is not described in any topical report.

The curves which our TM/LP circuitry must simulate are given in TS Figure 2-3. Our Til/LP circuitry is designed so that it can change the values of the constants A, B, and C at preselected cut points in reactor power. This feature gives our TM/LP circuitry the capability of accommodating the break points at 100% power in the curves of TS Fig-ure 2-3.

nul181-0173c-43 17

Q30 DEFINITIONS OF POWER PEAKING FACTORS Define all the symbols you use for power peaking factors.

A30 Assemblywise Pin Peaking Fi= (Maximum Pin Power) In the numerator, we may consider all the pins in j (Average Pin Power) the assembly or only a subset of the pins, such as interior pins, narrow water gap pins, or wide water gap pins.

_ _J Corewise LHR Peaking FQ = FQT = FQN = Fq -- FR A

• F..n (all *pins. in hot assembly)

• Fz Core, Assembly, or Pin Axial Peaking F (z) z

= (Power in an axial slice of core, assembly, or pin) * (Core height)

(Total power in core, assembly, or pin) (Slice thickness)

F = The maximum value of F (z) for any z z z Usually the (z) is suppressed in F (z), so that the same expression, F ,

z z refers either to the local value of F (z) at height z or to the maximum z

value of F (z) for any z. Usually it is obvious from context which z

definition applies.

Corewise Radial Peaking Factors (Including Tilt Allowance)

F xy = Fr - .F R - FAR

= (Maximum assembly power)

(Average assembly power)

F; = F~ = (Maximum pin power)

(Average pin power)

Flilir = Ft.HR -- FAH -- (M aximum

. power o f pin. interior

. . to an assem b 1y )

(Average corewise pin power)

FN = (Maximum power for pin in a narrow water gap) r (Average corewise pin power) null81-0173c-43 18

ATTACHME~lT 3 Reevaluation of the Steam Line Break Accident for Palisades Cycle 5 rleevaluation of the steam line break analysis for Palisades in XN-NF-77-18 has been performed for Cycle 5. Results of the analysis are summarized in Table 1 along with the previous results. The Minimum Critical Heat Flux Ration (MCHFR) for the hot zero power steam line break was calculated to be 1.40 versus the value of 1.41 previously reported. The MCHFR for );he full power steam line break was calculated to be l. 35 versus the 1. 30 value previously reported. *As in the previous analysis, these MCHFR results are sufficiently high that the number of rods which would be calculated to experience boiling transition in a steam line break accident at Palisades is negligible (<1%).

In the reference analysis, a Modified Barnett Correlation (1) was used to calculate MCHFR. In the present analysis, the same basic reference has been used, but with some ;important changes in application. In particular, the present analysis employs the M:>dified Barnett Correlation on an assembly cross-sectional average basis consistent with Reference 1, and consistent with the original work by Barnett (2). In the XN-NF-77-18 analysis, the r(-odified Barnett Correlation had been applied on an overly conservative subchannel basis. The excess convt!r-satism in the previous analysis is estimated to be about 25% to 30% in MCHFI\.

Eliminaticn of this conservatism in the present analysis has offset the increase in stuck rod total peaking calculated for Cycle 5 (approximately 22~ increase) shown in Table 1.

Two other factors have contributed to the improved MCHFR *for the full power steam line break case. These are:

(1) A reduction in the XTG calcula.ted radial peaking about the st:ick control rod for Cycle 5 versus the reference analysis.

(2) A further reduction in radial peaking by taking into account that a portion of the core power at the time of thermal mar3in limiting condi-tions is due to decay heat.

The reduced radial peaking for Cycle 5 relative to the previous analysis leads to sli8htly improved flow in the hi~h power region of the core, and yields a corresponding smali improvement; in MGHFH.

Differences between Cycle 5 neutronics and those for Cycle 2 necessitated the present reevaluation. Since Cycle 5 shutdown margin remains in excess of

.technical specifications limits, and since other factors in the steam line break analysis are not changed, it was not necessary to ~epeat the plant system transient analyses. It was necessary, however, to reevaluate power peaking about the stuck rod location with XTG neutronic calculations and to reanalyze core thermal.

hydraulic conditions at the time MCHFR occurs in the steam line break accidents.

The neutronic evaluation has shown that total peaking about the stuck rod location is increased relative to the reference analysis. Maximum radial peaking is calculated to be less than in the reference analysis, but axial peaking is higher.

The increase in axial peaking is a result of a difference in axial exposure distribution between Cycle 5 and the reference analysis. In .Cycle 5 the core will be composed of assemblies which have experienced one, two and three cycles of irradiation. In the reference analysis, all assemblies had experienced only one cycle of irradiation. Another reason.:or the changes in power peaking is that the location of the highest worth stuck rod has changed from a Group 3 to a Group 1 rod.

The.XTG neutronic calculations follow the same methodology as in the previous analysis. As in the previous analysis, additional conservatism is introduced by neglecting the hydraulic feedback effects associated with core flow distribution .

. Thus, reduced flow to high power core locations is conservatively not included in the XT.G calculations.

In performing the detailed core thermal hydraulic analysis* for the full power steam line break, the core power distribution included an account of the core power due to decay . heat (3.5~ of full

. power at 120 seconds). This accounting of decay heat, which was. also done in the FSAR analysis of the steam line break, yields a slightly flatter radial power distribution than results from the XTG calculations alone.

As in previous analyses, the modified Barnett CHF correlation has been used to establish thermal margin to boiling transition. However application of thi5 correlation in the pre:;ent analysis differs from the previous application .. Speci-

fically, the final form of the Eodified Barnett Calculation(l) has been employed on a bundle cros.s-sectional average basis, where the previous analysis e!

lployed an intermt:?diate form of the Modified Barnett Correlation on a subchannel bas is. The

!-bdified Sarnett CHF Correlation was derived on the basis of and intended for use

. with bundle cross-sectional averai;e parame~ers. The equivalent hydraulic and

heated diaraeters in the present analysis have been defined in accordance with the usage of Barrett (2)

• Finally, the correlation has been conservatively modified for application to non-uniforrn*axial heat flux profiles. Due to the removal of excess conservatisms in the application of the. Modified Barnett CHF Corre.].ation, the present analysis results in improved thermal margin in spite of increased total peaking in the vicinity of the stuck rod.

REFERENCES :

(1) E. Daniel Hughes, "A Correlation of Rod Bundle Critical Heat Flux for Water in the Pressure Range 150 to 725 psia", IN-1412, July 1970.

(2) P.G. Barnett, "A Correlation of Burnout Data for Uniformly Heated Annuli and its Use for Predicting Burnout in Uniformly Heated Rod Bundles",

AEEW-R463, 1966.

Table l Cycle 5 Steam Line Break Analysis Results Previous Cycle 5 Analysis Hot Zero Power Steam Line Break Max. Asse!llbly Radial Peaking 7.17 8.09 Max. Total Peaking 19.5 16.o Min. Critical Heat Flux Ratio l.40 l.41 Full Power Steam Line Break fl.ax. As serr:bly l\adial Peaking 6.22 8.87 l*lax. Total Peaking 22.4 16.2 Min. Critical l:eat l"lux Ratio . 1.35 1.30