ML20140G650
| ML20140G650 | |
| Person / Time | |
|---|---|
| Site: | Robinson  |
| Issue date: | 12/31/1985 |
| From: | Jo J, Rohatgi U, Yuelysmiksis BROOKHAVEN NATIONAL LABORATORY |
| To: | NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| References | |
| CON-FIN-A-3215, REF-GTECI-A-49, REF-GTECI-RV, TASK-A-49, TASK-OR BNL-NUREG-51946, NUREG-CR-4452, NUDOCS 8604030081 | |
| Download: ML20140G650 (63) | |
Text
_
1 N EG/CR-4452 B; =NUREG-51946 REVIEW 0F RELAP5 CALCULATIONS FOR H.B. ROBINSON UNIT 2 PRESSURIZED THERMAL SH0CK STUDY C. Yuelys-Miksis, U.S. Rohatgi, and J.J. Jo Date Published:
December 1985 l
i l
DEPARTMENT OF NUCLEAR ENERGY, BROOKHAVEN NATIONAL LABORATORY UPTON N.Y.11973
)
I ijl p-4
'I Prepared for the U S Nuclear Regulatory Comrmssion Office of Nuclear Regulatory Research
- ,, M tj bj Contract No DE-ACO2-76CH00016 3,,;
s,,, r,.-,
t et i 't
-' L b-8604030081 851231 PDR ADOCK 05000261 CR-4452 R PDR i.
.g
,,,a
..-,,,--------,--,,-y
-,---y
,y,.,-
y
--_,,w
NUREG/CR-4452 BNL-NUREG-51946 AN. R-4 REVIEW 0F RELAP5 CALCULATIONS FOR H.B. ROBINSON UNIT 2 PRESSURIZED THERMAL SH0CK STUDY C. Yuelys-Miksis. U.S. Rohatgi, and J.J. Jo Manuscript Completed: October 1985 Date Published:
December 1985 LWR CODE ASSESSMENT AND APPLICATION GROUP DEPARTMENT 0F NUCLEAR ENERGY BROOKHAVEN NATIONAL LABORATORY UPTON. LONG ISLAND NEW YORK 11973 Prepared for 0FFICE OF NUCLEAR REGULATORY RESEARCH UNITED STATES NUCLEAR REGULATORY COMMISSION WASHINGTON DC 20555 CONTRACT NO. DE-AC02-76CH00016 l
NRC-FIN A-3215
e l
l NOTICE This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any sgency thereof.or any of their employees, makes any warranty, expressed or implied, or assumes any legal liability or responsibihty for any third party's use, or the results of such use, of any information, apparatus, product or process disclosed in this report, or representa that its use by such third party would not infringe privately owned rights.
The views expressed in this report are not necessarily those of the U.R Nuclear Regulatory Commission.
Available from Superintendent of Documents U.S. Government Printing Office P.O. Box 37082 Wanhington. DC 20013-79M2 and National Technical Information Service Springfield, Virginia 22161 t
\\
L
EXECUTIVE
SUMMARY
Brookhaven National Laboratory was requested by the USNRC to review the thermal-hydraulic calculations performed by Los Alamos National Laboratory (LANL) and Idaho National Engineering Laboratory (INEL) as a part of an NRC program to study the Pressurized Thermal Shock safety issue. This report pre-sents the results of the BNL review of the selected calculations performed by INEL using the RELAP5/M001.6 code for the H. B. Robinson Unit-2 PWR plant.
A number of transient scenarios were simulated by INEL. The RELAP5 plant input deck and 11 of these scenarios were reviewed at BNL on the basis of information received before September 30, 1984*. No major discrepancies were found in the plant input deck or in the 11 transient calculations.
Six of these 11 transients were selected for an in-depth quantitative analysis. However, the detailed analysis of two of these transients, namely transients 1 and 4 are not presented here as they have since been recalculated by INEL and the new calculations could not be reviewed at BNL due to budgetary constraints. The review of these RELAP5 calculations was based on a simple method developed at BNL to predict the primary system temperature on the basis i
of mass and energy balances of the entire reactor system. This simplified approach applied the energy balance equation to a single volume. The entire reactor system, including the steam generator secondary sides and the metal structures (unless otherwise mentioned), was lumped into this single volume.
This volume was represented by a single average temperature and was compared to the RELAP5 results and the extrapolations provided by INEL.
The primary and secondary pressures were more difficult to estimate by simple analysis because they were largely dependent on the condensation and evaporation rates. Wherever possible the pressurizer pressure predictions were calculated by assuming adiabatic compression of the vapor and/or complete equilibrium of the phases in the pressurizer. The former assumed no condensa-tion or evaporation and was expected to provide the upper bound of system pressure when the pressurizer was filling and the lower bound during the emptying stage. The equilibrium approach provided the upper bound of the pressure during the emptying stages and the lower bound during the filling stage. The actual pressure was expected to fall between these boundaries.
However, because of the significant nonequilibrium effects and the many fac-tors affecting the condensation and evaporation rates that were difficult to quantify with this simple approach, it was more appropriate to compare the pressurizer water levels obtained using RELAP5 and the BNL simple approach.
Some uncertainites in the RELAPS calculations arose because of the multidimensional effects in some of the transients. The core model in RELAPS is one-dimensional and thus cannot adequately represent these multidimensional effects. The uncertainty in a one-dimensional analysis would affect predic-tions of the temperature distribution in the core, downcomer, and hot legs.
o Four of these transients (Nos. I through 4) were recalculated by INEL. How-ever, these new calculations could not be reviewed at BNL because of resource limitations.
o 4
Uncertainties are also associated with the modeling of condensation effects in the pressurizer, and further work is necessary to assess the accuracy of code calculations of the primary pressure when the pressurizer is filling.
However, in general, the RELAPS-calculated primary temperatures and pressures appear to be reasonable, although the INEL extrapolations might be somewhat conservative for certain transients.
a l
I I
1 i
f l
- iv -
t i
ew m
g
-mu--
j l
l ABSTRACT i
Idaho National Engineering Laboratory (INEL) used the RELAP5/M0Dl.6 code l
to simulate a number of transient scenarios for the USNRC PTS study of the H. B. Robinson Unit-2 PWR plant. Eleven of these scenarios were revir.eN >:
BNL on the basis of information received before September 30, 1984*.
~ 's 4
these eleven scenarios were selected for an in-depth quantitative e'c!1-1 performed on the basis of a simple method developed at BNL. Howevo,
detailed analysis of two of these transients, namely transients 1 e4' i
not presented here as they have since been recalculated by INEL and t.
m*
calculations could not be reviewed at BNL due to budgetary constraints.
simple method uses the mass and energy balance equations to predict the temperature and pressure of the reacter system. The results of these calcula-tions were compared to the RELAPS results and the INEL extrapolations.
In 1
general, the RELAP5 and INEL results appear to be reasonable, i
i i.
i h
f i
l I
't j
o Four of these transients (Nos. I through 4) were recalculated by INEL. How-ever, these new calculations could not be reviewed at BNL because of resource limitations.
{
j
-v-1 i'
f....
,m.-_
- - _-~-_ -.. - _.. _ _ _.
_.___.mm.._.m..__m__,m_.
..;__.m..m_....
- ~ - _
~
t T
I
+
f iif P
i F
t P
F t
t Y
i t
j p
I r
i h
I i
i 1
i i
i t
6
?
L w-e.-e-----
w
-m v v vw..
m.
i i
l 1
l CONTENTS 1
Section Page ABSTRACT...............................
v EXECUTIVE
SUMMARY
iii ACJ"' WL EDGEME NT............................
x ABBREVIATIONS.............................
xi 1.
INTRODUCTION..........................
I 2.
TRANSIENT 6: SMALL H0T LEG BREAK LOCA AT HOT FULL POWER CONDITIONS...........................
9 4
3.
TRANSIENT 8: SMALL HOT LEG BREAK LOCA AT HOT ZER0 POWER CONDITIONS...........................
14 4.
TRANSIENT 9:
STEAM GENERATOR TUBE RUPTURE AT HOT ZERO POWER i
CONDITIONS...........................
18 l
5.
TRANSIENT 11: LOSS OF SECONDARY HEAT SINK WITH PRIMARY SYSTEM
]
FEED-AND-BLEED REC 0VERY AT HOT FULL POWER CONDITIONS......
22 6.
SUMMARY
AND CONCLUSIONS.....................
26 i
7.
REFERENCES...........................
27 APPENDICES A - BNL REVIEW 0F RELAPS INPUT DECKS FOR H.B. ROBINSON FROM INEL..
29 i
8 - PRELIMINARY ASSESSMENT OF RELAPS THERMAL-HYDRAULIC ANALYSIS 0F PTS TRANSIENTS OF H.B. ROBINSON UNIT 2............
33 4
C EXTRAPOLATION OF EXISTING PTS CALCULATIONS WITH OR WITHOUT CHANGES IN BOUNDARY CONDITIONS.................
45 l
l 1
1 l
l
- vii -
1
FIGURES Figure No.
Page 2.1 Transient 6: Downcomer and Hot leg Liquid Temperature...
12 2.2 Transient 6: Liquid Temperature in the Downcomer.....
12 2.3 Transient 6: Normalized Pressurizer level.........
13 2.4 Transient 6: Downcomer Pressu re..............
13 3.1 Transient 8: Downcomer and Hot Leg Liquid Teaperature...
16 3.2 Transient 8: Liquid Temperature in the Downcomer...
16 3.3 Transient 8: Downcomer Pressu re..............
17 4.1 Transient 9: Liquid Temperature in the Downcomer.....
20 4.2 Transient 9: Normalized Pressurizer Level.........
20 4.3 Transient 9: Calculated and Adjusted Reactor Vessel Downcomer Fluid Pressure...........
21 4.4 Transient 9: Steam Generator Secondary Pressure......
21 5.1 Transient 11: Liquid Temperature in the Downcomer.....
24 5.2 Transient 11: Liquid Temperature in the Downcomer, 4000 - 11,000 sec.
24 5.3 Transient 11: Downcomer Pressure 25 i
l l
l l
1
- viii -
4 4
L I
s t
TABLES N
Table No.
Page i
1.1 H. B. Robinson PTS Transients'...............
4 1
1 1.2 Comparison Between RELAPS and Design / Plant Data of Hot Full Power Conditions.............,...
i 7
i 1.3 Comparison Between RELAPS and Design / Plant Data of l
Hot Standby Condit ions...................
8 2.1 Scenario for Transient No. 6...............'.
11 3.1 Scenario for Transient No. 8.....'...........
15 i
j 4.1 Scenario for Transient No. 9................
19 i
1 5.1 Scenario for Transient No. 11'...............
23 i
l i
]
~ <
1 2
Y t
1 1
4 V-f 1
I 1
i ix -
j
ACKNOWLEDGEMENTS The. authors wish to express their appreciation to Dr. P. Saha of BNL for his valuable comments and suggestions. They also would like to acknowledge l
the efforts of Ms. Ann C. Fort and Ms. Nancy Griffin in typing this report.
Thanks are also due to Don Fletcher of INEL for providing information and clarifications about their calculations.
j l
1 i
1 j
i i
t i
i r
I i
k l
-x-
ABBREVIATIONS A
Affected AFW
\\uxiliary Feedwater BNL Brookhaven National Laboratory ECC Emergency Core Cooling HFP Hot Full Power HPI High Pressure Injection HZP Hot Zero Power INEL Idaho National Engineering Laboratory LOCA Loss of Coolant Accident LPI Low Pressure Injection MFIV Main Feedwater Isolation Valve MFW Main Feedwater l
MSIV Main Steam Isolation Valve j
ORNL Oak Ridge National Laboratory PORV Power Operated Relief Valve PTS Pressurized Thermal Shock RCP Reactor Coolant Pump RCS Reactor Coolant System SDV Steam Dump Valve SIAS Saf ety Injection Actuation Signai SG Steam Generator TSV Turbine Stop Valve USNRC U.S. Nuclear Regulatory Commission
- xi -
1.
INTRODUCTION Rapid cooling of the reactor pressure vessel during a transient or acci-dent accompanied by high coolant pressure is referred to as pressurized ther-mal shock (PTS).
In late 1981 the U. S. Nuclear Regulatory Commission (NRC) designated PTS as an unresolved safety issue and developed a task action plan (TAP A-49) to resolve the issue.
The saf ety issue exists because rapid cooling at the reactor vessel wall inner surf ace produces thermal stresses within the wall.
As long as the f rac-ture toughness of the reactor vessel is high, overcooling transients will not cause vessel failure. However, NRC staff analyses (SECY-82-465) showed that certain older plants with copper impurities in vessel weldments may become sensitive to PTS in a few years as the nil-ductility transition temperature of the weld material gradually increases. The purpose of the thermal-hydraulic analyses presented in this report is to better understand the behavior of a plant during various kinds of postulated severe overcooling transients with multiple f ailures of equipment and without operator corrective action.
The understanding gained f rom these detailed calculations will be used to interpo-late coolant temperature and pressure responses in the downcomer for other postulated transients using a simplified mass-and-energy balance approach. For each of these postulated transients, Oak Ridge National Laboratory (0RNL) will then calculate the reactor vessel temperature distribution and stresses during the transient and the conditional probability of vessel f ailure if the tran-sient should occur. ORNL will publish a report that integrates these results to estimate the likelihood of PTS driving a crack through the reactor vessel wall and to identify important event sequences, operator and control actions, and uncertainties.
The inf ormation obtained in this series of analyses is intended to help the NRC staff confirm the bases for the screening criteria in the proposed PTS rule (proposed 10CFR 50.61) and determine the content required f or licensees' plant-specific saf ety analysis reports and the acceptance criteria f or correc-tive measures.
The U. S. Nuclear Regulatory Commission (USNRC) has selected three plants for detailed PTS study which represent the PWRs supplied by three ven-i dors in the United States.
These plants are Oconee-1 (Babcock & Wilcox),
Calvert Cliffs-1 (Combustion Engineering), and H. B. Robinson-2 (Westinghouse Electric). Oak Ridge National Laboratory (0RNL), the lead contractor for the entire PTS study, has identified several groups of transients with multiple failures of equipment and without corrective operator action which could lead to severe overcooling in these plants.
Los Alamos National Laboratory (LANL) 1 and Idaho National Engineering Laboratory (INEL) were selected to perform the thermal-hydraulic calculations for these transients using the latest versions of the TRAC-PWR and RELAPS codes, respectively.
The Oconee-1 transient cal-culations were apportioned between LANL and INEL, with some transients calcu-lated by both.
The Calvert Cliffs and H. B. Robinson transient calculations were perf ormed by LANL and INEL, respectively. _ _ _ _ _ _ _
Brookhaven National Laboratory (BNL) was requested by the USNRC to re-view and compare the plant input decks developed at LANL and INEL, and to re-view the calculation results. This report presents the results of the BNL re-view of the selected H. B. Robinson calculations performed by INEL. The re-sults of the BNL review of the Oconee-1 and Calvert Cliffs calculations have been reported earlier (Rohatgi,1984; Jo,1985).
Because of resource limitations, this review is based on the information received at BNL before September 30, 1984. Much of the review is based on the INEL draft report (Fletcher,1983) presenting the results of 11 transients calculated using the RELAP5/ MOD 1.6 code.
(The results presented in the INEL final report (Fletcher,1984) could not be reviewed at BNL because of funding limitations.) These transients, as shown in Table 1.1, were chosen to maxi-mize the understanding of the effects of various combinations of operator and equipment failures, regardless of their probability of occurrence, in a mini-mum number of calculations. The transient scenarios chosen were purely hypo-thetical and not necessarily probable.
Six of the transients were initiated from hot full power (HFP) conditions and the remaining five were initiated from hot zero power (HZP) conditions. A comparison between the RELAPS-calcu-lated initial conditions for the hot full power conditions at the time of the plant trip transient and the initial values representative of the H. B. Robin-son plant is shown in Table 1.2.
The standby hot zero power steady-state ini-tial conditions are compared in Table 1.3.
The RELAPS input decks for these calculations were reviewed at BNL and comments were transmitted to the NRC, ORNL, and INEL study members in March 1984. A copy of this memorandum is pre-sented in Appendix A. In addition, a preliminary assessment of the RELAPS thermal-hydraulic analysis of the 11 transients is presented in Appendix B.
This was also transmitted to the study participants in August 1984.
Six of the 11 transients that were selected for detailed review at BNL are Transients 1, 8 and 9 for the HZP condition, and Transients 4, 6 and 11 for the HFP condition. However, the detailed analysis of two of these transients, namely Transients 1 and 4 are not presented here as they have since been recalculated by INEL and the new calculations could not be reviewed at BNL due to budgetary constraints. Transients 6 and 8 are identical except that Transient 6 is initiated from the HFP conditions, whereas Transient 8 is initiated from the HZP conditions. These transients are representative of a small hot leg break LOCA and permit comparison between the two extreme power levels. Since Transients 9 and 10 are both steam generator tube rupture accidents, Transient 9 was designated the worst case scenario because it is initiated from the HZP conditions. Transient 11, which is a loss of secondary heat sink with a primary system feed-and-bleed recovery,is different from any other transient.
The quantitative review of these RELAPS calculations was based on a sim-ple method developed at BNL to predict the primary system temperature on the basis of mass and energy balances. This method was described in detail in a memorandum in January 1984 and is presented in Appendix C.
This simplified approach applies the energy balance equation to a single volume. The entire reactor system, including the steam generator secondary sides and the metal structures (unless otherwise mentioned), is lumped into this single volume.
However, the mass balance equations are applied separately to the primary system and to the secondary side of each steam generator. This simplified method assumes that there are relatively small differences between the hot and cold leg temperatures and t,' tween the steam generator primary and secondary side temperatures.it will be shown in this report that, for most of the tran-sients, the system temperature calculated with this simple method are reason-ably close to the primary temp tratures predicted by RELAPS.
The primary and secondary pressures are more difficult to estimate by simple analysis because they are largely dependent on the condensation and evaporation rates. Wherever possible the pressurizer pressure predictions are calculated by assuming adiabatic compression of the vapor and/or complete equilibrium of the phases in the pressurizer. The former assumes no condensa-tion or evaporation and is expected to provide the upper bound of system pres-sure when the pressurizer is filling and the lower bound during the emptying stage. The equilibrium approach provides the upper bound of the pressure dur-ing the emptying stages and the lower bound during the filling stage. The actual pressure is expected to fall between these boundaries. However, be-cause of the significant non-equilibrium effects and many factors that affect-ed the condensation and evaporation rates, but were difficult to quantify with this simple approach, it was more appropriate to compare the pressurizer water levels obtained with RELAPS and by BNL.
Although the pressurizer water levels predicted with RELAP5 tend to ap-proximate the pressurizer pressure trends, there are uncertainties in the code in modeling the condensation effects while the pressurizer is filling. During this time the condensation rate is unrealistically high because it is calcu-lated from the volume averaged void fraction, which yields a higher inter-f acial area than actually exists at the boundary between the two phases.
This results in an unrealistically high condensation rate and an underestimation of primary system pressure.
In addition, the high condensation rate in the pres-surizer causes fluid from the upper head to flow to the pressurizer. For this reason, the RELAP5 pressurizer level for several transients is higher than the level calculated at BNL.
Other uncertainties in the RELAP5 calculations arise because of the im-portance of multi-dimensional effects in some of the transients.
In the reac-l tor system, fluid of different temperatures could be passed to dif ferent hot legs. This would then af fect the primary pressure and flow rates, the pres-surizer condensation rate, and the voiding in the loops and upper head.
If hotter fluid is passed to the pressurizer loop, the condensation rate in this loop would be lower, thereby maintaining a higher primary pressure. The re-sulting reduction in safety injection flows and the earlier cessation of high pressure injection would cause the downcomer fluid temperature to remain at a higher level than predicted. The core model in RELAPS is one-dimensional and cannot adequately represent these multidimensional effects.
The uncertainty in a one-dimensional analysis would af fect predictions of the temperature distribution in the core, downcomer, and hot legs.
l Table 1.1 H. B. Robinson PTS Transients Trans Initial Plant Equipnent Failures No.
Descriptive Title State Initiating Event on Demand Operator Actions 2
2 1
1.0-f t steam line break Hot 0% Power 1.0-f t break in None 1.RCPs tripped if primary at hot standby.
(HZP) steam line A.
pressure below 1315
. psia and SIAS generated.
2.Stop AFW to SG "A" af ter 10 min.
2 Double-ended guillotine iiot 0% Power Full break in steam 1.AFW to the 1.RCPs tripped if primary steam line break at hot (HZP) line A.
aff ected SG is pressure below 1315 standby.
not isolated.
psia and SIAS generated.
- 2. Fails to isolate AFW 8
to SG "A".
l s-I 3
Stuck open steam line Hot 0% Power Steam line A PORV 1.AFW to SG "A" 1.RCPs tripped if primary PORV at hot standby.
(HZP) fails open.
is not isolated.
pressure below 1315 psia.
4 Three steam dump valves 100% Power Three out of five 1.The af f ected 1.RCPs tripped if primary f ail open at f ull power.
(HFP) main steam line dump SDVs will not pressure below 1315 valves are locked
- close, psia and SIAS generated.
open.
2.MSIV in Loop A 2.AFW to unaff ected SG fails to close stopped when liquid l
if trip signal carryover in main steam 1
generated.
line is observed.
3.AFW to SG "A" stopped 10 min af ter attempted MSIV closure or when carryover occurs.
f i
Table 1.1 (Cont.)
Trans Initial Plant Equipment Failures No.
Descriptive Title State Initiating Event on Demand Operator Actions 5
Overf eed with AFW at 100% Power SG overf eed with 1.AFW pumps fail
- 1. Restart AFW pumps to full power.
(HFP)
AFW.
to start.
provide max. flow at 8 min to all 3 SGs.
2.AFW stopped in main steam lines when liquid carryover observed.
6 Small hot leg break at 100% Power 2.5-i n.-d iameter None 1.RCPs tripped if primary full power.
(HFP) hole Loop C hot leg.
pressure below 1315 psia and SIAS generated.
2.AFW throttled to main-on tain 40% SG level.
7 Stuck open pressurizer 100% Power Primary PORV sticks 1.P0RV blocking 1.P0RV blocking valve shut PORV at full power.
(HFP) open.
valve does not after 10 min.
close until 10 2.RCPs tripped if primary min into the pressure below 1315 psia transient.
and SIAS generated.
8 Small hot leg LOCA at Hot 0% Power A 2.5-in.-diameter None 1.RCPs tripped if primary hot standby (HZP) in the pressurizer pressure below 1315 psia loop.
and SIAS generated.
2.AFW throttled to main-tain 40% SG level.
l Table 1.1 (Cont.)
Trans Initial Plant Equipment Failures No.
Descriptive Title State Initiating Event on Demand Operator Actions 9
Steam generator tube Hot 0% Power A single steam gen-None 1.RCPs tripped if primary rupture at hot standby.
(HZP) erator tube rupture.
pressure below 1315 psia and SIAS generated.
2.AFW throttled to main-tain 40% SG 1evel.
3.RCPs restarted after 10 min. if :(a) pressurizer level > 20% or in-creasing, (b) R.C. pres-sure > 325 psig and (c)
>40 F subcooled.
m 10 Steam generator tube 100% Power A single steam gen-None 1.RCPs tripped if primary rupture at full power.
(HFP) erator tube rupture.
pressure below 1315 psia and SIAS generated.
2.AFW throttled to main-tain 40% SG level.
11 Loss of secondary heat 100% Power Loss of secondary
SG wide range level be-f eed-and-bleed recovery.
panied by opening low 5%.
both pressurizer 2.HPI initiated and PORVs PORVs and initiating opened af ter RCP trip &
HPl.
loop A hot leg tempera-ture increased 5 F.
I
Table 1.2 Comparison Between RELAP5 and Design / Plant Data at Hot Full Power Conditions Parameter RELAP5/M001.6 Design / Plant PRIMARY SIDE Core Power 2300 MW 2300MW Pressurizer Pressure 15.5 MPa 15.5 MPa (2250 psia)
(2250 psia)
Hot Leg / Cold Leg Temp.
591.4/558.9 K 591.4/558.9 K (604.8/546.3*F)
(604.8/546.3*F)
Pressurizer Level 54.5%
53.3%
Pump Speed 1247.1 1190 RPM Mass Flow 12726 kg/s 12726 kg/s (28056 lb/s)
(28056lb/s)
Net Make-up Flow 0.001 gpm 0.0 gpm SECONDARY SIDE Steam Pressure 5.5 MPa 5.7 MPa (804 psia)
(828 psia)
Steam Generator Level 53.7%
52%
Steam Flow per SG 425 kg/s 424.2 kg/s (937lb/s)
(935.2lb/s)
Liquid Mass per SG 44253 kg 42302 kg (97560 lb)
(93260 lb)
Feedwater Temperature 500.4 K 500.6*K (441F)
(441.5'F)
Condenser Temperature 312 K 312 K (102 F)
(l02 F)
Feedwater Recirculation 0 gpm Not Available l
Table 1.3 Comparison Between RELAP5 and Design / Plant Data at Hot Standby Conditions Parameter RELAP5/ MOD 1.6 Design / Plant PRIMARY SIDE Core Power 8.29 MW 8.29 MW Pressurizer Pressure 15.5 MPa 15.5 MPa (2250 psia)
(2250 psia)
Average Temperature 560.17 K 559'K (548.65*F)
(547'F)
Pressurizer Level 23.6%
24.4%
Pump Speed 1226 1190 RPM Mass Flow 12629 kg/s 12626.8 kg/s (27841.9 lb/s)
(27837 lb/s)
Net Makeup Flow 0.0 gpm 0.0 gpm SECONDARY SIDE Steam Pressure 7.03 Pa 7.03 MPa (1020 psia)
(1020 psia)
Steam Generator Level 38.4%
39%
Steam Flow per SG 2.27 kg/s Not Available (5.0 lb/s)
Liquid Mass per SG 57788 kg 54432-61689 kg (127400lb)
(120000-136000 lb)
Feedwater Temperature 299.8 K Not Available (80'F)
Condenser Temperature 299.8*K Not Available (80'F)
Feedwater Recirculation 1500 gpm Not Available 1 -__-_____-_.
i
2.
TRANSIENT 6: SMALL HOT LEG BREAK LOCA AT HOT FULL POWER CONDITIONS This transient was initiated by a 0.0635 m (2.5 in) diameter break in the pressurizer loop hot leg during hot full power operation. All systems operated automatically and there were no equipment failures.
The transient scenario description is given in Table 2.1.
The RELAP5/ MOD 1.6 code was used by INEL to calculate the transient to 2800 seconds and the key parameters were extrapolated to 7200 seconds. Fig-ure 4.1 shows the downcomer and hot leg temperatures as calculated with RELAPS and the BNL system average temperatures without considering the ener-gy stored in metal structures. This BNL system average temperature assumes that the relatively slow wall heat transfer between the liquid and the metal structures of the reactor and other components will not affect the tempera-ture of the liquid in the early stages of the transient. A second BNL tem-perature was also calculated, but is not shown, with the assumption that the wall heat transfer is instantaneous, so that the metal and the liquid tem-peratures change simultaneously.
The actual temperature would initially be close to that calculated without accounting for the metal structures and eventually approach that calculated with the metal' stored heat accounted for since the metal cooling is significantly slower than the liquid cooling.
In this scenario, multidimensional effects were observed when two of the three loops experienced stagnation.
In general, the RELAP5 one-dimensional vessel model cannot adequately predict multidimensional effects such as radial and azimuthal temperature distribution in the core and downcomer.
Since different fluid temperatures will not be passed to different hot legs, the primary pressure and flow rates would also be affected.
If the hot leg temperature in the pressurizer loop were higher than calculated, then the rate of condensation in the pressurizer would be lower.
In addition, there might also be more voiding in this loop and in the upper head, which would maintain a higher primary pressure and would result in an earlier cessation of the high pressure injection. Consequently, the downcomer liquid tempera-ture might be higher than predicted. However, since this behavior was seen only in the early portion of the transient, the end results should not be appreciably affected by multidimensional effects. Natural circulation in the broken loop stopped at about 1000 seconds. At this time the cold HPI and makeup flow caused the large drop in the downcomer temperature because stagnation prevented hot leg fluid from reaching the cold leg. As seen in Figure 2.1, the BNL system average temperature agrees well with the RELAPS calculation until stagnation occurs at -1000 seconds. After stagnation, the BNL temperature is between the hot leg temperature and the downcomer temperature.
In the temperature extrapolation, shown in Figure 2.2, INEL calculated a sharp drop of the downcomer temperature at around 2700 seconds. Altnough
)
this drop levels off at the ECC fluid temperature, decay power input still exists and there are upper plenum to downcomer leakage flows; therefore, it is not likely that the downcomer temperature will decrease to the ECC tem-l l
perature in a relatively short time period. However, it is conservative to assume a sudden temperature drop, and the actual temperature probably will fall between the INEL extrapolated and the BNL system average values.
Figure 2.3 shows the pressurizer level.
In both the RELAP5 and BNL calculations the pressurizer rapidly lost inventory. After the pressurizer empties, it is expected that the downcomer pressure will be maintained at the saturation pressure corresponding to the hot leg temperature. This can be seen in Figure 2.4.
Since the BNL system average temperature is lower than the RELAPS hot leg temperature, its saturation pressure will also be lower, although it is expected eventually to converge to the INEL pressure.
However, since stagnation affects only the cold leg temperature, the pres-sure should follow a gradual decrease as opposed to the rapid drop predicted by INEL.
In summary, the RELAPS calculated results appear reasonable. The INEL extrapolations are conservative, but eventually converge to the expected values.
i l _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
)
Table 2.1 Scenario for Transient No. 6 Plant Initial State - Just prior to transient initiator.
General
Description:
100% Power steady state.
System Status:
Turbine: Automatic control Secondary PORV: Automatic control Steam Dump Valves (SOVs): Operative / Automatic control Charging System:
Automatic control Pressurizer: Automatic control Engineering Safety Features: Automatic control PORVs: Automatic control Reactor Control: Automatic Main Feedwater: Automatic control
(
Aux Feedwater: Automatic control MSIVs: Open, Automatic control MFIVs: Open, Automatic control.
Transient Initiator - A 2.5-in. hole appears in the hot leg.
Equipment Failures which occur during the transient if the equipment is demanded.
None.
Operator Reactions to Reported Information 1.
If SIAS signal is generated, the operator will trip the reactor coolant pumps when RCS pressure reaches 1300 psig.
2.
The operator will throttle AFW flow to maintain 40% SG level. _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _
C Auf lON. THE SCE'tARIOS $1MJLATED CONT AIN $1CNIFICANT CONSERbAfl$WS IN CPERATOR ACTIChS. EQUlPWENT FAILURES, OR BOTH.
600 i iiii iiiiiiiiiiiiiiiiiiiie i ii X
X x
. N x
s x
500 x
N, x
s
~~~~~~~____-s h.400 N
}
o f,300 r
200 Donncomer Temp. ( REL AP S)
~
x Loop A Hot Leg Temp. ( RELAP S)
- ------ System Average Temp. without Metal Structures (BNL) 100 O
SOO 1000 1500 2000 2500 3000 Time (s )
Figure 2.1 - Transient 6: Downcmer and Hot Leg Liquid Temperature CA JflON. TM( $C('GARIOS SiWUL ATED CONTAIN $1CNtFICANT CON $tRv4fl$h$ IN CPER ATOR ACTIOh$. (QUIPMENT F AILURES. OR BOTH.
600 i
i i
i i
i i
g Downcomer Temperature (RELAP 5)- 600 g
Ex tra polate d Downcomer o
Temperature (INEL )
j s --- System Average Temperature (BNL)- 500 ;
2 500
[
5 x'N
- 400 y jM E
P
,N 3
.?
e e"
300 3 3400
-a
-5
- 200 E
\\
2 N
s i
i i '-T--T--r--T-100 300 0
2000 4000 6000 8000 Time ( s )
Figure 2.2 - Transient 6:
Liquid Temperature in the Downcomer ________ _ _ __ _ _ __ _ _ _ _ _ _ _ _ _ _ _ _ _
CAufloot THE SCENARIOS SIWULATED CONTAIN SIGNIFICANT CON 5LRvAfl5W5 IN OPERATOR ACflohs. EQUIPWENT rAILUR(S, OR BOTH.
0.6 i
i i
i i
i i
(
Pressurizer Level ( RELAP S )
g
Pressurizer Level (BNL)
_J
, 0.4 -\\
l
'5
\\
3
\\
\\
2
.?
\\
- o.2 g
E 5
\\
z
\\
\\
'\\
I I
o.o O
40 so 12 0 160 200 l
Time ( s )
Figure 2.3 - Transient 6: Normalized Pressurizer Level C A Jflott THE SCt%ARIO5 SIWJLATED CONT AIN $1GNIFICANf CON $[RVAll$WS IN OPERATOR ACfloh5, (OUiPutNT r AILUR(S. OR BOTH.
I I
l l
l l
l Downcomer Pressure ( RELAP S )
En tropolated Downcomer Pressure (INELf 2500 2
~
15 o
lE Corresponding to System Average 2000 Temp. ( BN L) a
Saturation Pressure Corresponding to e
h10 H t Leg Temp. ( RELAPS) 1500 5 E
ct 1000 f5 -\\
_b
' N, M --
- w
- 500 $
7 0
0 0
2000 4000 6000 8000 Time ( s )
Figure 2.4 - Transient 6:
Downcomer Pressure.
3.
TRANSIENT 8: SMALL HOT LEG BREAK LOCA AT HOT ZER0 POWER CONDITIONS This transient, like Transient 6, was initiated by a small break (2.5 inch diameter) in the pressurizer hot leg. At the time of the break, the reactor was assumed to be at hot standby condition, as opposed to Transient 6 where the reactor was at hot full power operation.
The transient sce-nario is shown in Table 3.1.
The RELAPS code calculation was terminated at 1740 seconds because of large oscillations and the remainder of the tran-sient was extrapolated by INEL.
Figure 3.1 shows the RELAP5 downcomer and hot leg temperatures and the BNL system average temperature. The BNL temperature was calculated without considering the heat stored in the metal structures. This transient exhi-bits the same phenomena observed in Transient 6.
The hot leg temperature is very close to the BNL temperature. However, there is considerable stagna-tion in the loops and therefore the cold injection has a significant effect on the downcomer temperature.
4 The extrapolation of the downcomer temperature to 7200 seconds is shown in Figure 3.2.
The INEL extrapolation shows a sudden downcomer temperature drop to the HPI/LPI temperature at about 3000 seconds. This temperature would occur only if there was complete stagnation in the system. Even in this case the decrease should probably be more gradual and would take longer than 7200 seconds because of warm leakage from the upper plenum, downcomer bypass regions, and upper head mixing with the cold HPI in the downcomer.
Moreover, it is not clear how it was determined that this leakage caused a 6 K (10*F) difference between the HPI/LPI temperature and the final down-comer temperature.
Also shown in Figure 3.2 are the saturation temperature corresponding to the downcomer pressure calculated by INEL and the BNL system average temper-ature, which decreases gradually.
The saturation temperature is expected to be the highest system temperature. The BNL temperature is expected to fall between this saturation temperature and the downcomer temperature, as is the case except between 1200 and 4000 seconds. The BNL temperature was pre-dicted with metal stored energy considered, which implies thermal equili-brium between the fluid and the structures. However, it is possible that in the early part of the transient (<4000 seconds) the metal stored energy was not fully released to the fluid in the RELAPS calculation, resulting in lower temperatures than computed by the simple BNL approach.
The saturation pressure corresponding to the BNL system average tempera-ture is compared with the RELAPS downcomer pressure in Figure 3.3.
As ex-pected for a LOCA, these two pressures are very close.
In summary, it appears that the results calculated by RELAP5 are reason-able except for the extrapolation of the downcomer temperature which drops to a very low temperature very quickly.
Table 3.1 Scenario for Transient No. 8 Pl' ant Initial State - Just prior to transient initiator.
General
Description:
Hot 0% Power, 0% Power after 100 hr of shutdown.
System Status:
Turbine: Not latched, TSVs closed Secondary PORV: Automatic control Steam Dump Valves (SDVs): Automatic control Charging System: Automatic control Pressurizer: Automatic control Engineering Safety Features: Automatic control PORVs: Automatic control Reactor Control: Manual Main Feedwater:
In bypass mode, manual control to provide zero power level in SGs; 1 condensate pump, 1 MFWP operating Aux Feedwater: Automatic control MSIVs: Open, Automatic control MFIVs: Closed, Automatic control.
Transient Initiator - A 2.5-in. hole appears in the hot leg.
Equipment Failures which occur during the transient if the equipment is demanded.
None.
Operator Reactions to Reported Information 1.
If SIAS signal is generated, the operator will trip the reactor coolant pumps when RCS pressure reaches 1300 psig.
2.
The operator will throttle AFW flow to maintain 40% SG level.
i l
C AUll0N: THE SCENARIOS SiwutATED CONT AIN $1GNIFICANT CONSERVAll5MS IN CPER ATOR ACT10h5. EQUIPWENT FAILURES. OR BOIN.
600
%s
~
N %,Nk 50 0 N
s s%x m
5400
\\\\
5
~
\\
}300 V
}
2*
0..ac...,T..,iRELAPSi
- --- s.,
.. A..,..,. r.....,E L A P S )
Loop A Hot L.g T..p. ( R noof n..i ser......(suti
'.' 0
' = > ' ' ' ' '
iOO 10 0 1500 2000 O
S00 Tim (.i Figure 3.1 - Transient 8: Downcmer and Hot Leg Liquid Temperature C A J!l0N; THE SCENAeto$ SIWULATED CONT AIN $1GNIFICANT CONSERVAllSW5 IN CPERATOR ACflohs. (OUIPWENT FAILURES. OR BOTN.
600
- 600 DownComer Temperature (RELAP S) b4
~ ~ ~ ~ ~
550 Temperature (INEL )
--- System Average Temperature ( BNL)- 500 j
s x Soo
\\ ---
Corresponding Saturation Temperature
\\N to DownComer
[
Pressure ( RELAP S)
- 400 f450 q
g s
N, %
- 300 g a
E,a 400 E,
p _ _ _,n
- 200 350
\\
\\
t________________
ion o
2000 4000 6000 sooo Time (s )
Figure 3.2 - Transient 8: Liquid Temperature in the Downcomer 1
16 -
I
-)
i CAUTION: THE SCENARIOS SIWULATED l
CONTAIN SIGNIFICANT CONSERVATISWS IN OPERATOR ACTIONS, EQUIPMENT FAILURES, OR BOTH.
I 20 i
j Downcomer Pressure ( RELAP 5)
- 2500
Entro polat ed Pressure ( kN EL )
IS
--- Saturation Pressure Corresponding to System Average
]
g Temperature ( BNL )
- 2000 g 1
2 J
- to
- 1500 i
3 5
E
- 1000[
S s
- 500 7-h :_% _ m _____ ____
0
-~
0 0
2000 4000 6000 8000 1
Time ( s )
i 1
Figure 3.3 - Transient 8: Downcuner Pressure i
i i
l 1 !
i
4.
TRANSIENT 9: STEAM GENERATOR TUBE RUPTURE AT HOT ZERO POWER CONDITIONS This transient was initiated from hot standby conditions by a double-ended rupture of a steam generator tube at the cold leg end. There were no equipment failures but the operator failed to follow the correct procedures for recovery from a steam generator tube rupture and instead, responded as ex-pected for a small-break LOCA. This scenario is described in Table 4.1.
The transient was calculated to 7200 seconds using the RELAP5 code, so there was no need for extrapolation.
Figure 4.1 shows the downcomer fluid temperatures as calculated using RELAPS with two different safety injection locations. Also shown are the BNL calculated system average temperature and the saturation temperature corres-ponding to the RELAP5 calculated downcomer pressure. The RELAPS downcomer temperature for the base case (solid line in Fig. 4.1) shows large oscilla-tions in the later portion of the transient that were attributed to code limi-tations.
Since these oscillations were probably not physical, a second RELAP5 calculation was, performed to remove the oscillations during this latter part of the transient. This was achieved by injecting the cold HPI and make-up di-rectly into the vessel. The temperature calculated in this way was referred to as the adjusted temperature. Although this adjustment succeeded in damping out the oscillations, it neglected heat transfer from the cold leg wall to the cold injection water. For this reason the RELAPS adjusted downcomer tempera-ture should have been slightly higher than the temperature shown in the fig-ure.
It can be seen from Figure 4.1 that the saturation temperature corres-ponding to the RELAP5 calculated downcomer pressure is much higher than the BNL average system temperature. This could mean that voiding in the upper head is maintaining high temperature and pressure. Since the BNL temperature is a system average temperature it should fall between this saturation temper-ature, which should be the highest system temperature and the downcomer tem-perature, as it does.
Thus, the RELAP6 results appear reasonat,le and conser-vative from the PTS viewpoint.
i The pressurizer normalized levels calculated with RELAPS and by the sim-ple BNL method are shown in Figure 4.2.
In both calculations there was a rapid inventory loss through the break. As the inventory decreased, the down-comer pressure rapidly decreased, as shown in Figure 4.3.
Soon after the HPI flow started and the RC pump was tripped the pressure increased and then stabilized at just below the HPI shutoff head. At this time the pressurizer was empty and the break and injection volumetric flows were nearly equal. The break flow was maintained by the pressure differential between the downcomer and the affected steam generator (SGA). As can be seen in Figure 4.4, the SGA pressure remained elevated at the steam dump valve set points and thus the af-fected steam generator was the heat sink in the early portion of the tran-sient. By 200 seconds the pressurizer emptied and at 276 seconds SGB and SGC received the AFW. After the AFW to SGB and SGC was terminated at approxi-mately 600 seconds, these two steam generators became the primary heat sources and the secondary pressures dropped. During this time the mass inventory of SGA increased and since the liquid in this steam generator was subcooled, the nonequilibrium effects were expected.
In summary, the RELAPS-calculated results for this transient seem reason-able and the sequence of events appear to follow the expected trends. t
Table 4.1 Scenario for Transient No. 9 Plant Initial State - Just prior to transient initiator.
1 General
Description:
Hot 0% Power, 0% Power after 100 hr of shutdown.
System Status:
Turbine: Not latched, TSVs closed Secondary PORV: Automatic control l
Steam Dump Valves (SDVs): Automatic control Charging System: Automatic control Pressurizer:
Automatic control Engineering Safety Features: Automatic control PORVs: Automatic control Reactor Control: Manual Main Feedwater:
In bypass mode, manual control to provide zero power level in SGs; 1 condensate pump, 1 MFWP operating l
Aux Feedwater: Automatic control i
MSIVs: Open, Automatic control MFIVs: Closed, Automatic control.
1 Transient Initiator - A steam generator tube rupture on the cold leg side of l
tube sheet of SG "A".
1 Equipment Failures which occur during the transient if the equpment is demanded.
J
- None, l
Operator Reactions to Reported Information l
1.
If SIAS signal is generated, the operator will trip the reactor coolant pumps when RCS pressure reaches 1300 psig.
2.
The operator will restart reactor coolant pumps 10 minutes after all the following criteria are met:
A.
> 40*F subcooled B.
Pressurizer level > 20% or increasing
~
C.
R.C. pressure > 32T psig.
4 l
3.
The operator will throttle AFW flow to maintain 40% SG level.
l l
I 5
i 4 !
i
CAuf t(*. IHL $CENARIO$ $1WULAf(D CONT AIN $1GNiflCANT CON $tRVAfl5MS IN OPERAIOR ACilch5. (OulPMENT F AILURES. OR BOTN.
DOWNCOMER TEMP (REthP5) g
_x
--- ADJUSTED DOWNCOMER TEMP.
t C
(RELAP 5) g
--- SYSTEM AVERAGE TEMP.(BNL) a y
--- S ATURATION TEMP. CORRESPONDING oc TO RELAPS DOWNCOMER PRESS E
g 600 )
kJ
- 600g g
g, o
- 550 o 5 550 5
'N-.-
- 500$
m
~_
m 500 l0(/flfff
- 450f
~
O a
^
. 400 >
I I
I I
I I
J O
2000 4000 6000 8000 TIM E (s)
Figure 4.1 - Transient 9:
Liquid Tmperature in the Downcomer CAuf tom THE $CENan105 SiWULAf t0 COMf AIN $1GNif lCANT CON $tRVAllSW$ IN CPERATOR ACT10h$. (QUIPMENI F AILURES. OR BOfN.
O.3 i
i i
PRESSURIZER LEVEL (RELAP5)
_.s
PRESSURIZER LEVEL y
i (BNL) w 0.2
\\
O
\\
wd
\\
_a 4
]O.1 N
o
\\
Z
\\
\\
\\
i O
O 200 400 600 800 1000 i
TIME (s)
Figure 4.2 - Transient 9: Normalized Pressurizer Level l
CAUil0N: THE SCEstARIOS SIWULATED CONT AIN $1GNIFICANT CONSERbAllSW$ (N OPERATOR ACTIOhS. EQUIPMENT FAILUltES, OR BOTN.
16 i
i i
i i
- CALCULATED
-- ADJUST [0 2200 l
7
^
Ig '-
~-2000 %6 v
i E
I R
-1800 g
S 12 3
E 1
-1600 l
5 g
2
> 10 o
o 1400 >
1200 i
i i
0 1000 2000 3000 4000 5000 6000 7000 8000 Time (s)
Figure 4.3 - Transient 9: Calculated and Adjusted Reactor Vessel Downcomer Fluid Pressure C Aufl0N; THE SCENAA105 SIWuLAfE0 CONT AIN SIGNIFICANT CONSERVAll5WS IN i
CPERATOR AcilOh5, E0uthsENT FAILURi$. OR BOTN.
78 i
i i
i d*"
,2 : t c"
C
=-
O
- I 1020 7
i o
o I
~ 6.9 --
0
-1000
.i 2
S 6.8 j
-980 E
e E 6.7 l
3 O SG A i
a j
~
960 >
O SG B e
6.6 1
6.5 O
1000 2000 3000 4000 5000 6000 7003 8000 Time (s)
Figure 4.4 - Transient 9:
Steam Generator Secondary Pressure.
5.
TRANSIENT 11: LOSS OF SECONDARY HEAT SINK WITH PRIMARY SYSTEM FEED-AND-BLEED REC 0VERY AT H0T FULL POWER CONDITIONS This transient was initiated from hot full power conditions by a manual trip of the main feedwater pumps. This trip occurred when the water level in one of the three steam generators fell below 5% of full range resulting in a loss of secondary heat sink. As a result, the temperature differential be-tween the minimum and current temperatures at the inlet of SGA increased.
When this hot leg temperature increased by 5 F, the operator initiated the high pressure injection and opened both pressurizer power-operated relief valves to attempt a feed-and-bleed recovery.
In addition, the auxiliary feed-water pumps failed to start. The scenario is described in Table 5.1.
The RELAPS calculation was run to 8100 seconds and the key parameters were extra-polated up to 11000 seconds.
Figure 5.1 shows the downcomer and the Loop A hot leg temperatures calcu-lated with RELAPS and the system average temperature calculated by BNL. As expected, the BNL system average temperature falls between these two RELAP5 values. The sudden downcomer temperature drop in the RELAP5 calculation oc-curred after the high pressure injection was initiated and the PORVs were opened. The RELAP5 temperature dropped again soon after the RCPs were tripped in response to the low SG level at 3600 seconds. Assuming that the fluid flowing through the SDV was all vapor, it would have taken approximately 4700 seconds for the SG level to reach the low level condition. This indicates that water was entrained in the fluid during this period. Calculations showed that for the SGA level to reach 5% of full range at 3600 seconds, the average void fraction M the (Wid flowing out of the SDV would have to be approxi-mately 98% during this period. This agrees with the value predicted with RELAPS.
The INEL downcomer temperature extrapolation and the BNL system average temperature are shown in Figure 5.2.
As can be seen, both temperatures stead-ily decline at about the same rate. This temperature decrease is caused by the energy loss from the PORV and the cooling effects of the HPI, makeup flow and accumulator flow, which are greater than the heat addition due to the core decay power. The two temperatures shown in this figure are in agreement al-though the lower downcomer temperature in the RELAP5 calculation indicates that stagnation continues throughout the transient.
The RELAP5 calculated downcomer pressure and the saturation pressure cor-responding to the BNL system average temperature are shown in Figure 5.3.
Since the BNL system average temperature is slightly lower than the RELAPS hot leg temperature, its saturation pressure is also expected will be slightly lower than the downcomer pressure, as is the case. Both pressures follow the same trend and the INEL pressure prediction appears reasonable.
In summary, the primary system temperatures and pressures calculated by RELAPS generally follow the expected trends, and the results are reasonable.
l l
Table 5.1 Scenario for Transient No. 11 Plant Initial State - Just prior to transient initiator.
General
Description:
100% power steady state.
System Status:
)
Turbine: Not latched, TSVs closed Secondary PORV: Automatic control Steam Dump Valves (SDVs): Automatic control Charging System: Automatic control Engineering Safety Features: Automatic control Pressurizer PORVs:
Automatic control Reactor Control: Automatic control Main Feedwater: Automatic control Auxiliary Feedwater: Automatic control Main Steam Isolation Valves (MSIVs): Automatic control Main Feedwater Control Valves (MFIVs): Automatic control.
4.
Transient Initiator - Both Main Feedwater pumps trip simultaneously.
i Equipment Failures that occur during the transient if the equipment is demanded.
4 Auxiliary Feedwater pumps fail to start.
Operator Reactions to Reported Information 1.
Operator trips reactor coolant pumps (RCPs) when 1/3 Steam Generator (SG) wide range (WR) levels decrease below 5%.
1 2.
Operator initiates safety injection (HPI) and opens the pressurizer PORVs after RCP trip and when the A Loop hot leg temperature has increased 5'F.
4 1
i i l 4
C Auf l0N. THE SCENARIOS SIWULAILO CONT AIN $1GNIFlCANT CON $(RWAfl$W5 IN CP(RATOR ACit0h$. EQUIPWCNT FAILURES. OR BOTH.
600 550 x
u.
,s
[500 N
x s,'
s m
x o
!Tr 450 N'N s
e
,N w
$400 w
DOWNCOMER TEMP.(RELAP 5)
F
-- SYSTEM AVERAGE TEMP (BNL) 350 x
HOT LEG TEMP.(RELAP 5) 300 2000 4000 6000 8000 TIME (s)
Figure 5.1 - Transient 11:
Liquid Temperature in the Downcaner C AVil0N. THE SCENARIOS $lWVLAf(D CONT AIN $1GNIFICANT CON 5(RWAll$W5 IN CP(RATOR Atilohl. EQUIPW(NT F AILURES. OR BOTH.
600 1
i 00WNCOMER TEMP.
- 600 (REL AP 5)
--- EXTRAPOLATED 550 00WNCOMER TEMP.(INEL)
C
~ -
, N --- SYSTEM AVERAGE TEMR - 500 w q
E G 500 N,N (BN L) 3 ay
\\
e
=
w s
w N,
- 400g s,
N450
's W
s s N
s's
% 1 300.
400 4000 5000 6000 7000 8000 9000 10000 11000 TIME (s)
Figure 5.2 - Transient 11:
Liquid Tenperature in the Downcaner, 4000 - 11,000 sec. l l
l l
CAufiON: THE SCENARIOS SluulATED CONTAIN SIGNIFIC ANT CONSERVAll5MS IN CPER ATOR ACilch5. EQUIPWENT FAILURIS, OR 80iN.
20 DOWNCOMER PRESSURE (RELAP 5)
- 2500 15
EXTRAPOLATED j
00WNCOMER PRESSURE i
(INEL)
- 2000 3 s
j g
_.- SATURATION PRESSURE 3
I F3 CORRESPONDING TO SYSTEM < 1500 E m to AVERAGE TEMP. (BNL) a w
m m
w s,'
- 1000y ea n.
5
'N'N
- 500 O
O4000 5000 6000 7000 8000 9000 10000 11000 TIM E (s) 1 1
Figure 5.3 - Transient 11: Downcaner Pressure 1
I I
l I
r a
6.
SUMMARY
AND CONCLUSIONS Idaho National Engineering Laboratory used RELAPS/ MOD 1.6 to simulate a number of transient scenarios, specified by Oak Ridge National Laboratory, for the USNRC PTS study of the H. B. Robinson-2 PWR plant. Six of these tran-sients were initiated from hot full power conditions, and the remaining five were initiated from hot standby conditions. Three HFP and three HZP tran-sients were selected by BNL for in-depth review. However, only two HFP and two HZP transients have been presented here. The reactor system temperature and pressure were calculated at BNL using a simple method of mass and energy balances of the entire reactor system. The results of these hand calculations were compared to the RELAP5 results and the extrapolations provided by INEL.
As expected, the RELAPS code, because of its one-dimensional nature, had some difficulties in adequately representing transients with multidimensional effects. There are also uncertainties associated with the modeling of conden-sation effects in the pressurizer, and further work is necessary to assess the accuracy of code calculations of the primary pressure when the pressurizer is filling. However, in general, the RELAPS calculated primary temperatures and pressures appear to be reasonable, although the INEL extrapoiations might be somewhat conservative for certain transients.
i I
4 i
l l
i i
a 26 -
_ _ _. _... _ _ _.. - _. = _ _ _ _ _. _ _... _ _ _ _ _ _ _
i
- 7. REFERENCES i
FLETCHER, C.D. et al. (1983), "RELAPS Thermal-Hydraulic Analyses of Pres-surized Thermal Shock Sequences for the H. B. Robinson Unit 2 Pressurized Water Reactor," EGG-SAAM-6476, December 1983.
FLETCHER, C.D. et al. (1984), "RELAPS Thermal-Hydraulic Analyses of Pressur-ized Thermal Shock Sequences for the H. 8. Robinson Unit 2 Pressurized j
Water Reactor," NUREG/CR-3977, September 1984.
J0, J.H., and R0HATGI, U.S. (1985), " Review of TRAC Calculations for Calvert
]
Cliffs PTS Study," NUREG/CR-4253, BNL-NUREG-51887, April 1985.
ROHATGI, U.S. et al. (1984), " Assessment of Selected TRAC and RELAPS j
Calculations for Oconee-1 Pressurized Thermal Shock Study," NUREG/CR-3703, BNL-NUREG-51750, November 1984.
1 l
}
}
}
i i
i l
J t
1 f
5 1
[
i J
1 l
APPENDIX A l
l __
?
i :;
y' '!
BROOKHAVEN NATIONAL LABORATORY s
u mn7]
ASSOCIATED UNIVERSITIES, INC.
Upton. Long Island. New York 11973 (516)282s 2438 Department of Nuclear Energy F TS 666' March 28, 1984 Mr. J. N. Reyes Reactor Safety Research Branch Division of Analytical Evaluation Office of Nuclear Regulatory Research Mail Stop 1130 SS U. S. Nuclear Regulatory Commission Washi ngton, D.C.
20555
Subject:
Review of RELAP5 Input Decks for H. B. Robinson PTS Study
Dear Jose,
Enclosed please find a copy of the BNL memorandum by V. S. Rohatgi and C. Yuelys-Miksis on the review of RELAPS input decks that were used to per-form the thermal-hydraulic calculations for the H. B. Robinson PTS study. We did not discover any error in either of the two decks, and found the inputs acceptable. We are now reviewing the RELAP5 steady-state and transient calcu-lations.
In regard to the recommendation for TRAC calculations, we have always in-dicated that TRAC should be used for analyzing transients which cause asymmet-ric behavior in the reactor loops. Thus, recommended transients for TRAC cal-culations are:
1.
A break in one of the main steam lines 2.
A break in one of the hot legs.
Please feel free to contact either me or Dr. U. S. Rohatgi (FTS-666-2475) if you need any further clarification.
Sincerely yours, fdb Pa, Group Leader LWR Code Assessment and Application PS:ctf Encl.
cc: Attached Distribution Sheet -
BROOKHAVEN NATIONAL LABORATORY MEMORANDUM DATE:
March 22,1984 To:
P. Saha FROM:
U. S. Rohatgi and C. Yuelys-Miksis
SUBJECT:
BNL Review of RELAP5 Input Decks for H.B. Robinson From INEL This is a preliminary report covering only the review of input decks de-veloped by Idaho National Engineering Laboratory f or the H. B. Robinson plant, which consists of a Westinghouse PWR.
These decks were prepared f or the RELAP5/MODl.6/CY=16 code.
This review is based on the information obtained f rom the handouts receiv-ed in three meetings (8/23/83, 9/29/83,11/3/83) and the Robinson FSAR.
The input listings were received in the first meeting (8/23/83).
INEL has de-veloped two similar decks, one each f or the hot standby and f ull power con-ditions. We have checked the consistency (component nodalization and con-nections) and symmetry of loops in each deck.
We have also compared these two decks. The component dimensions have also been compared wherever possible with the information available f rom FSAR and the handouts.
In general, both the decks have been f ound to be consistent and the loops have been correctly represented.
We did not find any errors; there were only a f ew minor diff erences f rom the FSAR values.
Our comments on the input decks are summarized below:
1)
All the components and their junctions were consistent. and match with the nadalization provided.
2)
All the components except f or the accumulators used the non-homogeneous and non-equilibrium options...
3)
The total system water volume was very close to the FSAR value.
It was 9088 f t3 in the FSAR and 9105.7 f t3 in the INEL decks.
There were minor diff erences between individual com-ponent volumes and their corresponding FSAR values, but these are not significant f or PTS cor cern.
4)
All the loops are similar except f or the pipe lengths in the main loops and the injection system (accumulator line). These diff erences are probably plant specific.
5)
The hot standby and f ull power decks were also compared, and they have been f ound to be the same f rom a geometric viewpoint.
The dif-f erences in f eedwater lines such as only one condensate and only one main f eedwater.nump in the hot standby condition instead of two of each as in the f ull power case, are part of the transient scenario specified by Oak Ridge National Laboratory (ORNL).
af cc:
J. Jo _____ _
APPENDIX B l
)h BROOKHAVEN NATIONAL LABORATORY g
ASSOCIATED UNIVERSITIES, INC.
Upton, Long Island. New York 11973
($16)282' 2438 Department of Nuclear Energy FTS 666' Building 130 August 16, 1984 Mr. J. N. Reyes Reactor Safety Research Branch Division of Accident Evaluation Office of Nuclear Regulatory Research Mail Stop 1130 SS U. S. Nuclear Regulatory Commission Washington, DC 20555
Subject:
Review of RELAP5 Calculations for H. B. Robinson-2 PTS Study
Dear Jose,
Enclosed please find a section entitled, " Calculational Quality Assurance in Support of PTS" from the BNL Quarterly Progress Report for April-June 1984. This section presents the results of our preliminary assessment of the RELAP5 calculations for the H. B. Robinson-2 PTS study. All the eleven transients calculated by INEL have been assessed qualitatively. Quantitative assessment of six selected transients with the BNL simplified m3 del is in the final stage of completion, and will be reported soon.
We suggest that you and other PTS study participants review the enclosed material and inform us of any comments as soon as possible.
With best regards, Sincerely, PS:af Pradip Saha, goup Leader Encl.
LWR Code Assed5 ment and Application Calculational Quality Assurance in Support of PTS Su= mary Preliminary review of the RELAPS/ MODI.6 calculations and extrapolations of all the eleven transients for the H.
B.
Robinson-2 PTS study has been completed.
The calculations performed at INEL seem to be reasonable.
How-ever, there are uncertainties due to the pressurizer model, structure stored energy and multidimensionality of some of the transients.
Some of these transients are being reviewed in-depth using the simple method developed at i
BNL.
1 I,
a a 8.
Calculational Quality Assurance in Support of PTS
' /
f (P. Saha, U. S. Rohatgi and C. Yuelys-Miksis) f
.?
The objective of this project is to provide a peer review of the ther-
- f
^
mal-hydraulic calculations that have been performed at LANL (using the TRAC-PWR code) and INEL (using the RELAPS code) for the NRC Pressurized Thermal Shock (PTS) study.
Specifically, this includes a review of the plant decks and the calculations, and an assessment of the reasonableness of the re-sults.
The major activities performed during April to June 1984 are des-cribed below.
8.1 Preliminary Assessmer,t of RELAPS Thermal-Hydraulic Analysis of PTS y
Transients of H. B. Robinson Unit 2 (U. S. Rohatgi and C. Yuelys-Miksis)
Idaho National Engineering Laboratory has simulated eleven transients
[
for the H.
B. Robinson-2 PWR plant using RELAP5/MODl.6.
The scenarios for
]
these transients were specified by Oak Ridge National Laboratory.
Five of f
these transients are from the hot standby conditions and the remaining six are from the full power conditions.
Except for Transients 9 and 11, RELAPS calculations were made only for the early part (2000-4000 s); transient be-haviors out to 7200 s were obtained by extrapolation.
~I In the remaining sections, the BNL comments based on the assessment of these eleven transients, as documented in the INEL informal report (Fletcher, 1983), have been summarized.
,a
++
Transient 1: Main Steam Line Break from Hot Standby Conditions 2
~
The transient was initiated by a 0.093 m break in the steam line of Steam Generator A.
The code was successful in predicting the early sequence of events and the direction of heat transfer in the different steam genera-tors. The calculation was terminated at 1800 seconds and the remaining part of the transient was estimated by extrapolation.
The minimum downcomer fluid temperature of 386.2 K (235*F) occurred at 1026 seconds into the tran-sient.
The uncertainties in the results are due to the condensation model, structure stored energy and nultidimensional effects in the physical tran-sient.
Multidimensionality effects occur because each hot leg sees a slightly different fluid which may result in the hotter fluid going to the f
pressurizer, reducing the condensation there, and thereby slowing down the l?
e associated pressure drop.
This higher system pressure will reduce the HPI and other safety inj ection flows which will result in a higher downcomer
~-
fluid temperature. This effect cannot be modeled by RELAP5. There are some other phenomena which were predicted, but not explained.
These are: (1) the lag in pressure increase after the pressurizer level indicated a full pressurizer, (2) rapid drop in cold leg temperatures at 600 seconds, and (3) lack of oscillations in the colc leg fluid temperature in the presence of flow oscillations.
d p,.
En-
l Transient 2: Double Ended Steam Line Break at Hot Standby This transient was initiated by a 200% break in the steam line of Steam Generator A.
The transient is quite similar to Transient i except for a faster pressure drop and cooling.
The calculation was terminated at 1586 seconds and the rest of the transient up to 7200 seconds was predicted by extrapolation.
The minimum downcomer fluid temperature of 369.8 K (206*F) occurred at 7200 seconds.
The uncertainty in the prediction is also due to the same phenomena as described in the previous transient.
This calculation also showed. sharp changes in cold leg temperatures and flow rates between 500-600 seconds which were not explained in the INEL report (Fletcher, 1983).
Transient 3: Stuck Open Steam Line PORV at Hot Standby This transient is similar to the previous two steam line break tran-sients except for a smaller size break and additional operator failure which allowed continuous inj ection of auxiliary feedwater througnout the tran-sient.
Most of the events were the same as in the previous two transients but occurred much later due to the smaller break.
The code calculation was terminated at 3787 seconds and the remainder of the transient was predicted by extrapolation.
The lowest downcomer fluid temperature of 397 K (256*F) occurred at the end of the transient at 7200 seconds.
This calculation is also subject to uncertainties due to inadequacies in describing the condensation phenomenon in the pressurizer during pressurizer iill up.
Since the transient is slower and the code was run for a sufficiently long duration, the stored energy in the structure will oe properly accounted for.
The asymmetric behavior will still intluence the transient in terms of less condensation in the pressurizer as explained in Transient 1.
IIoweve r, the calculation and extrapolation seem reasonable.
Transient 4: Three Steam Dump Valves Stuck Open at Full Power This transient was also initiated by the secondary side blowdown as in the previous three transients, except that the plant was at full power conditions. The code calculation was terminated at 2500 seconds and the remainder of the transient was predicted by extrapolation.
The lowest downcomer fluid temperature of 373 K (212*F) occurred at 4837 seconds.
This transient is suited to a one-dimensional analysis since the secondary side of all three steam generators secondary sides was blown down, resulting in good natural circulation in all three loops.
The fluid conditions were also quite similar in all three loops.
The primary side would eventually reach the saturation temperature corresponding to the ambient conditions.
The code calculation and the subsequent extrapolation correctly predicted that behavior.
The calculatica incurred the usual problems with condensation as the pressurizer was filling up, but the final results were not affected.
Furthermore, it is not clear from the 1NEL report, why the nodalization was changed at 2157 seconds.
Transient 5:
Overfeed with Auxiliary Feedwater at Full Power This transient was initiated by the f ailure of the Auxiliary Feedwater
( AFW) to come on when the Main Feedwater (MFW) was tripped.
The operator manually started the auxiliary feedwater flow at 480 seconds. The code cal-culation was terminated at 3600 seconds and the remainder of the transient was extrapolated. The minimum downcomer temperature of 535 K (503*F) occur-red at 3027 seconds.
All steam generators were available and auxiliary feedwater flowed to all of them.
The transient was quite symmetric and amenable to the one-dimensional analysis of RELAPS.
There was no stagnation as the Reactor Coolant Pump ( RCP) stayed on and there were no large temperature variations in the primary side.
There were a few minor uncertainties wnich were not explained in the INEL report.
For example, why was there a variation in feedwater temperature for different steam generators?
Why did the Steam Generator A receive the coldest feedwater?
Why did the feedwater flow increase first in Steam Generator C? Despite these minor uncertainties, the results seem reasonable and the event sequence was as expected.
Transient 6:
Small Hot Leg Break at Full Power A small break (0.0635 m in diameter) at the bottom of the hot leg (Loop C with the pressurizer) initiated the transient.
The code calculation was terminated at 2800 seconds and the remainder of the transient was predicted by extrapolation.
The lowest downcomer fluid temperature or 310 K (100*F) occurred at 7200 seconds.
This transient exhibited strong multidimensional behavior in the begin-ning of the transient where a one-dimensional analysis would not be appro-priate.
However, since the minimum temperature occurred at the end of the transient when all the loops were stagnant, the RELAPS formulation was adequate.
The pressurizer remained empty, and so the system pressure was not affected by the condensation phenomenon in the pressurizer.
The system did not repressurize and the primary side temperature reached the ECCS temperature as correctly predicted by the code calculation and extrapolation.
There are a few phenomena which need more clarification.
Why is the Loop C steam generator secondary side pressure higher than the other two steam generators at the time of the initiation of motor driven auxiliary feedwater pump?
Furthermore, why does this steam generator (SG-C) get most of the auxiliary feedwater and fill up earlier?
Loop flows exhibit consid-erable oscillations before the loops are stagnant. What causes these oscil-lations?
In addition, why is the downcomer fluid temperature so much higher than the cold leg temperatures between 1000 and 2000 seconds? The differ-ence is on the order of 100 K.
l Transient 7: Stuck 0. pen PORV With Reactor at Full Power This transient was initiated by a primary side upset when the PORV valve was stuck open and the primary side started to blow down.
The operator action closed this valve at 600 seconds and ended the depressurization of the primary side. The remaining system worked as designed.
The calculation was terminated at 2200 seconds, when the primary side became liquid solid and there was no HPI or auxiliary feedwater flow into i
the system.
Beyond this point, the primary side pressure would be main-
~
l tained at 2535 psia (17.5 HPa) by the safety relief valve, and the primary side temperature would be controlled by the saturation temperature corresponding to the steam dump valve set pressure.
The INEL report correctly showed this behavior for the extrapolations. However, there would be minor oscillations in all the fluid conditions which cannot be predicted by extrapolation.
The minimum downcomer fluid temperature of 538 K (509'F) occurred at 947 seconds.
The calculation in the first 2200 seconds proceeded as expectea.
There were some differences in the flow behavior in three loops, wnica are not easily explainable.
Loop C and Loop B were quite close in flow rates, tem-peratures and timings of events, while Loop A showed delay in initiation anc termination of auxiliary feedwater (AFW).
Furthermore, between 900 and 1400 seconds, Loop A experienced the coldest cold leg temperature, highest flow, and highest steam generator secondary side pressure.
The natural circula-tion was maintained by the sinks which, in this case, were the steam genera-tors.
It is not clear how a larger flow rate was maintained in Loop A wnen the secondary side had probably the highest saturation temperature and the least capacity to absorb heat.
Transient 8:
Small Hot Leg Break at Hot Standby l
This transient is similar to Transient 6 discussed earlier.
This was a primary side upset with a small hot-leg break in the pressurizer loop and system blowdown from hot standby conditions, unlike Transient 6, where the reactor was initially at full power.
The remaining systems worked as de-signed and the operators followed the appropriate guidelines.
The calculation was terminated at 1740 s due to oscillations which would require smaller time steps and larger computer running time.
The final tem-perature would be close to the LPI/HPI fluid temperature in the extrapolated part of the transient as cooling was due to these injections only. The min-imum downcomer fluid temperature of 310 K (100*F) was estimated to occur at 7200 seconds (Fletcher, 1983).
There are a few questions about the extrapolating procedure. The sources of energy were the core power and heat transfer from the steam generator secondary side.
The extrapolation predicted a constant primary side pres-sure and temperature up to 2700 seconds, at which time the secondary side was cooled to the saturation temperature ( -464 K) and the primary side was cooled to -375 K.
Furthermore, the rate of heat transfer between the pri-mary and secondary sides was assumed to be a constant, but in reality it _ _ _ _ _ - _ _ _ _ __
will decrease.
Therefore, the primary side conditions will not remain constant until 2700 seconds but will vary.
This will delay the initiation of LPI and subsequent cooling.
The calculation up to 1740 seconds seems reasonable.
Although the code has the capability of modeling a break at different locations in the horizontal pipe, but to the best of our knowledge has not been assessed.
The break may receive some voids even when it is not covered as some vapor may be entrained into the break flow.
It is not clear if the code accounts for this entrained vapor.
In addition, the code predicted oscillations at the end of the calculation, and it is not clear what caused it and if they were physical?
It is inportant as these oscillations warmed up the downcomer fluid.
However, it only affected the lowest downcomer fluid temperature reached during the code calculation which is higher than the downcomer fluid temperature at the end of the transient.
Transient 9:
Steam Generator Tube Rupture at Hot Standby In this transient, a single tube in Steam Generator A ruptured near the cold leg side creating two breaks from which the primary side fluid flowed into the secondary side.
As the secondary side pressure was very close to the saturation pressure of the primary side, the flow was always single phase liquid and not choked.
The calculation was run for 7200 seconds and there was no need for extrapolation.
The minimum downcomer fluid tempera-ture of 465 K (378*F) occurred at the end of the transient at 7200 seconds.
The calculation exhibited considerable oscillations in the mass flow rates and temperatures after 3000 seconds. These are not physical as stated in the INEL report and are due to code limitations.
This prooles was circumvented by injecting the HPI flow in the vessel downcomer and the re-suits exhibited no oscillation.
The new flow parameters agreed well with the corresponding values in the initial calculation.
However, this adjust-ment in the model will preclude any warming up of the HPI fluid due to cold leg walls.
Loop A had the lowest flow rate, and it was mostly affected by the HPI flow.
The modeling change of directly injecting HPI into the vessel also had the most influence on Loop A fluid conditions.
In the real situation, however, there may be some delay due to the fluid transit time, and a finer nodalization between the HPI point and the vessel may account for it.
The original calculation predicted a lower downcomer fluid temperature, and additional sensitivity studies, such as finer nodalization, should have been made before the injection point was shif ted to the vessel.
Transient 10: Steam Generator Tube Rupture at Full Power This transient is similar to the previous transient (Transient No. 9) except that the reactor was at full power.
There was a single tube rupture near the cold leg side of the steam generator A.
However, the course of this transient was quite different from the similar transient at hot stand-by.
I J _ _ _ _ _ _ _ _ _ _ _ _
This calculation was terminated at 2400 seconds and the remaining pre-diction of this transient until 7200 seconds was by extrapolation.
The pumps were always running; so there was good mixing and less variation in the fluid conditions in different loops. Furthermore, the steam generators were not receiving any feedwater af ter 1200 seconds, and the secondary and primary side fluid conditions were governed by the steam generator dump valve setting (1 MPa).
The secondary side temperature at this pressure is 559 K and the primary side temperature will remain close to it.
The extrapolation seems reasonable.
This transient is quite mild as the temoerature variationa in different loops throughout the transient were less tha7 10K.
The only uncertainty is the increased heat transfer in the steam generator which is explained as flow reversal due to condensation in the secondary side.
However, it is not clear what caused it since there was no feedwater coming in.
This event only caused a temperature variation of 3 K.
The calculation, in general, looks reasonable.
Transient 11:
Loss of Secondary Heat Sink with Primary System Feed and Bleed Recovery at Full Power This transient was initiated by the failure of the secondary side f eedwater system to deliver any feedwater; the situation was made worse by by the operator action of feed and bleed, thereby cooling the pritary side.
This calculation was terminated at 8100 seconds and further prediction of the transient until 11000 seconds was by extrapolation.
The minimum downcomer fluid temperature occurred at the end of the transient at 11000 seconds and it was 422 K (300 F).
The extrapolation seems reasonable as pressure drop was due to con-traction of liquid and loss of fluid at the PORV.
The main energy loss was at the PORV and energy addition was due to core power and ECC inj ections.
As the system pressure was decreasing, the flow at the PORV decreased while injection flow increased.
At soc e pressure, there would be a balance between liquid contraction and the PORV flow against inj ection flow rate, which would determine the final pressure.
However, the system temperature continued to decrease.
The first 8100 seconds of the transient was computed by the code.
The sequence of events were as expected for this type of transient.
There are, however, a few phenomena which have not been clearly explained in the INEL report.
There are contradictory statements about the direction of heat transfer in the steam generator.
If the steam generator secondary sides are also the heat source, what maintains the natural circulation in Loops A and B?
Furthermore, if the heat transfer is negligible in the SG-A and SG-B, what is mainta'ining the natural circulation?
The cold leg flow in Loop C cannot be stagnant as stated in the report, but it should be at least equal to the HPI flow.
Furthermore, it is not cicar why the normalized level of the SG-B and SG-C secondary side had very little effect of feedwater header blowdown while there was a large effect on the SG-A secondary side level?
This transient will also have strong multidimensional ef fects due to PORV _ _ _.
flow affecting Loop C first.
However, the calculation predicted flow reversal of almost the same magnitude for all three loops due to PORV opening.
In reality, Loop C is expected to be affected the most.
Summary and Conclusion RELAPS is a one-dimensional code and it cannot model the asymmetric be-havior of some of the transients.
The code always predicts the same fluid conditions for all of.the hot legs.
In some of the transients the cold legs fluid conditions are quite different and the asymmetry may be carried through the core to the corresponding hot legs.
This usually occurs in the transients which have flow stagnation in some of the loops.
Most of the transients computed by INEL had asymmetric initiators but only the steam line break transients had flow stagnation and large differences in cold leg fluid conditions.
These transients should be assessed either with a multi-dimensional code or a limiting one-dimensional analysis with no mixing in the vessel for estimating the uncertainty due to multidimensionality of the transients.
Two other major items which are common to all the transients are the condensation effect in the pressurizer and steam generator secondary side, and the effect of stored energy in the structure.
INEL has made a reason-able compromise in modeling the pressurizer filling phenomenon.
The code generally overpredicts the condensation, but in the case of pressurizer filling, the calculation maintained a good condensation rate for the first volume and the remaining filling is by adiabatic compression.
There is a need to assess the code's ability to predict the pressurizer filling rate with separate effect tests.
It was observed f rom the review of tne Calvert Cliffs PTS transients (Jo, 1984) that the stored energy in tne structures influenced the course of transients between 2000 s to 4000 s.
This will be more significant for hot standby conditions where the core power is less and the structure stored energy is a much larger fraction of the total energy.
The calculation should be continued at least up to 4000 s before the results
[
are extrapolated.
Note that the rate of heat transfer from the structure varies as the transient proceeds.
1 1 I i
j
References
- FLETCHER, C.D., et al., (1983) "RELAPS Thermal-Hydraulic Analysis of Pres-surized Thermal Shock Sequences for the H. B. Robinson Unite 2 Pres-surized Water Reactor," EGG-SAAM-6476, December 1983.
JO, J., et al., (1984) " Review of TRAC Calculations for Calvert Cliffs PTS Study," BNL-NUREG report to be published, 1984.
1 _____
l l
APPENDIX C
BROOKHAVEN NATIONAL LABORATORY MEMORANDUM DATE:
January 12, 1984 TO:
P. Saha FROM:
U. S. Rohatgi and J. Jo
SUBJECT:
Extrapolation of Existing PTS Calculations With or Without Changes in Boundary Conditions NRC has requested LANL and INEL to compute primary side response to var-ious hypothetical accident scenerios using the advanced codes such as RELAPS and TRAC-PF1. However, there are many probable event sequences and only a few of them could be considered for detail calculations. These calculations will provide downcomer liquid temperatures and wall heat transfer coefficients which will be used in stress analysis code. However, the other possible tran-sients will not be calculated by advanced codes, but the downcomer fluid tem-perature, wall heat transfer coefficient and primary pressure will be estimat-ed using the results of other transients with similar features and simpli-fied balance equations. This memorandum describes some approaches to system-atically extrapolate the calculation from any time in the transient.
1.
Multi Volume Approach:
In this approach primary and secondary sides are modeled as separate vol-umes with heat transfer in the steam generator. The heat transfer calcula-tion in the steam generator takes into account the liquid level.
Primary Side Mass balance:
dM dt " "HPI
- N
-N (I) c BR Energy balance:
dg (M h +Mh)=Q*O+Omis
- N h
h HPI HPI + N h - BR BR pp mp d
p cc (2) n OSGI _________..
where the M, M, h are primary side total mass, system metal equiva-g m
p lent fluid bass and average enthalpy, and WHPI' W ' WBR' Od
- Q e
Qmis and QSGi are HPI, charging and break mass flow rates, decSy= heat, ptrip power, energy input through spray, and heat transfer to secondary side of the steam generator, respectively.
Secondar_y side Each steam generator will have the following set of balance equations:
Mass balance:
M
-W t 3j =Wfgj sti (3)
Energy balaace:
hsi)
- N (Msi h
h fwi fwj - Wsti gj + Q Gi (4)
S Here, M, h are secondary side total mass and average enthalpy, and s
s Wfw, h and W steamho,wrate,respectively.st, are feed water flow rate, feed water enthalpy and Msi = AL pg4 + A (Lt-l)P j
i gj M
h h
si si = ALjp$ gj t A(Lt - l )Pgj gi (6) i h
QSGi = P(L Hj gj + (Lt - l )Hgj)(Tp-Tsi)
(7) i Tp=Tp (h )
(8) p hj=hg (Tsi)
(9) g hj=hg (Tsi)
(10) g P
- P (Tsi) gj g
(11) gj g (Tsi)
A
=A (12)
Here A, L, L,
,T,Tsi' Nti and H j are flow area, liquid level, total j
t p
g height, perimeter, primary side temperature, secondary side temperature, liquid and vapor region heat transfer coefficients, respectively.
In general there are two steam generators and one may be blowing down. The three volumes in this situation are; primary side and two secondary side for two steam gen-erato rs.
So there are twenty-one variables; M, h, T for the pri, mary p
p o
side and M, h,, h, h, Q for each steam generator. There s
s' P t' Pg g
SG' are also twenty-one equations, i.e., (1), (2) and (8), and two sets consisting of Equations (3) to (7) and (9) to (12).
This system of equations can be further simplified,
) _ _ - _ _ _ _ _.
QSGi
- Hoi (Tp-Tsi)Li (13) where
,TSG,i,o g
Li,o(T -Tsi)o of p
Here H, QSG0' l and (T -Tsi) are heat transfer coefficient, steam generator g
o p
heat transer rate, liquid level and tenperature difference in stean generator at the time from where the extrapolation will begin.
It has been assumed that most of the heat transfer will be in the liquid phase and the net heat trans-fer will be proportional to the liquid level.
In this model, H represents o
an average heat transfer coefficient and assumed to be constant during the time period of the extrapolation. This approach implies that as the steam generator secondary side fills up the heat transfer will improve. The other important variable is the break flow rate and it will be estimated from the primary pressure which will be computed separately from the pressurizer analysis.
2.
Simple Approach Previous approach will require detailed analy:is of each steam generator and in many instances, that is not required either for extrapolation or check-ing the detail calculations.
This simple approach will apply to situations where the heat transfer between the primary and secondary side is small and variation in fluid temperature throughout the system is small. The system, consisting of primary side and all the secondary sides can be modeled as single volume.
In computing system energy the contribution due to secondary side steam energy can be neglected. The metal part of the system stores a significant amount of energy and it is accounted for by estimating equivalent liquid mass and adding it to the system fluid mass. The balance equations are:
h(M ) = Wgpg + Wc-WBR (14) p (Ms1) " Wfwi - Wsti (15)
((M +M +EMsi)h) = Q +0 +0 mis +W h
HPI ypg d p p m
(
)
+Wh
-W h
h BR BR + IW fwi fwj cc h
- EWsti sti where h is the average fluid enthalpy for the system.
The ot.her variables are the same as described in the previous approach.
For the system with two steam generators, there are 4 equations in four unknowns which are, M, Msl' p
Ms2 and h. _ _ _ _ _ _ _ -
3.
Boundary Conditions Both approaches described so far require high pressure injection (HPI),
charging and feed water conditions.
These are generally known as they are input to the system and in some instances are function of the pressure of the volume in which they are introduced. The flow through the breaks and valves are also functions of the conditions such as pressure and void fraction of the volume in which they are located. This will require modeling pressurizer and steam generator secondary side separately to estimate the pressures. The steam generator secondary side model has been described in this fomulation.
The secondry side pressure is assumed to be saturation pressure corresponding to its temperature.
In the cases where saturation pressure exceeds the TBV pressure setting the valve will open and steam will be released.
This steam flow will also depend upon the secondary side pressure.
4.
Pressurizer Model This conpanent controls the pressure in the primary side through sprays and heaters.
However, during the transient there is flow through the surge line which will affect the pressurizer pressure. The model described here will predict primary side pressure for cases where the primary side has no vapor except in the pressurizer. The surge line flow will depend upon the contraction or expansion of the liquid in the remaining primary side. The balance equations are as follows:
hML*Wsr + Wsp -
(17)
P My =-WBR + r dt (18)
I dg (M h )
- Oh+W b
h sr sr+Wsp sp - rh Ll (19) y
}
A- (M h ) = - W h
BR y + rhv y y dt (20)
WBR = f(P)
(21)
Vg=Mg+My = pressurizer volume (22)
A V
p
=p (P,h )
y y
y (23) hsr = hL or h, depending upon surge line flow direction.
p (24) _ _ _ _ _
l The unknowns of the model are M, h, M, h, WBR, r, Wsr, hsr, p and p L
t y
y which are liquid inventory and enthalpy, vapor inventory and enthalpy, break flow, vapor generation rate, surge line flow and enthalpy, pressure and vapor density. However, there are only eight equations for ten unknowns and two more equations are needed. The primary side liquid has much larger volume compared to the liquid volume in the pressurizer and small expansion or con-traction of primary loop liquid will have significant change in the liquid inventory of the pressurizer.
Wsr
- Y do /dt (25) p g
g (P, T )
p
=o p
g Here V, p, a nd T are primary side liquid volume, liquid density and p
temper 0ture. This still leaves this formulation short of one equation which will come from assumption on the processes of vapor generation.
In the first limiting case it can be assumed that there is no vapor generation and it is a frozen case, which implies that liquid could become superheated and vapor could become subcooled depending upon the direction of surge line flow.
In this case r=0 (26)
This completes the formulation.
However, it can be further simplified if the vapor expansion or contraction can be represented by some polytropic processes P/pk = constant (27)
This simplication will replace Equations (20) and (23) by Equation (27) and also vapor enthalpy b will not be needed. The variable k is 1.0 for iso-y themal process and is 1.33 for isentropic process.
The second limiting case is where the liquid and vapor are both saturated and any addition of mass or energy will change the pressure along the satura-tion conditions. The system of equations is as follows:
fL=Wsr + Wsp - F (28) dMV =-W8R + r (29) fE _ _ _
dh h
sr (h
-h ) + Wsp (h
-h ) - M O+N sr f sp 7 L dp r_
h (30) fg WBR
- WBR(P)
(31) o
=p f
f(p)
(32) p
- p g
g(p)
(33) hf = h (P)
(34) f hfg = hfg(P)
(35)
Vt"
= pressurizer volume (36)
V f
hsr = hf or h, depending upon the direction of flow.
(37) p Wr=V f
s p
(38)
So there are eleven equations and eleven unknowns which are, M
- M, r,
L v
WBR* pfpg, h, hfg, hsr> Wsr f
and P.
These two cases will provide limiting pressure histories for the primary side.
In case of flow into the pressurizer, the frozen case will predict
}
higher pressure than the equilibrium case, while for flow out of the pressuri-zer the frozen case will predict lower pressure as flashing of pressurizer liquid in equilibrium case will try to maintain the pressure. Equations (26) and (38) restrict this model to transients which have intact primary side with only vapor region in the pressurizer.
5.
Solution Procedure All the differential equations are linear and so simple Euler types of integration can be used. Most of the equations are essentially equation of state and can be replaced by steam table. As most of the changes during PTS transients occur over the long term, these equations can be finite differenc-ed.
This will make it possible to use hand calculations to estimate temperature and pressure history on the prinary side, af cc:
C. Yuelys-Miksis R. J. Cerbone _ _ _ _.
DISTRIBUTION LIST NRC Roy Woods, GIB Theof anis G. Theof anous Edward Throm, RSB Purdue University Newton Anderson, RIB 132 Pathway Lane Raymond Klecker, MTEB West Lafayette, IN 47906 Pryor Randall, MEBR Donald Kirkpatrick, EGRB Sol Levy Wayne Lanning, AE00 Sol Levy, Inc.
Demetrios Basdekas, ICBR 1999 So. Bascom Avenue Joseph Murphy, DRA Suite 2200 James Clif f ord, PSRB Campbell, CA 95008 Lambros Lois, CPR Novak Zuber, RSRB Daniel Speyer Malcolm Ernst, DRA 160 W. End Avenue Pradyot Niyogi, ASTP0 New York, New York 10023 Elpidio Igne, ACRS Hugh Thompson, DHFS R. L. Turner Louis Shotkin, RSRB Westinghouse Electric Corp.
Carl Johnson, RRBR (2)
Monroeville Nuclear Center Fuat Odar, RSRB Bay 319 Jose Reyes, RSRB (3)
P.O. Box 355 Paul Boehnert, ACRS Pittsburgh, PA 15230 Guy Vissing, ORB Sanford Israel, RRAB Jerry Phillips George Knighton, LB3 Carolina Power and Light James Buzy, LQB 414 Fayetteville St. Mall Edward Goodwin, PE Raleigh, NC 27602 Alf red Spano, RRAB Martin Virgilio, ICSB Donald Ogden Public Docur,ent Room EG&G Idaho, Inc.
James Van Vliet, ORB 4 550 Second St.
Steven Varga, ORB 1 Idaho Falls, ID 83401 Edmond Tourigny, ORB 3 Yih-Yin Hsu Other Organizations University of Maryland Department of Nuclear Eng.
Rudy Oliver College Park, MD 20742 Carolina Power and Light P.O. Box 1551 Douglas Selby 411 Fayetteville Street Oak Ridge National Lab.
Raleigh, NC 27602 Post Office Box Y Oak Ridge, TN 37830 Pradip Saha Brookhaven National Laboratory Upton, NY 11973 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _
'O""
U.S. NUCLE AR REGULATORY COMMISSION NUREG/CR-4452 BIBLIOGRAPHIC DATA SHEET pm-nupm ra 94r, 4 11TLE AN O SUUTIT LE LAad Volume No, of soproorratel
- 2. (Leave D!ms)
Review of RELAPS Calculations for H. B. Robinson Unit 2 Pressurized Thennal hock Study
- 3. R[ LENT'S ACCESSION NO.
I AU THOHIS)
$ ATE HEPORT COMPLE TED C. Yuelys-Miksis, U.S.
ohatgi and J. H. Jo
(
- N'"
l^"
J October 1985
!! Pt HF OHMING OllGANI/A TION N AME NO M AILING ADOHESS (/ncluae I,p Codel DATE HE POHT ISSUED Brookhaven National Labora cry I^"
Department of Nuclear Ener Upton, Long Island, New York 11973 8 (Leave danni I/. SPONSOHING OHG ANil AllON N AMk AND M A ING ADDHESS Itaciuor />p Co #
- 10. PROJECT /T ASK/WOHK UNIT NO Division of Accident Evaluation Office of Nuclear Regulatory Res rch
- 11. FIN NO.
U.S. Nuclear Regulatory Cunmissio Washington, DC 20555 A-3215 l
lJ l Y Yk OF Hk PO H T fPt RsOD COvt Rt O linclusive dates)
Technical Report (Formal) l l'.s buPPLEME N T AHY NO f t S 14 (Leave ntan A A IG AHSTH ACT (200 warns nr lessJ Idaho National Engineering Laboratory (If used the RELAP5f.1001.6 code to simulate a number of transient scenarios for the US C 'TS study of the H.B. Robinson Unit-2 PWR pl ant. Eleven of these scenarios were vie d at BNL on the basis of infonnation received before September 30, 1984.
S" of t se eleven scenarios were selected for an in-depth quantitative analysis perfo d on th basis of a simple method developed at BNL.
l The simple method uses the mass and ergy bala e equations to predict the temperature
'f tand pressure of the reactor system.he results f these calculations were compared o the RELAP5 results and the INEL trapolations In general, the RELAP5 and INEL result s l
appear to be reasonable.
l s
I / KE Y WOHUS ANO DOCUMEN T AN ALYSIS I la OE SC HIP T OHS y
Pressurized Thermal Shock RELAP5 H. B. Robinson Review INEL 1/Is IDE N TIFIE HS OPE N E N DE D TE RMS 5
fad AV AIL ABILITY STAT EME N T 19 SE CURI TY CLA5S (Tass reports 21 NO OF PAGES 10 dE CUHi TY CL ASN Ifans ows
// PH'Ch 4HC 5 0Hu 33% os sie
_ _ _ _ _ _