ML20215G547
| ML20215G547 | |
| Person / Time | |
|---|---|
| Site: | North Anna  |
| Issue date: | 06/17/1987 |
| From: | VIRGINIA POWER (VIRGINIA ELECTRIC & POWER CO.) |
| To: | |
| Shared Package | |
| ML20215G537 | List: |
| References | |
| NUDOCS 8706230294 | |
| Download: ML20215G547 (77) | |
Text
.,.
i l
. ' ~...
I' ',
ATTACHMENT 3 Safety Evaluation For Change In Methodology ' And Supporting Analyses 8706230294 070617 ADOCK 0500 B
gDR L:.
r P
I'
.f.
TABLE OF CONTENTS.
Page L.
f l
TABLE OF C0NTENTS............................................
2 LI T OF FIGURES..............................................
3 l
a l
e-LIST OFsTABLES...............................................
5 l
l
.c J5 #
1.0 INTRODU,CTION.............................................
6 srs q
[.,t' 7
h
'2.0, PRO, POSED TECHNICAL. SPECIFICATION CHANGES.................
y
(
L
- 3. 0 ' DISCUSSION AIO EVALUATION...............................
11 f
3.1 Yntroductioni....................................
11
' 3. 2 Stat i sti ca l DtG2 Limi t............................ 11 1;
~3.3 Safety Analysis..................................
26 i
j
- 4. 0 10 CFR 5 0. 59 EVAltiATI0N.................................
73 REFERENCES..................................................
74 i
I i
1
!j e.
j l'
l i.
1 i
0 i
,t s
I t
yt 4*
f i,*
'd!
t t
s 8.
t 2
I LIST OF FIGURES' Page 3.2.3.1' Total D E R Standard Deviation vs. Inlet ~~ Temperature. 25 3.3.2.1 Trip Reactivity Curve versus Position..............
33 3.3.2.2L Trip Reactivity Curve versus Time..................
34 3.3.2.3 Most Negative Doppler Only Power Coefficient.......
35 3.3.3.1.Effect of Reactivity Insertion' Rate on Minimum DER for a Rod Withdrawal Accident from 100% Power....... 39 3.3.3.1-2 Effect of Reactivity Insertion Rate on Minimum DER for.a Rod Withdrawal Accident from 60% Power.......
40 3.3.3.1-3 Effect of Reactivity' Insertion Rate on Minimum DNBR for a Rod Withdrawal Accident from 10% Power........ 41 3.3.3.2-1 Loss of Load Accident, with Pressurizer Spray and Power-operated Relief Valve, Beginning of Life
'(Sheet 1)..........................................
45 3.3.3.2-2 Loss of Load Accident, with Pressurizer Spray and
' Power-operated Relief Valve, Beginning of Life-r
.(Sheet 2)...................'......................
46 3,3.3.2-3 _ Loss of Load Accident,-with Pressurizer Spray and Power-operated Relief Valve, End of Life (Sheet 1). 47-3.3.3.2-4 Loss-of Load Accident, with Pressurizer Spray and
- Power-operated Relief Valve,.End of Life (Sheet 2). 48 3.3.3.2-5 Loss of Load Accident, without Pressurizer Spray and Power-operated Relief Valve, Beginning of Life
'(Sheet 1)...........................................-49 3.3.3.2-6 Loss of Load Accident, without Pressurizer Spray and Power-operated Relief Valve, Beginning of Life (Sheet 2)..........................................
50 3.3.3.2-7 Loss of Load Accident, without Pressurizer Spray and Power-operated Relief Valve, End of Life (Sheet 1).
51 3.3.3.2-8'. Loss of Load Accident, without Pressurizer Spray and Power-operated Relief Valve, End of Life (Sheet 2).
52 3
l
LISTOFFIGURES(Continued) 3.3.3.3-1. Feedwater Control Valve Malfunction, End of Life with Rod Control (Sheet 1)..........
.............. 56 3.3.3.3-2 Feedwater Control Valve Malfunction, End of Life with Rod Control (Sheet 2).........................
57 3.3.3.4-1 Excessive Load Increase With Manual Rod Control, End of Life.(Sheet 1)..............................
60 3.3.3.4-2 Excessive Load Increase With Manual Rod Control, End of Life (Sheet 2)..............................
61 3.3.3.4-3 Excessive Load Increase With Automatic Rcd Control, End of Life (Sheet 1)..............................
62 3.3.3.4-4 Excessive Load Increase With Automatic Rod Control, End of Life (Sheet 2).........................
.... 63 3.3.3.5-1 -All Loops Operating, All Loops Coasting Down, Core Flow versus Time..............................
66 3.3.3.5-2 All Loops Operating, All Loops Coasting Down, Flux Transient versus Time.........................
67
^
3.3.3.5-3 All Loops Operating, All Loops Coasting Down, DNBR versus Time...................................
68 I
4
LIST OF TABLES-Page l3.2.3.1L Monte Carlo Analysis Summary.........................
23-3.2.3.2 Applicability of Statistical DNBR Evaluation Methodology...............................
24 3.3.2.1 Key Analysis Assumption Details......................
32 3.3.3.1 Time Sequence of Condition II and III Events..........
69' 5
1.0 INTRODUCTION
The NRC approved a core power uprating for the North Anna Power Station from 2775 MWt. to 2893 MWt during 1986.
The supporting core thermal-hydraulics calculations were' performed by Westinghouse using the.
Improved Thermal Design Procedure (ITDP, Reference 1), which is a statistically based Departure from Nucleate Boiling (DNB) methodology.
During the past 'several - years, Virginia Electric and Power Company (Virginia Power) has'been developing an independent statistically based DNB methodology.
This methodology, documented in Reference 2, has been previously submitted to the NRC for-review (Reference 3).
The purpose of this Safety' Evaluation is to provide justification for implementation of Virginia Power's Statistical DNB Evaluation Methodology for the North Anna Power Station.
The new methodology results in additional analysis margin when compared to ITDP.
The result provides a very conservative standard against which to compare reload-specific parameters, thereby 1
. reducing the likelihood that transient analysis will be required on a 1
reload basis.
In support of this implementation package, the existing ITDP-based core thermal limits and related setpoints have been verified as bounding, and the North Anna Technical Specifications have been updated to reflect the change in the plant DNB licensing basis.
As additional support for this change, the principal UFSAR Chapter 15 DNB events have been analyzed under the new DNB methodology.
Some additional Technical Specifications l
changes have resulted from these analyses.
l 6
i j
- n
.2.0' PROPOSED TECHNICAL SPECIFICATION CHANGES
.Several. North Anna Technical Specifications need' to be changed to incorporate the revised DNB Ratio (DNBR). limit and the results of the associated transient analyses.
Proposed changes to the Units 1 and 2 Technical Specifications include:
I 1
TECHNICAL SPECIFICATION 3/4.1.1-MOST NEGATIVE MTC LIMIT l
i
.i I
The change in the Limiting Condition for Operation (LCO 3.1.1.4)'and the Surveillance Requirements (SR 4.1.1.4) for the moderator temperature coefficient. (MTC) are intended to provide an End-of-Cycle limit' and associated trigger values which are appropriate-for current North Anna fuel cycles. The revised limit and trigger values are based on a revised safety analysis of the UFSAR Chapter 15 transients which are sensitive to the most negative MTC parameter. The DE design limit was not violated
-in any of the analyzed transients. A further discussion of the basis for the specific values is included in Section 3.3.2.of this Safety Evaluation.
Additionally, ACTION a.1 of the Unit 2 specification was changed to correct an administrative error approved in Amendment No. 59 issued on December 13, 1985.
7
TECHNICAL SPECIFICATION 3/4.3.1 PRESSURIZER WATER LEVEL RESPONSE TIME The change to Table 3.3-2, which is referenced by LC0 3.3.1.1, reflects a new requirement to have the pressurizer water level response time be s2.0 seconds.
A response time requirement has been added to protect against filling the pressurizer prior to the actuation of the overtemperature AT reactor trip.
It was found that for very slow reactivity insertion transients the DNB limit was never violated but in some situations the pressurizer filled before the plant OTAT trip was actuated. When the same transients were analyzed assuming an active high pressurizer water level trip the pressurizer did not fill.
This phenomenon is discussed further in Section 3.3.3.1 of this Safety Evaluation.
TECHNICAL SPECIFICATION B 2.1.1 FULL CORE DNB PROBABILITY CRITERION The change in Bases Section 2.1.1 adds a new criterion in the establishment of the DNBR limit for the Virginia Power Statistical DNB methodology.
Previously, traditional analyses and ITDP analyses considered only peak pin DNB probability; plant operation required that, for normal operation and Condition II operation, the peak pin avoid DNB with 95% probability at a 95% confidence level. The new methodology retains this criterion, and adds an additional criterion that the DNB probability of every rod, when summed over the whole core, shows that at least 99.9% of the core is expected to remain in the nucleate boiling regime. A further discussion is presented in Section 3.2 of this Safety Evaluation.
8 L
-)
TECHNICAL SPECIFICATION B 3/4.1 MTC SURVEILLANCE REQUIREMENTS The changes to Bases section 3/4.1 are related to the new MTC limit and surveillance requirements.
Additionally, the reference to the moderator density coefficient (EC) is deleted because it is no longer related to the safety analyses performed' by Virginia Power.
The NC parameter was used in the previous safety analyses performed by our fuel vendor, Westinghouse Electric Corporation.
Since the safety analyses i
performed by Virginia Power use temperature instead of density to specify ll moderator : reactivity feedback, it is preferable to use MFC in the Technical Specifications.
An additional benefit of this approach is that the relationship between the Technical' Specification limit and the safety analysis limit can be more clearly defined.
In fact, the two limits difier only by the measurement uncertainty and a correction for Bank D insertion.
The revised bases section makes the connection between the two limits clearer.
A further discussion is given in Section 3.3.2 of this Safety Evaluation.
l l
TECHNICAL SPECIFICATION B 3/4.2.3 DNBR LIMITS i
Technical Specification B 3/4.2.3 has been modified to reflect the revised DNBR limit as obtained with the new methodology. The new safety analysis DNBR limit is 1.26; the addition of 13.7% retained DNBR margin I
yields a design DNBR limit of 1.46. Separate values were not derived for the typical and thimble cell, since the single limit was shown to be 9
bounding for both. The retained margin is used for such applications as compensation of the rod bow penalty, for example. Additional information on the derivation of these limits is pros th.' in Section 3.2 of this Safety Evaluation.
The rod bow penalty is Westinghouse preprietary information and is not listed in the FSAR. However, a thorough discussion of the rod bow penalty is provided in the FSAR, and the sources of the appropriate numerical values are referenced. Technical Specificrtion B 3/4.2.3 has been updated accordingly.
TECHNICAL SPECIFICATION 8 3/4.2.5 DNB PARAMETER SURVEILLANCE The change to Bases section 3/4.2.5 clariffes the treatment of measurement uncertainties.
The company has performed analyses to show that the measurement uncertainties on the DNB parameters can be offset by the retained DNB margin and need not be accounted for by the plant operations staff.
This action is discussed further in Section 3.2.4 of this Safety Evaluation.
10
fj j
-3.0 DISCUSSION AND EVALUATION 3.1 Introduction
'The ' implementation of Virginia ~ Power's Statistical DNBR Methodology was a two step process' in the analysis l phase. The first step 'was the derivation of the revised DNBR limit through a Monte Carlo process, employing plant-specific uncertainties.
In the second step,.the revised, limit was used in analyring the key FSAR Chapter 15 DNB events. Some of the accident analyses used more conservative assumptions. than the.
previous docketed. analyses, thus absorbing.some.of the analysis margin gain of the new methodology.
The result 'provides a very conservative standard against which to compare _ reload-specific parameters, thereby reducing the likelihood that transient analysis will. be required on a reload basis.
3.2 Statistical DNBR Limit 3.2.1 Background Information' In. late 1985, Virginia Power submitted to the NRC a topical report (Reference 2) describing a proposed methodology for Statistical DNBR Evaluation. This methodology differed in method and philosophy from the Westinghouse ITDP, the' statistical DNBR methodology upon which the 1986 North Anna core uprating was based. Reference 2 provides a DfER analysis margin gain of approximately 10% over ITDP, and both provide substantial margin gains when compared to the deterministic DNBR methodology which 11 i
had been used as a part of the original plant licensing basis.
Coupled with the implementation of the Statistical DNBR Methodology, Virginia Power has also used the Westinghouse WRB-1 Critical Heat Flux (CHF) correlation (Reference 4), which is typically used by Westinghouse in their ITDP implementation. The WRB-1 CHF correlation has been qualified (Reference 5) for use in the Virginia Power thermal-hydraulics code COBRA (Reference 6).
In a deterministic methodology, a COBRA /WRB-1 DNBR limit of 1.17 provides protection from DN8 with 95% probability at a 95%
confidence level (95/95 protection). The comparable W-3 95/95 DNBR limit is 1.30 (Reference 7).
l 3.2.2 Methodology Review When used in the Statistical DNBR Evaluation Methodology, the COBRA /WRB-1 data statistics are used to determine a revised DNBR limit.
This revised limit includes the DNBR sensitivities of principal input parameters such as temperature and power, and is greater than the 1.17 CHF correlation limit. As a result, transient analysis under the revised j
methodology for Condition II DN8 events does not require that the uncertainties be accounted for in initial conditions; instead, nominal values may be used.
l J
l The Statistical DNBR Evaluation Methodology develops a revised DNBR l
limit by means of a Monte Carlo process.
Thousands of statepoints are
{
1 produced with random number generators which conservatively model the
)
[
parameter measurement uncertainties. These statepoints are then supplied to Virginia Power's core thermal-hydraulics code COBRA, which calculates 1
12
the local DNBR.
Each DNBR' is further randomized by a correlation uncertainty factor:
DNBR DNBR' = ---------------------
(3.2.2-1) 1 + 0.0871
- RANN0R I
in which RANN0R is a normal random number with mean zero and standard deviation one, and 0.0871 is an upper 95% confidence limit on the 1
i COBRA /WRB-1 measured-to predicted (M/P) Critical Heat Flux ratio standard deviation.
This formula differs slightly from equation (2.4.1) of I
Reference 2 because its original assumption of a normally distributed qualification DNBR data base was found to be incorrect. Reference 5 found I
the reciprocal of the DNBR to be normal, but not the DNBR itself; the randomizing factor was thus re-written to reflect the normality of the reciprocal DNBR.
The standard deviation of the resultant DNBR distribution is increased by a small sample correction factor to obtain a 95%' upper confidence limit, and is then combined Root-SunrSquare with code and model uncertainties to obtain a total DNBR standard deviation. The Statistical DNBR Limit (SDL) is then SDL = 1 + 1.645*s(total)
(3.2.2-2)'
in which the 1.645 multiplier is the z-value for one-sided 95% probability of a normal distribution.
This SDL thus ensures that the peak pin will avoid DNB on a 95/95 basis.
13 l
As an additional criterion, the SDL.is tested to determine the full core DN8' probability when the SDL is reached by the peak pin.
This process is performed by summing the DNB probability of.each rod in the core, using a bounding rod census curve and the DNBR sensitivity to rod 4
power.
If necessary, the SDL is increased until 99.9% of.the core is expected to remain in nucleate boiling even at the SDL; or, in other words, no more than 0.1% of the core is expected to be in DNB.
The addition of the 0.1% requirement means that the North Anna core is assured both of 95/95 peak pin protection and 99.9% full core protection from DNB.
14
3.2.3 Implementation Analysis 3.2.3.1 Uncertainty Analysis-i The magnitudes and functional forms of the uncertainties which were statistically treated were derived in a rigorous analysis of plant -
hardware. and measurement / calibration procedures.
These analyses were first performed in support of the North Anna core uprating, supporte'd by the Westinghouse ITDP. Virginia Power's Statistical DNBR Methodology and ITDP combine the same variables at virtually identical uncertainty levels:
the former-through Monte Carlo calculations, and.the latter by-means of a Root -- Sum ' Square combination of DNBR sensitivities.
The statistically treated variables are core
- power, pressure, inlet temperature, vessel mass flow, core bypass flow, and the nuclear and engineering enthalpy rise factors.
l The uncertainties-for pressurizer pressure, Tavg, core thermal power and vessel flow are quantified in ' Reference 8.
This analysis was performed under the standard Westinghouse uncertainty analysis methodology, quantifying all sensor, rack and miscellaneous uncertainty components of a total uncertainty and combining them in a manner consistent with their relative dependence or independence.
For the purpose of this implementation analysis, the Reference 8 uncertainties l
were treated as 1.645o
- values, corresponding to one-sided 95%
probability..This treatment is more conservative than the level which could be justified by Reference 8, to allow for any future changes in l
plant calibration procedures or hardware.
j 15
i The nuclear enthalpy-rise factor uncertainty was quantified in an f
I l
in-house analysis of available measurement / prediction data.
These data
)
l consisted of more than 11,000 points taken over nine of the operating I
cycles for both units, including recent cycles. The error of prediction relative to the measured value, (P-M)/M, yielded a mean of 0.1% with a j
1 standard deviation of 1.55%.
The non-zero mean is conservatively 1
positive, and as a conservative measure a standard deviation of 2% was employed in the implementation analysis.
The data were shown to be normally distributed.
t L
As discussed in FSAR Section 4.4.3.1.1, total core bypass flow consists of separate flow paths through the thimble tubes, direct leakage to the j
outlet nozzle, baffle joint leakage flow and upper head spray flow. These components were each quantified, their uncertainties conservatively modeled, and the flows and uncertainties summed. The total flow and its uncertainty were confirmed as being bounded by the assumed North Anna core uprating ITDP analysis values.
Because of the difficulty in characterizing the form of the uncertainty distribution, the implementation analysis assumed that the probability was uniformly l
smeared over a much larger range than was justified by the sum of the components.
The engineering enthalpy rise uncertainty factor consists primarily of the uncertainty in hot channel power and flow.
These factors were quantified by means of a closed-channel calculation, in which bounding values of high hot channel power and low flow were enployed. A uniformly 16
i
)
distributed 2% uncertainty was found to conservatively bound the results-by a large margin.
3.2.3.2 CHF Correlation I
l i
This implementation analysis used the Westinghouse WRB-1 correlation j
(Reference 4) while the topical report _ analysis utilized. the W-3 correlation.
The WRB-1 correlation is the current licensed design i
correlation for North Anna.
The replacement of W-3 with WRB-1 for this implementation is acceptable because the methodology philosophy is fully independent of the correlation, as noted in Reference 2.
3.2.3.3 Monte Carlo Calculations The Monte Carlo analysis itself consisted of ten sets of 2000 calculations each, performed over the full range of normal operation and anticipated transient conditions. These conditions spanned the pressure range between the high and low trip setpoints, inlet temperatures between a bounding cooldown event and a maximum heatup, powers up to the 118%-
i overpower limit and a bounding low flow event.
A nuclear design radial 1.56, rather than the current Technical power factor of FAh
=
Specification value of 1.49, was used in anticipation of a possible future Technical Specification increase of that parameter.
i Consistent with Reference 2, the DNBR standard deviation at each Nominal Statepoint was augmented by a correlation uncertainty factor (equation 3.2.2-1), a small sample correction factor and code and model 17
{
- uncertainties to' obtain a total DNBR standard deviation.
Each value'was Jused first in equation (3.2.2-2) to find the peak pin DER: limit, and then in a full-core DE probability summation to determine' whether' the 0.1%
criterion.had also been met.
A summary of. the. Monte Carlo Statepoint analyses is presented in Table 3.2.3.1.
The steps'in the SDL derivation analysis may be summarized as follows.
'At the limiting Nominal Statepoint J,7the standard deviation of the 2000
. COBRA DNBR's was found to be 0.0991. The' application of the correlation randomizing' factor (equation 3.2.2-1)'to each COBRA DNBR increased the.
j i
' standard deviation to.0.1501.
At this point,- a.1.0268 multiplier was applied to obtain a 95%' upper confidence limit on the standard' deviation, yielding a value of 0.1541_. This value was then combined Root Sum Square with code and model: uncertainty standard deviations of 0.0304 and 0.0047 respectively to obtain a total DER standard deviation of 0.1572, as
~
listed in Table 3.2.3.1.
The use'of this number in equation (3.2.2-2)
-yields a peak pin DE R. limit of 1.26.
The total DNBR standard' deviation
-from the worst statepoint, Nominal Statepoint J, was then used in the full core DNB probability summation to find a total of 0.08% of the ' core expected to be in DNB when the peak pin reaches the SDL, which meets the 0.1% criterion. No increase in the SDL was required to meet the the full core probability requirement.
Similarly, the 0.1% criterion was met for all of the other Nominal Statepoints.
l 18
4 i
3.2.3.4 Model Error Term Reference 2 included a model error ' term to account' for differences between the 6-channel COBRA model, which was'used to" perform the Monte Carlo calculations, and the COBRA production model in doing DNBR calculations. The'6-channel COBRA model is used only to estimate the DNBR.
standard deviation.as ~ it would be calculated.by the COBRA production model, and. comparisons showed that the 6-channel model DNBR standard deviation.is consistently much larger than its production model counterpart.
The larger standard deviation is quite. conservative, eliminating the need for a model error term.
However, the model error term was requantified based upon the WRB-1 correlation at 0.47%, and this value was used in accordance with the specifications of Reference 2.
3.2.3.5' Verification of Existing Protection Setpoints The existing - North Anna core thermal limits, overtemperature and
- overpower AT trip functions and the f(61) function were shown to be~
bounding and will not be changed.
Nevertheless, several other Technical Specifications changes are needed to support the implementation of.
i
- COBRA /WRB-1 with Reference 2.
The changes will be in the Bases Section B 3/4.2.3, where the DNBR limit is given (1.46 for Virginia Power), as
- wellastheadditionalprotectionphilosophy(full-coreDNBprobability).
19
E 3.2.3.6 Applicability of Methodology In order to demonstrate that the methodology. i_s valid at any conceivable statepoint in the intended range of application, Figure 3.2.3.1 is presented.
This figure is a plot of the total DNBR standard devi6 tion versus the inlet temperature at which it was derived. The total
~
DNBR standard deviation is clearly seen to decrease sharply with increasing temperature.
Linear regression was employed to correlate the DNBR standard deviation as a function of temperature, after which the residuals were plotted as a function of pressure, power and flow.
No trends whatsoever were observed in the residual plots, substantiating the fact that the DNBR standard deviation had been conservatively maximized by the cooldown statepoint for any conceivable Condition I, Condition II or low flow DNB event.
Specifically, the Statistical DNBR Evaluation Methodology will be applied to all Condition I and II DNB events, and to Loss of Flow analysis. The accidents to which the methodology is applicable are listed
]
in Table 3.2.3.3.
This methodology will be applied neither to steamline break analyses, because WRB-1 has not been qualified at MSLB conditions; nor to the Rod Withdrawal from Subcritical, since present physics predictions of FAh are less accurate at zero power.
Further, since the locked rotor accident is a Condition IV event, the statistical methodology o
will not be applied, although the accident occurs at conditions which lie both within the uncertainty and the methodology ranges of applicability.
20 kN u i er 19mi p-rr 11ri 1i ri-Lam
_..,_n-u,m
_..mm-__.-_,__________au
3.2.4 Retained Margin and the Rod Bow Penalty l
1 The implementation of the Statistical DNBR Evaluation Methodology with the WRB-1 correlation means that all retained DNBR margin must be quantified anew.
Rather than continue the original practice of using j
excessively conservative values for some parameters (e.g., the Thermal Diffusion Coefficient) in the core thermal-hydraulic analyses and quantifying a niinimum DNBR sensitivity as a credit, all retained margin was defined as a penalty upon the design DNBR limit:
instead of working with.the allowable 1.26 DNBR limit, a design limit of 1.46 was used. The difference is 13.7% retained margin (M):
SDL 1.26 Design DNBR Limit = ------- = ----------- = 1.46 (3.2.4-1) 1-M 1 - 0.137 The 13.7% retained DNBR margin easily accommodates the rod bow penalty, which is based upon Reference 9 and has been quantified at less than 3%.
As required by Reference 10, the penalty is applicable to a maximum burnup of 24,000 MWD /MTV.
When the North Anna core uprating was implemented, a statement was added into Technical Specification B 3/4.2.5, concerning periodic surveillance of certain DNB parameters, requiring that measurement uncertainties be accounted for during the surveillance.
The parameters in question were Tavg, pressure and flow.
For the implementation of Virginia Power's Statistical DNB Methodology, a penalty will be taken out I
of the retained DNBR margin to acconrnodate these measurement 21
. uncertainties. This penalty was found to be less than 2.0% in DNBR. This.
action permits clarification of the uncertainty accounting requirement in TS.B 3/4.2.5. 'The DNB Parameter Surveillance penalty and the rod bow penalty are both-easily absorbed by the available 13.7% retained DNBR j
l l
margin.
]
l 4
i 22
M 1
- l.,
'\\
Table 3.2.3.1 Monte Carlo Analysis Summary Total Nominal State-Power LInlet Pressure Flow DNBR:
Standard-point-(%)
Temperature
'(psia)
(%)
Mean Deviation *-
(degrees F).
(-)
~(-)
a A.
105' 558.6 2250 73 1.25' O.1496 q
B
~118-549.6 1860 2100' 1.26-0.1548 J
C 116.
562.6
.2000.
100-1.24 0.1514
.D 118-559.3~
~2000 100 1.24 0.1473 i
E 110
.586.6.
2250 100 1.24 0.1389 F'
118 574.1~
2250 100
-1.25 0.1503 G
106' 601.0' 2400 100 1.25 0.1413 H
118 583.0 2400 1100-1.25 0.'1465 I
100 553.6 2250 65
- 1.24
-0.1535 j
J 118 538.6 1860 92 1.28 0.1572 4,
- T'he etotal DNBR standard deviation includes the effects of the
' Monte. Carlo calculations, equation (3.2.2-1),' the 93% upper.
confidence limit multiplier, and a Root Sum' Square combination
)
with the code and model uncertainties.
1 23 l
Table 3.2.3.2 Applicability of Statistical.DNBR Evaluation Methodology l
FSAR Section Accident VEP-NE-2 Use?
l 15.2.1 RCCA Bank Withdrawal from Subcritical No 15.2.2 RCCA Bank Withdrawal at Power Yes 15.2.3 RCCA Misalignment / Dropped Rod Yes 15.2.4 Boron Dilution No (non-DNB) 15.2.5 Partial Loss of Flow Yes 15.2.6 Startup of Inactive Loop Yes 1
15.2.7 Loss of Load / Turbine Trip Yes 15.2.8 Loss of Normal Feedwater.
No (non-DNB) 4 15.2.9 Loss of Offsite Power No'(non-DNB) 15.2.10 Excessive Feedwater Yes 15.2.11 Excessive Load Increase Yes
-15.2.12 Depressurization of RCS
.Yes 15.2.13 Credible Steamline Break No 15.2.14 Spurious Safety Injection Yes 15.3.1 Small Break LOCA-No (non-DNB) 15.3.4 Complete Loss of Flow Yes 15.3.6 Single RCCA Withdrawal at Full Power Yes i
15.4.1 Large Break LOCA No (non-DNB)
]
15.4.2 Hypothetical Steamline Break No I
Feedline Break No (non-DNB)
]
15.4.3 Steam Generator Tube Rupture No (non-DNB) 15.4.4 Locked Rotor No 15.4.6
' Rod Ejection No (non-DNB) 24
~
0 1
6 R
U B
i N
D 0
L i95 A
a C
)
F I
(
S TT i
RE L
SU U
I S T
E A
1 TR U
R 3 E
AO 0
P 2 L
i7 M3 TR 5
A E er SC T ua E
T i
F T
U EL AN N
O U
I i
NM g
NA 0
i5 a
5 H
T R
O i
0 N
03 5
1 498 87054321 598765432'41 44444444 33 555555551 1 1 1 1 1 1 1 1 0 1 1 1 1 1 1 1 1 1 O 1 1 0000O000 00000O000 00 Z
aA ak$
s 0
q s
3.3' Safety; Analysis-1
-1
.3.3.1 Introduction l
l
-As a part of the program to implement Virginia Power's Statistical DNB i
methodology, for North Anna, many of the DNB-limited events of the UFSAR j
were reanalyzed with the new methodology. These analyses were performed using modified key input parameters including:
most negative MTC, trip reactivity shape and the absolute value of the most negative Doppler Power
' Coefficient (DPC).
3.3.1.1 Most Negative MTC The most negative MTC is used in five ANS Condition II accidents in the UFSAR (See Table 15.1-2): rod withdrawal at power (E0C case), startup of an inactive reactor coolant loop, loss of load (E0C case), excessive heat removal due to feedwater system malfunction and excessive load increase (E0C case).
The UFSAR also lists the major rupture of a feedwater pipe. (feedline break) as using the most negative MTC (most positive FOC).
However, the text of the UFSAR states that the cooldown portion of a feedwater break is bounded by the Main Steam Line Break which uses a temperature dependent MTC.
This group of transients bounds the possible condition II and III cooldown scenarios which rely on a constant
'MTC.
In view of these considerations the following accidents should be analyzed with an increased (more negative) MTC:
uncontrolled RCCA bank withdrawal at power, loss of load, excessive heat removal due to feedwater system malfunction, and excessive load increase.
Reanalysis of the 26
inactive loop startup is not required since the current operating license precludes N-1 loop operation.
3.3.1.2 Trip Reactivity Shape Reanalysis with a _ revised trip reactivity curve impacts _ several l
\\
accidents.
Any transient which relies on reactor trip and subsequent
-l I
control rod insertion is. impacted by a revision to the trip reactivity l
curve. A limiting set of these transients consists of those which occur
(
rapidly.
For example, the loss of flow events which are bounded by a complete loss of flow transient, require rapid rod insertion to preserve
)
margin to the DNB design limit.
Other transients which are affected by a
the trip reactivity shape (e.g.
ejected rod, rod withdrawal from i
subcritical) are not included in this package because they are not analyzed with statistical DNB methods.
3.3.1.3 Most Negative DPC I
1 1
The most negative Doppler-only power coefficient is a key safety parameter for numerous events, including dropped rod, BOC loss of load, I
accidental depressurization of the RCS, feedline break and locked rotor.
{
1 Since so many accidents are effected reanalysis of all the accidents affected by this parameter was not performed at this time.
However, to limit the number of potential future analyses, those analyses which were i
required for other reasons assumed a Doppler-only power coefficient ranging from -20 pcm/% at HZP to -14 pcm/% at HFP.
27
i Those transients which are affected by the ' input parameter chan:ges ar.d are DNB-limited have been reanalyzed:
1.
Uncontrolled Rod Cluster Control Assembly Bank Withdrawal at Power 2.
Loss.of External Electric Load and/or Turbine Trip 3.
Excessive Heat Removal Due to Feedwater System Malfunction 4.
Excessive Load Increase Incident 5.
Complete Loss of Reactor Coolant Flow i
3.3.1.4 Computer Codes
)
.The Virginia Power RETRAN models were used to determine the reactor coolant' system (RCS) transient response (Reference 11). RETRAN simulates the neutron kinetics, reactor coolant system, pressurizer, pressuri,zer relief valves, pressurizer spray, steam generators, and stes jenerator relief valves. The program computes pertinent. plant variables including temperature, pressure, and power level.
1 s
The Virginia Power COBRA models (Reference 4) were used to perform'ai detailed therma 1-hydraulic analysis of the reactor core.
COBRAsolvbs l
the governing conservation and state equations to resolve Ehe flow and energy fields within the reactor core geometry. These results are used in. turn to calculate DNBR with the WRB-1 CHF correlation.
COBRA can perform either steady state DNBR calculations or transient DNBR analyses with forcing functions supplied by the RETRAN code.
k 28 4
m.
q M.,4;g.
- +.
n w
,9 3
'. 4.O y
p l
n 3.3.2 Modified Safety Parameters 7.p a1
\\
\\
- \\
3s'
\\
~\\
\\
f; *
~
i;.
i
3.3.2.1 General. Discussion.
j
-4
(
\\\\
(j
-s FV.
Recent reload cycles indicate that the'most. positive calculated EOC q
, Moderator Density Coefficient (M)C), which is related to the most negativ'e.
sC a,
67 MTC specified in Section 3/4.Li.4! of the Technical Specificatiorb, Thas.
g,
.s>
et n
a 6
j$
beenapproh.ching the Ifmit value due to increasing cycle average bur'nup.,
V In order \\ X' to bound future reloads a WC. of +0.47. Ak/k/gm/cc, equiva, lent
.O j
-y
.c to an Ph,C'of -59.27 pen /'F, was used'as input to the ' analyses. e Section 1,
.s
.t-r s jl/
3.3.2.2 epre:ents:a detailed, discussion of-these two parameters andahow o
[
5 h
- s. -.
Y H-
)N 4) 4
%Q they ariused.
i,
,4 t
' d
- k
/{ I is i
$ q(3 i
\\_-
+
7 q'.
(
h L
- c. t
.s 3
- y
. [A cdeve of trip reactivityjosertion (%Ak/k93 as a fbction of position
/,
v L:
. ~,; C '\\ '
a is peerated and contparedto a current limpt cueva fde. each reload.
A
, w
-]t.
s shows{a, proposed revision toi the.gtrip' reactivity vs.:
a s Figure 3.,3.2.1 i
n d
Ms position curve which bounds the historical reload QAtas Figure 3.3.2.2 7.,
shows the corresponding trip reactivity versus time curviir.
Although the m
7 s
a impact of the revised 4 rip reactivity curve is bounded by the loss of flo), y
+
3 a;
a
. transient, the revised trip retictivity curve is used in all the reanalyzed n
.t
?q, r'
n[
d UFSAR Chapter 15 events to be consistent.
- h J{
6 3
p-c ;;r a
j.
a:.
n,
The.most negative DPC (c2D pcm/% at HZP 'to -14 pcm/% at HFp) was' w.
'modifi
'as shown. in Figure '3.3,2.3.
However, it should be made clear h
a tbat t he values of OkC-were{Nsed hn anticipati bof a future ubmittal
\\t
~
s X,
because got all of the transients which need to be analyzed to jusMfy 9 s
A m
change to the allowable parameter,1Vm.it have poen included in this s
a
, 8
'kJ 29 i,
o 2,
f.
+
di
d
+
package.
As mentioned earlier,. only the ' DNB-limited transients are presented in this package.-
j 3.3.2.2 Redefinition of Maximum Moderator Reactivity Feedback Parameter 1
The maximum moderator reactivity feedback safety parameter can be specified in terms of either E C or MTC. The original North Anna safety analysis performed by the fuel vendor used WC as the safety parameter.
The calculations performed by Virginia Power actually use temperature to determine moderator reactivity feedback.
The Technical Specifications are worded in terms of MTC.
However,-since the original FSAR analyses used density to determine moderator reactivity feedback, the E C is still reported in the bases section of the Technical Specifications.
Because the EC is not used in reload analyses performed by Virginia Power, it is replaced with the more standard MTC.
The limits being incorporated into the Technical Specificatioet are summarized below:
- Safety Analysis Limit
-59 pcm/'F
- Plus Measurement Uncertainty and Correction for Bank D partial insertion 9 pcm/'F
- End of Cycle limit (LCO 3.1.1.4)
-50 pcm/'F
- Plus Corrections to 60 ppm 3 pcm/'F
- 60 ppm trigger (SR 4.1.1.4)
-47 pcm/'F
- Plus Corrections to 300 ppm 7 pcm/'F
- 300 ppm trigger (SR 4.1.1.4)
-40 pcm/'F 30
ll[.
l t-The safety analysis limit was used..in'each of the Chapter-15 events
~
1 that were' reanalyzed.
These corrections clearly define the transition from the safety analysis limit to the. technical specification limit while l
maintaining inherent conservatism.
{
1 k
u 1
t 1
4 t
L i
31 j-
n 4
~
Table 3.3.2.1 KEY ANALYSIS ASSUMPTIONS DETAILS Initial Conditions Power-2893.0 MWt Average Temperature 586.82 'F RCS Flow Rate 289,200 gpm Pressure 2250 psia Reactor Kinetics Moderator Temperature Coeff.
-59.2686 pcm/'F (EOC)*
+6.0 pcm/'F (BCC)
Doppler Temperature Coeff.
-2.9
.pcm/'F_(E0C)
-1.4 pcm/'F (BOC) 0 0.0075 (BOC)
O.0043 (EOC) l l*
26.0 ysec l
Trip Reactivity Shape Figure 3.3.2.2*
Most Negative DPC Figure 3.3.2.3*
4 1
- These values differ from those which were used to support 1
the North Anna core uprating (Reference 13).
32
w._
i FIGURE 3.3.2.1 I
10.0 1
4 1
1.0
<a*
C E
.1 HO(WE Q.
.01 i
t f
.001 O
.2
.4
.6
.8 1.0 l
l FRACTION INSERTED l
1 TRIP REACTIVITY VS. POSITION I
I i
33
1 FIGURE 3.3.2.3 6
1 5
i 4
a b
a
=
R 2
1 i
t i
1 1
0 0 0.4 0.8 1.2 1.6 2
2.4 2.8 3.2 TIME (SECONDS)
TRIP REACTIVITY VS. TIME 34
1 FIGURE'3.3.2.3
-20 18 NOTE 1 i,
16 E -14 2
12 g
2 m
10 4.
NOTE 2 Er 3
s N
-6 NOTE 1: " UPPER CURVE" MOST NEGATIVE DOPPLER
-4 DOPPLER ONLY POWER DEFECT = 1.70% a K NOTE 2: " LOWER CURVE" LEAST NEGATIVE DOPPLER ONLY POWER DEFECT = 0,843% o K
-2 0
0 20 40 60 80 100 PERCENT POWER DOPPLER POWER COEFFICIENT USED IN ACCIDENT ANALYSIS I
35
3.3.3 Condition II and III Transients 3.3.3.1 Uncontrolled Rod Cluster Control Assembly Bank Withdrawal at Power 3.3.3.1.1 Introduction The purpose of the rod withdrawal at power (RWAP) analysis is to verify the adequacy of the high neutron flux and overtemperature AT reactor trips i
for core protection.
The RWAP transient was reanalyzed with the safety parameters listed in Table 3.3.2.1.
The core thermal limits and AT l
setpoint equation constants are the same as those used in the core uprating analysis which has been shown to be conservative with respect to core thermal limits determined using Virginia Power's statistical DNBR methodology. See Section 3.2.3.5 fcr details.
An RWAP transient produces a mismatch in steam flow and core power, resulting in an increase in reactor coolant temperature.
3.3.3.1.2 Method of Analysis The RWAP transient was analyzed using RETRAN and COBRA.
Virginia Power's statistical DNB methodology was employed.
Initial operrating conditions are assumed to be at the nominal, steady-state value for the power level being analyzed.
Three different initial conditions, analyzed for both minimum and maximum reactivity feedback parameters, are presented in Section 15.2.2 of the North Anna UFSAR:
100% power, 60% power and 10% power.
Each of 36
i 1
I L
the power levels were reanalyzed over a' wide range of reactivity insertion l
rates.
i l
1 4
' All of the minimum reactivity Teedback cases assumed a constant +6 l
l l
pcm/'F.MTC.
The assumption that a positive MTC exists at full power is conservative since.at full power, the MTC must actually be zero or i.
negative. The maximum rmctivity feedback cases were reanalyzed assuming l
a constant -59.27 pcm/'F which is more negative than could he achieved 1
at full power because it assumes that bank D is almost fully inserted.
At reduced powers, the MTC is inherently less negative than full power.
so the conservatism of the value is guaranteed.
- 3.3.3.1.3 Results Although a Hi-Flux or OTAT. trip setpoint was always obtained, the pressurizer filled prior to trip in the following cases:
1.
60% power with maximum feedback at a reactivity insertion rate of 15 pcm/sec.
l-2.
10% power with minimum feedback at a reactivity insertion rate of 3 L
pcm/sec.
3.
10% power with maximum feedback at reactivity insertion rates of 20, 40, and 60 pcm/sec.
Therefore, operation of the high level pressurizer trip was assumed in the safety analysis even though the DNBR limit was never violated without this assumption.
The assumed operability of this trip channel requires that Table 3.3-2 of the Technical Specifications be modified to include a response ' time of s2.0 seconds for the high pressurizer water level channel.
37
i
+,l
- A sequence 'of events for a typical fast and a typical slow reactivity insertion rate.is' presented in Table 3.3.3.1.
Figure 3.3.3.1-1 shows the
. minimum DER as. a function of reactivity insertion rate from' initial full power.. operation for the minimum and maximum. reactivity feedback-r scenarios.
It can ' be seen - that two ' reactor trip - channels provide prote.-tion'over the whole range of reactivity insertion rates. These are the high neutron flux and overtemperature AT trip' channels..The minimum DER is never less than the limit value.
i I
Figures 3.3.3.1-2 and 3.3.3.1-3 show the ' minimum DNBR as a function i
of reactivity insertion rate fo-RWAP incidents starting at 60% and 10%
power, respectively.
The results-are similar to the 100% power case, except that as'the initial power is decreased, the range over which the overtemperature AT trip is effective is increased.
In neither case does the DNBR fall belev the design limit value.
3.3.3.1.4 Conclusions The high neutron flux and overtemperature AT-trip channels provide adequate protection over the entire range of possible reactivity
' insertion rates. Additionally, the high pressurizer level trip channel provides protection against water solid operation.
The 10NBR is always
' larger.than the design limit value.
38
l
\\
'l FIGURE 3.3.3.1-1
)
2.s Minimum Feedback
--- Maximum Feedback 2.3 1
2.1 5.
i E
1.9
~* Neut n Flux 7,
~ ~
I
/
1.7 OTAT s/
\\
1.E.
0.000001 0.00001 0.00010 0.00100 REACTIVITY INSERTION RATE, AK/SEC EFFECT OF REACTIVITY INSERTION RATE ON MINIMUM DNBR FOR A ROD WITHDRAWAL ACCIDENT FROM 100% POWER l
39
l FIGURE 3.3.3.1-2
)
i i
I 2.5 Minimum Feedback
--- Maximum Feedback 2.3 High WI Neutron Flux /
2.1
/
~
/ ',
/
OTAT
's
/
g
/
,7
\\,/
l 1,9 sv 1.7 i
1.5 0.00001 0.00010 0.00100 0.01000 MEACTIVITY INSERTION RATE, A K/SEC EFFECT OF REACTIVITY INSERTION RATE ON MINIMUM DNBR FOR j
A ROD WITHDRAWAL ACCIDENT FROM 60% POWER
)
1 40 J
1 i
i FIGURE 3.3.3.1-3 j
l 3.1 Minimum Feedback
--- Maximum Feedback 2.9 2.7 I
2.5 gm I
/
I
l'52*by' I
\\
l 2.1 OTAT
/
\\
f
,/
g
- s I
r i
1.9
/
'V 1.7 1.5 0.00001 0.00010 0.00100 0.01000 REACTIVITY INSERTION RATE, A K/SEC EFFECT OF REACTIVITY INSERTION RATE ON MINIMUM DNBR FOR A ROD WITHDRAWAL ACCIDENT FROM 10% POWER 41
'a m
3.3.3.2 Loss Of. External Electric Load And/0r Turbine Trip u
l; 4
!3.3.3.2.1 Introduction The loss'of load. transient is characterized by the rapid reduction in-steam flow from. the steam generator' and a resultant rapid rise in L
secondary ' side pressures.
Consequently, primary side temperature and -
pressure increase. The transient.is terminated by a direct. reactor trip-or in the limiting case by the.high pressurizer pressure trip. Two cases for'both beginning and end of life conditions are' presented in the UFSAR.
i All four cases were analyzed:
i 1.
Reactor in manual rod control with operation of the pressurizer spray r
and the pressurizer power operated relief valves; and 2.
Reactor in manual rod control with no credit for. pressurizer spray or power operated. relief valves.
l 3.3.3.2.2 Method of Analysis
~
The loss of' load transient was analyzed using RETRAN. The assumptions?
and initial conditions are consistent with those specified in Table'
)
3.3.2.1. The Virginia Power statistical DNB methodology was employed.
.]
The initial condition is assumed to be at the full power steady-state.
J The behavior of the unit is evaluated for a complete loss of steam load.
Jl from ful1~ power without a direct reactor trip and the steam dump system.
j i
u Therefore,'this transient represents one of the most severe transients, a
with respect to overpressurization, of the reactor coolant system and the j
The transient shows the adequacy of the 42 3
- pressure-relieving devices and demonstrates margin to the DNB design l
i H
limit.
3.3.3.2.3 Results A sequence of events for each of the cases analyzed is ' presented in
-Table 3.3.3.1. The transient responses for a total loss of load from 100%
full power operation are shown in Figures 3.3.3.2-1 through 3.3.3.2-8 for-two cases at beginning of core life and two cases at end of core life.
i
.i l
l Figures 3.3.3.2-1 and 3.3.3.2-2 show the transient response for the i
total loss of load at beginning of_ life with a +6 pcm/'F MTC assuming full i
credit' for the pressurizer spray and pressurizer power-operated relief valves.
'No. credit is taken for steam dump actuation.
The reactor is l
tripped by the high pressurizer pressure signal. The minimum DNBR is'well above the limit value.
Figures 3.3.3.2-3 and 3.3.3.2-4 show the response for the total loss l
All other plant of load at end of life, assuming a -59.27 pcm/'F MTC.
parameters are the same as the case above.
The reactor power decreases due to the negative reactivity effect of the increase in-core water temperature.
No trip setpoint is reached.
The pressure initially increases, and then slowly decreases after about 10 seconds as a result
- of the ' reduction in neutron flux.
The DNBR increases throughout the i
transient and never drops below its initial value.
43-
i The total-loss of load accident was also studied, assuming the plant to be initially operating at 100% of full power with no credit taken for the pressurizer spray, pressurizer power-operated relief valves, or steam dump actuation. The reactor trips on a high pressurizer pressure signal.
Figures 3.3.3.2-5 and 3.3.3.2-6 show the beginning of life transient with a +6 pcm/'F MTC. The neutron flux increases to 104% of full power until the reactor is tripped.
The minimum DNBR remains well above the limit value.
In this case the pressurizer safety valves are actuated.
Figures 3.3.3.2-7 and 3.3.3.2-8 show the transient at the end of life with the other assumptions the same as in Figures 3.3.3.2-5 and 3.3.3.2-6.
. Again, the DNBR increases throughout the transient and the pressurizer safety valves are actuated.
3.3.3.2.4 Conclusions The integrity of the core is maintained by operation of the reactor protection system, i.e., the DNBR will be maintained above the limit value.
Thus there will be no cladding damage and no release of fission products to the reactor coolant system.
The pressurizer and steam generator safety relief valves are adequate to maintain system pressure below the design limit.
44
1.2 7 1.0 4
5
.8 d
i
$8 N,.6 zO
.4 E
.2 0
Y 2600 i
2400 i
$L E
N 2200 i
E E
2000 1800 10 j
l 8
i
^
4 I
I O
10 20 30 40 50 TIME (SECONDS)
LOSS OF LOAD ACCIDENT, WITH PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE BEGINNING OF LIFE j
(Sheet 1) i FIGURE 3.3.3.2-1 45 i
FIGURE 3,3.3.222 1400 l
1300 v
{
1200 1100 g>r
~1000
-l 900 E
l 3
a 800
'I 700
.00 620 i
1 610 mg
'600 1
2 g
590 F
Eg
.580 W<
570 560 550 0
10 20 30 40 50 TIME (SECONDS)
LOSS OF LOAD ACCIDENT, WITH PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE BEGINNING OF LIFE (Sheet 2)
1.2 1.0 7
$~
s5
.a d
"',~6 8
6
.4 e*2 0
2600 Ig 2400 m
n<
2200 m
@mg 2000 j
1800 10 i
8 E
4 l
i 0
0 10 20 30 40 50 TIME (SECONDS) f LOSS OF LOAD ACCIDENT, WITH PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE END OF LIFE (Sheet 1)
FIGURE 3.3.3.2-3 g
j
c.
('
FIGURE 3.3.3.2-4' 1400 e
1300 p
I l
1200 m
8 g
1100 cc N*
3:
1000 e
N g.
900 1
I su 1
[
800 700 F
600 l
620 i
610 g
y E
I p
600
)
E 3
5 590 Y
i 8
1
[
580 j
W<
1 E
570 Oo 560 550 1
0 10 20 30 40 50 TIME (SECONDS)
LOSS OF LOAD ACCIDENT, WITH PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE END OF LIFE i
(Sheet 2) l 48 j
1.2
-~
J 1.0 sgl
.8
]
I d F.
E8
.6 5g
.4
- b
.2 I
l 2600 E
2400 Ie 2200 m
N 2000 l
t i'
1800 l
1G i
8 6
1' 4
2 L
'O 10 20 30 40 50 TIME (SECONDS)
LOSS OF LOAD ACCIDENT, WITHOUT PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE BEGINNING OF LIFE (Sheet 1)
FIGURE 3.3.3.2-5 49
FIGURE 3.3.3.2-6 11400 i
1300 1
6 1200
- f
.Y,1100 z
h1000 Rw d '900 Em 600 I
j 700
.]
,i 1
600 i
620 l
610 l
600 W
l2
$.590 I
N 580 8
e 570
. o 560 2
550 0
10 20 30 40 50 TIME (SECONDS)
LOSS OF LOAD ACCIDENT, WITHOUT PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE BEGINNING OF LIFE (Sheet 2) 50 1
FIGURE 3.3.3.2-7 1.2
\\
1.0
)
7 l
5 i
~
dM Eg
.6,
h4 gg
.2
)
L 0
2600 4
2400 i
N 2200 m
.s d
f2000 m
E a.
1800 10 i
'j s
i I
6 E
E 4
2 O
10 20 30 40 50 TIME (SECONDS)
LOSS OF LOAD ACCIDENT, WITHOUT PRESSURl2ER SPRAY AND POWER OPERATED RELIEF VALVE END OF LIFE (Sheet 1) 51 l
u.
FIGURE 3.3.3.2-8 1400 f
- /.
E L-
~ c1200 f1100 mL 1
l
< 1000 B
g 1
900 a
ic 800 1
700 600 -
)
620 610 j
E i
l'c 5'
590 h
580 m
-i 570 1
560
.{
550 0
16 20 30 40 50 TIME (SECONDS)
, LOSS OF LOAD ACCIDENT, WITHOUT PRESSURIZER SPRAY AND POWER OPERATED RELIEF VALVE END OF LIFE (Sheet 2) 52-
3.3.3.3 Excessive Heat Removal Due To Feedwater System Malfunction 3.3.3.3.1 Introduction Excessive heat removal due to a feedwater system malfunction can result from 'a reduction in feedwater temperature or an increase in feedwater flow. Either malfunction can cause an increase in core power by lowering reactor coolant and fuel temperature.
Excessive ~feedwater additian due to a control system malfunction or operator error which allows a feedwater control valve to open fully is considered.
.3.3.3.3.2 Method of Analysis The excessive heat removal due to a feedwater system malfunction transient was analyzed using RETRAN.
The assumptions and the initial conditions are the same as those presented in Table ~3.3.2.1.
Two cases were analyzed assuming a -59.27 pcm/'F MTC, which bounds end of cycle conditions:
1.
Accidental. opening of one feedwater control valve with the reactor just critical at zero load conditions.
2.
Accidental opening of one feedwater control valve with the reactor in automatic control at full power.
i 3.3.3.3.3 Results In the case of an accidental full opening of one feedwater control valve with the reactor at zero power and the above-mentioned assumptions, J
the maximum reactivity insertion rate is less than the maximum reactivity insertion rate for the uncontrolled control rod assembly withdrawal from 53 (L
a subcritical condition, and therefore the results of the analyses are not presented.
It should be noted that if the incident occurs with the unit just critical at no-load, the reactor may be tripped by the power range high neutron flux trip (low setting) set at approximately 25%.
A sequence of events for each of the case analyzed is presented in Table 3.3.3.1.
The full power case (end-of life with control) gives the largest reactivity feedback and results in the greatest power increase.
When the steam generator water level in the faulted loop reaches the high-high level setpoint, all feedwater control and isolation valves close, and the feedwater pumps are tripped which prevents the continuous addition of feedwater. With no incoming feedwater and steam still being generated, the steam generator water level decreases.
When it reaches the low-low level setpoint in an unfaulted loop, the reactor is tripped.
This generates a turbine trip. The assumption of a reactor trip in this sequence is conservative because the engineered safety features actuation system (ESFAS) is designed to trip the turbine and the reactor, in addition to isolating feedwater, when steam generator level reaches the high-high setpoint.
Assuming the reactor to be in the manual control mode results in a slightly less severe transient.
Transient results given in Figures 3.3.3.3-1 and 3.3.3.3-2 show the nuclear power, core heat flux, pressurizer pressure, Tavg, loop AT, and DNBR associated with the increased thermal load on the reactor. The DNBR does not drop below the limit value.
54
7 y
ti i
3.3.3.3.4 Conclusions E,
R The above results show that the DIER encountered for excessive
. feedwater addition at power is well above the-limit value. Thus, the heat i'.,
removal capability is adequate to prevent clad damage.
e r' ;
?
c o
l.
--p:
)
55 p.
V FIGURE 3.3.3.3-1 120 100
.E I
2 g
60 1-20 k
0-120 100 E
s d
W 80 l
20 0'
5.0 4.0 3.0 1
2.0 7
1.0 0
0 20 40 60 80 100 TIME (SECONDS)
FEEDWATER CONTROL VALVE MALFUNCTION, END OF LIFE WITH ROD CONTROL (Sheet 1) 56
2300 2200
~
2150 m
h 2100 y
2050 a.
2000 800 E
l 50 j
I
{
580 N
570
- 4 560 550 100 80 E
80 G
a 20 0
0 20 40 60 80 100 TIME (SECONDS)
FEEDWATER CONTROL VALVE MALFUNCTION, END OF LIFE WITH ROD CONTROL (Sheet 2)
FIGURE 3.3.3.3-2 57
I 3.3.3.4 Excessive Load Increase Incident 3.3.3.4.1 Introduction An excessive load increase incident is defined as a rapid increase in the steam flow that causes a power mismatch between the reactor core power and the steam generator load demand.
The reactor control system is designed to accommodate a 10% step increase or a 5%/ min ramp load increase in the range of 15 to 100% of full power. Any loading rate in excess of these values may cause a reactor trip actuated by the reactor protection.
3.3.3.4.2 Method of Analysis This accident was analyzed using RETRAN.
The assumptions and the initial conditions are the same as those in Table 3.3.2.1.
At end of life the MTC has its highest absolute value resulting in the largest amount of reactivity feedback due to changes in coolant temperature.
Two bounding cases are analyzed to demonstrate the plant behavior:
1.
Manually controlled reactor at end of life.
2.
Reactor in automatic control at end of life.
The core uprating analysis for the beginning of life cases remains bounding for the minimum reactivity feedback scenario.
58
3.3.3.4.3 Results The sequence of events for.the cases analyzed is presented in Table 3.3.3.1.
Figures 3.3.3.4-1 and 3.3.3.4-2 illustrate the transient with the reactor in the manual control mode. For the end of life, manually controlled case, there is a much larger increase in reactor power due to the moderator feedback than there would be at beginning of life.
A reduction in.DN8R is experienced, but the DNBR remains above the Ifmit value.
Figures 3.3.3.4-3 through 3.3.3.4-4 illustrate the transient assuming-the reactor is in'the automatic control mode. The end-of-life cases show that core power increases, thereby reducing the rate of decrease in coolant average temperature and pressurizer pressure.
For the end-of-life automatically controlled case, t.
minimum DNBR remains above the limit value.
3.3.3.4.4 Conclusions It has been demonstrated that for an excessive load increase the minimum DNBR during the transient is above the design limit.
- Also, equilibrium conditions of core power and core average temperature are reached.
59
FIGURE 3.3.3.4-1 1.20 k
1.15 mz
@B l
1.10 edz 89 1.05 b
z<
b 1.00 0.95 610 i
E M
600 E
5 g b 590 W
580 g
8 570 0
50 100 150 200 250 300 TIME (SECONDS)
EXCESSIVE LOAD INCREASE WITH MANUAL ROD CONTROL END OF LIFE (Sheet 1)
W 60
2400 7
4 2300 I
x E
2200 W
n.
b 2000 E
I 55 88 1900 E
a.
1800 73 71 j
C L
F-69 67 65 3.0 2.5 e
d g
g 2.0 1.5 0
50 100 150 200 250 300 TIME (SECONDS)
EXCESSIVE LOAD INCREASE WITH MANUAL ROD CONTROL END OF LIFE (Sheet 2)
FIGURE 3.3.3.4-2 61 i
i FIGURE 3.3.3.4-3 l
1.20 k 1.15 mE E5 g z 1,10 E8
$z d 9 1.05 EW b1.00 0.95 610 E
Ey 600 W
t
- ( 590 g
a W"
580 E
8 570 0
50 100 150 200 250 300 TIME (SECONDS)
EXCESSIVE LOAD INCREASE WITH AUTOMATIC ROD CONTROL END OF LIFE i
(Sheet 1) 62 i
A 2400 E
( 2300 g
s -
g 2200 E
2000 N
1900 1800 73 71 E
g 69 a
67 6C 3.0 2.5 i
E 2.0 1.5 0
50 100 150 200 250 300 TIME (SECONDS)
EXCESSIVE LOAD INCREASE WITH AUTOMATIC ROD CONTROL END OF LIFE (Sheet 2)
FIGURE 3.3.3.4-4 63
3.3.3.5 Complete Loss Of Forced Reactor Coolant Flow 3.3.3.5.1 Introduction A complete loss of forced reactor coolant flow may result from a
. simultaneous loss of electrical power to all reactor coolant pumps.
If the reactor is at power.at the time of the accident, the immediate effect of a loss of coolant flow is a' rapid increase in the coolant temperature.
resulting in a decrease in the DNB with a potential for subsequent fuel damage.
Undervoltage or - underfrequency setpoints from the reactor coolant pump power supply buses is provided to protect the reactor against l
a loss of coolant flow accident.
The present analysis assumes reactor trip on the undervoltage signal after the simultaneous loss of all three reactor coolant pumps.
The UFSAR demonstrates this scenario to be the limiting loss of flow event.
3.3.3.5.2 Method of Analysis This transient was analyzed using RETRAN employing the assumptions and initial conditions found in Table 3.3.2.1.
First, RETRAN was used to calculate the core flow and time of reactor trip as well as the nuclear power and heat flux transients.
COBRA was then used to calculate the minimum DNBR during the transient based upon the heat flux and flow from RETRAN.
64
3.3.3.5.3 Results l
l The calculated sequence of events is shown in Table 3.3.3.1 for the
{
l
\\
the case analyzed. Figures 3.3.3.5-1 through 3.3.3.5-3 show the core flow l
coastdown, the neutron flux changes, the core average heat flux changes and the DNBR. The reactor is assumed to trip on the undervoltage signal.
t The DNBR curve is not less than the limit value.
3.3.3.5.4 Conclusions l
The analysis performed has demonstrated that for the complete loss of l
forced reactor coolant flow, the DNBR does not decrease below the limit
)
value during the transient, and thus there is no clad damage or release
}
l of fission product-s to the reactor coolant system.
l 65
i FIGURE 3.3.3.5-1 1.2 1.0 7(
.8 i
S
.6
[
=0
.4
.2 I
I I
I 0
o 2
4 6
8 10 TIME (SECONDS)
ALL LOOPS OPERATING, ALL LOOPS COASTING DOWN, CORE FLOW VERSUS TIME 66
~
FIGURE 3.3.3.5-2 l
1.2 1.0<
7 4
Core Average 18 Heat Flux q
8 Neutron Flux
.6 0
i l
.4 5
t l
.2 l
i f
I 0
0 2
4 6
8 10 TIME (SECONDS)
ALL LOOPS OPERATING, ALL LOOPS COASTING DOWN, FLUX TRANSIENTS VERSUS TIME 67
l
)
FIGURE 3.3.3.5-3 1
2.6 3
)
I i
2.4 i
l 2.2 E
2.0 1.6 1.6 1
I I
I 0
1 2
3 4
5 TIME (SECONDS)
ALL LOOPS OPERATING, ALL LOOPS COASTING DOWN, DNBR VERSUS TIME 68
I TABLE 3.3.3.1 TIME SEQUENCE OF EVENTS FOR CONDITION II AND III EVENTS Accident Event' Time (sec) a) Uncontrolled RCCA Bank Withdrawal at Power 1.
Case A Initiation of uncontrolled RCCA withdrawal at a high' reactivity
.J insertion rate (75 pcm/sec) 0' l
Power range high neutron flux high trip point reached 1.7 Rods begin to fall into core 2.2 Minimum DNBR occurs 2.9 2.
Case B Initiation of uncontrolled RCCS withdrawal at a small reactivity insertion rate (20 pcm/sec) 0 Overtemperature AT reactor trip signal initiated 53.2 Minimum DNBR occurs
~57.0 Rods begin to drop into core 57.2 b) Loss of External Electrical Load 1.
With pressurizer control (BOL)
Loss of electrical load 0
- High Pressurizer Pressure Trip Point Reached 8.9 Initiation of steam release from steam generator safety valves 9.0 69
TABLE 3.3.3.1 (Continued)
TIME SEQUENCE OF EVENTS FOR CONDITION II AND III EVENTS Accident Event Time (sec)
Rods begin to drop 10.9 l
l-Minimum DNBR occurs 12.25 Peak pressurizer pressure z
occurs 12.25 l
2.
With pressurizer control (EOL)
Loss of electrical load 0
Initiation of steam release from steam generator safety valves 9.1 Minimum DNBR occurs Peak pressurizer pressure occurs 10.25 3.
Without pressurizer control (BOL) l Loss of electrical load 0
High pressurizer pressure reactor trip point reached 5.3 Rods begin to drop 7.3 Minimum DNBR occurs 8.25 l
Peak pressurizer pressure occurs 9.0 Initiation of steam release from steam generator safety valves 9.0
- DNBR does not decrease below its initial value 70
1
- TABLE 3.3.3.1 (Continued)
TIME SEQUENCE OF EVENTS FOR CONDITION II AND III EVENTS Accident Event Time-l 4.
Without pressurizer control (EOL)
Loss of electrical load 0
High pressurizer pressure reactor trip point reached 5.4 Rods begin to drop 7.4 Minimum DNBR occurs Peak pressurizer pressure occurs 8.75 Initiation of steam release from steam generator safety valves 9.0 c) Feedwater Malfunction at Full power One main feedwater control valve fails fully opens 0.0 Minimum DNBR occurs 50.0 Feedwater flow isolated due to high steam generator level trip 51.0 Reactor trip on low-low steam generator level in non-faulted loop 75.0
- DNBR does not decrease below its initial value 71
__j
TABLE 3.3.3.1 (Continued)
TIME SEQUENCE 0F EVENTS FOR CONDITION II AND III EVENTS Accident Event Time (Sec)~
d) Excessive Load Increase
- 1. Manual Reactor Control 10 % step load increase 0.0 (maximum feedback)
Equilibrium Conditions Reached (approximate.
I time only) 51.0
- 2. Automatic Reactor Control 10% step load increase 0.0 3
(maximum feedback)
Equilibrium Conditions Reached (approximate i
time only) 61.0 e) Complete loss of forced Coastdown begins 0
reactor coolant flow, j
I all three loops operating, Rod motion'begins 1.2 all three pumps coasting down Minimum DNBR occurs 3.3 i
I 1
1 I
i 72 L.
1 4.0 10 CFR 50.59 EVALUATION 4
The proposed changes have been reviewed against the criteria of 10 CFR 50.59 resulting in the conclusion that an unreviewed safety question does I
not exist.
This determination was reached based on the following specific considerations:
1.
Neither the probability nor the consequences of any UFSAR Chapter 15 j
event are affected by the implementation of the new methodology Accident probability is independent of the DNB evaluation methodology and thus is unaffected by the changes which are documented in this package.
With regard to accident consequences, the DNB protection probability and confidence level requirements have been retained, each at a 95% level; thus the consequences of any accident can be no more severe under the new Statistical DNBR Methodology sith WRB-1 than j
they are under the present ITDP methodology with the WRB-1 correlation.
In fact, the new methodology is even more restrictive in that a full-core DNB probability criterion has been added, further limiting the potential consequences of any DNB event as compared to the previous methodology.
Since the proposed changes involve parameters which are not accident initiators they will not increase the probability of occurrence of any accident previou-sly analyzed.
The transients analyzed herein verify that operation under the proposed technical specification changes would not result in an increase in accident consequences.
2.
No new or different accident type is generated as a result of these proposed changes.
Specifically, because no hardware changes accompany this new methodology, the range of accident initiators evaluated in the FSAR remains valid for the new DNBR niethodology.
Similarly, the proposed technical specification changes are based upon the use of the safety parameters listed in Table 3.3.2.1.
The parameters involved are physical attributes of the fuel and therefore do not create the possibility of an accident of a different type than evaluated previously because no system changes are involved.
3.
No reduction in the margin of safety occurs. The principal DNBR safety j
criterion has been, and continues to be, that DNB shall be avoided j
with 95% probability at a 95% confidence level. The margin of safety, which in this case is the margin between the DNBR limit and cladding
)
failure, thus retains the same basis. In fact, the DNBR limit is now j
based upon the additional criterion of the sum of non-DNB probability for every rod in the core remaining above the 99.9% level, which previously had not beer, required.
73 o
REFERENCES 1.
Chelemer, H.,
et al.:
Improved Thermal Design Procedure," WCAP-8567 (July,1975).
2.
Anderson, R.C.:
" Statistical DNBR Evaluation Methodology," VEP-NE-2 (July,1985).
3.
Letter from W.
L.
Stewart (Vepco) to H.R.
Denton(NRC), " Virginia Electric and Power Company Surry Power Station Unit Nos. I and 2 North Anna Power Station Unit Nos. 1 and 2 Statistical DNBR Evaluation Methodology," Serial No.85-688, dated October 8, 1985.
4.
Motley, F.
E., et al. :
"New Westinghouse Correlation WRB-1 for Predicting Critical Heat Flux in Rod Bundles with Mixing Vane Grids,"
WCAP-8762-P-A (proprietary) and WCAP-8763-A (non proprietary) (July,-
1984).
5.
Anderson, R.C. and N. P. Wolfhope:
" Qualification of the WRB'1 CHF Correlation in the Virginia Power COBRA Code," VEP-NE-3 (November, 1986).
6.
Sliz, F. W. and K. L. Basehore: "Vepco Reactor Core Thermal-Hydraulic Analysis Using the COBRA IIIc/MIT Computer Code," VEP-FRD-33-A (October,.1983).
- 7. '" Reference Core Report 17x17," WCAP-8185 (December, 1973).
8.
- Tuley, C.
R.:
" Improved Thermal Design Procedure Instrument Uncertainties for North Anna Units 1 & 2 Core Uprating," WCAP-11203 (July,-1986).
9.
Skaritka, J., Ed. :
" Fuel Rod Bow Evaluation," WCAP-8691, Rev. 1 (July,1979).
- 10. Letter from C. Berlinger (NRC) to E. P. Rahe, Jr. (Westinghouse),
" Request for Reduction in Fuel Assembly Burnup Limit for Calculation of Maximum Rod Bow Penalty," (June 18,1986).
- 11. Letter from W. L. Stewart (Vepco) to USNRC, " North Anna Units 1 and I
2 Locked Rotor Reanalysis," Serial No.85-150 (April, 1987).
- 12. Smith, N. A. :
"Vepco Reactor System Transient Analysis Using the RETRAN Computer Code,"
VEP-FRD-41A (May, 1985) and supplement transmitted in letter 85-753 from W.L.
Stewart (VEPCo) to H.
R.
Denton(NRC) (November 19,1985).
- 13. Letter from W. L. -Stewart (Vepco) to H.R. Denton (NRC), " Amendment Requesting a Core Thermal Power Uprate of 2983 MWt," Serial No.85-077 (May,1985).
)
74
P.
W l
)
1 ATTACHMENT 4 l
10 CFR 50.92 EVALUATION c
'i 1
g 10 CFR 50.92 EVALUATION i
The proposed changes do not involve a
significant hazards consideration for North Anna Units 1 and 2 for the foll a reasons:
1.
The probability or consequences of any UFSAR Chapter 15 event do not increase. Accident probability is independent of the DNB evaluation methodology and thus is unaffected by the changes to T.S.
82.1.1, 83/4.2.3 and B3/4.2.5 documented in this-package.
With regard to accident consequences, the DNB protection probability and confidence level requirements have been retained, each at a 95% level; thus the consequences of any accident can be no more severe under the new Statistical DNBR Methodology with WRB-1 than they are under the present ITDP methodology with the WRB-1 correlation.
In fact, the new methodology is even more restrictive in that a full-core DNB probability criterion has been added, further limiting the potential consequences of any DNB event as compared to the previous methodology.
The change to the most negative moderator temperature coefficient affects T.S. 3/4.1.1.4 and 83/4.1.1.4.
The principal change is to expand the limit and associated trigger values.
The transients affected by this change have been analyzed and shown to not violate any design limit. Specification B3/4.1.1.4 is reworded to remove the discussion of the moderator density coefficient since Virginia Power does not use this' key input parameter in safety analysis.
As discussed in the safety evaluation, the MTC limit is more clearly related to the safety analysis limit as a result of this change.
Therefore the consequences of any UFSAR chapter 15 event are not increased. Furthermore, accident probability is not increased since the MTC limit change meets all the design limits.
Administrative changes were also made to T.S.
3/4.1.1.4 and B3/4.1.1.4.
First, all references to change in reactivity are consistently specified in terms of Ak/k.
Second, the Unit 2 Specification B3/4.1.1.4, ACTION a.1, is changed to correct an administrative error so that it is consistent with the equivalent Unit I specification.
The high pressurizer water level reactor trip was assumed to be operative in the rod withdrawal at power analysis.
This assumption was not required to maintain margin to the DNBR limit but to prevent the pressurizer from filling. Therefore Table 3.3-2 of T.S. 3/4.3.3.1
]
was changed to include a response time of $2.0 seconds for this channel. This change represents an additional functional requirement of the reactor protection system so it does not increase the consequences of any accident nor increase the probability of any accident occurring.
2.
No new or different accident type is generated as a result of the proposed. changes to the DNB limit in T.S.
B2.1.1, B3/4.2.3 and B3/4.2.5. Specifically, because no hardware changes accompany this 1
i new methodology, the ratige of accident initiators evaluated in the UFSAR remains valid for the new DNBR methodology.
Similarly, the remaining proposed Technical Specification changes 1
(i.e. T.S. 3/4.1.1.4, 3/4.3.3.1 and B3/4.1.1.4) are based on the use j
of the key safety parameters which are physical attributes of the fuel,'and therefore do not create the possibility of an accident of-a different type than evaluated previously.
Thus, the proposed J
changes do not involve alterations to the physical plant which would introduce any new or unique operational modes or accident precursors.
3.
The margin of ' safety is not reduced.
The principal DNBR safety 1
criterion has been, and continues to be, that DNB shall be avoided with 95% probability at a 95% confidence level. The margin of safety, which in this case is the margin between the DNBR limit and cladding failure, thus retains the same basis.
In fact, the DNBR limit (i.e.
T.S.
B2.1.1, B3/4.2.3 and 83/4.2.5) is now also based upon the additional criterion that at least 99.9% of the core must avoid DNB, which previously had not been required.
The MTC limit change (i.e. T.S. 3/4.1.1.4, and B3/4.1.1.4) has been analyzed for each of the affected transients from Chapter 15 of tne UFSAR confirming that niargin remains to the DNB design limit.
The pressurizer level response time requirement (i.e. T.S. 3/4.3.3.1) imposes. an additional surveillance requirement on the reactor protection system and therefore does not reduce the margin of safety.
In conclusion, the DNBR analyses which are performed with the Statistical DNBR Evaluation Methodology, including transient analysis with modified key safety parameters, pose no significant hazards as defined by the criteria of 10 CFR 50.92.
1