ML20117E447
| ML20117E447 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 08/22/1996 |
| From: | James Knubel GENERAL PUBLIC UTILITIES CORP. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| 6710-96-2230, NUDOCS 9608300002 | |
| Download: ML20117E447 (14) | |
Text
- ,
i
{
GPU Nuclear,Inc.
(
Route 44i South NUCLEAR Post Office Box 480 Middletown. PA 17057-0480 Tel 717-944-7621 (717) 948-8005 August 22, 1996 6710-96-2230 U. S. Nuclear Regulatory Commission Att: Document Control Desk Washington, DC 20555 Gentlemen:
Subject:
Three Mile Island Nuclear Station, Unit 1 (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 Response to Request for Additional Information - Core Reload Methodology In response to NRC questions discussed at the May 30,1996 NRC/GPU Nuclear meeting on the TMI-l Core Reload Methodology effort, GPU Nuclear is providing the attached information.
This additional information is related to GPU Nuclear Topical Repon TR-078, Revision 0, "TMI-1 Transient Analyses Using the RETRAN Computer Code," submitted on March 17,1995 for NRC review and approval for in-house GPU Nuclear core reload design, If any additional information is required, please contact Mr. David J. Distel, GPU Nuclear Regulatory Affairs at (201) 316-7955.
Sincerely, l
J v
(
J.
nubel l
ice President and Director, TMI I
l JK/DJD/ pip Attachment ACOII L
c:
Administrator, Region I i
~
NRC TMI Senior Resident Inspector NRC Senior Project Manager, TMI 9608300002 960822 PDR ADOCK 05000289 P
PDR I
I Attachment 1
j Response to NRC Request for Additional Information (RAI) on TR-078 l
QUESTION 1.0:
l l
Identify how (e.g., explicit analysis, qualitative assessment, vendor supplied, etc.) each l
Chapter 14/15,5.2.2, and/or other non-LOCA event is evaluated, and if the methodology would be used to explicitly evaluate it.
l
RESPONSE
l Chapter 14 accident analyses, hereafter referred to as the ' reference safety analyses', along with the facility Technical Specifications, establish the bases and conditions for safe operation of the plant. The reference analyses include values of parameters assumed in the analyses, performance j
characteristics of the mitigating systems, and the analytical models used. Values of parameters l
selected in the reference analyses are chosen to bound values expected during the life of the plant.
r Performance characteristics of the mitigating systems are modeled to give conservative i
performance characteristics and produce bounding consequences for each of the accidents.
l For core reload design analysis, the key parameters that have the greatest effect on the outcome
~
l of a transient can typically be classified into three major areas: core thermal parameters, thermal-l hydraulic parameters and reactor kinetics parameters, including the reactivity feedback j
l coefficients and control rod worths.
i The reference values for these parameters are compared to those calculated for the particular cycle. If key parameters are within the envelope of values assumed in the reference analysis or can be shown by evaluation to not significantly reduce margins to the safety criteria, then the analysis is valid and no reanalysis is necessary. If, however, one or more of the plant parameters assumed in the reference analysis are found to be nonconservative for the reload cycle to the extent that safety criteria could be compromised, the core design must be revised or those accident analyses which are afTected by the nonconservative parameters must be reanalyzed and submitted to the NRC for review and approval prior to implementation if the change results in an unreviewed safety question.
1 QUESTION 2c0 i
l
. Identify the criterion or criteria (some events could require evaluation versus more than one conDicting criteria) each of the events will be evaluated against.
RESPONSE
The acceptance criteria each of the events will be evaluated against are based on the TMI-l FS AR and summarized in Table 1.
1 i
1 l
Table 1: TMI-l FSAR Chapter 14 Transients and Criteria TRANSIENT / ACCIDENT ACCEPTANCE CRITERIA i
1.
Startup/ Rod Withdrawal a)
The RCS pressure will be limited to less than 110% of design 1
pressure or 2750 psig; b) The reactor thermal power shall not cycced 112% of rated power.
2.
Moderator Dilution a)
The reactor thermal power will be limited to less than the design overpower; b) The RCS pressure will be limited to less than i10% of design pressure or 2750 psig; c)
The reactor minimum shutdown margin of 1% delta-k/k subcritical will be maintained.
3.
Pump Startup a)
The RCS pressure will be limited to less than 110% of design pressure or 2750 psig; b) The reactor thermal power shall not exceed 112% of rated power.
4.
- 1. ass of Coolant Flow The minimum DNB ratio will not be less than 1.18 with the BWC C5fF correlation.
5.
Locked Rotor The minimum DNB ratio will not be less than 1.0.
6.
Stuck-out Stuck-in or Dropped Control Rod a)
The RCS pressure will be limited to less than i10% of design pressure or 2750 psig; b) The reactor thermal power shall not exceed 112% of rated power; c)
The reactor minimum shutdown margin of 1% delta-k/k subcritical will be maintained.
7.
Loss of Electric Power a)
The RCS pressure will be limited to less than i10% of design pressure or 2750 psig; b) The reactor thermal power shall not exceed 112% of rated power; c) Doses will be within 10 CFR 100 limits.
8.
Steam I ine Break a)
The core will remain intact for effective core cooling; b)
No steam generator tube break or separation from the tube sheet will occur due to a loss of secondary side pressure and the resultant temperature gradients.
c)
Doses will be within 10 CFR 100 limits.
9 Steam Generator Tube Failure a)
Additional loss of reactor coolant boundary integrity shall not occur; b)
Doses will be within 10 CFR 100 limits.
- 10. Rod Ejection a)
A control rod ejection accident shall not cause further violation of the reactor coolant system integrity; b)
Doses will be within 10 CFR 100 limits.
I1. Loss of Normal Feedwater a)
The RCS pressure will be limited to less than i 10% of design pressure or 2750 psig; b) The reactor thermal power shall not exceed i 12% of rated power; c)
Pressurizer does not become water solid.
I QUESTION 3.0:
Identify important analysis assumptions for each event and how each assumption is biased in the input to achieve conservative results (some event assumptions may vary depending on which criterion is being addressed).
l l
RESPONSE
In analyzing licensing calculations, assumptions consistent with those documented in the FSAR would be used. The deterministic approach assumes that all components are simultaneously at i
their most adverr.e values. These especially include the more sensitive parameters such as initial conditions, reactivity parameters and system performance assumptions.
The initial system power, temperature and pressure would include errors which minimize core thermal margin or margin to other plant design criteria. The reactivity parameters are chosen in a manner which tends to maximize the nuclear power during the transient. In many instances, the mitigating effect of various system design features on postulated transients are ignored.
In order to adequately account for the impact ofinstrumentation errors and signal delays, conservative protection system characteristics are assumed when performing accident analyses.
Thus, expected instrument errors and system response times are conservatively bounded in the 9
analysis assumptions, thereby adding to the previously discussed conservatisms employed in a transient analysis. These assumptions assure that the deerministic or licensing calculations conservatively ' bounds' the actual expected system performance. Table 2 summarizes the important analysis assumptions (with emphasis on reactivity parameters) for each event and how each assumption is biased in the input to achieve conservative results.
l
TABLE 2:
SUMMARY
OF IMPORTANT ASSUMPTIONS USED IN TRANSIENT ANALYSIS
~.
ACCIDENT /
CONSERVATIVE OTHER MITIGATION FUNCTIONS NQT TRANSIENT KEY PARAMETERS DIRECTION CREDITED (Assumed in Analysis)
- 1. Stantup Reactivity Addition Rate Most limiting (assumes
- 1. Only one control rod group can be
~
simultaneous multiple rod withdrawn at a time.
withdrawal)
- 2. Control rod withdrawal rate is limited.
Moderator Coeflicient Most positive
- 3. Source range startup rate withdrawal stop Doppler Coeflicient Least negative and alarm.
Tripped Rod Worth Minimum
- 4. Intermediate range startup rate withdrawal stop and alarm.
- 2. Rod Withdrawal Reactivity Addition Rate Most limiting (assumes
- 1. Only one control rod group can be simultaneous multiple rod withdrawn at a time.
withdrawal)
- 2. Control rod withdrawal rate is limited.
Moderator Coeflicient Most positive
- 3. High temperature, pressure and level alarms.
Doppler Coeflicient least negative
- 4. High reactor coolant outlet temperature Tripped Rod Worth Minimum trip.
- 3. Moderator Dilution Initial Boron Concentration Maximized
- 1. Integrated control system action Inverse Boron Worth Minimized
- 2. Interlocks and alarms to prevent improper operation Dilution Rate Maximized
- 3. Flow of dilution water is automatically Moderator Coefficient Most positive stopped at a preset value Doppler Coeflicient Least negative
=
TACLE 2:
SUMMARY
OF IMPORTANT ASSUMPTIONS USED IN TRANSIENT ANALYSIS (CONT'D)
ACCIDENT /
CONSERVATIVE OTHER MITIGATION FUNCTIONS NOT TRANSIENT KEY PARAMETERS DIRECTION CREDITED (Assumed in Analysis)
- 4. Pump Startup Moderator Coeflicient Most negative
- 1. Pump controlinterlock prevents accident' from occurring; Doppler Coemcient Least negative
- 2. Pumps cannot be started above 30% power,
- 3. Administrative controls prevent the simultaneous start of two reactor coolant pumps.
- 5. Loss of Coolant Moderator Coemcient Most positive Flow / Locked Rotor Doppler Coemcient Least negative Hot Channel DNB Peaking Maximized Factors System Flow Minimized
- 6. Stuck-out, Stuck-in er Reactivity Addition Rate Maximized
- 1. Integrated control system action; Dropped Control Rod Moderator Coemcient Most negative
- 2. Alarms.
Doppler Coemcient least negative Tripped Rod Worth Minimum i
TACLE 2:
SUMMARY
OF IMPORTANT ASSUMPTIONS USED IN TRANSIENT ANALYSIS (CONT'D)
~
ACCIDENT /
CONSERVATIVE OTHER MITIGATION FUNCTIONS NOT TRANSIENT KEY PARAMETERS DIRECrlON CREDITED (Assumed in Analysis)
- 7. Loss of Electric Power Moderator Coemeient Most positive Doppler Coemcient Least negative Defective Fuel Assumed Primary to Secondary Leak Assumed 8.
Steam Line Break Moderator Coemcient Most negative Doppler Coemcient Least negative Tripped Rod Worth Minimized Steam Generator Secondary Maximized Inventory Break Flow Maximized
- 9. Steam Generator Tube Break Flow Maximized Failure Defective fuel Assumed
- 10. Rod Ejection Ejected Rod Worth Maximized Moderator Coeflicient Most positive Doppler Coefficient least negative
- 11. Loss of Normal Moderator Coeflicient Most positive Feedwater Doppler Coeflicient Least negative
i QUESTION 4.0 Identify RETRAN options which are exercised for each analysis.
i
RESPONSE
Applications of the TMI RETRAN model utilize the phenomenological models built into the RETRAN code. User options which are exercised for all analyses are addressed
)
below:
Non-equilibrium Pressurizer Model for Pressurizer The non-equilibrium pressurizer model is used to predict the non-equilibrium effects 4
(different temperatures) between the vapor and liquid in the steam-water interface region of the pressurizer. The non-equilibrium pressurizer model allows the vapor region to superheat as the RCS fluid expands into the pressurizer. Because of this, the use of this model will produce a higher RCS pressure and is thereby more conservative than the default thermal equilibrium assumptions. In addition, the pressurizer spray is conservatively assumed not to be available.
For most reload transients, the pressurizer never fills or empties. However, if that should occur, it was demonstrated in report NSAC-78, 'RETRAN-02 Analysis of LOFT ATWS Experiments', October 1984, that the pressurizer model behaved appropriately under j
- those conditions.
Algebraic Slip Model The algebraic slip model is used to determine the phase velocity differences for all applications. The algebraic slip model in RETRAN is based on the Zolotar-Lellouche void fraction model. This void fraction modelis based on the drift flux approach, with the empirical coeflicients determined for steady-state, two-phase flow in tubes, channels and bundles. The basic model including a series of qualification analyses, is described in the report, ' Mechanistic Model for Predicting Two-phase Void Fraction for Water in Vertical Tubes, Channels and Rod Bundles', EPRI NP-2246-SR.
i
Th.is RETRAN-02 algebraic slip model has been used extensively in the industry for two-l phase conditions. Most, if not all, organizations submitting topical reports for BWRs use j
this model and have qualified it by nmning several system comparisons with data.
Dubble Rise Model The bubble rise model is used in regions of a RETRAN-02 model when a distinct level is present or may become established. In the TMI model, a bubble rise model is used in the pressurizer and in each steam generator downcomer. In the pressurizer, the initial level is conservatively selected for each specific transient. In the steam generator downcomers, no initial level is specified. A level may become established during the transient as would be expected to simulate post trip or low power situations.
Enthalpy Transport The integral energy balance equation includes a term involving the enthalpy at the flow path or junction between two volumes. If energy is exchanged between the fluid within a volume and the surfaces enclosing the volume, the enthalpy at the junction of the volume 1
may have a different value than the volume averaged enthalpy. In this instance, the change in enthalpy at thejunction will depend on the rate of energy exchange with the 1
surrounding surfaces and the flow rate of the mixture and the phasic velocity slip ratio.
This is accounted for by the enthalpy transport model.' In the TMI RETRAN model, enthalpy transport is applied to all thejunctions associated with the steam generator tube region and the core. This provides a more accurate initial temperature distribution in the i
primary and void fraction distribution in the secondary.
Irmperature Transport Mode]
The temperature transport model in RETRAN-02 allows for changes in system fluid temperature to move as a front. This phenomena occurs in portions of the system where there is no significant mixing such as piping. The default option in RETRAN-02 mixes all incoming fluid with the existing contents of a fluid volume, thereby ' smearing' any temperature front movement. The main influence of the temperature transport model will be in the timing of temperature changes reaching the core and the effect it may have on kinetics feedback.
],.
-y 1
4 I
In the TMI RETRAN model, the transport delay model is applied to volumes with a i
dominant flow direction such as hot leg and cold leg piping volumes only. Most reload transients will not be affected by this phenomena.
t i
Critical Flow When the projected mass flow rate using thejunction thermodynamic properties and the l
Bernoulli equation exceed the critical mass flow rate, choked flow is assumed to occur, The TMI RETRAN model uses the extended Henry-Fauske and Moody choking models t
for the subcooled and two phase regions. The primary safety and relief valves are modeled asjunctions connected to a containment volume at atmospheric pressure.
Contraction coefficients are used on valvejunctions to get the specified flow at the l
reference pressure.
k 1.,
4 j
Sleadv State Initialization The RETRAN steady state initialization option was used to initialize the TMI model. The i
j steady state initialization option provides a consistent set ofinitial conditions so that the transient response is not biased by unbalanced conservation equations. The FSAR values p
for the system flow and core inlet flow were specified. The steam generator aspirator flow
{
E was adjusted to get the correct initial steam generator liquid mass 'as specified by the y
vendor, The steady state initial conditions from RETRAN were compared to information p
i U
transmitted by the vendor, the FSAR and plant data and found to be in close agreement.
L f
4 8
1 A
4
-nm,y-,
ww ve--p-w e.e---,.
w w
+
l
QUESTION 5.0 Provide benchmark analysis comparisons demonstrating appropriate modeling of phenomena.
RESPONSE
REi RAN-02 is a general purpose thermal-hydraulic code which may be used for i
performing operational and licensing analysis of a light water reactor system. The phenomenological models in the RETRAN computer program have been subject to extensive verification and validation, resulting in acceptance of RETRAN-02 by the NRC for licensing applications.
]
The RETRAN model for TMI-1 documented in the topical report is based on a model which has been in use within GPU Nuclear since 1978. The model was design verified by an independent consultant. As verification of the model, appropriate results and comparisons have been performed for a representative series of analyses oflicensing and best estimate plant transients. The results of these predictions are intended to demonstrate the adequacy of the TMI-l RETRAN model and GPU Nuclear's ability to apply the RETRAN code for performing operational support and licensing calculations.
The topical report contains comparisons of a variety of the TMI-l RETRAN model predicted results to experimental test / operational data. These comparisons include pump coastdown tests of various configurations, a reactor / turbine trip and the TMI-2 accident, and demonstrate that the model provides an accurate representation of the reactor plant i
and associated systems and components.
In addition, a comparison to many of the vendor predictions in the FS AR are also contained in the topical report. These demonstrate that for a representative series of transients, the modeling oflicensing conservatisms as applied to the TMI-l FSAR are understood and implemented correctly. The comparisons also demonstrate that the TMI-l RETRAN model provides expected system response for a wide range ofinitiating events.
Topical Report 078, was submitted to the NRC to demonstrate GPU Nuclear's technical competence to use RETRAN 02 and analyze results obtained by the use of the code for applications to TMI-l plant transients. In support of this, a representative series of analyses oflicensing and best estimate plant transients was performed. Table 3 summarizes which benchmark analysis comparisons have been performed. While it was not the intent of the topical report to provide a comprehensive benchmark of every application, but rather a representative series of analyses, it can, however, be seen from Table 3 that most of the reload-related analyses contained in the TMI FSAR have been benchmarked. The two events that were not are now discussed.
The moderator dilution event, which is one of the transients not benchmarked, is a reactivity insertion transient and is bounded by the startup/ rod withdrawal accidents. The positive reactivity insertion is driven primarily by reduction in coolant boron concentration and a small part by the possibility of RCS moderator cooldown from the cold injection water. The FSAR assumes a reactivity insertion rate based on an average boron concentration change for the enti:e system. A benchmark of this approach is well within the capability of the RETRAN 02 model.
The main concern for the steam generator tube rupture event is the radiological doses based on the mass released to the steam generator secondary side which is assumed to pass to the condenser. The mass released in the FSAR was based on a hand calculation assuming a constant leak flow rate over the time period required to cool the reactor coolant system to the temperature at which the affected generator could be isolated. This primary-to-secondary mass release can easily be computed by RETRAN with conservative assumptions to maximize the leak flow, such as by using the Moody critical flow model with a discharge coeflicient of 1.0.
TABLE 3: BENCHMARK ANALYSIS COMPARISONS OF FSAR TRANSIENTS / ACCIDENTS TRANSIENT / ACCIDENT BENCHMARK 1.
Startup/ rod withdrawal FSAR comparisons 2.
Moderator dilution Bounded by startup/ rod withdrawal accidents 3.
Pump startup FSAR comparison 4.
Loss of coolant flow Plant data comparison FSAR comparison 5.
Locked rotor FSAR comparison 6.
Stuck-out, stuck-in or dropped FSAR comparison control rod 7.
Loss of electric power Plant data comparison 1
8.
Steam line break FSAR comparison 9.
Steam generator tube failure Mechanistic dose cateulation 10.
Loss of normal feedwater Plant data comparison