ML20247H427
| ML20247H427 | |
| Person / Time | |
|---|---|
| Site: | 05200002, 05000470 |
| Issue date: | 03/30/1989 |
| From: | ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY |
| To: | |
| Shared Package | |
| ML20247G537 | List:
|
| References | |
| NUDOCS 8904040457 | |
| Download: ML20247H427 (20) | |
Text
--
APPENDIX 15A LOSS OF PRIMARY COOLANT FLOW METHODOLOGY DESCRIPTION 1
8904040457 890330 PDR ADOCK 05000470 K PDR Amendment No. 7-March 31, 1982 1
EFFECTIVE PAGE LISTING CHAPTER 15 i APPENDIX 15A Table of Contents Page Amendment i 7 l ii 7 k iii 7 Text Page Amendment 15A-1 7 15A-2 7 15A-3 7 15A-4 7
+
15A-5 7 Tables Table No. Amendment 15A-1 7 Figures Figure No, Amendment 15A-1 7 15A-2 7 15A-3 7 15A-4 7 15A-5 7 O
' Amendment No. 7 March 31, 1982
V O
O THIS PAGE INTENTIONALLY BLAl1K, O
)
e f x TABLE OF CONTEllTS CHAPTER 15 APPENDIX 15A Section Subject Page fio.
15A.1 INTRODUCTION 15A-1 15A.2 COMPUTER CODES 15A-1 15A.2.1 Data Transfer 15A-1 15A.2.2 C0AST 15A-1 15A.2.3 CESEC II 15A-2 15A.2.4 HERMITE 15A-2 15A.2.5 TORC 15A-3 ,
l 15A.3 COMPARISON WITH PREVIOUS METHODS 15A-3 (S i 15A.4 TOTAL LOSS OF PRIf1ARY COOLANT FLOW ANALYSIS 15A-4 ,
t !
'd 15A.4.1 Identification of Causes 15A-4 l l
15A.4.2 Sequence of Events and Systems Operation 15A-4 15A.4.3 Analysis of Effects and Consequences 15A-5 I 15A.4.3.1 Mathematical Models 15A-5 15A.4.3.2 Input Parameters and Initial Conditions 15A-5 15A.4.3.3 Results 15A-5 4
15A.4.3 Conclusions 15A-5 4
l l
\
r m. 1
\
i Amendment No. 7 March 31, 1982 l
LIST OF TABLES CHAPTER 15 APPENDIX 15A Table Subject 15A-1 Assumed Initial Conditions for the Total Loss of Primary Coolant Flow Case Presented l
l O{
4 l
O ii Amendment No. 7
.1 arch 31,19??
I LIST OF FICURES CHAPTER 15 APPENDIX 15A Figure Subject 15A-1 Data Transfer Between Computer Codes for the Space-Time Kinetics Loss of Flow Method 15A-2 Total Loss of Forced Reactor Coolant Flow Core-Average Inlet Flow vs Time 15A-3 Total Loss of Forced Reactor Coolant Flow-Core Power vs Time
, 15A-4 Total Loss of Forced Reactor Coolant Flow-Hot Channel l
Heat Flux vs Time ,
1 15A-5 Total Loss of Forced Reactor Coolant Flow-Minimum DNBR l
vs Time
'i O
V 1
4 O
l %
Amendment No. 7 iii March 31. 1982
O Tills PAGE INTENT 10f1 ALLY BLAllK.
g i
l Ol
\
l APPENDIX 15A LOSS OF PRIMARY COOLANT FLOW METHODOLOGY DESCRIPTION ,
l 15A.1 INTRODUCTION !
This appendix describes the analytical methods.used to determine the NSSS response to a loss of primary coolant flow (LOF) which could occur as a i result of a loss of electrical power to the four reactor coolant pumps. I This method, referred to hereafter as the Spa.ce-Time Kinetics LOF (ST-LOF) method, is used to support the conclusions and results in Section 15.3.1.
l A sample analysis of the four pump LOF using the ST-LOF method is provided l
in Section 15A.4. .
l The computer codes used in the ST-LOF method are; C0AST, CESEC II,'ERMITE H I and TORC. These codes are described in topical reports which are referenced in Section 15.0. The principal time dependent parameters calculated are the primary coolant flow rate, reactor core power, hot bundle heat flux and limiting channel Departure from Hucleate Boiling Ratio (DNBR).
15A.2 COMPUTER CODES
)
15A.2.1 Data Transfer n Given the postulated initiating event the C0AST code is used to compute the
( core inlet volumetric flow rate as a function of time. This data is in;;ut to the CESEC code which is used to predict the overall system response.
CESEC calculates Plant Protection System responses and valve actuations for assessing the long term consequences of the LOF. CESF.C also computes the time dependent core inlet mass flux, core inlet coole.1 temperature and i
reactor coolant system pressure (however no credit is taken for pressure increases in computing the DNBR transient) which can be used for input to HERMITE for those cases where a reactor trip occurs so rapidly that only the cociant flow rate changes, CESEC is bypassed and the flow coastdown is input directly to HERMITE.
HERMITE is used to predict the reactor core response during a LOF. HERMITE ,
, calculates the transient core power, core average heat flux and hot bundle J l heat flux. The time dependent core average and hot bundle heat fluxes l along with the core inlet coolant mass flux, core inlet coolant temperature l and reactor coolant system pressure are input to the TORC computer code. '
TORC computes the core average.and limiting channel coolant conditions and the limiting channel DNBR. Figure 15A-1. depicts the transfer of data between the computer codes.
15A.2.2 C0AST The C0AST code is used in the same manner as described in CENPD-183 (Section 15.0 Reference 20). COAST analy:es reactor coolant flow under any combination of active and inactive pumps in a two-loop four pump plant. The equation of conservation of rinmentum is written for each of the flow paths of the Amendment No. 7 15A-1 March 31,1982
C0AST model assuming unsteady one-dimensional flow of an incompressible fluid. The equation of conservation of mass is written for the appropriate nodal points. Pressure losses due to friction, bends and shock losses are assumed proportional to the flow velocity squared. Pump dynamics are modelled using a head-flow curve for a pump at full speed and using four quadrant curves, which are parametric diagrams of pump head and torque on coordinates of speed versus flow, for a pump at other than full speed.
The COAST code has been verified by measurements of the flow coastdown at the Palisades plant. Additionally, the plant specific coastdown is measured and compared with the predicted values during the initial startup test program for each plant to verify that the coastdown used in the analysis is conservative. A further description of C0AST is contained in CENPD-98 (Section 15.0 Reference 13).
15A.2.3 CESEC II The CESEC II code is used to determine the long term response of the NSSS to primary coolant flow reductions resulting from postulated LOF events.
Also, CESEC II may be used to predict the change in core inlet coolant temperature if this parameter changes before the time of minimum DNBR.
CESEC II computes key system parameters during a transient including core heat flux, pressures, temperatures, and valve actions. A partial list of the dynamic functions included in this NSSS simulation includes: point Linctics neutron behavior, Doppler and moderator reactivity feedback, boron and CEA reactivity effects, multi-node average channel reactor core thermal hydraulics, reactor coolant pressurization and mass transport, reactor coolant system safety valve behavior, steam generation, steam generator water level, main steam bypass, secondary safety and turbine valve behavior, as well as alarm, control, protection, and engineered safety feature system actions. The steam turbine and its associated controls are not included in the simulation. Steam generator feedwater enthalpy and flow rate are provided as input to CESEC II. For a further description of CESEC 11 see Section 15.0.
15A.2.4 HERMITE One application of the HERMITE code is to determine the reactor core response ,
during postulated LOF events. HERMITE can accept as input the transient boundary conditions of: coolant flow rate, inlet coolant temperature, i reactor coolant system pressure and CEA position. In this application, l HERiilTE solves the few-group, space and time dependent neutron diffusion I equation including feedback effects of fuel temperature, coolant temperature, coolant density and control rod motion for a one-dimensional average fuel i bundle. The fuel temperature model explicitly represents the pellet, gap, j and clad regions of an average fuel pin and representative hot bundle fuel pin. The hot bundle fuel pin power density is related to the average fuel pin power density by time dependent planar radial power peaking factors, which are discussed in more detail below. For the calculation of heat flux, heat conduction equations are solved by a finite difference method.
Continuity and energy conservation equations are solved in order to determine the coolant temperature and derisity for the average and hot bundles. A further description of HERMITE is contained in CillPD-188-A.
Amendment No. 7 15A-2 March 31, 1982
i n The hot bundle fuel pin power density is equal to the core average fuel pin i
[V power density multiplied by the planar radial power peaking factor, Fr(z).
for times prior to the insertion of CEA's and for regions of the core that the CEA's have not penetrated, the F (2) is equal to a conservatively chosen initial value. AstheCEA'spassablaneofthecore,theradialpower
/
peaking factor of that plane is increased as a functicn of time from the ;
initial value to a final maximum value. This final maximum value of F J has been at least a factor of 5 greater than that predicted by 2-0 traEs(z) ient J HERMITE calculations assuming the worst stuck CEA, over the time of interest.
The radial peaking factor representation leads to the non physical, but conservative, result that the local hot channel power density rises as the CEA's pass each plane.
The synthesis of the axial power distribution and the planar radial power peaking factors provides a conservative representation of the hottest fuel l assembly during the LOF transient including maximum 3-D power peaking effects. This technique yields a conservative prediction of the minimum DNBR which can occur as a result of the LOF transient.
I 15A.2.5 TORC 1
The 10RC code is used to calculate the limiting channel DNBR transient.
TORC receives the core average fuel bundle heat flux, hot bundle heat flux, I core inlet coolant mass flux, core inlet coolant temperature, and reactor I coolant system pressure at selected times during the LOF transient. The cnde is used to perform static calculations of the axial coolant enthalpy O distribution and DNDR at these times. No credit is taken for reacter coolant system pressure increases in calculating the DNBR. TORC solves the conservation of mass, energy and momentum equations for a 3-dimensional I
representation of the open-lattice core to determine the local coolant conditions at points in the core average fuel bundle and hot fuel bundle.
Lateral transfer of mass, momentum and energy between neighboring flow channels (open-core effects) are accounted for in the calculations of local coolant conditions. These coolant conditions and the HERMIIE calculated l
hot bundle heat flux are then used with the CE-1 critical heat flux correla-tion to compute the minimum DNBR value. A further description of 10RC is contained in CENPD-161-P and CENPD-206-P.
15A.3 COMPARISON WITH PREVIOUS METHODS CENPD-183, Appendix A describes the methodology used to predict the conse-quences of postulated LOF events far many previous Combustion Engineering NSSS designs. This section summarizes the fundamental differences between the ST-LOF method and that described in CENPD-183.
The primary difference between these methods is in the calculation of the core power. The CENPD-183 method uses the QUIX code to compute reactivity as a function of CEA position assuming the neutron flux and delayed neutron precursors are in equilibrium. Combining CEA position versus time data with the reactivity versus CEA position data produces the time dependent !
reactivity function which is input to the CESEC point kinetics equations, j Amendment No. 7 March 31, lag 15A -3 L______-__________
The ST LOF mnthod uses HERMITE to calculate core power directly from CEA position versus time. HERMITE calculates the time dependent neutron flux in one dimension (axial) with the few group diffusion equation explicitly accounting for fission, absorption and transport cross section variations.
Other differences exist in the calculation of the hot channel heat flux.
In the CENPD-183 mothcdulogy, it is cssumed that the hot chan~ normalized heat flux decay is equivalent to the core average normalizea maat flux decay for conputing the time of minimum DNBR. Furthermore, it is assumed that the axial heat flux distribution is constant in time. The minimum DNBR value calculated With the CENPD-183 methodolog'y assumes no cecay of the hot channel heat flux.
In tnc ST-LOF method it is assumed that the not bundle normalized power decay is equivalent to the core average normalized power decay, however the hot bundle power decay is modified upon the insertion of CEA's in that the planar radial power peaking factors are increased as the CEA's enter the core. The hot bundle and core average axial heat flux distributions are each time dependent. The rainimum DNBR value calculated with the ST-LOF method is based on the decayed heat flux calculated by HERMITE at the time of mininum DNDR.
CLNPD-183 describes both static and dynamic methods for computing the DNBR.
The ST LOF method uses the static method for calculating the DNBR as described in CEfRD- 103, Appendix A, except that TORC is used in place of COSMO. '
ILA.4 TOTAL LOSS OF PRIMARY COOLANT FLOW ANALYSIS 15A.4.1 J_dentifi. cat, ion of Causes A total Loss of Primary Coolant Flow (LOF) is caused by a simultaneous loss of electric power to all four reactor coolant pumps (RCP). A LOF will occur as part of the sequence of events following a loss of all AC power.
The LOF is currently used as a design basis in determining the required margin to DNB for the Com Operating Limit Supervisory System (COLSS).
Results of an analysis of a LOF are presented herein.
15A.4.? Sequence of Events and Systems Operation _
A loss of electric power to all reactor coolant pumps produces a reduction of coolant flow through the reactor core. The reduction in coolant flow rate causes an increase in the core average coolant temperature with a concurrent decrease in the margin to DND. A low DNBR reactor trip is generated by the core protection calculators, as described in Section 7.2.
The CEAs begin to drop into the core 1.09 seconds af ter the loss of electric power to the pumps. The 1.09 second delay conservatively includes the largest possible times for sensor celays, CPC calculation period, CEDM dead tiue, and CEDM coil decay time. The minimum DNBR o' l.19 occurs 2.56 seconds after the initiation of the event.
O Amendment No. 7 Ma rch 11,10R7 15A-4 l
I 1
15A.4.3 Analysis of Effects and Consequences l 15A,4.3.1 Mathematical Models v
The NSSS response to a LOF was simulated using the methods described pre-viously in this Appendix.
15A.4.3.2 Input Parameters and Initial Conditions The input parameters and initial conditions used to analyze the NSSS response to a LOF are discussed in Section 15.0. The parameters, which are unique to the analysis discussed below, are listed in Table 15A-1.
The principal process variables that determine thermal margin to DNB in the core are monitored by COLSS. COLSS computes a power-operating limit which ;
assists the operator in maintaining adequate thermal margin in the core. l When this margin is initially available, the CPCs will prevent the minimum l DNBR frot being less than 1.19 during a LOF, COLSS is described in Section '
7.7. The set of initial conditions chosen for the analysis presented in this section is one of a very large number of combinations within the reactor operating space given in Table 15.0-5 which would provide the minimum thermal margin required by the COLSS power operating limit. Th consequences following a LOF initiated from any one of these combinations of conditions would be no more adverse than those prasented herein.
i 15A.4.3.3 Results The time dependent behavior of the core flow, reactor core power, hot bundle heat flux, and limiting cFannel DNBR is presented in Figures 15A-2 theough 15A 5. The core coolant inlet temperature does not change prior to the time when the minimum DNLR is reached. For conservatism no credit is taken for the slight RCS pressure increase in computing this minimum DNBR.
15A.4.3.4 Conclusions ;
The minimum DNBR does not fall below 1.19. The maximum pressurizer pressure is less than 2500 psia, and there are no significant radiological releases.
l 1
l (n]
v
~
Amendment No. 7 March 31. 1982
, 15A-5
O THIS PAGE INTENTIONALLY BLANK.
O i
O
(9 TABLE 15A-1 V ASSUMED INITIAL CONDITIONS FOR THE TOTAL LOSS OF PRIMARY COOLANT FLOW CASE PRESENTED Parameter Assumed Value Core Power Level, MWt 3876 Core Inlet Coolant Temperature, F 567 Core Mass Flow Rate,106 lbm/hr 157.4 l Reactor Coolant System Pressure, psia 1800 Initial Core Minimum DNBR 1.51 Maximum Radial Power Peaking Factor 1.62 Maximum Axial Power Peak. 1.47 i
Axial Shape Index .269 l
Doppler Coef ficient Multiplier 0.0 1 CEA Worth for Trip,10-2 a -10.0 I
l l
l l
1 i
i l
3
. (0 l
l l
Amendment No. 7 March 31, 1982
O l l
1 l
l 1
j I
THIS PAGE INTENTIONALLY BLANK, i
i O;
1 l
NSSS DESIGN DATA O -
COAST l O(t) v m
. .Y WILL CF 1 BEFOREt n l CESEC 7 YES U
CESEC P(t), TIN (t)
G(t)
LONG TERM !
NSSS RESP, !
CEA,(t)
VyV l HERMITE 4 AVG (t),qHB(t)
LEGEND CEA, - CEA AXIAL liElGHT DNDR - DEPARTURE FROM NUCLEATE DOILING RATIO P -
RCS PRESSURE O - REACTOR VESSEL VOLUMETRIC V FLOW RATE 4 AVG . CORE AVERAGE HEAT FLUX TORC '
ggg -
HOT BUNDLE HEAT FLUX DNBR(t)
T IN - CORE INLETCOOLANT TEMPERATURE G -
REACTOR VESSEL MASS FLOW RATE t TIME ,
tm TIME OF MINIMUM DNBR .
SHORT TERM NSSS RESP.
(DNBR)
Amendment No. 7 O March 31,1982 C-E Figure DATA TRANSFER BETWEEN COMPUTER CODES FOR SDEF8f / ,
THE SPACE-TIME KINETICS LOSS OF FLOW METHOD 15A-1
___.-___---___w
'l
( 1 l
i i i 2.8 - -
l N
tz h2.6 - -
3 o
- . 2. 4 3:
S u- ;
0 N 2.2 0
1 83 D
w 2.0 - -
o <
1 1 I I 1.0 2.0 3.0
- TIME, SECONDS
'l Amendment No. 7 March 31.1982
\ !
C'- E iT' TOTAL LOSS OF FORCED REACTOR COOLANT FLOW Figure SEl78w / ! CORE AVER AGE INLET FLOW vs TIME 15A-2
y I
O ;
I i g 1
'100 _
90 - _
[ 80 - _
s s 70 s
b60 - _
m !
O u.
Q 50 w
a.
40 - -
w 1
@ 30 )
o A 20
- J 10 I i g 1.0 2.0 3. 0 i TIME, SECONDS Amendment No. 7 March 31, 1982
\ _
TOTAL. LOSS OF FORCED REACTOR COOLANT FLOW- - Figure S 3.b CORE POWER vs TIME 15A-3
O I I i
- p 100 -
C i5 i $ I u
5 c.
x' 90 a
b
=
w s
5 1 E3 z 80 i
i l i
- 1. 0 2. 0 3.0 TIME, SECONDS Amendment No. 7.
March Fi,1982 C-E i TOTAL LOSS OF FORCED REACTOR COOLANT FLOW Figure EM@P8/7/ ! HOT CHANNEL HEAT FLUX vs TIME 15A-4 3
O i I i
- 1. 8 - -
i 1.6 - -
E 5
s 5 1.4 -
l l
5 2
/
N 1.2 -
N -
1.0
, , n 1.0 2.0 3.0 TIME, SECONDS
' Amendment No. 7 March 31, 1982 C-E / TOTAL LOSS OF FORCED REACTOR COOLANT FLOW Figwe 88F_fdclu/ ! MINIMUM DNBR vs TIME 15A-5
- - - - -