ML20080C253
| ML20080C253 | |
| Person / Time | |
|---|---|
| Site: | Oconee  |
| Issue date: | 07/31/1983 |
| From: | Bolander M, Fletcher C, Stitt B EG&G, INC. |
| To: | NRC |
| References | |
| CON-FIN-A-6047, REF-GTECI-A-49, REF-GTECI-RV, TASK-A-49, TASK-OR EGG-NSMD-6343, NUDOCS 8402080032 | |
| Download: ML20080C253 (252) | |
Text
{{#Wiki_filter:EGG-NSMD-6343 July 1983 RELAPS THERMAL-HYDRAULIC ANALYSIS OF PRESSURIZED THERMAL SHOCK SEQUENCES FOR THE OCONEE-1 PRESSURIZED WATER REACTOR C. D. Fletcher M. A. Bolander l B. D. Stitt
, M. E. Waterman Idaho National Engineering Laboratory Operated by the U.S. Department of Energy
^ /;$WW 4 jyS[ ' ' ~~
;- 4 2
N{ (
'\ .
i M N mummune )I/M 'yf , "
~ .
, M .$C M m W ~lii - ;---~ -'d C O-w.9x __
W,'t,2 W - _. 4p'
$ 1lI sM m wL~'?uh y y p%, ::~.*; %= ~
. hm% '~ C n'n % n'W ~% ~R Q m m --- :=
- .x
-- ;; $5,j"lf '_
't %. _;nm = - __][_ lMM I
~
s - "%rt. ' -- Q. This is an informal report intended for use as a preliminary or working document E1 ( 6l/ . Prepared for the
- U.S. NUCLEAR REGULATORY COMMISSION II l Under DOE Contract No. DE-AC07-761001570 pg E E E b idaho FIN No. A6047 8402080032 830731 PDR ADOCK 05000269 P PDR
i i ABSTRACT
~
Thermal-hydraulic analyses of pressurized thermal shock sequences for the Oconee-1 pressurized water reactor were performed at the Idaho National Engineering Laboratory (INEL) using the RELAPS computer code. This final 1 report summarizes the results of previously reported calculations and presents the results of recently completed calculations. Comparisons of the two counterpart calculations performed using the RELAPS code at INEL and the TRAC code at Los Alamos National Laboratory, are included as appendices. The results of these thermal-hydraulic analyses will be used as ( boundary conditions for fracture-mechar.ics calculations to be performed at Oak Ridge National Laboratory. i FIN No. A6047--Code Assessment and Applications (Transient Analysis) 11 t -' F NW'-'--^*'T "" - - 'E
-"*9t **r- 3-c-7? --
?? "^"='T' * * +94 s- -M-W7----=--7rWeM' # -
w7w'N7 Fr '-w'T- --WF*-"*r*
SUMMARY
To support the U.S. Nuclear Regulatory Commission's investigation of the pressurized thermal shock (PTS) unresolved safety issue, thermal-hydraulic analyses were performed at the Idaho National Engineering Laboratory (INEL) . using RELAPS computer code calculations. The ten event sequences to be investigated were defined at Oak Ridge National laboratory (ORNL), the integrator of the PTS' study. Four of the ten sequences analyzed were previously reported in Reference 1. These sequences were: (a) main steam line break, (b) steam generator overfeed, (c) hot leg small break (stuck-open power opersted relief valve), and (d) Oconee-3 turbi e trip plant transient. For each of these, a sequence description and summary of analysis results are provided in this report. The remaining six sequences were: (a) revised main steam line break, (b) maximum sustainable steam generator overfeed, (c) failure open of four turbine bypass valves at reactor hot standby, (d) pressurizer surge line small break, (e) reactor coolant pump suction small break, and (f) steam generator tube rupture. For esen of these, a detailed sequence description and a discussion of the soecific model changes required are provided in this report. Also provided are details and analyses of calculation results. For those calculations analyzed to be of significant concern with regard to pressurized thermal shock, limited estimates of uncertainty are given for the reactor vessel downcomer fluid temperature and pressure. For two sequences, the revis+d main steam line break and pressurizer surge line break, counterpart cal.ulations were also performed at Los Alamos National Laboratcry using the TRAC computer code. These calculations were compared with those performed at INEL using the RELAP5 code and the results of these comparisons appear as appendices r' this report. These comparisons proved invaluable in the estimating of uncertainties in the RELAP5 calculations. i kii
. _ _- __ _- - _ . ~
An extensive computer model of the Oconee-1 pressurized water reactor
'(PWR) was developed specifically to perform these calculations.' The model contains detailed representations of the pertinent PWR primary and secondary system components including the feedwater train. Control system models are also included for the turbine bypass, emergency feedwater, and
~
feedwater functions of the Babcock and Wilcox integrated control system. A table summarizing the lowest fluid temperatures and highest subsequent pressures in the reactor vessel downtomer for each sequence is presented in the conclusions section. The maximum estimated effect of uncertainty on both parameters is also shown for the most severe sequences. With respect to the PTS concern, the most severe of the sequences investigated were found to be, in order of severity:
- 1. Failure open of fo'Jr turbine bypass valves with the reactor at hot standby,
- 2. Two and one-half inch diameter reactor coolant pump suction break,
- 3. Main steam line break with reactor coolant pumps restarted at the time subcooling margin is obtained.
The 2-in. pressurizer surge line break was found to produce the coldest reactor vessel downcomer fluid temperatures, however, in this
~
sequence there is no mechanism to repressurize the primary system and thus precludes the potential for pressurized thermal shock. The results of the thermal-hydraulic analyses presented in this report will be used as boundary conditions in PTS fracture mechanics analyses to be performed at ORNL. , iv
ACKNOWLEDGEMENTS The authors wish to acknowledge the timely efforts of text processors Anna May White, Brenda Murrieta, Glada Gatenby and Joan Mosher in the compilation of this document and of Operation Technician Erma Jenkins in the development of the graphics presented here.
/ \
V
,- ~
i CONTENTS ABSTRACT ........... ................ .................:....... .. .. ii
SUMMARY
................................................... . ......... IIi
- 1. INTRODUCTION ....... ...... .................. ........ ...... ... I
- 2. MODEL DESCRIPTION .............................................. . 3 2.1 Steady State Model ........... .......... .................. 3 2.2 Primary System . . ................. ... ................. .. 3 2.3 S e c u n da ry Sy s t em . . . . . . . . . . . . . . . ...... ............... II 2.4 Feedwater System ....... . ...... .. . .... .......... ...... 18 2.5 Integrated Control System ........ ... ........... .... .. 22 2.6 Documentation Control of Codes and Models . ..... . . . . . . . 25
- 3. OVERVIF4 0F PREVIOUSLY REPORTED ANALYSES . .. . . . . . . . . . . . . . . . . . . . 28 3.1 Main Steam Line Break Transient . . . . . . . . . . . . . . . . . . . . . . . . . 28 3.2 Steam Generator Overfeed Transient ..... . . . . . . . . . . . . . . . . . 28
( 3.3 Hot Leg Small Break Transient ......... 33 3.4 Oconee-3 Plant Transient ............... . . . . . . . . . . . . . . . 38 , 4. MAIN STEAM LINE BREAK REVISED TRANSIENT . ...... . ........... . 39 4.1 Transient Scenario Description ................... .... . 39 4.2 Model Changes ................... . .... ... . . . . . . . . . . . . 39 l 4.3 Transient Results ...................... .. ............... .42 I 4.4 Conclusions .... . . . . . . ..... ... ...... .. .... .... .... 67
, 5. MAXIMUM SUSTAINABLE OVERFEED TRANSIENT .... . . . . . . . . . . . . . . . . . 70 5.1 Transient Scenario Description .... .. . . . . . . . . . . . . . . . 70 5.2 Model Changes .... ... . .... ... .. . .......... ...... 70 1
vi l I Ls _,y ,-y,, .
, ..r---- - - - - - --e---------~-------- = - ' - - ~ ~ *
- 5.3 ' Transient Results .. .................... ............ ..... 72 5.4 Conclusions ............ ........ ................ ......... 86
- 6. TURBINE BYPASS VALVE FAllVRE AT REACTOR HOT STANDBY ..... ... .... 90 6.1 Transient Scenario Description .......... .. ...... .. ..... 90 6.2 Model Changes ......... ........................ .... ...... 90 ,
6.3 Transient Results .................... ..................... 98 6.4 Conclusions ......... ...... ............. ............... 117
- 7. PRESSURIZER SURGE LINE SMALL BREAK TRANSIENT ........ ............ 118 7.1 Transient Scenario Description . . . . . . . . . . . . . . . . . . . . . . . .... 118 7.2 Model Changes .... .............. .. ....................... 118 7.3 Transient Results ....................... .... ....... . .. 120 7.4 Conclusions ............................. ... ............ 145
- 8. REACTOR C00LANl PUMP SUCTION SMALL BREAK TRANSIENT .. ........... 146 8.1 Transient Scenario Description ....... .. .... ...... . ....
~./
146 8.2 Model Changes ...... .. ............ ..... .... ... . 146 8.3 Transient Results .................... ...... . ...... 146 8.4 Conclusions ..... .............. ... ... .................. 167
- 9. STEAM GENERATOR TUBE RUPTURE TRANSIENT .... .......... . ....... 169 9.1 Transient Scenario Description . . . . .. ........ . ....... 169 9.2 Model Changes ............................... . ... .. . .. 169 9.3 Transient Results ............. ....... . . ..... ..... 171 .
9.4 Conclusions .... .. ................ ....... ........ ..... 184
- 10. OVERVIEW AND CONCLUSIONS ....... . .. .... .. .... ....... .
188
- 11. REFERENCES ......... ... ........ ........ . ......... . ... ..
191 APPENDIX A--COMPUTER RUN TIME STATISTICS . ................ ...... .... A-1 s vii
APPENDIX C--MAIN STEAM LINE BREAK REVISED TRANSIENT COMPARISON OF B ", COUNTERPART TRAC AND RELAP5 CALCULATIONS . ......... ...... .. ... APPENDIX C--PRESSURIZER SURGE LINE SMALL BREAK TRANSIENT COMPARISON OF COUNTERPART TRAC AND RELAPS CALCULATIONS . . . . . . . . . . . . . . . . . . . . . C-1 FIGURES
- 1. RELAP5 Oconee-1 model; Primary Loop A .............. . . . . . . . . . . .
4
- 2. RELAPS Oconee-1 model; Primary Loco B . . . . . . . . . . . . . . . . . .
5
- 3. RELAP5 Cconee-1 model; pressurizer system ................. . . . . . . 6
- 4. RELAPS Oconee-1 model; vessel, core flood tank and LPI system .... 10
- 5. RELAPS Oconee-1 model; Loop A steam generator and main steam line ............. . .. ...... .. . . . . . . . . . . . . . . . . . . . . . . . . . .
15
- 6. RELAPS Oconee-1 model; Loop B stea.2 generator and main steam line .. .. ... . ..... ......................... .. . . . . . . . . . . . . . 16
- 7. RELAPS Oconee-1 model; feed train from hotwell to startup and main feed valves .. ...... ...... ..... ... . . . . . . . . . . . . . . . . .
19 ( 8. RELAP5 Oconee-1 model; feed train from startup and main feed valves to steam generator, crossover, and emergency feedwater system .. ... .............. .. . . ............... . .... .... 20
- 9. B&W integrated control system organization ...... . . . . . . . . .
23
- 10. Main steam line break, R.V. downcomer pressure ... . . . . . . . . 29
- 11. Main steam line break, R.V. downcomer fluid temperature ... . . . 29
- 12. Main steam line break, R.V. downcomer inside surface heat transfer coef ficient . .. ......... . . . . . . . . . . . . . . . . . . . . . . . . 30
- 13. Steam generatar everfeed, R.V. downcomer pressure . . . . . . . 34
- 14. Steam generator overfeed, R.V. downcomer fluid temperature . . .. 34
- 15. Steam generator overfeed, R.V. downcomer inside surface heat transfer coefficient .. . . ... ........ . . . . . . . . . . . . . . . . . 35
- 16. Hot leg small break, R.V. downcomer pressure . . . . . . . . . . . . . 35
- 17. Hot leg small break, R.V. downcomer fluid temperature .. . . . . . . 36 viii h
- 18. Hot leg small break, R.V. downcomer inside surface heat transfer coefficient ........... ... ......... ... .:.... .. . 36
- 19. Revised MSLB, primary and secondary system pressures ........ 36
- 20. Revised MSLB, void fraction at the break junction ... . ... .. 43
- 21. Revised MSLB, break mass flow rate ... ..... . ................ .. 44 .
- 22. Revised MSLB, steam generator operating levels .. ........... ... 44
- 23. Revised MSLB, cold leg A-2 mass flow rate at reactor vessel ... .. 47
- 24. Revised MSLB, cold leg A-1 mass flow rate at reactor vessel ...... 47
- 25. Revised MSLB, cold leg B-2 mass flow rate at reactor vessel .. . . . . 48
- 26. Revised MSLB, cold leg B-1 mass flow rate at reactor vessel . ... 48
- 27. Revised MSLB, pressurizer collapsed liquid level ... ......... .. 50
- 28. Revised MSLB, reactor vessel upper head void fraction .. .... .. . 50
- 29. Revised MSLB, void fraction at top of affected loop hot leg ... . S '.
- 30. Revised MSLB, void fraction at top of unaffected loop hot leg .. 51
- 31. Revised MSLB, main feedwater flow rates at steam generator ,'
entrances ... .... ... ................ ........ . ... ..... .... 52
- 32. Revised MSLB, flow rates at emergency feedwater header . . .... . 54
~
- 33. Revised MSLB, fluid temperatures in emergency feed headers ... . 54
- 34. Revised MSLB, Loop A-1 high pressure injection flow rate .... . 56
- 35. Reviseo MSLB, qualities in affected steam generator secondary boiler .. .......... ............ .. . .. .. . .. .. ... ...... 59 .
- 36. Revised MSLB, phasic velocities through uppermost tube support plate of affected steam generatcr ... . ..... .. .... ...... . 58
- 37. Revised MSLB, affected steam generator secondary water mass .. 60 -
- 38. Revised MSLB, affected steam generator heat removal rate ...... 60
- 39. Revised MSLB, affected loop cola leg fluid temperatures at reactor vessel ................ ....... .. ....... .... . ..... 61
- 40. Revised MSLB, unaffected loop cold leg flu!d temperatures at reactor vessel ..... .................. ...... ... .. .. . . . 61 ix
A
- 41. Revised MSLB, fluid temperature in reactor vessel downcomer .. . 63
- 42. Revised M5LB, fluid pressure in reactor vessel downcomer . . . . .. . . . 66
- 43. Revised MSLB, extrapolated pressure in R.V. downcomer ..... .... 68
~
- 44. Revised MSLB, extrapolated fluid temperature in R.V.
downcomer .... .. ... .. ................... . . .. ....... 68
- 45. Revised MSLB, extrapolated neat transfer coefficient on inside surface of reactor vessel downcomer wall . . .. ..... ......... .. 69
- 46. Maximum sustainable overfeed, steam generator secondary pressures .... ................... .............................. 73
- 47. Maximum sustainable overfeed, hot leg pressures ..... ........... 73
- 48. Maximum sustainable overfeed, steam generator safety -elief valve flows ....................... .... .... . ......... ... . 74
- 49. haximum sustainable overfeed, turbine bypass flows . ... ... ... 74
- 50. Maximum sustainable overfeed, steam generator heat removal rates ... .. ..... .. . .. .... .. ... .. . .... . . 75
- 51. Maximum sustainable overfeed, Steam Generator A primary side
( s inlet and outlet fluid temperatures . . . . . . . . .. ... ... .... 79
- 52. Maximum sustainable overfeed, Steam Generator B primary side inlet and outlet fluid temperatures . . . . . . . ... . .... ... 79
- 53. Maximum sustainable overfeed, pressurizer collapsed liquid level . . ....... .. .. ....... ... .. .. ..... .. .. . . 80
- 54. Maximum sustainable overfeed, Loop A high pressure injection rates . ..... . ... .... ........... . . .. ........ .. 80
- 55. Maximum sustainable overfeed, Loop B high pressure injection .
rates .. ... . .... .... . ..... . .. ... .. .. .... 81
- 56. Maximum sustainable overfeed, liquid temperatures at top of steam generator boilers ..... ... . . .. . .. . ... ... .. 81
- 57. Maximum sustainable overfeed, hot leg mass flow rates . .. . . 82 l
i
- 58. Maximum sustainable overfeed, Steam Line.B void fractions .. .... 84 l
- 59. Maximum sustainable overfeed, Steam Line A void fractions . ... 84
- 60. Maximum sustainable overfeed, reactor vessel upper head void fraction ....... . . . ....... ... .... . . ..... 85 I
x i I
- 61. Maximum sustainable overfeed, mass flow rates into reactor vessel upper head and pressurizer surge line .... .. .... .. 87
- 62. Maximum sustainable overfeed, vapor generation rate in reactor vessel upper head ................................................ 87
- 63. Maximum sustainable overfeed, reactor vessel downcomer fluid pressure ................... .............. .. . . ..... ... .... 88
- 64. Maximum sustainable overfeed, reactor vessel downcomer fluid temperature .......... ............... .. .... ...... .......... . 88
- 65. Maximum sustainable overfeed, reactor vessel downcomer inside surface heat transfer coefficient .... . . ... . ....... . .. .. 89
- 66. Hot standby model, main feedwater train .... ................... . 91
- 67. Hot standby model, Steam Generator A feedwater header .. .... . 92
- 68. Hot standby model, Steam Generator B feedwater header ............ 92
- 69. TBV failure at HSB, Steam Generator A secondary pressure . . . . . .. 99
- 70. TBV failure at HSB, Steam Generator B secondary pressure .... .. 99
- 71. TBV failure at HSB, Steam Generator A secondary pressure and feedwater flow rates .......... . ..... ....... .......... .. . . 100 i
- 72. TBV failure at HSB, Steam Generator B secondary pressure and
'J feedwater flow rates .................. .. .. ....... ......... . 100
- 73. TBV failure at HSB, feed train pressures and normalized startup valve areas' ......... ....... . . ..... ......... . .... ... .. 103
- 74. TBV failure at HSB, Steam Generator A secondary pressure 0-1500 s ... ........... ........................ ................ 103
- 75. TBV failure at HSB, Steam Generator B secondary pressure ,
0-1500 s .. .. ... ................ . ...... ... ... .... ... 104
- 76. TBV failure at HSB, Loop A fluid temperatures ... ..... . ...... 106
- 77. TBV failtre at HSB, Loop B fluid temperatures .. .. . . .. .. 106
- 78. TBV failure at HSB, Loop A fluid temperatures, 0-2500 s ... ..... 107
- 79. TBV failure at HSB, Loop B fluid temperatures, 0-2500 s . .....
107
- 80. TBV failure at HSB, Loop A hot leg mass flow rate ..... ... ...... 110 Xi
- 81. TBV failure at HSB, Loop B hot leg mass flow rate ................ 110
- 82. TBV failure at HSB, Loop A and B ccid leg discharge fluid temperatures ... . . .. ......................................... 111
- 83. TBV failure at HSB, reactor vessel downcomer fluid pressure . . . . 113
- 84. TBV failure at HSB, reactor vessel downcomer inside surface heat transfer coefficient ........................... . . . . . . . . 113
- 85. TBV failure at HSB, reactor vessel downcomer fluid temperature ..... ..... ...... ..... ... . . . . . . . . . . . . . . . . 115
- 86. Pressurizer surge line break, break mass flow rate ..... ... ..... 124
- 87. Pressurizer surge line break, hot leg pressure . . . . . . . . . . . . . . . . . 124
- 88. Pressurizer surge line break, steam generator secondary pressures ......... . ... ... ... ............... .. ......... .. 125
- 89. Pressurizer surge line break, steam generator startup levels ..... 125
- 90. Pressurizer surge line break, R.V. downcomer fluid temperature . ........... ....... ........ . . . . . . . . . . . . . 126
< 91. Pressurizer surge line break, hot leg fluid temperatures . . . . . .. 126
('
- 92. Pressurizer surge line break, Loop A cold leg fluid temperatures ..................... ..... .... ................... 127
- 93. Pressurizer surge line break, Loop B cold leg fluid temperatures ... ........ .. ... ....... .. ....... ....... . . . 127
- 94. Pressurizer surge line break, void fraction in R.V. upper head ... 129
- 95. Pressurizer surge line break, total feedwater flow rates . . . . . . . . 129
- 96. Pressurizer surge line break, void fraction in top of R.V. .
downcomer annulus .. . ....... ..... . .. . ... ...... . . . . . . 131
- 97. Pressurizer surge line break, Loop A cold leg flow rates ... .
131
- 98. Pressurizer surge line break, Loop B cold leg flow rates . . .
133
- 99. Pressurizer surge line break, core flooo tank injection rate . .
137 100. Pressurizer surge line break, low pressure injection rate .... . . 139 101. Pressurizer surge line break, R.V. downcomer fluid temperature with and without cold leg circulation ........... .. . . . . . . . . . . . 141 xii
--'m -
102. Pressurizer surge line break, vent valve mass flow rate .......... 141 103. Pressurizer surge line break, total HPI mass flow rate ... . .. ... 142 104. Pressurizer surge line break, extrapolated R.V.. downcomer pressure .................. ... .....e. ................... . .. 142 105. Pressurizer surge line break, extrapolated R.V. downcomer . fluid temperature .... ...... ......... ... ... . .......... .... 143 106. Pressurizer surge line break, extrapolated R.V. downcomer inside surface heat transfer coefficient ...... .. ............. 143 107. Pump suction break, hot leg pressures ............. .. .......... 151 108. Pump suction break,-steam generator secondary pressures .......... 1 51 109. Pump suction break, normalized main feedwater valve area ........ 152 110. Pump suction break, normalized startup valve area .... ........... 152 111. Pump suction break, .eactor vessel upper head void fraction ...... 153 112. Pump suction break, emergency feedwater flow rates . ........... 155 113. Pump suction break, turbine bypass flow rates ... . ......... .. . 155 114. Pump suction break, Loop A hot and cold leg mass flow rates . .. . 157 , 115. Pump suction break, Loop B hot and cold leg mass flow rates .... . 157 116. Pump suction break, Loop A steam generator primary inlet and outlet fluid temperatures . . . . . . . . . . . ......... ..... ..... . 158 117. Pump suction break, Loop 8 steam generator primary inlet and outlet fluid temperatures ..... .......... .... .............. 158 ! 118. Pump suction break, Loop A steam generator primary inlet and . j outlet fluid densities . .............. . .. .. ........ .... . 159 l 119. Pump suction break, Loop B steam generator primary inlet and i outlet fluid densities .. . ..... ........ .. ... . ......... . 159 i . 120. Pump suction break, break and HPI volumetric flow rates . . .. . 160 121. Pump suction break, pressurizer collapsed liquid level ..... .... 160 122. Pump suction break, break and HPI integrated mass flow rates ...................... ...................... . . 161 i
,_/
xiii t
.~
123. Pump suction break, break and HPI mass flow rates . . . . . . . . . . . . . 163 124. Pump suction break, cold leg B-2 mass flow rates in the region of the HPI injection site ......... .. ........................... 163 125. Pump suction break, fluid densities in loop B cold leg pump
, suction and steam generator outlet regions . ................... . 164 126. Pump suction break, reactor vessel downcomer fluid temperature ......... .................. .. .. ........ .......... 166 127. Pump suction break, reactor vessel downcomer fluid pressure ...... 166 128. Pump suction break, reactor vessel downcomer inside surface heat transfer coefficient ................... ... ............... 168 129. Steam generator tube rupture, hot leg pressures .................. 172 130. Steam generator tube rupture, tube side break mass flow rate ... . 172 131. Steam generator tube rupture, steam generator secondary pressures ............................................ ........... 173 132. Steam generator tube rupture, turbine bypass mass flow rates ..... 177 ,, 133. Steam generator tube rupture, steam generator relief valve mass
( flow rates ............... ...... ...... ...... ...... . .. ...
)77 134. Steam generator tube rupture, tubesheet side break mass flow rate .............. .. . ........ ............... .. . ..
178 135. Steam generator tube rupture, pressurizer collapsed liquid 1evs3 .. ........ ... ...... .................. ............. .... 178 136. Steam generator tube rupture, steam generator operating levels ............................. ....................... ... I79 137. Steam generator tube rupture, steam generator startup levels . ... 179, 138. Steam generator tube rupture, steam generator heat removal rates ... . ........ . .. ....... . ... ... ... . .. ... ... ... 180 ~ 139. Steam generator tube rupture, EFW header mass flow rates (sum of EFW and MFW delivered at the EFW header) ...... . . . . 182 140. Steam generator tube rupture, total break and total HPI mass flow rates .. .. ....... . .. . ... . . ........ ... . ..... 182 141. Steam generator tube rupture. Loop A hot and cold leg mass flow rates .... .. . ...... ...... . .. ... .. ... 183 xiv
142. Steam generator tube rupture, Loop B hot and cold leg mass flow rates ............................ .. .. . .. . . . . . - 183 143. Steam generator tube rupture, Loop A hot and cold leg fluid temperatures .. ....... . .......... ..... 185 144. Steam generator tube rupture, Loop B hot and cold leg fluid temperatures . ......./................. . . . . . . . . . . . . . . . . . 185 . 145. Steam generator tube rupture, extrapolated R. V. downcomer fluid pressure . .... . ......... *
. ... .. . . . . . . . . . . . . . . . . . 186 .
146. Steam generator tube rupture, extrapolated R. V. downcomer fluid temperature ...................... . . . . . . . . . . . . . . . . . 186 147. Steam generator tube rupture, extrapolated R. V. downcomer inside surface heat transfer coefficient . .. . . . . . . . . . . . . 187 A-1. Differentiated computer run time, revised MSLB ............. ..... A-3 A-2. Differentiated computer run time, maximum sustainable overfeed . ........ ...... ................ ... . . . . . . . . . . . . . A-3 A-3. Differentiated computer run time, TBP failure at hot standby ... A-4 A-4. Differentiated computer run time, pressurizer surge line break .... . ......... .. .............. .. ................ . . . A-4 , A-5. Differentiated computer run time, pump suction creak ..... .. .... A-5 J A-6. Differentiated computer run time, S..G. tube rupture . . . . . . . . . A-5 B-1. TRAC and RELAPS comparison, revised MSLB, hot leg pressures . . . . . . B-3 B-2. TRAC and RELAPS comparison, revised MSLB, reactor vessel downtomer fluid temperature . ......... . . .... . . . . . . . . . . . B-3 B-3. TRAC and RELAPS comparison, revised MSLB, affected S.G. heat - removal rates . ... ... .... .......... . . .. . . . . . . . . . . B-6 B-4. TRAC and RELAPS cc.nparison, revised MSLB, break mass flow rates .. ..
.... . .... ..... .. .. . . . . . . . . . . . . . B-6 B-5. TRAC and RELAPS comparison, revised MSLB, Loop A hot leg flow rates ...... .... .. ...... . .. . . . . . . . B-7 B-6. TRAC and RELAPS comparison, revised MSLB, Loop B hot .
leg flow rates ..... .... .... ........ ... . . . . . . . . . . . . . B-7 B-7. TRAC and RELAP5 comparison, revised MSLB, void fractions at top of Loop B hot legs . ..... . . .. . . . . . . . . . . . B-8 XV
--a- - - _ m__ _ _ __ _ _________m
B-8. TRAC and RELAP5 comparison, revised MSLB, Loop B hot leg liquid temperatures .. ........ ...... .. .... ....:... .. .... B-ll B-9. TRAC and RELAP5 comparison, revised MSLB, Loop A hot leg liquid temperatures ............................. ............... B-ll B-10. TRAC and RELAPS comparison, revised MSLB, affected S.G. secondary pressures . . .. ..... ....... ...... ......... ... . G-14 B-11. TRAC and RELAP5 comparison, revised MSLB, affected S.G. main feed header mass flow rates ... .. .... . ............ . B-14 B-12. TRAC and RELAPS comparison, revised MSLB, reactor vessel upper head void fractions .. .......... ........... .... ........ B-15 B-13. TRAC and RELAPS comparison, revised MSLB, affected S.G. emergency feedwater header flow rates .................... ... . B-18 B-14. TRAC and RELAP5 comparison, revised MSLB, void fracticns at top of affected S.G. boiler . ......................... ......... B-18 B-15. TRAC and RELAP5 comparison, revised MSLB, volumetric flow rate of liquid at uoper tube support plate of affected steam generators ........ .... ......... ........... . ...... .. ..... B-19 B-16. TRAC and RELAP5 comparison, revised MSLB, affected S.G. heat [s removal rate, 0-1000 s ........... ........ ..... .... ...... ... B-21 2-17. TRAC and RELAP5 comparison, revised MSLB, liquid temperature at top of affected 5.G. boiler .............. ........... . .... . B-21 C-1. TRAC and RELAPS comparison, pressurizer surge line break, hot leg pressures .. . . ... . ........ ... ....... ........ .... C-3 C-2. TRAC and RELAPS comparison, pressurt:er surge line break, reactor vessel downcomer fluid temperature . . . . . . . . . .. .. .. C-3 C-3. TRAC and RELAP5 comparison, pressurizer surge line break, Steam Generator A secondary pressures .. ....... . .. .. ...... C-6 C-4. TRAC and RELAP5 comparison, pressurizer surge line break, reactor vessel upper head void fractions .. .. .. . .. ....... C-6 C-5. TRAC and RELAPS comparison, pressurizer surge line break, void fractions at top of Loop A hot leg .... . . .. . . .. C-7 C-6. TRAC and RELAPS comparison, pressurizer surge line break, void fractions in cells upstream of break . . .. ... ... . ... . C-7 C-7. TRAC and RELAP5 comparison, pressurizer surge line break, break mass flow rates ... ... ....... .... ..... . .. ... .. C-8 xvi.
4 C-8. TRAC and RELAPS comparison, pressurizer surge line break, cold leg A-2 mass flow rates at reactor vessel .. ................ C-8 C-9. TRAC and RELAPS comparison, pressurizer surge line break, cold leg A-2 liquid temperature at reactor vessel ................ C-10
', C-10. TRAC and RELAPS comparison, pressurizer surge line break, Loop A hot leg liquid temperatures ..... .... ...... . ....... C-10 -
C-11. TRAC and RELAP5 comparison, pressurizer surge line break, total vent valve mass flow rates ... ..... .. ..... . .... .. C-ll . TABLES
- 1. Full power steady-state conditions calculated by RELAPS com with plant nominal conditions .............. . ......... .. pared . ..... 7
- 2. Correspondence between the physical and mathematical components in the prif1ary loop . ............ . ........ .... ..... ...... 8
- 3. Correspondence between the physical and mathemati;al components in the vessel ...... . ................ .. .. . ... . ........, 12
- 4. Junctions using the abrupt area change model . .. ... ........... 13
- 5. Correspondence between the physical and mathemati:al components in the A and B steam generator secondaries . . .. .... .......... 17 ,/
- 6. Correspondence between the physical _and mathem4ti:al components in the feed train ........ ........... . .... .. ...... .. ... . 21 ,
- 7. Documentation control of code versions and models ............... 26
- 8. Main steam line break transient scenario ..... ........ ......... 31
' 9. Sceam generator overfeed transient scenario ...... .. .......... 32
- 10. Hot leg tmall break transient scenario . .... . .... .......... 37
- 11. Revised main steam linc. break transient scenario .... .......... 40
- 12. Revised main steam line break sequence of events ... ... .... 45 -
- 13. Maximum sustainable overfeed transient scenario . ............ . 71
- 14. Maximum sustainable overfeed transient sequence of events .. ..... 76
- 15. Turbine bypass valve failure at reactor hot standby trannent scenario................................................... 94
- 16. Hot standby steady-state initial conditions . ... .. . . .
97 xvii
~
- 17. Turbine bypass valve failure at reactor hot standby sequence of events ................................ . .......... ........ . 101
- 18. Pressurizer surge lins creak transient scenario .. . ... ... 119
- 19. -Pressurizer surge line break transient sequence of events ... .... 121
- 20. Pump suction break transient scenario .................. ........ 147
- 21. Pump suction break transient sequence of events ....,............. 148
- 22. Steam generator tube rupture transient scenario . ................ 170
- 23. Steam generator tube rupture transient sequence of events ........ 174
- 24. Summary tabulation of Oconee-1 PTS RELAP5 calculation results .. . 189 A-1. Timing Statistics ............................................... A-6 B-1. Comparison of TRAC and RELAPS Sequences of Events, Main Steam Line Break Revised Transient ....................... ............ B-4 B-2. TRAC ard RELAPS Sequence Phase Timing ........................... B-9 C-1. Comparison of TRAC and RELAPS Sequences of Events, Pressurizer Surge Line Break Transient ...................................... C-4 xviii
RELAP5 THERMAL-HYDRAULIC ANALYSES OF PRESSURIZED THERMAL SHOCK SEQUENCES' FOR THE OCONEE-1 PRESSURIZED WATER REACTOR
, 1. INTRODUCTION The U.S. Nuclear Regulatory Commission (NRC) is investigating the pressurized thermal shock (PTS) unresoived safety issue (Number A49).
PTS refers to plant transients in which the welded reactor vessel walls of a pressurized water reactor (PWR) are subjected to a rapid cooldown at interior surfaces and coincidently, or subsequently, to high pressures as well. The concern centers on reactors that have been operating for long periods of time and have reactor vessels that were welded with high-copper-content weld rod. Transients of interest must include the potential for the aforementioned cold reactor vessel downcomer fluid temperatures with high primary system pressures. The NRC has identified a group of PWRs for which PTS is of near-term concern and the Oconee-1 PWR was selected as the first plant in the group to be investigated in detail. The Oconee-1 FWR is a Babcock a'nd Wilcox, lev-loop design reactor and has been operated bv Duke Power Comoany since 1973. The NRC has contracted with Oak Ridge National laboratory (ORNL) to integrate the investigation into the PTS unresolved safety issue. The NRC has also contracted with the Idaho National Engineering Laboratory (INEL)
~
and Los Alamos National Laboratory (LANL) to support ORNL by performing thermal-hydraulic analyses using state-of-the-art computer codes and plant-specific models of the PWRs being investigated. The NRC has contracted with Brookhaven National Laboratory (BNL) to provide a quality-assurance audit function on the work performed at INEL and LANL. , Functions retained at ORNL include: probabilistic risk assessment which is used to define potential transients of interest, definition of the specific [ sequences to be analyzed by INEL and LANL, fracture mechanics analyses for the vessel wall, and integration of the overall study, l t l
The purpose of this report is to document the Oconee-1 thermal-hydraulic analyst s >per'ormed at ~INEL to support' the P' TS study. A description of the RELAP5 model of the Oconee-1 PWR is p. resented in Section 2. Analyses of ten transient sequences have been performed. Detailed aralysis results for the first four of these sequences appeared in Reference 1 and a summary of the results for these is presented in Section 3. The first four sequences were: (a) a main steam line break, ' (b) a steam generator overfeed, (c) a hot leg small break, and (d) a turbine trip plant transient that occurred in the Oconee-3 PWR for which calculated and measured data are compared. Sections 4 through 9 contain the detailed analysis results for the remaining six transients: a revised main steam'line break sequence (Section 4), a maximum sustainable steam generator overfeed (Section 5), turbine bypass valve failure at reactor hot standby (Section 6), a pressurizer surge line small break (Section 7), a reactor coolant pump suction small break (Section 8), and a steam generator tube rupture (Section 9). An overview and conclusions for all ten sequences analyzed are presented in Section 10 and references are listed in Section 11. Appendix A presents computer run time statistics for the ' calculations performed. A comparison between the two counterpart calculations performed at INEL using the RELAPS cede and LANL using the TRAC-PF1 code are presented in Appendix B for the revised main steam line break sequence and Appendix C for the pressurizer surge line small break sequence. I e 2
- 2. MCDEL DESCRIPTION This section describes the RELAP5 Oconee-1 FWR model used for the steady state initialization calculation and subsequently for each of the pressurized thermal shock (PTS) transient calculations. The transient models were modifications to the steady state model, and transient calculations were restart calculations from the steady state conditions.
Specific changes from the model described here that were required to perform the transient calculations are described under the "Model Changes" headings in Sections 4 through 9. 2.1 Full Power Steady State Model The RELAP5 Oconee-1 PWR steady state model is a detailed model of the Oconee-1 PWR power plant describing all the major flow paths for both primary and secondary systems, including the main feed train. Also modeled are power operated relief valves (PORV), safety valves, and the emergency ( core cooling system (ECCS). Secondary side features include turbine bypass and turbine stop valves, safety valves, and the emergency feedwater (EFW) system. Another feature modeled was the integrated control system (ICS). The model contained 220 volumes, 232 junctions, and 208 heat structures and is shown schematically in Figures 1 to 8. A description of the primary system, secondary system, feed train and the control system are presented in che following sections. Full power steady state conditions calculated by the model are compared with nominal plant conditions in Table 1. 2.2 Primary System The Oconee-1 PWR plant is a 2 by 4 configuration, i.e., two loops each containing one hot leg and two cold legs. The loops were designated as Loop A and Loop B. Each loop contained one hot leg, one steam generator, two pump suction legs, two reactor coolant pumps, and two cold legs (refer to Figures 1 and 2). Attached to the hot leg in Loop A was the pressurizer surge line and pressurizer (see Figure 3). Each cold leg contained a high pressure injection (MPI) port. Table 2 summarizes the relationship between the physical components of the Oconee-1 A and B primary loops and the 3
t 3 j
-o 9 -3 tt e l m
T: 1
.3 m rh r,
,' e e i
i
------ To avaa s s,ana =
,4 I -
2 I tes i T , i e -E g M MI ,
- I
. . . . . . . .! . . . . ' :ee ,, ,
j j i M itei a ave! ,--;--sl
, (ses ,
/
j 8" t es l si g l -i : -a -3 : , I- . 4
= no
.se ..e
-a IIII i -'
vEssEr, vEssEr.
.. . A D O W N C O MERIP P E R PLENUM
.< -< ne ns
\ -
nasV )8" , VV
"- "' RIGH PRESSURE -
INJECTION Figure 1. RELAPS Oconee-1 model; primary Loop A. 4
4 i -a
~ ~
8 7 -3 20 31 3 E 3 j. i 20$ T , ,
,208-2i 308-t, En s_ .
I Is7oi w avs; l -i ; -a : -s ; ["r e ... ..
' sea; ;sesl l -t ! - a ! - 3 ,'
1 73 724 see ses l , -d l -e vgsggt ygssgn i
.. .. n DOWNCOWEEPPER PLENUM
-4 -4 730 72S g
\ 325 -
See 230 QQ HICH PRESSURE . 4 _, ,p IN J ECTIO N l t
~
Figure 2. RELAP5 Oconee-1 model; primary Loop B. i l s . A s 5
SAFETY V ALVES POWER O PER ATED RELIEF V ALVES 803 8 01 s 803 800 m
/ *t s N
-e K ae
-s 3
-7
-e g -5 m ..........
O -4 5 ..........
=
; ......3....
.e 3
= .......
e -1 !
.J e 100 s......
A =- 820
-1' ~4 e
5 I
-3 -3 e i
5 e e e . e e I
' - - - FRO M INLET 170 l
l l l Figure 3. RELAP5 Oconee-1 model; pressurizer system e 6
TABLE 1. ' FULL POWER STEADY-STATE CONDITIONS CALCULATED BY RELAP5 COMPARED WITH PLANT NOMINAL CONDITIONS
- Parameter Plant RELAP5 Core power (MW) 2568 2546 ,
Average of hot and cold 577 577 leg temperatures (K) Hot leg pressure (MPa) 14.96 14.93 Steam generator pressure (MPa) -6.377 6.380 Secondary mass per steam 17700 17100 generator (kg) Feedwater temperature (K) 511 511 Steam exit superheat (K) 33.3 32.8 Mass flow per hot leg (kg/s) 8820 8819 i Feedwater flow per steam 680.4- 689.9 generator (kg/s) Pressurizer level (m). 5.59 5.71 l e f [. . , _ _ . , - _ . . -
TABLE 2. CORRESPONORNCE BETWEEN THE PHYSICAL AND MATHEMATICAL COMPONENTS IN THE PRIMARY LOOPS - physical Component RELAPS Component (s) Loop A Hot leg 100, 101, 105, 110, 111 Steam generator inlet plenum 115, 116 Steam generator tubes 120 , Steam generator outlet plenum 125 A-1 pump suction leg 160-A-2 pump suction leg 130 A-1 reactor coolant pump 165 A-2 reactor coolant pump 135 A-1 cold leg 170, 175, 180, 181 A-2 cold leg 140, 145, 150, 151 A-1 HPI 710, 711 A-2 HPI 715, 716 Loop B Hot leg . 200, 201, 205, 210, 211 Steam generator inlet plenum 215, 216 Steam generator tubes 220 Steam generator outlet plenum 225 ; B-1 pump suction leg 260 / B-2 pump suction leg 230 B-1_ reactor coolant pump 265 B-2 reactor coolant pump 235 B-1 cold leg 270, 275, 280, 281 8-2 cold leg 240, 245, 250, 251 B-1 HPI 720, 721 B-2 HPI 725, 726 Pressurizer Surge line 600, 605 Pressurizer 610 Pressurizer dome 615 Spray line 620 Spray valve 616 PORV 800, 801 Safety valve 800, 803 Low pressure injection system 727, 728 Core flood system 700 o 8
I i-corresponding mathematical components of the RELAPS model. RELAP5 components were numbered between 100 and 199 for the A loop, 200 and 299 for the B loop, and 600 and 630 for the pressurizer.
, The RELAP5 vessel model shown in Figure 4, consisted of several components describing the various vessel flow paths. The RELAPS vessel included an inlet annulus, downcomer, lower plenum, core, core bypass, upper plenum, upper head and a vent valve.
The eight reactor vessel internal vent valves between the upper plenum and downcomer inlet annulus were modeled with a single RELAPS servo valve. A servo valve uses the output of a RELAPS. control variable to determine the open area of the valve. The control variable was designed to calculate a valve flow area as a function of the differential pressuro across the valve. Differential pressures at which the valves start to open and are full ooen were obtained from Reference 2. Because the range of differential pressure over which the valve position changes is small, it was found necessary to smooth the calculated differential pressure to [ prevent valve chattering. Without smoothing, the valve position responded instantaneously to the differential pressure across it. With smoothing, the change in valve position in a time step was limited thus preventing excessively quick opening and closing of the valve and qualitatively timulating the response time of the flapper valves. An alternate method of modeling the vent valve, using the RELAPS inertial valve model, was considered, a test model was developed, and checkout runs performed with the valve inserted in a one-dimensional piping string. Offficulties were - encountered in obtaining proper valve performance in these checkout runs and this, along with concerns about applying the inertial valve where multi-dimensional effects are expected, led to a decision to use the servo valve method described above. The model af the vent valve used in the RELAP5 model is the best usable representation of the Oconee-1 reactor vessel vent valve presently available. Discussions of vent valve behavior ( and possible uncertainties involved are addressed in the transient results in Sections 4 through 9. l 1 l 9
e i ri i e i i i Im 850 accum sta to n l 700 7 604 536 S35 y> '
/
,I l , Z , l
^
l
^
gud etPsessvan s cTio 8 727
/ { y --
y' 500 y 2 840 / j 530 hj 4 &oo P 728 set Las100 200 g NT,g o towt,,)"R et * ** j c.100[s o6o o s
- - 2 51 i i
- r 845 645 f J
62$ / J l l 870- 520 f l CORE
's 670-2/
j
/
R O-8
/
/
/ S15-8 y ......- A. ....... / .. ... ....
j, 570-3 fi. 510-S f,
/. ...... . . . . . ..
'S16-6 ,/
j 670-4
~......- / .
f 610-4 h p ...... ...... si s-4 p 670-0 ff. . 81o-3 f
/. ...... ......
61S.3
/
870-of. c ......- St 0-8 f<. . . . . . . . . . . . . a13-2
, p
/ 670-7 j 61 0 -8 f 61 5 -1 i l I r
~
j r
- 570-0 f 6CS f/
670-e /
/ /
ri l l
~
57S lii!iif Figure 4. RELAP5 Cconee-1 model; vessel, core flood tank and LPI system. 10
The RELAPS component numbers for the vessel were between 500 and 599. Table 3 summarizes the relationship between the physical' components of the Oconee-1 vessel and the corresponding mathematical components of the RELAP5 model. Also included in the primary system were the low pressure injection (LPI) system, the core flood system, PORV, and safety valves. The LPI and core flood systems were connected to the vessel inlet annulus. The PORV and safety valves were connected tr the top of the pressurizer. Table 2 summarizes the relationship between the physical components of the Oconee-1 LPI, core flood systems, PORV, and safety valves and the corresponding mathematical components of the RELAPS model. The RELAPS nonequilterium model was applied to all volumes in the primary system except for the accumulator where the accumulator model was I applied. The wall friction model was applied in all primary volumes. Centrally located junctions for horizontal stratification were applied to all horizontal junctions. The choking and two-velocity models were applied at all primary junctions. An inertial solution was calculated at all junctions. The smooth area change model was applied at all junctions except those listed in Table 4. Heat structures were used to represent heat transfer and stored energy from fuel rods, steam generator tubes, loop piping, vessel wall, vessel [ internals, pressurizer wall, pressurizer surge line, and pressurizer heaters. ~ l 2.3 Secondary System i The Oconee-1 steam generators are Once-through type steam generaters and are oriented vertically. Between the outer shell and the heat exchange tube bundle is a cylindrical baffle, forming a downcomer section. A gap in the baffle allows steam to be drawn from the boiler region into the downcomer to heat the incoming feedwater. Af ter falling throagh the downcomer the feedwater enters the tube bundle and flows upward, vaporizing to saturated steam in the nucleate boiling region. Dry saturated steam is l 11
. . _ . . . ~ . . . _ . - . - - -.--- . - - _ ,
TABLE 3. CORRESPONDENCE BETWEEN THE PHYSICAL AND MATHEMATICAL COMPONENTS IN THE VESSEL - Physical Component RELAPS Component (s) Inlet annulus 555, 560, 565 Downcomer 570 Lower pier.um 505, 575 , Core 515 Core bypass 510 Upper plenum 520, 525, 530, 535, 540, 5*2, 545 Upper head 550 Vent valve 536 O e 4 12
I 7 TABLE 4. JUNCTIONS USING THE ABRUPT AREA CHANGE MODEL r Junction Number
- Description ___
116 Steam Generator A inlet plenum to tube bundle 12501 Steam Generator A tube bunole to outlet plenum 216 Steam Generator B inlet plenum to tube bundle l 22501 Steam Generator B tube bundle to outlet plenum 34501 Steam Generator A to main steam line 44501 Steam Generator B to main steam line i 536 Reactor vessel vent valve i 811 Steam Generator A turbine bypass valve 821 Steam Generator A turbine stop valve 911 Steam Generator B turbine bypass valve
- (s 921 Steam Generator B turbine stop valve
- a. Junctions with three digit numbers represent RELAPS components and are either valves or single junctions. Junctions with five digit numbers are
, internal junctions of either a branch or pipe component. The first three [ digits of the internal junction number specify wnich component the junction is associated with. The last two digits are sequence numbers which uniquely define the internal junction. 4 I f 13
produced in the film boiling region and then raised to the final steam temperature in the superheat region. The steam flow then enters the steam outlet section which is between the outer shell and the cylindrical baffle i above the feedwater inlet port. The superheated steam then exits the steam j generator via the main steam line. The RELAP5 model of the steam generator secondary is shown , schematically in Figures 5 and 6. Table 5 summarizes the relationship a between the physical comoonents of the Oconee-1 A and B loop steam generator secondaries and the corresponding mathematical componants of the RELAPS model. RELAP5 component numbers between 300 and 399 were used to represent the A loop steam generator secondary and between 400 and 499 for the B loop steam generator secondary. Also described in Table 'S are the emergency feedwater syster.: and the main steam line components for each loop out to the turbine stop valves including the turbine bypass and safety valves. Since the initial calculations were performed using this model (Reference 1), the manner in which secondary levels are calculated in the p model was revised to better represent the actual method used in the plant. Originally the startup and operating levels were calculated based on the collapsed downcomer liquid level. .For calculations presented in this report the levels were based on calculated static pressures and fluid temperatures at the locations corresponding to pressure and temperature taps in the Oconee-1 steam generators. The calculated temperatures and differential pressures were smoothed in a manner consistent with the actual - instrumentation system. The improved method of level calculation significantly affected the sequence of events in the revised =ain steam line break transient, as discussed further in Section 4.3, and, in general, improved the behavior of emergency feedwater injection which is controlled on level. The RELAPS options used in the secondary system follow. The l nonequilibrium option was applied to all volumes in the secondary system. This was a change from the equilibrium cption used to perform the initial calculations with this model. Difficulties encountered with secondary behavior _ during the initial portion of the main steam line break calculation were overcome, allowing the nonequilibrium option to be used. t-14 D
l i _ 44 4 d I see. e.e 5804
.... ....... 4 ..........
en e SGG Gee.t tee.s 800 9 Eg 609
.....p....
... . -oo-- _
, ).. ;
b 944 ,- 904 0 199*8
..t...." .....,... J
( m/ S GS M i I ES 889*4 l j&4 I E9*G 006 0 009+ 1 59*8 SSS.e 4994P = . 809 6 584 gas.e See 800 4
..... - - i l
1 Figure S. RELAPS Oconee-1 model; Loop A steam geileratur secondary side, main steam line. 15
e e l-se See gg lSGS I ld l 1 h g i >f ae - 809 l 406*B ett.O 988'I a0 009 ,
,, _. L....,.....
eee + ....., ... 9
, L.L. ... .. 4
- r * *
- a 's i fd see-o
. . . . .r . . . .
l 09*qM ] f {
... p ,
seg. gg 8.D*. pr-------- 436 5 69 8 889 4 este I G S *G 899 4
- m 9 9 f &P - =
tot.l e64 seset i n C V g .......... Irt
.....J -
n. Figure 6. RELAP6 Oconee 1 model; Loop B steam generator secondary side, .. main.steem line. 16
TABLE 5. CORRESPONDENCE BETlEEN THE PHYSICAL AND MATHEMATICAL COMPONENTS IN THE A AND B STEAM GENERATOR SECONDARIES - Physical Component RELAP5 Component
. Steam Generator A .
Downcomer 3C5 Tube bundle Nucleate boiling region 206, 310 Film boiling region 315, 320 Super heat region 320, 325 Aspirator region 315 , Superheat steam outleg region 330, 335, 340 Main steam line 345, 350 Turbine stop valve 820, 821 Turbine bypass valve 810, 811 Safety valves 806, 807 Emergency feedwater system Turbine-driven 812, 813 Motor-driven 803, 804
/
( Steam Generator B Downcomer 405 Tube bundle Nucleate boiling region 406, 410 Film boiling region 415, 420 Super heat region 420, 425 l Aspirator region . 415 Super heat steam outlet region 430, 435, 440 Main steam line 445, 450 . Turbine stop valve 920, 921
- Turbine bypass valve 910, 911 l 56fety valves 906, 907
- Emergency feedwater system Turbine-driven 912, 913 l-Motor-driven 903, 904 17 ,
~ As will be described later, this change resulted in a more satisfactory behavior of the 1cendary system during periods of emergency' feedwater injection. The wall friction model was applied to all secondary volumes. Centrally located junctions for horizontal stratificatidn were applied to all horizontal junctions. The choking and two-velocity models were applied to all junctions. An inertial solution was calculated at all junctions. ' The smooth area change model was applied at all junctions except those listed in Table 4. - Heat structures were used to simulate the stored energy contained in the secondary shell wall, piping and internals. 2.4 Feedwater System The Oconee-1 feedwater system is described in this section. This portion of the Oconee-1 model was originally supplied by SAI-Oak Ridge (Reference 3) and has been extensively revised. The various components of the feedwater system model are shown in Figures 7 and 8. Table 6 summarizes the relationship between the physical components of the Oconee-1 / feedwater system and the corresponding mathematical components of the RELAPS model. The purpose of the feeowater system is to supply demineralized / deaerated subcooled water to the A and 8 steam generator secondary inlets. The water is heated and pressurized from 305 K (90 F) and 0.01 MPa (14.5 psia) to 511 K (460'F) and 6.55 MPa (950 psia). - The RELAP5 options used in describing the feedwater system include tre nonequilibrium and wall friction models. Centrally located junctions were applied to all horizontal junctior;. Choking and two-velocity models were also applied at all junctions. . Heat structures were used to represent the high and low pressure heaters and the piping wall metal masses. The feedwater heaters were modeled using the appropriate tube bundle surface areas. The energy contributed to the feedwater from the heater secondaries was modeled using 18
A B P U S T E 1 R NDV. ADIEI 1 1 gMFV AEA 1
,s
, 2 1 5 1 7 o 1 8 t 6 l 7 l N
e I A w t R o D R
, , h E
T 8 4 m64 1 e f m A 7 1 7 f E , 1 1 e o l i
,
- r r
f D 1 , N
, 4 ,
6 6 n 1 6 i I A , 7 , 7 a R 1 , r D t . R 2 - d E 4 - e T A 7 1 f e E 3 n, H - 1
;l?
6 l E 0 7 e 1, d 1 4 7 o. 1
, mse 2W 6 1 lv PR 8 E 7- e a ev 3 SP nd f 7 OM OU N 1 6
7 R oe ce DP E T A Of W 5n i i D E Pa A
, 6 E
F B Lm 5 NP E d 6 7 ( IW AU WP Rn
. a
, 3 R 7 p
, E 7 T A e rt u
W 8 ur D 5 ga t
- E E A 7 it
- L F Fs NP L
4 E Ik 3 WP AU TW WP f7 OU HP A H 0 3 7 G
- l ,
m WOTO R DRif ts EMERGENCT f tEDT ATER . l OI' TURBINE D RIVRN gg STI AM CENER ATO R a ED !g .. ... % ... V g((gg yfE M DOME l
-D4 irf!!,c'0"llt ro k 4 n.< j n.-. ... .
? , , , , C
_pq_ .n . 734 "a a.4 STR Roo A.M nc0 O EN.ER R ATO R 4 ...< ,,..... ,.... n. l N- STR AW G ENER ATO R MOTORDRtVEN ...
!!!f!!?t!!o0W.
!!!!!!"r!! a. a. -N- a.
I Tunstwe o Rives
- a. EMIRGENCY FEEDW ATER Figure 8. RELAPS Oconee-1 model; feedtrain from startup and main feed valves to steam generator, crossover, and emergency feedwater system.
l 1
/ ,
1 20
t TABLE 6. CORRESPONDENCE BETWEEN THE PHYSICAL AND MATHEMATICAL COMPCNENTS IN THE FEED TRAIN . Physical Comoonent RELtPS Component
. Hotwell 730 Hotwell pump 734 Demineralizer and deaerator system 736 Booster pump 738 Low pressure F and E heaters 740 E heater drain 742, 744 1
Low pressure D heaters 746 D heater drain 748, 750 Low pressure C heaters 752 A main feedwater pump 754 A main feedwater pump discharge line 756, 757 B main feedwater pump inlet line 758 B ..:ain feedwater pump 760 Main feedwater pump discharge line 761, 762 Main feedwater pump header 763 A and E heater train number 1 764 f A ar.d B heater-train number 2 766 I ( High pressure heater header A and B startup and main feedwater 768 770,- 771, 772, 773 control valves Steam Generator A control valve header 774, 775 Steam Generator B control valve header 776, 777 Steam Generator A feedwater header 778, 782 Steam Generator B feedwater header 780, 784 Steam Generator A feecwater crossover 850, 851, 852, 853, 854 Steam Generator B feedwater crossover 950, 951, 952, 953, 954 e 21 4 1- < - , - - - . ,, y -
the heat structure energy source option. The piping was modeled by first calculating the applicable metal volume, then adding the equ'ivalent metal thickness to the appropriate heat structure. Heat structures repr9senting several components were combined and added to selected components to reduce the computer storage requirements. 2.5 Integrated Control System The Babcock and Wilcox Integrated Control System (ICS) is a comprehensive system that controls all major portions of the Oconee-1 plant. A schematic of the system is shown in Figure 9. The models developed for the Pressurized Thermal Shock program study of the Oconee-1 plant comprise only three portions of the overall control system. These are
- 1. Turbine bypass control
- 2. Feedwater control
- 3. Emergency feedwater control. ,
The model of the ICS was developed by SAI-Oak Ridge (heference 4). The unit load demand and reactor control subsystems have not been modeled except to provide the demand signal for the feedwater control subsystem and the reactor control demand signal. The reactor demand signal is required for the neutron power error cross-limit signal to feedwater control from reactor control. The integrated master has not been modeled except for the - turbine bypass valve control. Specific limitations of the model are:
- 1. Operator hand control stations are not modeled. The sequences modeled do not require hand control actions.
- 2. The turbine header pressure is assumed to be a boundary condition. Consequently, turbine control is not modelec in the integrated master. This was not a significant limitation because in the sequences evaluated turbine trips o'ccur early in the transients.
s 22
~~
LOAD DISPATCH SYSTEM . h UNIT LOAD DEMAND l
+ ,
- INTEGRATED MASTER 1F 1F 1F PRESSilRE CONTROL REACTOR CONTROL STEAM GENERATOR FEEDWATER CONTROL ro W _
1F 4F iF 4F 1F 1p 0'" TURBINE FW SG A FW BVPASS RD FW SG 8 GOVERNOR VALVE PUMPS VALVE SYSTEM DRIVES INTERFACE IWiEEFALti INTERFACE INTERFACE tNTf rtFACE INTERFACE
, 1F 1F 4F 1F 1F 1F TURBINE ROD TURalNE SG A FEED SG B VPASS DRIVES GOVERNOR yggy VALVES PUMPS VALVES Figu're 9 II & W integrated control system organization,
- 3. Unit load demand is assumed to transfer to " track generated megawatts" upon - reactor / turbine trip. This initia'tes a -20% full power (FP)/ min runback. If a reactor coolant pump or a' main feedwater pump is tripped before the unit load demand has reached its lower limit of 10% FF, the ramp rate should be switched to
-50% FP/ min. Since the Stu limits or the neutron cross-limit are the dominant control in the transients to be simulated, this is not a significant lim'.tation. -
4 Small subsystems withir. the feedwater control were not modeled. These include:
- a. Ratio Controller--This subsystem is designed to split the total feedwater demand between the two steam generators so as to maintain equal reactor coolant cold leg temperatures and total feedwater demand. This system was not modeled because it is blocked during transients or if eitSer steam generator is on level control. Since these conditions are exoected throughcut most of the transients to be studied, it was not felt necessary to model this effect. If operational upsets are modeled in the future, this subsystem will be required,
- b. Total Feedwater Controller--This sebsystem is designed to aid in quick recovery of large feedwater/ demand mismatches which are possible during operational upsets such as loss of
- a reactor coolant pump. This controller is blocked if the primary loop flow mismatch is <10%. Since no transient specified asymmetric reactor coolant pump status, this system was not simulated.
- 5. Calibrating integrals are not modeled. These integrating controllers are provided for long term control of system mismetches and are consequently blocked during transients.
24
All other portions of the feedwater control and the turbine bypass control are faithful implementations of the ICS control lo'gic. Gains for proportional and proportional plus integral control modules have been taken i from generic Bailey Meter Company Integrated Control System manuals
. (Reference 5) where available. Proportional gains for the feedwater control manuals have been adjusted so that 1*4 flow error provides approximately 1/2*4 main feedwater valve motion and 1" water level error provides approximately 2*J startup feedwater valve movement. These i
. sensitivities were provided by Mr. Luther Joiner of Babcock and Wilcox.
Mr. Joiner also provided ranges of integral controller gains that are typical of operating plants. The mid-range values of these gains were selected or values from Bailey literature were used, if available and consistent with Mr. Joiner's recommendations. The Btu limits were provided by Duke Power Company for Oconee-1. These curves indicated a design flow 6 of 2.4 x 106kg/h (5.3 x 10 lb/n) for each steam generator. The nominal feedwater flow specified for the simulation model is ! ( 2.45 x 106 kg/h (5.4 x 10 lb/h).6 The Stu curves were adjusted upward to provide the same margin at the higher flow. In summary, the model is an accurate representatior of the ICS functions necessary to model the specified transients. For the long response time transients under consideration, the controller gains are sufficiently accurate to provide reasonable response. 2.6 Documentation Control of Codes and Models - Input decks and the code versions used to perform the Oconee-1 j pressurized thermal shock RELAPS calculations are documented in Table 7. Input decks are controlled on tape at INEL under configuration control number F01216. 25
F TABLE 7. DOCUMENTATION CONTROL OF CODE VERSIONS AND MODELS Record Number- Steady State on Configuration From Which RELAP5/M001.5 Control Tape Transient Was Calculation Cycle Used _ F01216__ _ _, I n i t i a t_e.d. ., - Steady state number 1 27 1 -- (hnt standby) Steady state number 2 16 2 -- (full power) Steady state number 3 27 3 -- (full power) Steady state number 4 30 4 -- , (full power) Main steam line break a 23 5 2 Steam generator 23 6 2-overfeed" Hot leg small break 23 7 2
~'
(PORV)* , Oconee-3 plant 23 8 2 transient
- Revised main steam 27 9 3 line' break Maximum sustainable 25 10 2 steam generator -
overfeed Turbine bypass failure 27 11 1 at hot standby Pressurizer surge line 30 12 4 small break
'd 26
TABLE 7. (continued) Record Number Steady State on Configuration From Which RELAPS/M001.5 Control Tape Transient Was
- Calculation Cycle Used F01216 Initiated R.C. pump suction small 30 13 4 creak Steam generator tube 31 14 4 rupture
- a. Results from these calculations are detailed in Reference 1.
t e 27
l
- 3. OVERVIEW OF PREVIOUSLY REPORTED ANALYSES Detailed results from the first four sequences analyzed were presented in Reference 1. A brief overview of the results for these sequences is presented in Sections 3.1 through 3.4 along with plotted results of key parameters to De used as boundary conditions in tracture mechanics ~
calculations. 3.1 Main Steam Line Break Transient A description of the transient, as defined by ORNL, appears in Table 8. The sequence is identical to that for the revised main steam line break presented in Section 4, except for ~the reactor coolant pump (RCP) restart and hich pressure injection (HPI) throttling criteria. For this sequence, one RCP per loop is te be restarted 10 min after attainment of 28 K (50*F) subccoling in both hot legs and Hol is not throttled. In the revised sequeri:e, the RCP restart is immediate upon attaining 42 K (75"F) subcooling and HPI is to be throttlea to raintain 28 to 56 K (50* to 100*F) subcooling. ,
)
Results of the calculation, and extrapolations to 2 h are shown in Figures 10 thrcugn 12. Figure 10 shows the reactor vessel downcomer pressure response. The minimum pressure calculated was 5.02 MPa (728 psia) and the maxirum was 16.99 MPa (2465 psia), the opening setpoint of the power operated relief valve. Figure 11 shows the reactor vessel downcomer fluid temperature. The minimum temperature calculated was 487 K (407"F). - L
'The reactor vessel wall inside surface heat transfer coefficient is shown l in Figure 12. The low values (for example at 500 s) correspond to periods of loop flow and high values correspond to periods when reactor coolant
! pumps are operating. 3.2 Steam Generator Overfeed Transient A description of the transient as defined by ORNL appears in Table 9. The scenario differs from that of the maximum sustainable overfeed
- transient, presented in Section 5, in the availability of feed train protection functions. For this sequence, the protection functions, -
28
20 , , , , , , , CALCL".ATED EXTRAPOLATED m ^ o p....._._...---. 2300 o . y . 15 - - 3 y m 2000 g W . W e E 6 Q- c. w g1. 1300 , 2 E D l :s J I : - o ' o 1000 5 r 0 1000 2000 3000 4000 50C0 6000 7000 5000 ( Time (s) Figure 10. Wotn steem line break, R.V. downcon er pressure. t
$80 , , , , , , ,
a ^ M CALCULATED
- l'-
- -- EXTRAPOLATED
- e
' $60 - - 550 6 3 :3 0 -
- O s u e . . - e g540 ,.
c. s -
..- 500 5
? ~
l l
?
=
3 CT $20 ,F b' h j'il = 3 c-
=l ) 450
\ ESCO 3 {
- -2 i
i O o j > ' ' ' ' ' ' 480
. 0 1000 2000 3000 4000 5000 6000 7000 80C0 Time (s)
Figure 11. Wain steam line break, R.V. downcomer f'uld temperature. l 29 7 m _ , - - - , - _ - - <-
N. I I t i t 10 1 i . . .
- CALCULATE 0 5000
. . --- EXTRAPOLATED E c l W e 6 g 1 4000 O m.
Om 20 ' a g 3x u,) d l _. _ __ ___ _ _____ ____ 3000 0O o~ e
-k ws -.]
2000 =3 8"J = _ <-
.m =
s v= a - 1000 J 4
' -A ' ' ' '
0 0 0 8000 2000 3000 4000 5000 60C0 7000 5000 Tim. (s) Figure 12. Weln steam line break R.V. downcomer inside surf ace . heet transf er coefficient. 1 I
./
. 30 1
._ . . . .. = _ .-. _ __ - - . . -
1 s . TABLE 8. MAIN STEAM LINE BREAK TRANSIENI SCENARIO
- 1. Reactor trips, coincident with break of 0.86 m (34 in.) steam line
- 2. Turbine trips; TSVs close .
- 3. ICS functions as designed 4 Protection systems on hotwel'1, condensate booster; and MFW pumps function as oesigned
- 5. HPI actuates at setpoi.nt of 10.34 MPa (1500 psig)
- 6. Operator trips RC pumps 30 s after HPI actuation a
- 7. EFW pumps start when low MFW pump discharge pressure is sensed.
a
- 8. MFW/EFW system attempts to maintain 6.1 m (240 in.) steam generator leve6 a
- 9. Core flood tanks actuate at setpoint -
- 10. LPI actuates at setpoint
! 11. Operator isolates feedwater to affected steam generator 10 min D into the transient c
- 12. Operator restar's one RC pump in each loop 10 min after attaining 28 K (30*F) subcooling in the hot legs ,
- 13. PORV opens at setpoint' of 16.9 MPa (2450 psig) 8
- 14. SRVs open at setpoint of 17.2 MPa (2500 psig)
- 15. PORV/SRVs reseat at their setpoints 16.5 MPa (2400 psig)
- 16. Pressurizer goes water-solid
*17. EFW surge tank capacity of 272544 liters (72,00 gal) is exhausted; and the two motor-driven EFW pumps trip a
- 18. Turbine-criven EFW continues to draw from hotwell
- a. Event is phenomenologically dependent.
- b. Five times a typical time for operator action, based on simulator experience.
- c. Two times a typical time for operator action, based on simulator experience.
31
. - , ~ - - , , . , , - - - , - - , , - - , -
, - , , , . - , , - , , , . - -,-,,,,we, ,- , , , , - ,-g -,
7-..y-- ---, --, , , - -
TABLE 9. STEAM GENERATOR OVERFEED TRANSIENT SCENARIO
- 1. Reactor trip turbine trip, TSVs close
- 2. TBVs open and safety valves open and reseat at setpoints
- 3. Full MFW flow continues -
4 HPI actuates at setpoint of 10.34 MPa (1500 psig), but operator does not trip RCPs ,
- 5. MFW pump trip on steam generator high level fails to occur
- 6. Steam generators fill
- 7. Core flood tanks actuate at setpoint'
- 8. LPI actuates at setpoint"
- 9. Protection systems on hotwell condensate booster, and MFW/EFW pumps function as designed
- 10. Main steam lines fill with water
- 11. TBV closed on loss of condenser vacuum, at its setpoint
- 12. EFW oumps start, per low MFW pump discharge pressure setpoint ,,
- 13. EFW flow is controlled at steam generator level setpoint of 0.61 m (24 in.)
- 14. PORV opens at setpoint of 16.9 MPa (2450 psig)
- 15. 3RVs open at setpoint of 17.2 MPa (2500 psig)a
- 16. Pressurizer relief valves reseat at setpoint of 16.5 MPa (2400 psig)
- 17. Pressurizer goes water-solid
- a. Event is phencmenologically dependent.
,ma 32
a
~
primarily low suction' pressure trips on the feed train pumps, were available and as a result the main feedwater pumps were tr'ipped short after the transient was initiated. This sequence further differs from the sequence in Section 5 in that reactor coolant pump power is not tripped off
. by the operator in this sequence.
Results of the calculation, and extrapolations to 2 h are shown in Figures 13 through 15. Figure 13 shows the reactor vessel downcomer fluid pressure response, the minimum pressure calculated was 8.92 MPa (1293 psia). The maximum pressure calculated was 16.99 MPa (2465 psia) which corresponds to the power operated relief valve opening setpoint pressure. The minimum calculated downcomer fluid temperature was 505 K (450*F) as shown in Figure 14. The reactor vessel wall inside surface heat
- transfer coefficient, shown in Figure 15, represents forced convection conditions throughout the calculation. The variations observed in the coefficient are primarily due to changing fluid densities and pressures as the transient proceeded.
l
\
3.3 Hot Lee Small Break Transient The sequence is initiateo by a stuck open power operated relief valve. The operator is assumed not to trip power to the reactor coolant pumps. A description of the transient, as defined by ORNL, is described in Table 10. Results of the calculation, and an extrapolation to 2 h, are shown in
- Figures 16 through 18. Figure 16 shows the reactor vessel downcomer fluid pressure response. The minimum calculated pressure was 8.63 MPa (1259 psia). As shown in Figure 17 the minimum downcomer fluid temperature l was 545 K (521*F) and occurred at 2 h; the end of the extrapolation. The extrapolated primary pressure at 2 h is 11.38 MPa (1650 psia). The heat transfer coefficient on the inside surface of the reactor vessel downcomer wall is extrapolated to drif t downward as fluid properties change as shown in Figure 18. The coefficient represents forced-convection conditions throughout the calculation.
33
20C00 , , , CALCULATED . EXTRAPOLATED n 18000 - - O a ........................................... 2500 6 .n w w000 M
- a NN<" ~
2000 =[
=.
0 n m y 12000 - s 3 c. d < 1500
> 10000 -
8000 O 1800 3600 5400 7200 Time (s) Figure 13. S.C. overf eed, R.V. downcomer pressure. 580
)
, , . /
n ^ X CALCULATED b
-- - EXTR APOLATED e e h- 6 E $40 '- -'550
- O O 6 6 e e i
4 c. E E
- 540 - -
.e.
2 .. ,. .................. soo 3 . s a
=
7 , c'
,' =
520 - * - e , o E ,' E 3 , 3
- o * -......-
450 0
> t I f 0 1000 3600 5400 7200 Time (s) .
Figure 14 S.G. overf eed, R.Y. downcomer fluid tempera ture. 1-
/
W 34
, 40 i 4 a 7000 CALCULATED
.... EXTRAPOLATED -
e 6500 yp
~
,._f..........................................
*n 35 . 1i .
e
$y uee 6000 g;l uE
, 'o w i
* *c 5500cm\
"2 h 30 - - OE u
2 -s s000 O E i x 25 O 1800 3600 S400 7200 Time (s) Figure 15. S.C. overf eed. R.V. downcomer inside surf ace beat transf er I coe f ficien t.
\
16000 , , , , , , , CALC %A TED
-- -- EXTR APOLA TED 2200 m
O CL v* 14000 ,- -2000 m o w a m
- G.
@ 1800 m 12000 - . e w
ua g Q"
------.............-*--~.....-............. 2 1600 E ta 2 $
3 10000 - - O 1400 [
, , , , , , , 1200
- 0 1000 2000 3000 4000 5000 6000 7000 8000 Time (s) Figure 16. Hot leg smotl break, R.V. downcomer pressure. 35 e
580 , , , , , , , 5+' CALCULATED 580 *F ,
- - - . EXTR APOL A TED
=
W w 570 - - Qi 4 4 6 560 u. 2 d Sec A..,," W w Q * ...., c_. 3- 540 a O
' ..,'.. - o.
J ; taJ 550 - y 2 3 y
' ..' .. 3 p
520 g
> e. , i g e i ,
0 1000 2000 3000 4000 5000 6000 7000 8000 Time (s) Figure 17. Hot leg small break. R.V. down omer fluid temperature.
.)
I 30 i , , , , , , CALCULATED
* --- EXTRAPOLATED .-
J 5000 c e -
- 28 l-u ^
C .e. h O m e gM 18 -
.'%. - 0,.I un 8
.N., 4500 " ;:-
uE e s.
... -n
- 2. m 24 -
EN o2
@v i ' ... 4000 g
22 - j z Z 20 O 1000 2000 3000 4000 5000 6000 7000 8000 Time (s) , Figure 18. Hot leg small break, it.V. downccmer inside surf ace heat transf er coefficierit. 1 36
- - - . - -
- us- e-
L l TABLE 10. HOT LEG SMALL BREAK TRANSIENT SCENARIO
- 1. Small break loss-of-coolant accident
- t
- 2. Reactor trips, turbine trips, TSVs close
- 3. HPI actuates at setpoint of 10.34 MPa (1500 psi)
- 4. TBVs/SRVs in secondary function as designed
- 5. ICS fails to run back MFW
- 6. MFW pumps trip on high steam generator level
- 7. EFW system functions as designed
- 8. Core flood tanks dump; LPI system actuates!
- a. Size: 2.1368 x 10-3 ,2 (0.023 ft2 ); Location: (pres's'iriz'e r relief valve).
- b. Event is contingent on size of break.
4 37
3.4 Oconee-3 Plant Transient On March 14, 1980 the Oconee-3 pWR experienced a transient from full power which was initiated by a turbine trip followed by a partial steam generator overfeed. The Oconee-3 PWR is a virtually identical sister plant to the Oconee-1 PWR. By performing a RELAPS calculation duplicating the transient sequence and comparing calculated data with that measured in the plant, a limited assessment of the computer model was pe-formed. - Since the plant data measurea during the transient is proprietary to the Duke Power Company, specific comparisons between computed and measured data are not presented here since their inclusion would necessitate the same classification of this report. 'These comparisor,s, presented 11 Reference 1, generally indicate good agreement between calculated and measured data. Differences, where noted, were found to be caused by minor modeling problems or suspect plant data.
.]
e 38
- 4. MAIN STEAM LINE BREAK REVISED TRANSIENT The following subsections contain the main steam line break revised transient scenario description, modeling changes effected to perform the transient calculation, detailed analysis of th,e transient results, and conclusions drawn from the analysis.
4.1 Transient Scenario Description A description of the revised main steam line break sequence analyzed appears in Table 11. This sequence definition was developed at Oak Ridge National Laboratory (ORNL). The sequence is initiated by a 200% double-ended rupture of a main steam line on one steam generator. All automatic plant functions are assumed to be operative. Operator actions are assumed to trip reactor-coolant pump power 30 s after initiation of high pressure I.
\
injection, terminate all feedwater and turbine bypass on both steam generators at 10 min, and reactivate emergency feecwater and turbine bypass o the unaffected steam generator at 15 min. 4.2 Model Changes The basic RELAPS model used to perform the main steam line break transient calculation is described in Section 2. The double-ended rupture of the steam line en Loop A was ,imulated by the simultaneous insertion of the break flow path, at the downstream end of Component 345, to an atmospheric pressure sink anc isolation of downstream steam line volumes (Components 350, 820, and 821). Component numbers correspond to those shown on the nodalization diagram in Figure 5. The unaffected loop steam generator was assumed to Le immediately isolated from
, the effects of the break due to the very rapid closure time of the turDine stop valves.
39
~
i
l TABLE 11. REVISED MAIN STEAM LINE BREAK SCENARIO Summary
Description:
The accident sequence begins with the break of a 34 in. steam line, coincident with reactor and turbine trips. The forcing function for the accident is the delay by the operator in isolating the FW flow to the - affected steam generator, coupled with delay in throttling the HPI flow and restarting one RC pump in each loop after 50 F subcooling is attained. Initial Conditions:
- 1. Full reactor power
- 2. Nominal temperatures and pressures in primary / secondary
- 3. Decay heat: 1.0 times the ANS stardard
- 4. Pressurizer spray / heaters operate as designed Sequence of Events:
- 1. Reactor trips, coincident with break of 34 in. steam line
- 2. Turbine trips; TSVs close
- 3. ICS functions as designed ,/
- 4. Protection systems on hotwell, condensate booster, and MFW/EFW pumps function as designed
- 5. HPI actuates at setpoint (1500 psig)
- 6. Operator trips RC pumps 30 s after HPI actuation
- 7. EFW pumps start when low MFW pump discharge pressure is sensed *
- 8. MFW/EFW system attempts to maintain 240 in. steam generator level a
- 9. Core flood tanks actuate at setpoint' ,
- 10. LPI actuates at setpoint"
- 11. a. Operator isolates feedwater to both steam generators 10 min .
into transient (close MFW, start-up-FW, EFW and TBV systems)
- b. Operator begins refilling unaffected steam generator to 240 in. level with EFW at maximum rate 15 min. into transient 3
- c. Turbine bypass system on unaffected steam line opens at 15 min. and maintains 1000 psi pressure.
.,l 40 a
TABLE 11. (continued)
- 12. Operator restarts one RC pump in each loop after attaining 75'F subcooling and throttles HPI to maintain 75 25'F subcooling
. 13. PORV opens at setpoint (2450 psig)
- 14. SRVs open at setpoint (2500 psig)*
- 15. PORV/SRVs reseat at their setpoints (2400 psig)
- 16. Pressurizer goes water-solid
- 17. EFW surge tank capacity (72,000 gal) is exhausted, and the two motor-driven EFW pumps trip"
- 18. Turbine-driven EFW continues to' draw from hotwell a
- a. Event is phenomenologically dependent.
! i 9 41
For this transient, where the secondary pressure in the unaffected steam generator far exceeds that in the affected steam generator, all of the turbine-driven emergency feedwater (EFW) capacity was assumed to be available to the affected loop steam generator. Motor-drive'n EFW was continued to be split evenly between the two steam generators since one motor-driven EFW pump is specifically dedicated to each steam generator.
- 4.3 Transient Results -
A sequence of events for this calculation is summarized in Table 12. At zero time a 0.864 m (34 in.) diameter break openeo in the steam line. A 200%, double-endec~' rupture of the steem line in Loop A was simulated by the simultaneous insertion of the break flow path and isolation of downstream steam line volumes (Components 350, 820, and 821 as shown on the nodalization diagram, Figure 5). Turbine and reactor trips were also assumed to occur at zero time. As a result, the turbine stop valves were closed over 0.5 s, the feedwater heater power and feedwater heater drain flows were terminated at 5 s, and an instantaneous reactor ,,) shutdown to the ANS decay heat was assumed. The primary and secondary system prassure responses are shown in Figure 19. The affected secordary system depressurized rapidly and the resulting cooling depressurized the primary system. The unaffected secondary system was initially pressurized to the secondary relief valve opening setpoint at 3 s and the valves remained open until 13 s. - Depressurization in the affected secondary due to the break caused the secondary liquid inventory to flash rapidly and the resulting vapor velocities were sufficient to entrain liquid and sweep it out of the generator to the break. Figures 20 and 21 show the void fraction and the mass flow rate at the break, respectively. Because the flow in the affected loop secondary was accelerated as a result of the break, the indicated operating level increased dramatically during the initial phase of the transient as shown in Figure 22. This increase was observed because the operating level indication is based on 42
20 , , , PR IMARY
--- UNAFFECTED SECCNSARY 2500 g -- AFFECTED SECOhCARY Q
- a. -
2 E 2000 m w
- m D
M fn m-- . 1300 3 e W e E 6 Q- c. ht -! 1000 e s_4 4
*,,. ,. - 2 E
O (
> .,'m.s._. . . "* .- 500 >
o
\
**~~~~- '
0 O O 1000 2000 3000 40C0 Time (s) Figure 19. Revised MSLB, pr'mory and secondary system pressures. I
\
I f
& I I
O.99 - 2 0 b Q: 0.98 - - L. Q . O 0.97 - - gg , e i 0 1000 2000 3000 4000 Time (s) Figure 20. Revised MSLB, void fraction at its breon j unc tion. 43 e
I s x 5000 , , ,
'0000 m aC00 - -
Q N N E m 6 e W 2000'- - 1 o C 3 e o ' O 3 0-- wa $
-2000 O 1C00 2000 3000 4000 Tim. (s)
Figure 21. Revised WSLB, break mess flow ra t e.
/
20 , , ,
.J go UNAFFECTED LOOP - lI
--- AFFECTED LOOP m ~
n , 5 15 '- so { a 5 40 w$ w J
.J o w -
3o o z z NAFFFCTED LOOP 20 M s - E o o AFFECTED LOOP ' C. 0 0 1000 2000 30C0 4000 Tim. (s) Figure 22. Revised WSL9, steam generator operating levels. 44
TABLE 12. REVISED MSLB SEQUENCE OF EVENTS Time from Start of Transient
- Event (s)
Reactor trip, turbine trip break opens O Main feedwater pumps tripped on high level in affected steam 0.3 generator Emergency feedwater tripped on based on low main feedwater pump 4.4 discharge pressure HPI tripped on based on low hot leg pressure 5.3 RC pumps tripped (30 s after HPI initiation) and main feedwater 35.3 rerouted to emergency feedwater header 75'F fubcooling attained in both hot legs, 2 RC pumps restarted 300 and HPI throttled to maintain 50-100*F subcooling unaffected steam generator level recovered to 240 in. Main 320 l feedwater to tnis steam gener& tor is terminated and emergency i feedwater is throttled to mairtain 240 in. level. Restarted RC pumps are up to full speed 360 Hotwell surge tank is empty, motor driven emergency feedwater 513 is terminated to both steam generators Per scenario, all main and emergency feedwater and turbine 600 bypass capability is terminated ! Per scenario, turbine bypass and turbine-driven emergency 900 feedwater is made available to the unaffected steam generator . Last time when HPI is injected because subcocling is greater 1992 than 100*F after this time Pressurizer level reaches top 2354 l PORV opening setpoint pressure reached 2432 l j Calculation terminated 2697 l l 45 > 1 l
the differential pressure between the lower-downcomer and mid-boiler regions. When the flow was accelerated and. liquid was entrained, both the differential pressure sensed and the operating level dramatically increased at the start of the transient. The affected loop secondary operating level remained above its initial, steady state, value for the first 5 s of the transient and at 0.3 s the level exceeded the 90'. of range limit, thus causing a main feedwater pump trip. , Emergency feedwater was initiated to both steam generators at 4.4 s as a result of the main feedwater pump discharge pressure falling below H5.27 MPa (765 psia). This was an error in the calculation in that emergency feedwater initiation st,uld have been delayed until the startup levels fell below 0.61 m (24-in.) with reactor coolant pumps powered or 6.1 m (240 in.) with reactor coolant pumps tripped. Had the <=lculation been performed correctly, emergency feedwater would have been ir.itiated At 8 s to the affected steam generator and 15 s to the unaffected steam generator. The effect on the overall calculation results of this early initiation of emergency feedwater is expected to be minor. s' At 5.3 s high pressure injection (HPI) was initiated when the hot leg pressure fell below 1.24 MPa (1815 psia). As defined by the se uence description, power to the repctor coolant pumps was tripped ofi and main feedwater was rerouted to the emergency feedwater headers at.35.3 s. Mass flow rates in the cold legs, between the HPI injection nozzles and the vessel are shown in Figures 23 through 26. The increased flow rates during the initial portion of the transient reflect the increased densities - associated with the icwering fluid temperatures calculated following the
' break of the main stean line. Af ter the reactor coolant pump power was tripped the cold leg f. low rates decreased in conjunction with the coastdown of pumps. Loop flow in all four cold legs continued following p;mp coastdown due to natural loop circulation set up by core heat addition and heat removal to the steam generator secondaries. Flow rates during this period were slightly larger in the affected loop cold legs than in the.9 of the unaffected loop because of superior cooling to the affected loop steam generator and because the HPI injection rates in the A loop were higher than in the B loop.
46
. 5000 , , , ,
n 2 N 1
' \O2500 -
5000 E d O v D o I C .3 0-- - O n 2 0 2 L
, , , -5000 0 1000 2000 3000 4000 Time (s)
Figure 23. Revised WSLS, cold leg A-2 mose . low rate of reactor vessel. 8000 . . . , i 15000 me 6000 - # 7 g N E m
.x
* -y J2 10000 O l 4000 -
3
= _a
$ 1- -S000 $
2000 2 0 O O 1000 2000 30C0 4000 l Time (s) Figure 24. Revised MSLS, co!d leg A-I mass flow rate of rocciar vessel. 4 I 47
4 l Ml"/ s . . 10000 e . m
- n \
\ -
" 2500.
- 5000
- _f D +
o .0 c = n ' n 0<- .-- 0 m. O O 2 i 2 . f. L
-5000
,g- , , ,
! 0 1000 2000 3000 40C0 Time (s) Figure 25. Revised MSLS, cold leg B-2 mess flow rate at reactor
! vessel
/
E g i f IS000 f I 6000 - - _ Q 7 - N N E o a I .s i -l 10000 0 3
= 2 n v.
n n O 5000 2 2000' - 0 O O 1000 2000 3000 4000 l Time (s) Figure 26. Revised WSL8, cold leg 8-1 mess non rate of reactor vessel l 48
-n--- --g--e ,-- -.-- - ---n y---u--+-- - - - -w -
= ,
The pressurizer collapsed liquid level, shown in Figure 27, initially fell to the bottom of the pressurizer due to the primary Tiquid volume shrinkage resulting from the cooldown. HPI injection, however, added liquid volume to the primary and the pressurizer level began to recover. The _ shrinkage also caused the formation of a large void in the upper head of the reactor vessel as shown in Figure 28 and smaller voids in the hot legs as shown in Figures 29 and 30. The first repressurization of the primary system, shown in Figure 19, resulted when the upper head and hot leg voids had been collapsed and continuing HPI flow started to compress the steam inside the pressurizer. Due to one dimensional modeling limitations the calculated collapse rate of reactor vessel upper head voids is likely overstated. When the upper head starts to refill, the liqu'd entering is cooler than that already there due to the accumulated effects of addition of cold HPI fluid and cooling to the steam generators. As 2 result, steam in the upper head is condensed causing a localized depressurization there which accelerates the refilling, causing more condensation. The process continues at a nonphysical, ever increasing rate until the upper head is refilled completely with liquid. The upper head void thus collapses too quickly and is accompanied by a rapid ( drop in primary system pressure such as indicated in Figure 19 at 600 s. Since the magnitude of the primary system pressure cecline is smail, the poorly calculated upper head refill behavior has only a minor effect on the overall results of this calculation. The main feedwater flew rates, calculated at the entrances to the steam generators, are shown on an expanded scale plot in Figure 31. In the
- unaffected loop, main feedwater delivery was ramped off smoothly over the -
first 3 s because the pressure in the steam generator secondary quickly exceeded that at the main feedwater pump discharge header. The flow decrease shown represents only the blowdown of fluid between the discharge header and the unaffected steam generator. In the affected loop, the main feedwater flow decreased during the first second because the main feedwater valves initially closed in response to the high secondary level indication previously discussed. As the effects of the break were propagated back to the feed train, however, the affected loop main feedwater flow increased dramatically , and remained at high levels with flashing occurring in the main feed lines until main feedwater realignment to the emergency feedwater headers at 35.3 s. 49
4 m . . . -
^
O 5 , TOP OF O a to - PRESSURIZER - a
$ 30 $
M b o o N a. 20 a. 4 $ . ~ 4 1 d d ' O O o BOTTOM OF PRESSURIZER 10 0 0 O O 1000 2000 3000 40C0 Time (s) Figure 27. Revised WSL9, pressurizer cotopsed tiquid level.
~
0.8 ,
, i i c
o 0.s - E o ah 2 0.4 - o 6 o
>I 0.2
, 0.0 I O 1000 2000 30C0 4000 Time (s) Figure 28. Revised W$L9, reactor vessel upper head void frac tiori. f 9 1 i 50 t
O.04 i + i c o 0.04 - ~ 8 3 0.04 - - o b o
$ 0.02 ~
g,g k t t 0 1000 2000 3000 4000 Time (s)- Figure 29. Revised W5L8 void fraction of top of offected loop hot
, l *9-C 7 10 , , .
c E o b 3 to - ~ o b o 0.0 O 1000 2000 3000 4000 Time (s) Figure 3w Revised W5LB void fraction of top of unoffected loop Mo t leg. 51
b 4 1500 , , , MFW REALIGNED TO ArrEcito . cop sc00 EFW HEADERS --- uwArrtcTED toop
.\
m 1000
- l1 N 2000 N
- I E 6 i V e
D .i
$ soo~ [ , -AFFECTED LOOP a
j
- y ECTED LOOP
- s 0 . ........................... , y s
.g'- ' t f r -10C0 0 20 40 50 80 10 0 ; Time (s)
- FIpere 31. Revised W5LB. moln f eedwater flo. rotes of S.C. entrcnces.
O i 52
I l 1 Figure 32 shows tne flowrates calculated in the emergency feedwater headers at toe entrances to the steam generators. After 3'5.3 s these , curves represent the combined flow rates due to emergency feedwater and main feedwater fed through the startup valves to the emergency feedwater
, headers. Before 35.3 s the curves represent only the emergency feedwater flow rates. Figure 33 shows the temperatures of the fluids entering the steam generators through the emergency feedwater headers. In the unaffected loop only emergency feedwater was delivered until about 80 s when the pressure in the unaffected secondary fell below the head of the condensate booster pump allowing main feedwater to be injected. At 320 s the main feedwater to the unaffected loop was terminated when the level had recovered to 6.1 m (240 in.). Motor-driven emergency feedwater to the unaffected secondary was terminated at 513 s when the hotwell surge tank
'was calculated to empty. In the affected loop, the secondary pressure was much lower than in the unaffected loop. This resulted in a much larger flow to the affected loop steam generator as shown in Figure 32. At 513 s the motor-driven emergency feedwater was also terminated to the affected loop steam generator, however full turbine-driven emergency feeawater flow
( continued until all feedwater was terminated at 600 s. Subsequent to the performance of this calculation, new information was obtained which indicated the emergency feedwater was improperly modeled in two ways. First, emergency feedwater control should have been based on startup levels rather than as calculated using operating levels. Second, the hotwell and hotwell surge tanks are isolated from each other by a j closed valve such that the only fluid removed from the surge tank is due to
- motor-driven emergency feedwater flow. The effect of using operating level instead of startup level is minor for this calculation as the levels closely track each other. Because the calculation assumed the hotwell to tank valve was open during the transient, an early surge tank emptying time was calculated and this resulted ir. termination of motor-driven emergency feedwater at 513 s. Had this been properly modeled, motor-driven emergency feedwater flow termination would have occurred at 600 s and this would have resulted in an estimated mixed downcomer fluid temperature about 5 K (9^F) lower than that calculated.
53
Soo i i i
- AFFECTED Loop M0TER DRIVEN --. cNArrterED Loop EFW TERMINATED . coo n 4oc -
l . O (es N E 6 j ALL MFW AND f g 2oo '- 4-EFW TERMINATED . 5M }o c ,' 0 i C 2 o 2.I l . . . . . p. . . . . .. . . . [.' -
.o y
/MFWREAL!GNED TO EFW HEADERS MFW TO UNAFFECTED LOOP TERMINATED
-2o0 o 2oo too 6Co 800 MCo Time (s)
Figure 32. Revised MSLS, f:owrotee at emergency f eedwater hoocer. S5o - -
, , _l 5 - AFFECTED LOOP Soo *b
--- CHAFFECTED LOOP s soo -
. u MOTOR EFW s 2 400
~
o MFW REALIGNED TO TERMINATED 5 EFW HEADERS 5 p g42 -
= ,
~
- 9 soo -
! m 2 4w - ! .
8 y MFW TO UNAFFECTfD j - k ,;l
= =
S.G.TERMINATEDI 200 83so - l -
- . \. E 3
2 : O ! ' C
...r...................
20 > soo o 2Co too Goo Time (s) j Figure 33. Revtsed W5LS, fluid temperatures in emergercy f eed hooders. l i l f
/
54 l
Figure 34 shows the Loop A-1 high press'>re injection (HPI) rate. The rate for Loop A-2 was identical and for Loops S-1 and B-2 'the trends were the same but the rates were lower due to a lower HPI capacity available to Loop B. The figure shows that from 5.3 s, when low hot leg pressure
, tripped the HPI on, until 300 s, when 42 K (75"F) hot leg subcooling was obtained, full HPI flow based on cold leg pressures was injected. A'fter 300 s HPI was throttled by applying a multiplier based on subcooling in the hot legs such that no HPI flow was allowed when subcooling exceeded 55 K (100 F) and full HPI was allowed when subcooling was less than 28 K (50*F). Between these extremes, 40". of full flow was allowed at 50 K (90*F) subcooling and 60". of full flow was allowed at 33 K (60*F) subcooling. By controlling in this way the operator action associated with throttling was approximated by assuming only minor throttling as long as subcooling remained well within range (33 to 50 K). Major throttling only occurred when subcooling approached either extreme (28 or 55 K).
Subsequent to perform:ng the calculation it was learned that, with reactor coolant pumps (RCPs) tripped the subcooling margin for throttling HPI is cased on core exit ter.peratures rather than hot leg temperatures. Furthermore, the subcaoling margin for restarting RCPs is based on the minimum of all hot an.1 cold leg subcoolings. The RELAPS calculation would not have been differe t if these different means of measuring subcooling , margins been employcd because, with loop flow continuing, core exit and hot leg fluid temperatures were virtually identical. j An initial calculation of the Oconee-1 main steam line break sequence l was reported in Reference 1. Subsequent to that calculation the sequence - description was modified by changing the reactor coolant pump restart criteria. Also subsequent to that calculation an analysis of the affected steam generator heat rnoval rate was performed which indicated the calculated rate was likely understated. The primary purposes of recalculating the main steam line break were to update the sequence description and obtain a more representative heat removal rate to the affected steam generator secundary, i In the original main steam line break calculation the combination of main and emergency feedwater injected at the top of the affected steam
~
generator boiler was totally bypassed to the break. As a result the heat 55
- , . . .- . . . = -
', 1 t , i 1 t i 20 , , 40 4
^
^ is - ~
THROTTLING
'30 l ) BEGINS E 6 -
2 ge - - y 1 = 20 .9
- $ SUBC00 LING EXCEEDS .
j s - l 56 K (100*F) AFTER; 0 '8
- I I THIS TIME t' 1 /
i 0 i lh , . 0 0 1000 2000 3000 40C0 Time (s) 1 Figure 34. Revised MSL8. loop A-1 high presst re injection flow rate. i i l l i j. 4
,A
'f 56 - j.
A B
~
removed from the' primary was limited to the uppermost volume of the boiler. As reported in Reference 6 the heat removal rate'was about 40 MW. Based on one-dimensional countercurrent flow limiting (CCFL) phenomena at the uppermost tube support plate which was reported in Reference 6, upward
. steam velocities associated with an 68 MW heat removal rate below the plate would allow an equal amount of liquid mass downflow as vapor mass upflow and therefore an 88 MW neat removal rate (including 20 MW above the plate) represented a steady state operating condition. The conclusion was that, in the original calculation, the heat removal rate was less than that expected based on one dimensional CCFL phenomena. Furthermore, if three-dimensional effects are considered, such as liquid downflow at the perimeter and vapor upflow at the center of the boiler, then the underprediction of the heat removal rate would be even more severe.
In the recalculation of the main steam line break a downflow of liquid was calculated throughout the boiler section of the affected steam generator when feedwater was injected at the emergency feedwater header. ( Figure 35 shows qualities at five locations in the affected steam generator secondary. From 35 to 600 s feedwater enters only through the emergency feeawater header at the top of the boiler section. The injection rate during this period is about 400 kg/s as shown in Figure 32 and the temperature of the injected fluid decreased during this period from about
- 440 K (332*F) to 405 K (270 F) as shown in Figure 33. At the operating i
pressure of the affected steam generator secondary, about 0.35 MPa (50 psia), the injected fluid temperature is about 28 K (50*F) superheated l at the beginning of the period and 6 K (10*F) subcooled at the end. ' Curve 1 in Figure 35 shows the uppermost cell of the boiler operated at I about 8*.' quality during the injection period. Vapor in this cell was produced by boiling the injected fluid on the tubes and flashing during periods when the injected fluid was superheated. At a pressure of 0.35 Mpa (50 psia) an 8*." quality corresponds to a void fraction of about 98*.'. Thus the void fractions in the affected steam generator boiler are very high during the injection period. l Figure 36 shows the vapor and liquid velocities at the uppermost tube support plate. This junction is between Cells 1 and 2 shown on Figure 35. l A vapor upflow velocity of about 15 m/s (50 f t/s) and a liquid downflow l t 57 r,,,,. ,,- .-~,. - -- - -
- - - . , , - - - - - - - , . - - - - - , ,,--e , . . , . , - . _ , - , - , - - - - - --
e e s E4 5 5 s) E FW w o3 [ Y G - STEAM 14-- I . LINE 8 L' 4-2 C.6 o
= (-3 2m 1 l
l ' 0.4 00WNCOMER p4 - g 3 o i (5
> 0.2 i. g p : A -
> a y 2
0 O 1000 2000 3000 40C0 Titae (s) Figure 35. ;4evised MSL9. cuallfles in off ected $.C. secondary boiler. 60 . , , _ 4
?. - VAPOR UPFLOW 7 50
--- LICul0 CONPFLCW .
, ! '4, 15 0 v
,e 40 - ' . .D_ ,
e N l u b 30-. . 10 0
> I w jd2 j#;
J
~
- '. Ta e l c
> l ' ,*, i ,o e !
, $' (
O-- ;.............................. .
""" F 0 ?
ALL FEEDWATER TERMINATED 4
- 10 O 200 400 600 800 1000 Time (s)
Figure 36. Revised wSLB phcsic velocities throug* uppermost tube . suppo rt pla t e o f a f f ec t ed S.G. . 58 m a m m m=he p ._
velocity of about 1 m/s (3 ft/s) are shown. The corresponding vapor mast upflow rate is about 43 kg/s (95 lbm/s) and the liquid dow'nflow rate is about 48 kg/s (105 lbm/s). As the liquid falls below -this junction some of it is vaporized un tubes located in the lower cells and this causes the
. increasing qualities from Curves 1 to 4 in Figure 35. The net mass downflow of about 4.5 kg/s (10 lbm/s) rt:sults in a pooling of liquid in the bottom cell of the boiler as indicated in Figure 35 by the decreasing quality in Cell 5. This pooling of liquid is also indicated by an increasing affected steam generator secondary mass during the injection period as shown in Figure 37.
As a result of the tnermal-hydraulic processes just described, heat was removed from the primary system to the affected steam generator secondary at rates significantly higher than in the original main steam line break calculation. This heat removal rate is shown in Figure 38. Note that the spike in heat transfer shortly after termination of all feeawater at 600 s was due to a blowout of the liquid pooled at the bottom of the boiler. This blowout is also observed in the tube support plate ( velocities in Figure 36 and in the boiler qualities in Figure 35. The rangs of the calculated affected steam generator heat removal rate was 80 to 180 MW during the injection period. Reference 6 indicated that a heat removal rate of 68 MW below the plate and 88 MW total is predicted for a steady-state downflow of liquid and upflow of vapor through the uppermost tube support plate. This heat removal rate was based on one-dimensional Kutateladze CCFL correlation. The RELAPS-calculated heat removal rate - generally exceeded that predicted using the CCFL correlation and this represents an improvement over the original main steam line break calculation in which the opposite was true. Fluid temperatures in the four cold legs are shown in Figures 39 and 40. Before 300 s, when the reactor coolant pumps were restarted in the A-1 and B-1 cold legs, the greater cooling by the affected steam generator caused the affected cold leg temperatures to decrease at a faster rate than those in the unaffected cold leg. After reactor coolant pump restart the cooldown rates of the two loops were nearly the same due to the better 59
l l 20000, , , e '.4000-- - 40000 se00c - -
^ e .
(o W.- " 300C0E\ 3 1000 - - v e g 10000 - - c -20000 o a000 . . N ALL FEE 0 WATER ~ a
$ TERMINATED iOOc0 2 E
2000 - 0' O O SCO 1000 1500 Time (s) Figure 37. Revised MSLS, off ected stocri generator secondary water moes. 400 . . . .- m
- MOTOR EFW TERMINATED
. .00 -
0 n. 0 - r 2 ; ,
- RC PUMPS j RESTARTED 'ALL FW lERMINATED
-200 -
C 1000 2000 30C0 4000 Time (s) Figure 35. Revised W5LS, off ected stocr generator hoot removal , r a t e. j 60
.c , . . . . _ _ , ,
. MO , , ,
a e M - A-2 COLD LEG b
--- A-1 COLD LEG e e
. 550 w
's 560 -- ;, 3 O O n-e w e
Q 540 - - Q. E E e 500
.e.
? 520 l- '
?
3 ' s
& e A 450
{ ', ' h500 s I '. o _ o g i 0 1000 2000 30C0 4000 Tim. (s) Figure 39. Revised WSLB, affected loop cold leg fluid temperatures of reactor vessel M0 , , , m ^ l
$ - B-2 CCL? LEG I 60C b B-l COLD LEG e 5sc - -
o , s u 2 O
.s.
! O
' 540 <
e
- 550 w e
S C-e E e
- 540 -
.e.
.o 500 y 3 s 7 520 - ~
q ~
- , =
e ' 430 e i E 500 s
.. - E
.s.
O O 480 ' O 1000 2000 3000 40CC
- Time (s)
Figure 40. Revised MSLB, unaf f ect ed loop cold leg fluid j temperatures of reoctor vessel I s 61 v - ,-
A fluid mixing associated with higher locp flows. After all feedwater was
- terminatedoat 600 s, fluid temperatures in all cold legs increased because the heat sink for the primary system was lost.
The fluid temperature in the reactor vessel downcomer at the elevation , of the first,circumferential weld below the cold leg nozzles is shown in Figure 41. In this calculation the reactor coolant pumps were restarted before loop natural circulation was lost. As a result, vent valve opening - was not calculated and the reactor vessel downcomer fluid temperature was ! simply a mass ' flow weighted average of the cold leg fluid temperatures shown in Figures 39 and 40. The minimum calculated downcomer fluid temperature wa s 494 K (429'F). This minimuir, in temperature occurred at 600 s, the tira at which all feedwater was terminated. 4 As shown in Figure 41, uncertainty bars, and a shifted mean, have been added to the RELAPS-calculated reactor vessel downcomer fluid temperature curve. A comprehensive study of the ur. certainties in the calculation is beyond the scope of this work. However, an estt;nate of uncertainty in the calculation was required by the analysts performing fracture-mechanics _, / calculations who will use the downcomer temperature-time history shown in Figure 41 as a boundary condition. Therefore a limited uncertainty estimate has been made based primarily on insights gained from the comparison of counterpart TRAC and RELAPS calculations of the main steam line break sequence presented in Appendix B. The uncertainty was addressed by first shifting the calculated data to - i account for the known feficiencies in the RELAp5 calculation as detailed here and in Appendix B This shifted mean appears as the dashed line on Figure 41 and represents a best estimate response which (a) compensates l directly for known deficiencies, and (b) uses an arithmetic average of I extremes where uncertainties in phenomena have been identified. The upper - and lower ends of the uncertainty bars were then constructed by (a) again
< compensating for the known deficiencies but (b) using the upper or lower extreme of uncertain phenomena behavior, s
62 i
. _ _ _- - ~. . . . . .
800 , , m a M v 600 y v e e < u u I s
- - n. - .s.
O o 500 b o '
, e E
<\
/
E
- 500
\
, . ,,, ( --
o "s - i 400 -3 3 7 q % , s g - o-
= W - =
l e 5 -- 8 o
.. .s00 -
o 400 8 0 1000 2000 3000 Time (s) Figure 41. Revised W5LB f1uld temperature In reactor vessel downcomer. 4 e 5 P e 63
. . ._, -- - - . _ - - .. _ _ _ . _ _ _ . ~ . _ . . . . _ - . _ _ . _ _ _ _ . , _ - . _ .-- _ -._ _ _ _ -
Different phenomena are observed early in the transient between the
-TRAC and 'RELAPS calculations as detailed in Appendix B. It w'as observed the TRAC cooldown rate exceeded that for RELAP5 up to the time of RCP trip and the time of RCP trip was at 51.2 s with TRAC and 35.3 s with RELAPS.
The conclusion,was that the RELAP5 calculation, with its prototypical PFW
~
pump trip, better represented the cooldown rate up to the time of RCP trip. However, there is an uncertainty in tne RCP trip time that was unresolved. The shifted mean (dashed line on Figure 41) assumes a RCP trip time of 43.25 s; the average of the TRAC and RELAPS trip times. The uncertainty bar upper and lower extremes during this period assume RCP trip times of 35.3 and 51.2 s, respectively. Both the mean and the extremes were generated using the RELAPS cooldown rate. Another difference between the TRAC and RELAPS calculations was the time of RCP restart. With TRAC this occurred at 526 s and with RELAPS at 300 s. This difference was caused by a combination of asymmetric hot leg fluid temperatures, calculated with TRAC but not with RELAP5, and different reactor vessel upper head flashing behavior. The conclusion of the comparison was that the asymmetric hot leg behavior calculated with TRAC , better represents the plant behavior than the calculation with RELAP5, but
- it is likely the RELAPS calculation better represents the upper head behavior. For the shifted mean (dashed line on Figure 41) it was assumed the RCP restart occurs at 450 s which is the average of the RELAPS calculated time (300 s) and the latest time at which the minimum downcomer temperature is sensitive to the RCP restart (600 s). The uncertainty bar upper extreme assumes the RCP restart is at 300 s and the lower extreme -
assumes it is at 600 s. Between the times of RCP trip and restart Appendix B identifies different affected steam generator heat removal rates between TRAC and RELAP5. While some of the difference could be explained, most of it was found to be due to uncertainty in the heat transfer processes calculated above the uppermost tube support plate during periods when combined MFW and EFW were injected through the EFW header. Furthermore, a sensitivity to the modeling of the height of the uppermost boiler section calculational cell was identified and this affected both calculations To account for 64 _Q
these differences an average of adjusted TRAC-related and adjusted RELAPS-related heat removal rates above was used to genera'te the shiftec mean between the times of RCP trip and restart. The TRAC and RELAP5 heat removal rates above the plate were adjusted to compensate for the discrepancies between calculational cell heights and the distance from the lower surface of the upper tubesheet and the upper tube support plate. The kELAPS rate was further adjusted for the RELAPS vapor superheat problem discussed in Appendix B. The adjusted heat removal rates above the plate were calculated as: 95 MW for TRAC, 10 MW for RELAPS for a 53 MW average. To these amounts were added the bist-estimate below-the-plate heat removal rate of 70 MW based on Reference 6. The shifted mean thus assumes a total affected steam generator heat removal rate of 123 MW while the upper uncertainty bar limits assume 80 MW and the lower uncertainty bar limits assume 165 MW. It is noted that the uncertainty bars include the effects of different above-the plate heat removal rates between the two code calculations. T'e bars do not, however, account for possible multi-dimensional effects of liquid downflow and vapor upflow at the i
\
uppermost tube support plate. Multi-dimensional effects would likely increase the affected steam generator heat removal rate over that predicted using one-dimensional correlations such as used in Reference 6. The minimum temperature reached by the shifted mean, best estimate dashed line on Figure 41 is 462 K (372*F). The lowest temperature included within the uncertainty bars is 415 K (287*F). Figure 42 shows the calculated fluid pressure in the reactor vessel - cowncomer at the elevation of the first circumferential weld below the cold leg nozzles. Uncertainty bars and a shifted mean have been added to this figure in a similar 5. inner as they were added to Figure 41. The primary uncertainty just discussed for the downcomer temperature response which ciso affects the pressure response is the reactor vessel upper head flashing phenomena. The snifted mean, which appears as the dashed line in Figure 42 represents the pressure response expected if upper head flasning phenomena were the average of those calculated with TRAC and RELAP5. The extremes of the uncertainty bars represent the results of the calculations themselves. 65
. . = . _ . . .. .. . . ~ . - .. ~ _ .__ . .~ _ _ _ - . . . - . . _-..
f i I
. E i
i l i i 20 , . 3 2500 E g n i V W is e e m 2000 g to # . m w < e 5 / u Q. i. / ch 10 ' s' -' 00 Y f C s o f - __
, - 1000 5
O 1000 2000 3000 Time (s) Figure 42, Revloed he$LS, nuld pressure in reactor vessel downeemer. l l l k I t
/
a n/ i 4 G6
.,--r_..-.,.s., ,. _ - . , . . , , . , , , , , , , _ , _ , _ , _ _ , , _ . , . _ , , . , , - , ,,-,my-. --_-
Figures 43, 44 and 45 provide extrapolations of reactor vessel downcomer pressure, fluid temperature, and inside surface " heat transfer coefficient from the end of the calculation at 2697 to 7200s. The dc.ncomer pressure was estimated to continue to be controlled by the power
, operated relief valve throughout the extrapolated period. The downcomer fluid temperature was estimated to continue increasing until it reached 571 K (568"F) and remain near that temperature for the remainder of the extrapolated period. Above this temperature the subcooling margin would be less than 56 K (100*F) and the injection of cold HPI fluid would be recommenced, thus preventing the primary fluid temperature from increasing further. The heat transfer coefficient was estimated to continue drifting lower until the fluid temperature stabilized then remain constant thereafter.
4.4 Conclusions The calculated minimum reactor vessel downcomer mixed fluid temperature was 494 K (429*F) and the calculated maximum subsequent fluid ( pressure was 16.99 MPa (2465 psia). The calculation of downcomer fluid temperature was particularly sensitive to the requirement in the scersrio to restart reactor coolant pumps upon attaining 42 K (75*F) subcooling in both hot legs. This requirement was met at 300 s into the transient and the restarting of the pumps caused the minimum downcomer temperature to be much warmer than if the pumps had not been rcstarted. Since the attainment of subcooling is - sensitive to hot leg asymmetry and because a one-dimensional computer model will not predict this asymmetry, the timing of the pump restart is uncertain. The effect of this and other specific uncertainties on minimum d0wncomer fluid temperature has been estimated to reduce the best estimate minimum downcomer fluid temperature to 462 K (372 F) with a lower uncertainty bound of 415 K (287*F). 67
I i % I. . 1 4 3 A
- CALCULATED
--- EXTRAPOLATED 2soo g .................................. ,
n n O - . O
- a. 'SC00 -
.s 2000 S l e e w w 3 s a e e =
e e w socoo . .-1s00 6-Q. G. 1000 M00 O 2000 4000 6000 8000 Time (s) . Figure 43. Revised W$LS. er..coolated pressure in R.V. downcomer.
}
soc , , ,
/
n a 6 - CALCULATED 400 I EXTRAPOLATED
- 540 w
- - e
.s. ............................ 3 o ..- g g se0 -- ,-
,. _.sso .
e o. E
~ S@ ~ -
v 500 y 3 - . l ,g S20 - - S
- =
! ,43o e
- E
.$. m ~m o "
.440 O 2000 4000 6000 8000 Time (s)
Figure 44. Revised WSLB. entropolated fluid temperature in R.V. downcomer. 1 _s 68
i s t 30 * . ,
'5000
/ H C4LCULATED . F * ** EXTR APOLATED c 25 - - c l e e 4000 g^ CY O 20 - - e eQ ai oI u ,, ..~' " ~ ------------- - - --- * *** " ": uE 15 - - u eu: i (\ .oN m
- s 3
a .w b 10 - _ 2000 f)- u
- "O e-C 3 ~ _ 1000 $
J
~
z O O 0 2000 4000 6000 5000 l! Time (s) Figure 45. Revised WSLS. entropolated heat transfer coeffletent inside surf ace of reactor <essel downcomer won. 6 9 e
. 69
. =.
5. T MAXIMUM SUSTAINABLE CVCRFEED TRANSIENT ) The following subsections contain the maximum sustainable overfeed transient scenario description, modeling changes effected to perform the transient calculation, detailed analysis of the transient results, and conclusions drawn from the analysis. 5.1 Transient Scenario Description A description of the maxinium sustainable overfeed sequence analyzed appears in Table 13. This sequence definition was developed at Oak Ricge National Lrboratory. The transient was initiated from full power steady state cond t. ions (nominal temperature and pressure). The pressurizer heaters and sprav operate as designed. The transient was initiated by a turbine trip which tripped the reactor and closed the turbine stop valves. Decay hett was astumed to be at the ANS standard rate. Instead of running back flow, the main feedwater system continued to supply the maximum sustainable fl:.w to _, both steam generators without a trip from the component protection faature to the main feedwater pumps. It was assumed the main feedwater pump trip on steam generator high level failed to occur allowing both steam - generators to completely fill. 5.2 Model Changes r The basic RELAPS model used to perform the maximum sustainable [ overfeed transient calculation is described in Section 2. The following changes were made to the integrated control system model to initiate the transient. The steam generator nigh level signal to trip main feedwa,er pumps was disengaged. The steam generator low level signal controller was set at a maximum. This maximum signal overrode the Btu cross-limit and neutron power cross-limit signal to control the start up and main feed !' - valves in the feed train. The signal held these valves full open to orovide the maximum possible flow through the feed train. i 70 1
, _ - . . .-m_,, _ _ _ _ _ , _ , . , _ , . _ . - _ . . . , - . , . ~ , , . . - _ . . . . ____.,-m, , , , _.m . . - , .,
TABLE 13. MAXIMUM SUSTAINABLE OVERFEED TRANSIENT SCENARIO
- 1. Reactor trip, turbine trip, turbine step valves 0 4se.
- 2. Turbine bypass valves and safety relief valves open and reset at setpoints.
- 3. Instead of running back, main feedwater system continued to supply
. maximum sustainable flow to both steam-generators.'
4 High pressure injection actuates at setpoint, and operator trips all 4 reactor coolant pumps 30 s after high pressure injection actuation (per procedures).
- 5. Main feedwater pump trip on steam generator high level fails to occur; both steam generators fill completely.
- 6. Core flood tanks and low pressure injection system fuction as designed.D
- 7. Tt.:rbine bypass valves close on loss of condenser vacuum, at its setpoint.
( 8. Main feedwater and turbine-driven emergency feedwater pumps trip on low steam quality.
- 9. Emergency feedwater system functions as designed.
- a. To be determined by trial and error or ar, automated search for maximum main feedwater flow that will not result in *. rip of main feedwater pumps due to suction pressure or high discharge pressure limits being exceeded.
- b. Event may or may not occur, depending on phenomena encountered.
l
- l 71 l
i
s Following reactor trip, the Btu cross-limit and neutron cross-limit signal normally run back the main feedwater pump speed to a [ninimum value of 447 rad /s (4270 rpm). In the RELAP5 analysis, that minimum value was inc. eased, by trial and error, to 490 rad /s (4675 rpm) to obtain the maximum sustainable main feedwater flow without tripping the main feedwater pumps on low suction er high discharge pressu.e. At this pump speed, the suction pressure steadily decreased toward the low suction pressure trip setpoint. The main feedwater pumps tripped eventually en low void in the ' steam line following complete refill of the steam generator secondaries. 5.3 Transient Result _s This section presents the results of the steam generator overfeed transient with maximum sustainable feed flow. A sequence of events for the transient is presented in Table 14. The initiating event was a turbine trip, wnich tripped the reactor and closed the turbine stop valves over a 1 s period. Upon closure of the turoine stop valves, the secondary pressure, shown in Figure-46, rapidly . increased to the turbine bypass and safety relief valve setpoints which stabilized the pressure. The rapid secondary system depressurization resulted'in a cooldown and consequently a depressurization of the primary system as shown in Figure 47. At approximately 25 and 28 s the secondary side pressure for Steam Generators 8 and A, respectively, dropped below the safety relief and turbine bypass setpoints, closing the valves as shown by the mass flow rates in Figures 48 and 49. Also at this time the primary Ifquid temperature approached the secondary side saturation temperature and primary to secondary heat transfer was reduced to a minimum, resulting in the decrease of the primary side depressurization rate shown in Figure 47. Between 28 and 300 s, orimary and secondary pressures were strongly influenced by the turbine bypass valve flow. As the generators filled, primary to secondary heat transfer was enhanced as shown in Figure 50, causing steam generation and steam flow out the bypass valve. At 150 s the boiler region of the steam generator secondaries had become liquid full 72
a 7.5 , , , Reactor Coolant Pumps - STEAM oENERAToR A 8"*" GE"E"8 0" ' g , Trip feadwater realigned-" ,,o3a g
~
g I g i in Feedwater Pumps Trip
~
h r i . O h, f[ l' 1000 ;
; Turbine,4 ,! ;
2 Bypase ij ,. g o- Valves 4 ;,..~..._,_._._._. o. e - sso , E e.s ".Close i E 2 2 O O
> . . goo Condensation effects as Steam Line Fills 8
o Soo 1000 15 0o 2000 rim. (s) Figure 46. Waxirrom sustainable overf eed. S.o. secondary pressures. 18 , , ,
- Loop A --
2500
-- Loop 9
^ K _ T 2 "I Hign Pressure Injection Power Operated 7 3 onttiation Relief Valve ,S:
e Set Point ,
'u.
j Vessel Upper Wand - 2000 $ tegins to Void e
- E l
i 1 12 - Reactor Coolant o.
. Pumps Trip Condensation Effects frem ,
E Pressurizer Filling E 2 / l j io . - f -.isoo 2 8 Vessel Upper Head Void Collapsed a a 500 1000 tsoo 2 coo l Time (s) Figure 47. Mcximam sustaleable overf eed, hot leg pressures. f 1 73 j i
t l M 4
- 4 STEAM CENERATCR A 600 STEAW CENERATOR 8 O 500 7 -
.l N
\ 200 -
w r400 w2 E B o D C 300 o
=
E *
~-200 2
2 10 0 0 O O 500 1000 1500 2000 Time (S) Figure 48. Woulmum sustelnoble overf eed. S.O. saf ety relief volve fl o ws. 500 . 4 s - I
- i - STEAM GENERATOR A 0 00 l -- STEAM GENERATOR 0 v.
400 - p " m 7 i 800 ( N I
= E
.x
- 300 - 4 - o l
[ D
') 600
- l C I D C I o
200 _ C i M i 4C0
- M gg C yn 2 o 1 2 10 0 1
~
i 200 h 1 0 i O O 500 '000 1500 2000 i Time (s) Figure 49. .ooxtroom sustainable overf eed, turbine bypcss flows. . l s l 74 _ - - _ _ --_-_ _m . . -
b s g , , ,
- STEAM GENERATOR A m
-- STEAM GENERATOR 8
*0.00 "O. ^p 3 \
v 3 0 .
.- 0.00 S,
-10.0090*mI 3
. 5 - O g -
- 2.00 *10*
. o.
i la -n . .
- 3.00 *10' E e
\ q E
> - 4.00 *10' 3 o
,g. i ' i
-S.CO*10' i 0 500 1000 1500 2000 Time (s)
Figure S0. Womimum sustainocle overf ood. S.G. hoot removoble roles. l O l i I 75
TABLE 14. MAXIMUM SUSTAINABLE OVERFEED TRANSIENT SEQUENCE OF EVENTS Time Event (s) Turbine trip, reactor trip, turbine stop valves close, main 0.0 feedwater pumps run back to 490 rad /s (4675 rpm). Secondary - pressure increased to turbine bypass and safety relief valve setpoints. Feed train , heater drains began to close. Pressurizer began to drain. , Heater drains closed. 5.0 Steam Generator B seconaary pressure dropped below safety reitef 25.0 valve setpoint, valves close. Steam Generator A secondary pressure dropped below safety relief 28.0 valve setpoint, valves close. Pressurizer liquid level oropped below heater cutoff setroint; 35.0 heaters latched off. Steam Generator B secordary completely liquid full and steam line 240.0 commenced to fill. Steam Generator A seconcary completely liquid full and steam line 250.0 commenced to fill. _,/ Primary pressure droppeo below high pressure injection setpoint and 269.0 high pressure injection flow was initiated. Vessel upper head began to void. 275.0 Reactor coolant pumps trip; feed flow realigned to EFW header; main 299.0 feed valves close. , l Steam Generator B turbine bypass valves close. 341.0 , i Steam Generator A turbine bypass valves close. 350.0 t Void in Steam Line B dropped below 20%, main feedwater pumps trip 417.0 off. Void in vessel upper head collapsed. 434.0 Pressurizer began to refill. 448.0 - Primary system began to repressurize. 524.0 l l l . 76
TABLE 14. (continued) Time Event (s)
, Steam Generator A secondary pressure reached turbine bypass 930.0 valve setpoint and valves crack open.
Primary pressure reached power operated relief valve opening 1054.0 setpoint and cycles around setpoint. Pressurizer becomes liquid solid. 1079.0 Steam Generator 8 secondary pressure reached turbine bypass valve 1424.0 setpoint and valves crack open. Transient terminated; system pressure stabilized at power operated 1695.0 relief valve opening setpoint. Primary temperature was slowly increasing.
\
~
l. 4 l i' m 77 e m -,-q -- - e g
s s and the steam lines began filling. At 180 s the volumetric liquid inflows to-the steam generators from the feed line exceeded the volumetric outflow through the turbine bypass valves and the secondary pressures increased as shown in Figure 46. Shortly thereafter low quality steam began exiting the turbine bypass valves, the valve mass flow rates increased significantly as shown i'n Cigure 49, and the secondary pressures stabilized. This higher
~
mass flow rate enhanced primary to secondary heat transfer, further cooling the primary system as shown in Figures 51 and 52. The increased cooling of - the primary side caused the liquid inventory to shrink as illustrated by the decrease in pressurizer liquid level shown in Figure 53 and increased the depressurization rate as shown in Figure 47. At 269 s the primary pressure had dropped below the high pressure injection (HPI) initiation setpoint and cold HPI liquid flow to the cold legs was initiated as shown in Figures 54 and 55. The introduction of the cold HPI fluid further increased the depressurization rate in the primary system as shown in Figure 47. At 275 s the primary pressure hac dropped below the saturation temperature in the vessel upper head and voiding of , the upper head began. The primary depressurization rate decreased due to ,, ) the volumetric expansion of the steam generated in the upper head. Per the scenario description, the reactor coolant pumps were tripped at 299 s, 30 s after the initiation of HPI. Realignment of the main feedwater flow to the emergency feedwater headers was coincident with the reactor coolant pump trip. When the realignment occurred, cold fluid that had been in the crossover lines was pushed into the steam generators - resulting in a slight momentary increase in the primary to secondary heat transfer rate shown in Figure 50 and a sharp local decrease in liquid temperature at the point of injection into the secondary side shown in Figuce 56. The increased primary to secondary heat transfer rate, coupled with the establishment of natural circulation in the primary system (see Figure 57), resulted in better cooling of the primary liquid as it passed through the steam generator as shown in Figures 51 and 52. Also, as a result of injecting the cold liquid into the secondary, the secondary pressure decreased as shown in Figure 46. 78 e
800 , , . a ^
- w.
* - INLET b
-- CUTLET 600 *
- e
. 6 3
58C + - s 3 O w
,3 c
- a e ft e a 560 '-' s - S50 c.
E E e i e
*= \ =
? SM -
-I 2
1 u 500 3 c' i
'. \ '
l520 p 3 3 O o 450 y 500 C 500 1000 1500 2000 Time (s) FI;are St Wasimum sustsinable overf eed, steem geisrator A PRlWARY SIDE INLET AhD OUTLET FLUID TD#ERATURES. { s 600 . , n a M
! INLET b
, -. OUTLET 600 v e i e
w 3 S80 qr
- 6 ,
3 O L:,
- O w -
g e \ e Q. 340 - .'
- 550 c.
E E e \, e ! \ 3 S40 -
\
3 3 o
- 7
\ 500 c
! '\, - l520 g 3 3 l
- o o 450 >
l COO l 0 500 1000 1500 20C0 l
. Time (s)
Figure $2. hianirrum sustolnoole overf eed, steam generator B l primary side intet and ouHet fluid terrporature. l i ( 79 S a r- - rm - --
.~- . . . - . . . . -
1S , , , . 4 40 v 10 - - O - 30 -
- e e
T.
. 20 2 3 s e 5 -
-e J J 10 0 O O SCO 1000 ?SCO 2000 Time (s)
Figure $3. Wonin un sustainable overf eed, pressuriser casopsed IIquid level 1 i l 20 , , , ,
, - COLD LEG Al .e g
- COLD LEG A2 m is - -
7 (o 3o N. E 6 e v 3
-$W 20 o m
o E 2 O S.- -. go 2 0 O O $00 1000 1500 2000 Time (s) Figure 54. Wowimum sustainable overf eed, !=:;: A highpressare
- Inloction rotos.
.?
4 80
] 15 . . . LOOP B1 3o
-+ LOOP 82 25 2 N
) to
.x 20 4 E
O 8 y 2
- 15
_o S." ".io E 2 5 3 0 O O 500 1000 1500 20C0 Time (s) Figure 55. Wasimum suefoinable overteed. loop 8 high pre:ssure
, inlection rotes.
600 , , . n ^ M - STEAM GO ERAT 04 A 600 ad-STEAM GDERATCA 8 e e w 6 s 550 .s-0 0
* $50 - - 6 e e 9- Q-e 500 E*
F
= ,
'5 -
f e50 ~5 . 7 %0 - ~
- \ .c 1
e - e b o
.'i 2 o
450 O 500 1000 1500 2000
. Time (s)
Figure 56. Maximum susicinchie overf eed. 13 quid temperotures ai top of stoom generator boilert. 81
i t l-f l - M i , , i
- LCOP A 200C0 2 -- LCO2 B j 8000 - .,
m m ! es a
\ 15000c\
m
$ 6000 -
_2 , 3 '
+
0 y l C 10000
, _o a
n 0 $ 2 0 acoo '. _ S000 3 > i 0 '
-0 0 500 1000 1500 2000 Time (s)
Figure $7. Womimum sustof noole overf eed, het leg mass flow rates. l ( l S 6
)
82 l
At 341 and 350 s the turbine bypass valves in Steam Generator B and A, respectively, had closed because the secondary pressure ha'd decreased below the valve setpoint. With the valves closed, the steam line rapidly filled with liquid due to the incoming feed flow. At 417 s the B steam generator
. steam line had filled enough to trip the main feedwater pumps on low steam line void, isolating the steam generators. The steam lines are described by four volumes in the RELAPS model. At the time of the main feedwater pump trip, the void in the third volume of the B steam generator secondary was about 20*4, while the void in the A steam generator steam line was about 50*4 as shown in Figures 58 and 59. Thermal expansion of the liquid in the secondary further decreased the void in these volumes in the steam lines.
In the A steam generator steam line the expansion of the liquid compressed the steam which resulted in an increase in secondary pressure as shown in Figure 46. The expansion of the liquid in the B steam generator steam line condensed the steam which resulted in a slight pressure decrease as shown in Figure 46. At 930 s the pressure in the A steam generator reached the turbine bypass valve setpoint, the valves cracked open, and the pressure remained at the setpoint for the duration of the transient. It was not ( until 1100 s that the pressure in the B steam generator began to increase. At this time the third volume in tne steam line had completely filled and filling the fourth volume began. Condensation ceased and the pressure increased due to steam compression. At 1424 s the secondary pressure in the B steam generator reached the turbine bypass valve setpoint, the valves cracked open and the pressure remained at the setpoint for the duration of the transient. During the period of bubble formation in the vessel upper head, liquid was forced out of that volume, due to the volumetric expansion of the steam generated in the upper heiad. At 3PJ s the volumetric hPI inflow rate exceeded the volumetric shrinkage rate of the primary system liquid and the system pressure slightly increased. Also at this time cooler liquid entered the upper head due to the pressure increase from HPI inflow, and the bubble size began to decrease as shown in Figure 60. When the cooler liquid entered the upper head, the model overstated the condensation rate. The pressure in the upper head locally decreased, resulting in an increase in the mass flow rate into the upper head such that it exceeded the HPI 83
1P g -C Q' ,0 y ,
- STEAM LINE VOLOWE 1
-'. -- STEAM LINE YOLUWE 2 l C STEAM LINE VOLLWE 3
$ i O STEAW LINE voltWE 4 ,
=y i.
O 1 e i . 3o o.s -
!g I
i g i Main Feedwater Pumps Trip
> 1 i.
o 0 (! ' 300 2': 1000 C O C
'SCO 2000 Time (s) rigure 58. Wazircum sustatnoble overf eed, steam line B void fractions.
1 C C, -; ; ; ; ; ; ; ;- ; j
.O.,
.g - STEAW LINE VOLUWE 1 STEAW LINE VOLUWE 2 3 O STEAW LINE V0'.UWE 3 U O STEAM LINE YOLOWE 4 3 t-u I Q
? 0.3 ~
3 I
.. cc: a c ca L
o !
> i
.1 0
0 500 t000 1500 2000
. Time (s)
Figure 59. Maximum sustaincble overf eed, stoom line A void . fr ac tions. 84 _ _ --p
0.3 c l c i
=
y 0.2 - l - s t o (
\ $
c. 41 - I o l l 0.0 --l * ' l O SCO 1000 15C0 20C0 Figure 60. Time (s) Woximum sustainchte overf eed, reactor vessel wpper head. 0 9 85
flow rate and reversed the flow within the pressurizer surge line as shown in Figure 61. The overstated condansattor, rate resulted in 'a reduction in primary pressure as shown in Figure 47. The incoming liquid in the upper head was heated to saturation which stopped the condensation and the vapor generation was reestablished as shown in Figure 62. Liquid was then forced , back out of the upper head and liquid flow reestablished into the pressurizer surge line as shown in Figure 61. At 395 s the vapor generation rate in the vessel upper head had decreased enough that liquid - again flowed into the upper head and the cycle was rapeated. This behavior was repeated six times before the bubble in the upper head had collapsed completely. After the bubble in the vessel upper head collapsed, all of the incoming HPI flow was directed toward the pressurizer surge line and the pressurizer. At 524 s the primary system began to repressurize due to the compression of the steam in the pressurizer. Again, the pressure response was erratic due to condensation and vapor generation effects as the pressurizer filled (see Figure 47). At 1054 s the pressure had reached the , power operated relief valve operating setpoint and commenced cycling around j that setpoint. At 1079 s the pressurizer became liquid solid and the pressurizer continued to cycle around the power operated relief valve setpoint. At the termination of the calculation (1695 s) the primary system pressure was established at the power operated relief valve setpoint and the system temperatures were slowly increasing. 5.4 Conclusions - At the termination of the calculation, system behavior had been established such that extrapolation of the dowrcomer pressure, temperature - and vessel inside wall heat transfer coefficient to 7200 s was possible. as shown in Figures 63, 64, and 65. The icwest downcomer liquid temperature - in this time frame was 500 K (440 F). Subsequent to this minimum temperature the maximum downcomer pressure was 16.99 MDa (2465 psia). 86
. 300 ,
- tJPPER HEAD Flow 600 PltES$UlllHR FLOW n 200 - -
7 (cm i 400 N E d ! 200 2o
-!5 s A.
is A i t,, / '.\ .- - Q , , j\ j !~' ! o ; .- - 3 0<t ,,. ... - "'~ g s . ',f ' gl 0 5 t ,t 6 y t ,e g su Ig t,' 1.
,g < , N -200 300 400 SCO Time (s)
Figure 61. localmum suetalcoble overteed. moss flow rates into reactor vessel upper hood and pressurtzer surge !!ne. l 2 , 0.1 e e o 1 - - o 6 . b n I c$ ce 6 0 .- 0.0 E g 6s
.N c
e0 t-cE e mm v g 04-6
) 6 a
o
-i -
l 1 I a o 4.1 2 L 300 400 500 Time (s) Figure 62. uculmurn sustolnoble overf eed, vapor generoflon rate in reactor vessel 6pper Neod. l l , 87 i l l
N
! I I I 2500 m ^
- 4 "
- CALCULATED 1 2 ,
I w
-- EXTR &PCL A TE0 E v
- e 14 2000 $n
- 69
- j e 6- w L t2 . . 1
- e E E 3 s O
1500 -
>g . -
o
,, ! , i i , i 0 1200 2400 3000 4400 6000 7200 Time (s)
Figu.e 63. Woulmum sustelnoble overf eed reactor vessel downcomer flutd pressure. Y S40 , . . . , n ^ M
* - CALCULATED l'-
-- EXTRAPOLATED
- e
- 6. - 550 s.
- 3. Soo - 3 O o n- 6 e e .
g gS40 500
.e .e.
2 520 -
*~'~'I 3 ,
=
- g. . ... . -
=
3 450
' S00 - -
b
- .2 O o 440 0 f200 2400 3400 4800 6000 7:00 Time (s)
Figure 64. Woxim6m sustolnocle overf eed. reactor vesset downcorae r . fluid temperature. e 88
4 4 5 4
- CALCUt. ATED
~* "XTRAPOLATED
- 80C0 c e, e -
g 30 - - on I % Eo'
~m o
# oI (M 6 4000 01 ~
sC 20 - -
'I
*h na
-N sv a
B3 s a h e 2000 -O O O e I Z'
, ~.-- - -.~. . .-. . . . . . ,._._._.,
0 0 0 1200 2400 3600 4800 6000 7200 Time (s) Figure 65. Womimum sustainchie overf eed, reactor vessel dowecomer inside surfoce heat transf er coe fficien t. e l 89
- 6. TURBINE BYPASS VALVE FAILURE AT REACTOR HOT STANDBY The following subsections contain the Transient Scenario Description, modeling c5anges effected to perform this calculation, detailed analysis of the transient _results, and conclusions drawn from the analysis.
6.1 Transient Scenario Description A description of the sequence analyzed appears in Table 15. This sequence was developed at Oak Ridge National Laboratory. The transient was initiated with the reactor at hot standby conditions. - At zero time all four ~ turbine bypass valves (two on each steam line) are assumed to fail open. All automatic plant systems are assumed to be operative. Operator actions are assumed for tripping off reactor coolant pump power 30 s after high pressure injection begins, and closing ! turbine bypass block valves at 10 min. 6.2 Model Chances ,/ r The basic RELAPS model used to perform the turbine bypass valve failure at reactor hot standby calculation is described in Section 2. To perform this calculation, the basic model was modified, a RELAPS hot ! standby steady-state was run, and the transient calculation was initiated from the new steady-state conditions. This section describes these modifications and the hot standby steady state conditions obtained. - The model nodalization used for the turbine bypass valve failure at [ i reactor hot standby (TBF/HS) calculation is similar to that used in the - other calculations described in this report, with the exception of the feed train model. The nodalization of the feed train is shown in Figures 66, 67 ~ ' and 68. The heater drain sources indicated in Figure 66, components 742, 744, 748 and 750, were deleted because the turbine is not rolled at het standby, s 90
4 i r .
. e DELETED j 742 744 748 750: ;
4 ................................. 752 : 740i 730 734 736 l. 738 746 ! E i s
* . r r 1 7 1 1 ? 1 757 : .
(754 ; 756 N . 764 l. -
................) DELETED .! 763 768
. ' ' ~ '
762: i i s 758 760 761 %. - 766 l. - 6................................................ Figure 66. Hot stanby model, main feedwater train. 91
m 4
,ti 4
s 805 81 2 804 81 3 864 853 N 4 t . iN . , 862 . l l 325-1 8 51 : l DELETED.;
)
CLOSED l T7 ; r y l ll [ 850 ; l 770 '
..... ............)
W 108 774 7,,,,,,y,,,,,, 771 775 . - 782 X X 778-1 ! 775-2 ! 778-3 N -
- j VWV '
305-1 Figure 67. Hot stanby model. Steam Generator A feedwater header. 92 9
*- ,m- -w-- - -- ,- - - - - - - - - ~-
e, -- , - - ,, --,--,-,,,m ,,-,,w-- , - - - , , ,,
8 905 91 2 904 91 3 954 953 N N. 962 l 425-1 l : l 951 :
!. DELETEDj
(' : l CLOSED l T7 l l 960 l l 7 ! ! 772 '.................s X 768 776 ,,,,,,,77,7,y4 773 777 . 784 , X X 780-1 f780-2 780-3 .N , W ' I 405-1 Figure 68. Hot stanby model. Steam Generator B feedwater header.
~
93
E - = - A - X TABLE 15. TURBINE BYPASS VALVE FAILURE AT REACTOR HOT STANDBY TRANSIENT SCENARIO - Initial Conditions:
- 1. Hot standby conditions [T,y,'= 532 F, 2155 psig at top of hot leg riser, 1 MFW pump, 1 condensate booster pump, I hotwell pump running]
- 2. Four RCPs running
- 3. Decay heat: 9 MW 4 No steam supply to FW heaters
- 5. FW manually controlled (FW control portion of ICS deactivated)
- 6. FW supplied through startup line (no flow through main line)
Sequence of Events:
- 1. All 4 TBVs suddenly fail full open
- 2. MFW pump running at 45% speed and startup valve controlling on steem generator level
- 3. HPI initiates j
- 4. Operator trips all 4 RCPs 30 s after HPI initiation
- 5. Steam generator level set point changes to 240 in.
- 6. FW system fails to realign to auxiliary header
- 7. -EFW system actuates a
- 8. Core flood tanks inject"
- 9. PORV, SRVs open and close at their set points
- 10. LPI actuates (not likely, as RCS pressure would have to fall to 180 psi)a
- 11. Operator closes all TBV block valves 10 min. into the transient
- 12. No further operator actions -
e 94
~
TABLE 15. (continued) Failures:
- 1. Four TBVs fail open
- 2. FW system fails to realign to auxiliary header
- 3. Operator fails to throttle HPI and restart RCPs (procedural violation) 4 Basis for Transient Selection:
Determine possible system response differences resulting from transient initiation from hot s:andby, rather than fall power, conditions (FW system alignment is differe:it, most of ICS is deactivated, reactor temperatures / pressures are sligntly different).
- a. Event may or may not occur, depending on phenomena encountered.
1 i l l 95 t
which means there is no steam being supplied to the feedwater heater secondaries. The* deletion of these components also means th'ere are no heat sources for the various feedwater heater heat structures. The deleted components between Components 752 and 763, shown in Figure 66, were removed from the base model. This was done to reduce the site of the model and is justified because both feedwater pumps are controlled by the same control variable. Hence, no differences in operation between the two pumps Components 754 and 760, are expected. Appropriate conditions of Component 754 were doubled to account for this deletion and the check valve, Component 756/757, was removed.
'The components linking the main feedwater headers with the auxiliary feedwater headers, as shown in Figures 67 and 68, were removed as specified in the accident scenario. Additionally, the main feedwater valves, Components 770 (Figure 67) and 772 (Figure 68) were latched closed according to hot standby operating procedure.
Operation of the startup valves, Comoonents 771 (Figure 67) and 773 ,- I (Figure 68), was controlled cased on the corresponding steam generator
- startup level error, instead of control based on the plant Integrated Control System (ICS). This also represents hot standby operating procedure. The setpoint level in the steam generators was set at 0.91 m (36 inches) of indicated startup level.
The RELAP5 hot standby steady state calculation was performed with one ' hotwell and one condensate booster pump operating. One main feedwater pump was operating at 45% of rated speed. Manual feedwater control was assumed with the operator controlling the startup valves based on the startup level. The level control portion of the integrated control system was deactivated. All steam exited the secondary system through the turbine , bypass valves. Table 16 shows the not standby initial conditions obtained in the RELAPS steady-state calculation compared with desired plant conditions as supplied by ORNL. 96 t
l TABLE 10. ciOT STANDBY STEADY-STATE INITIAL CONDITIONS Parameter Desired RELAPS
- Core power 9 MW 9 MW RCP inlet temperature 551 K (532*F) 552 K (533.6 F)
Temperature rise across RCP 0.24 K (0.44*F) 0.23 K (0.41*F) Hot leg pressure 14.82 MPa (2150 psia) 14.81 MPa (2148 psia) Steam generator outlet 6.21 MPa (900 psia) 6.21 MPa (900 psia) pressure Steam generator outlet 551 K (532*F) 551 K (532*F) temperature Steam generator startup 0.91 m (36 in.) 1.02 m (40 in.) level Feedwater temperature 305.4 K (90*F) 305.4 K (90*F) (. t G 9 97
T 6.3 Transient Results The results of the Turbine Bypass Valve Failure at Reactor Hot Standby calculation (TBF/HS) are present in this section. A sequence of calculated events for this transient appears in Table 17. The transient was initiated
~
with the opening of the four turbine bypass valves, which were modeled with a single valve component per steam generator. The calculation was run for two hours of transient time. The pressure responses of the two steam generator secondaries are shown in Figures 69 and 70. The steam generator secondaries behaved similarly, as expected. At transient initiation the secondary pressures decreased from 6.21'MPa (900 psia) to approximately 0.41 MPa (60 psia) in 1100 s. The pressures then increased to 0.48 MPa (70 psia) by 1300 s. The pressures then gradually decreased for the remainder of the calculation, and were 0.29 MPa (42 psia) and 0.32 MPa (47' psia) at the end of the transient in Steam Generators A and B, respectively. The pressure response and feedwater flowrates of Steam Generator A ,, (SGA) and B (SGB) secondaries for the first 600 s are shown in Figures 71 and 72. The oscillating pressure behavior calculated over the first 250 s is due to changing feedwater demands caused by the calculated startup (SV) level indication. The feedwater flow demand was controlled by the deviation of calculated SU level from the desired setpoint level of 0.91 m (3 ft). When the TBP valves opened the indicated levels began decreasing, causing a corresponding increase in feedwater demand. The increased
- injection of 305 K (90 F) feedwater resulted in an accelerated depressurization in the steam generator secondaries which continued until the setpoint SU levels were reestablished. The feedwater flow demand then decreased as the SU levels became greater than the desired setpoint, and the secondaries then pressurized due to the heat transfer from the primary and choked flow at the TBD valves. These oscillations continued until the main feedwater source decreased to 0 kg/s at 190 s due to excessive depressurization of the main feedwater train.
98
6000 , , , i , i
- 800 M ~
Q
~
. o
- a. .
. 4000 t-6 3 3
; 3000 . . ::
6
" 400 &
c. ['2000 3 g s 3 200 -
> 3000 _ _
s 0 0 0 iOOO 2000 3000 4000 5000 60C0 7000 80C0 TIME (s) Figure 69. TSV fofture at HSB. stsom gensraior A secondery pressur e. 6000 , , , , , , . 800 m s000 . . n e O G- 5
- 4000I -#
s "3 e
; 2000 - - C 6 400 s' 1 c.
f3 2000 g m o 200 -
> iOOO - _
0 O O 1000 2000 3000 4000 5000 6000 7000 8000 TIME (s) Figure 70. TSV f ailure at HSS. steam generator 8 pressure. 99 e- -- . , _. .n
7000 , , , . 500
- FPESSURE 6000 --- WAIN FEEDWATER .
^ X AUX FEEDWATER 0
Q. .'. 6 sc00 -
- :' . 2 g
9 .'
'l a
$- e . . a
.. .w
~
a <000 . :. :
*f.. .
g ,
; '............................. ............ 0 x 0
6
- a. 3000 - - C e a E O" s 200C 2
C 1000 - - 0 -500 0 10 0 200 300 400 500 600 TIME (s) Figure 71. TBV f alture at HSB. steam generator A secondary pressurs and feedwater flawrotes. 7000 . . . . . 500 _,
- 8RESSURE 4000 --- WAlN FEEDWATER m . X AUX FEEDWATER 0 s.
- a. *.
6 $000 - *
.fI: ,. ., - 7
.l : ** N
- e . ;. , : ; a w . . .. . - , , , .x
~8 4000 ,' : - . e v n .
.g ' p. '
'-l .
......................................... -0 I
$ 3000 -
)
e E
\ a 3 2000 - -
j o 3000 . i 0 -500 0 10 0 21.',O 300 400 500 600 TIME (s) Figure 72. TBV f ailure el HSS. steem generator 8 secondary pressure and feedwater flow r o t os.
- e-100
d TABLE 17. TURBINE BYPASS VALVE FAILURE AT REACTOR HOT STANDSY TRANSIENT SEQUENCE.0F EVENTS Tizae Event (s) Turbine bypass valves fail open 0 MFW pump discharge pressure decreases to 765 psia 27.3 HPI initiated 125.1 , RCPs tripped--steam generator low level limit changed from 24 in 155.1 relative to SU level taps to 240 in relative to operating level taps EFW initiated in Steam Generator A and Steam Generator 8 due to change in low level setpoint Condensate booster pump trips on low suction pressure 163.4 MFW pump trips on low suction pressure 168.5 MFW train isolated from steam gererators after feedwater flow 228.6 decreases to Zero k' Core flood tank injection initiated 383.5 Core flood tank injection ends 391.6 Turbine bypass valves closed 600.0 , Pressurizer filled 900.0 PORV setpoint reached 949.6 Steam generator operating level at 240 in. 1010.2. Steam generator EFW sources tripped off for last time 1030.4
. Steam Generator A operating level at 240 in. 1070.5 Steam Generator A EFW sources tripped off for last time 1074.6 Primary Loop B cold leg suction temperature exceeds Loop B hot leg 1620.5 temperature Primary Loop A cold leg suction temperature exceeds Loop A hot leg 1892.2 temperature Downcomer wall temperature reaches 265"F 4600.0 Transient terminated 7200.0
, 101
The depressurization of the main feedwater train was caused by the loss of pressure in the steam generator secondaries coupled'with the operation of the SU valves and the tripping of the reactor coolant pumps (RCPs). Operation of the SU valves in response to steam generator SU level-demand resulted in depressurization of the main feedwater train. This cepressurization was halted when the SU levels were recovered and the
- SU valves reclosed (closure of the SU valves enabled the repressurization of the feed train by the feed train pumps). This cyclic depressurization s.nd repressurization in response to SU valve operation is shown in Figure 73; which depicts the feedwater train pressures upstream of the condensate booster pump and main feecwater pump and the SU valve area demands. Tripping of the RCPs'at 155 s changed the steam generator low level setpoint from 0.91 m (3 ft) to 6.1 m (20 f t). The new low level setpoint resulted in a sustained maximum SU valve area demand, with a consequential large feedwater train depressurization. The condensate booster pump suction pressure trip setpoint of 0.21 MPa (30.7 psia) was reoched 3 s later (158 s) and was sustained for the requisite 5 s, thus /
resulting in a trip of the condensate booster pump at 163 s. Loss of the condensate booster pump resulted in an accelerated depressurization of the feedwater train downstream of the pump, and consequently caused a main feedwater pump trip on low suction pressure at 168 s. The feedwater train was removed from the calculation at 228 s when it was determined that additional feedwater could not be delivered to the secondaries. The SGA and SGB pressures continued to steadily decrease over +'e next 1300 s, with three exceptions, as shown in Figures 74 and 75. The first exception involved core flood tank injection into the , reactor vessel downcomer from 383 to 392 s, which resulted in a temporary decrease in primary-to-secondary heat transfer as the primary steam generator tube bundle flow reversed direction. This decrease in heat transfer resulted in a slight increase in the temperature-induced rate of depressurization of the secondaries.
. 102
6000 i . . 2 WW PU@ FRESSURE BCCSTER PCWP PRES $URE e 5000 ld' h
. . X SGA SU VLV AREA -
4 2 2 A SGB SU VLV AREA $ v
- 400C - -
W _J 5
$ 3000 -
0
~
--I u o a
i zow 'i 2 5
- - . 1 m
2 .
/t l O 3000 -
, l ll z
)
ll
'" ~ ' '" '" "'
- 0- ~
O O 10 0 20C 300 400 500 TIME (s)
, figure 73. TSV f ailure of HSB. f eedtrain pressures cid normcl!!ed
( startup volve areas.
\
5000 . . . 800
.000 m
C Q-6 m
.A g v I.
v e 4000"- -800 u I e 3 u 85 3 m E 3000 - - g u 4C0 s a a . l3 2000 - g 3 3 200 g
> 1000 - -
0 0 O 500 1000 1500 2000 TIME (s) Figure 74. TBV f ailure of hSS, steam ser'eref or A secondary pressure 0-1500 s. 103
l e 4000 , , , 800 m S000 - ^ A O ?w 2 CL
" w g -
600
- e b b 3 s
" e e 3000 - -
- w 400 '
u a- a
/
j :000 - - g 3 3
; :00 -
> =0 - -
f _ o -0 0 S00 1000 1500 2000 TIME (s) Figure 75. TSV f ailure at HSS steam generator 8 secondary pressure 0-1500 s. 9 e
.s 104 .
ines
The temperature of primary coolant in the hog legs and reactor vessel continued to increase during this period of flow reversal'and accompanying stagnation. Consequently, the resumption of a positive direction coolant flow resulted in a temporary increase of the primary-to-secondary heat
, transfer rate due to the hotter primary coolant entering the tube bundle regions, with a suosequent temperature / pressure increase in the steam generator secondaries. This increase ended as the result of continued emergency feedwater (EFW) injection and resumption of the primary system cooldown.
The second delay in the secondary system depressurization occurred when the TBP valves were. closed at 600 s. The resulting increase in pressure was overcome by the continued injection of EFW into the upper boiler section; however, the rate of secondary depressurization decreased due to the TBP valve closure. The third delay in the secondary system depressurization occurred
/
around 1620 s in SGA and 1580 s in SGB. The EFW sources to tne steam (' generators were controlled based upon the low level setpoint of 6.1 m (20 ft) such that EFW would be terminated when the calculated level exceeded the setpoint level. Termination of the EFW resulted in a partial
" loss of secondary heat removal capability. The primary-to-secondary heat transfer rate was still positive and the secondaries began to heat /up, which resulted in the slight repressurizations. These increases in pressure ended as the secondary temperatures approached those on the primary side and consequently decreased the heat transfer rates to 0 MW at -
120 s. The temperature and pressure responses of the primary coolant system l- , are presented next. The temperature responses of the two cold legs in a given coolant loop were identical, thereby obviating a separate discussion of the calculated behavior in all four cold leg sections. The temperature responses of the Steam Generator A (SGA) and B (SGB) primary loops are shown in Figures 76 and 77, respectively. The same parameters are plotted in Figures 78 and 79 with an expanced time scale to better show detail at the beginning of the transient. 105
600 , , , , , , , n 600 m . M - RCP DISCHARGE b
--- HOT LEG X RCP SUCTION e $
- 500 O
O
- 500 - - w s e I 1 400 % e o
.e M --.-....... n..<_,....,, ,,---~....... A 300 3
} , , .g
= =
e 200 e E E 3 s O o
> ' ' ' ' ' ' ' 10 0 >
300 O 1000 2000 3000 4000 50C0 6000 7000 SOCO TIME (s) Figure 76. T8V f ailure at HSB. loop A fluid t empera t u r es. 600 , , , , , , , ., m 600 ^ M - RCP DISCHARGE b
... HCT LEG
- X RCP SUCTION e 3 500 :s O g 6 500 - - w e e a '. 4C0 a E , E o e
- a,, -
-...X...,
~
....r...c...z......,,,,,,,j 300 7-w 400 - -
= .o_-
e 200 e
- 2 O o
. i , 10 0 >
g , , i 0 1000 2C.^0 3000 4000 50C0 6000 70C0 80C0 TIME (s) Figure 77. TSV f ailure at 48. loop 9 fluid t emperclur es_ s' 106
600 , , n 800 ^ x
* - RCP DISCHARM b
--- HOT LEG e 3 X RCP SUCTION
- 500 O O
'e 500 a,
d 400 E .
e
. .e.
3 -- - ,...- - = ..:: . . . . . . . . . . .; sw .' s - 3 [*00
~
-M Y .j e 200 e E E 3
s O o
> ' ' 10 0 >
300 O 1000 2000 3000 TIME (s) Figure 78. TSV f ailure of HS8, loop A fluid temperatures. 0-2500 s. t 600 . . m 600 ^ M
" - RCP DISCHARGE i'-
HOT LEG
- 3 X RCP SUCTION
- w w a -
500 2 O O
'e 500 ,
'e g ", -
400 g e , e V
' .- . . s300 _V w 400 -
-W* f_o-e 200 e E E a o o o
> ' ' 10 0 >
300 O 1000 2000 3000
, TIME (s)
Figure 79. TSV f ailure of HSS, loop B fluid terrporatures 0-2500 s. 107
The temperatures in the hot and cold legs were essentially identical for the first 200 s of the transient, and then began diverging as the steam generator secondaries continued depressurizing and the primary-to-secondary heat transfer rates increased due to emergency feedwater injection into the steam generator upper boiler sections. A reactor coolant pump trip at-155 s resulted in reduced primary loop flow, consequently decreasing the rate of cooldown in the hot legs while increasing the HPI-induced cooldown rate in the cold legs. The hot leg temperatures in both loops began-
- increasing as the reactor vessel upper head began voiding, thereby contributing additional energy to the hot legs. The upper head was completely voided by 340 s and the hot leg temperatures resumed their cooldown.
The hot leg temperature in the A loop increased at 383 s due to core flood tank injection accompanied by temporary reverse flow. The core flood tank injection was initiated by an overprediction of condensation in the reactor vessel upper head, which resulted in a raoid depressurization of
'the primary system. The reverse flow caused a brief period of flow stagnation with an accompanying increase in primary temperatures in the hot ,/
legs and vessel. The associated decrease in cold leg temperatures was due to the effects of high pressure injection and the loop flow stagnation, and core flood tank injection. The reactor vessel upper head refilled by about 392 s and positive loop flow was reestablished. The effect of the core flood tank injection may be seen in the temperature response of the cold legs in relationship to their corresponding hot leg temperature responses. The loop A cold leg pump suction temperatures exceeded the hot leg - temperature about 100 s after core flood tank injection. This time delay corresponds weli with the loop transit time. A similar effect is seen in the B loop temoerature resoonie. I The hot leg and cold leg pump suction temperatures remained 7 approximately the same between Loops A and B until about 1000 s. The temperatures diverged as the HPI and core flood tank inventcry cooled the pump suction legs. The hot leg temperatures continued to decrease and finally became cocler than their corresponding cold leg suction temperatures as the net heat tran;ier from primary to secondary became negative. The apparent heatup cf the cold leg discharge sections occurred - 108
i l when the pressurizer became water solid at 911 s. The resulting increase in primary pressure causec a reduction in HPI . flow rates u'ntil, at 952 s, the primary pressure reached the PORV setpoint of 16.65 MPa (2415 psia). The HPI flow rates were essentially constant at 8.8 kg/s (19.51 lbm/s) in
. the_ A loop and 6.8 kg/s (15.1 lem/s) in the B loop for the remainder of the calculation.
.The reduction _in HPI flow rate, in conjunction with the pump discharge leg temperature increases, resulted in two distinct temperature zones in the primary loops. The first zone consisted of the coolant that was not heated during the initial 200 s of the HPI flow rate reduction, while the second zone cor.sisted of the coolant that was heated during the same period. The effect of these two zones on the temperature response of the p ri mary , in conjunction with the secondary-to primary beat transfer caused the oscillating temperature responses at the cold leg pump discharges. The mass flow rate in the vessel downcomer was driven by the density head of the inlet coolant. As the cooler zone of water entered the top of the cowncomer the density head increased, which resulted in an increased m.ss
(
- flow rate down the downcomer, and subsequently throughout .the rest of the primary. The increased flow rate through the core resulted in a cooler core exit temperature. The density of the warmer zone of water was j responsible for the flow rate reduction tnrough the primary. As this water entered the vessel downcomer, the density head decreased and caused a decrease in the loop flow rate. The decreased primary flow rate resulted l
in a greater temperature increase across the vessel. The temperature oscillations were damped out in the lower volumes of the two steam - l I generators, which maintained primary-to-secondary heat transfer throughout the transient. However, the effect of constant HPI flow rate coupled with the oscillating primary mass flow rate continued to dominate the flow , response of the primary for the remainder of the transient, as seen in Figures 80 and 81.
! The effect of the HPI flow rate on the A and B loop temperature responses may be seen in Figure 82, which depicts the temperature responses of the A and B loop cold leg pump discharge volumes at the points of l 109 L
B B 4 5 I I '
./
F 2C000 15000 m , 7 b 50= .- ~
.* .g s e c.
D o 5000 3 E o i G
$ 0-- -- -
+0 g O
2 O 2
-5000
, -10000 0 1000 2000 3000 4000 5000 6000 7000 5000 TIME (s)
Figure 80. TSV f ailure et HSS. loop A hat leg mass ficarote. 10000 , . . > > -
./ 20000 13000 m 7
) == . - -
=00 J w
D o 5000 D C o
= .
E 0 . -_- ------ O g O 2 o 2
-5000
, , , , , , , - M 00 0 1000 2000 3000 4000 50C0 6000 70C0 6000 TIME (s) -
Figure 81 TBV f ature at HSB. loop B hot leg mass flowrote. 1 i l 110 I 1
,- -e-T
- - - - - . =
t 4 1-i
#
- a s ,
m ,
- w: 600 m
- LOOP A M=
e- --- LOOP B. e d
'3 S00 i
O .$
- o
'e 500 .
5 E o E 400 e
" e 7 i
( 3 300 1 400 - E
. .. e 2 3 E o _3 i
> ' ' 10 0 >
O 300 t i i , , 0 20 2000 3000 4000 5003 6000 7000 8000 TIME (s) Figure 32. TSV f ailure of H03. loop A and 8 cold leg discharge fluid temperatures. f a f 9 e 111
-. - . - . , , . . . - , , - - ~ - . - . . - . . - - - ' -, - .- -
HPI injection. The larger temperature oscillations in the B loop were due to the lower HPI injection flow rates. The HPI tended to damp out the A loop oscillations, .ut was not of sufficient magnitude in the 8 loop to sicjnf ficantly affect the overall loop temperatures. The response of the B loop suggests that the B loop controlled'the total primary temperature
~
response; however, the net result is not significant in that the temperatures were not varying riarkedly during this period. The pressure response in the reactor vessel downcomer at the location of the first circumferential weld below the cold legs is shown in Figure 83, overlaid with an estimate of uncertainty due to sensitivities. These sensitivities will be discussed shortly. The primary pressure decreased at the initiation of the transient as the result of the rapid depressurization of the steam generator secondaries, the initiation of HPI, and tripping of the reactor coolant pumps. The first decrease in the rate of primary depressurization occurred at about 170 s as the pressurizer emptied. Further changes in the rate of primary depressurization occurred at aceut 800 s when the vessel upper head began to void. The i depressurization at about 900 s was due to condensation effects in the j vessel upper head. This depressurization ended when the core flood tank injection setpoint was reached and the upper head refilled. The combination of HPI and core flood tank injec' tion resulted in the refilling of the pressurizer, with a corresponding increase in primary pressure from about 900 s until 911 s, when the uppermost pressurizer volume filled. The pressure in the pressurizer upper head then increased from 6.5 MPa j (943 psia) to 16.65 MPa (2415 psia) by 952 s. The p0RV maintained the ' primary pressure at this point for the remainder of the transient. The response of the reactor vessel heat transfer coefficient (HTC) in - the vessel downcomer at the same weld location is shown in Figure 84. The 2 HTC decreased from 29300 W/m -K (1.43 Btu /s-ft2 *F) to 27000 W/m 2_g (1.32 Btu /s-f t2- F) by 155 s, when the reactor coolant pumps tripped. The subsequent reduction in vessel downcomer coolant velocity caused a significant reduction in the HTC to values around 1700 W/m2-K
/
112
HEAT TRANSFER COEFFICIENT Volume pressure (kPa) 2 (kW/m - K) w 2 2 8 j 8 8 2 o 6 3 8 1 o 8 o 8 8 2 . , # ;o , i-
- c. -a " a '
5 _i o 5 o
- r m am g< ;<
=~ M 4- M
{= 8 _ _ jh 8 _ So 8o o u , g', 3 o _
~*
o o - l i 3 M8 - i w ,ro ,ro 8 g mo . . g no _ _ .
,vw
? ;?w
,v
< 8 o
8
. g _ o _ _
I 8
. ?
a 8 -
- e. o >
_ g a _ _ 9 2 '
, R ir a b $ a 2,
a 8 f _ 8 _ _ 8 k"b E8 o s
. ' n. t
- w o
- 5 G "
8 a a a f 8 8 8 3 s Volume pressure (pslo) - HEAT (Blu/hr-f TRANSFER t 2_COE.FFICIENT p)
~
2 (0.08 Btu /s-ft 'F) at 400 s. The depressurization caused by the upper head condensation and subsequent core flood tank injection is seen at 400 s when the HTC initially decreased due to flow stagnation and then increased as the core flood tank injection mixed with the coolant in the downcomer region. The minor increase in the HTC at 900 s was due to the reduction cf HPI flow induced by the high primary system pressure after the pressurizer filled. The reduction in HPI flow resulted in a reduction in 'the . percentage of cold leg fluid contributed by the 283 K (50*F) HPI source. The oscillations beginning at about 2600 s are the result of the temperature and velocity oscillations previously discussad in this section. The decrease in downcomer temperature, which was coincident with the density induced increase in velocity, resulted in an increase in the HTC. Parameters or occurrences that may have significantly affected the calculation of the downcomer temperature are addressed next. The parameters to be discussed are (a) the operation of the startup valves, (b) the condensation depressurization that occurred at approximatelv 880 s, i and (c) the calculation of the steam generator low level setpoints. The / calculated downcomer temperature response overlaid with an estimate of uncertainty due to sensitivities is shown in Figure 85. The lower limit of the temperature uncertainties is overlaid by the temperature response, as will be described in the following discussion. Operation of the feedwater startup valves determined the rate of feed train depressurization. A faster than normal valve closure rate would allow the feed train to repressurize in the early phases of the calculation. As described previously closure of the startup va'lves prior to sufficient feed train de' pressurization during the early phase of the transient resulted in a recovery of the feed train pressure due to continued operation of the condensate booster pump and the main feedwater pump. Had the startup valves remained open, the feed train would have depressurized to the condensate booster pump suction pressure trip setpoint. The amount of depressurization required was about 0.69 MPa
'(100 psi), which occurred during the first 20 s of the transient. Had the condensate booster pump tripped, the main feedwater pump would have tripped 114
4 t 600 . , . . . , . m M 600 ^ v P-v e a w 330 3 6
- s 0 -500 -
- 6. 0 e 6 e
g500 - g 2 .=0 e
- 450 - . -
3 7 3
= 7 l . 300 =
j
~
\ -
$-{m i f 1 I f f f f 2
200 y C 7000 2000 3000 4000 5000 6000 7000 8000 TlWE (s) Figure 85. T8V f ailure of HSS, reoctor vessel downcomer fluid t ernpera t ur e. i 115 e
~_
due to the loss of pump suction pressure. Loss of the booster pump and main'feedwater pump would have reduced the amount of feedwater flow into the steam generators. The loss of the main feedwater source would have resulted in a low-level actuation of the Emergency Feedwater System, which would have resulted in feedwater injection into the upper tube bundle region. A sensitivity calculation was performed to determine the effect of initiating emergency feedwater injection at the start of the transient. The calculation was stopped when the pressurizer became liquid filled, since no further system perturbations occur after this event. The difference in downtomer temperatures between the sensitivity calculation and the transient calculation at 1100 s was about 2 K (3.6 F). If the model startup valve opening time had been too fast, the feed train would
'have'still depressurized at the time calculated in the transient. The feed traia depressurized to the condensate bo'oster pump trip setpoint when the steam generator low level setpoint was changed from 0.91 m (3 ft) to 6.1 m (20 ft) due to the reactor coolant pump trip. The change in low level setooi.it caused the startup valves to completely open, and remain open until feed train isolation. The rate of valve stem movement was modeled as 2 s from full closed to full open. Had the valve been allowed to open instartaneously, the maxirum difference in times for the full open condit on would have been 2 s. The rate of pressure decrease 1,n the feed train during. this period of time was 0.48 MPa/s (70 psi /s); therefore, the time of the condensate booster pump trip would have been approximately 1 s earlier, which is insignificant. It is concluded that early depressurization of the main feed train due to (a) a slower than normal startup valve response time, or (b) a faster than normal valve opening rate
- would not significantly affect the downcomer temoerature response.
An overprediction of reactor vessel upper head condensation occurred - at 880 s. The resulting depressurization of the prima *y resulted in core flood tank injection of 305 K (90'F) water into the vessel downtomer. The initial effect of the depressurization was a brief period of reverse flow in the hot legs, pressurizer surge line and tank, and the B loop cold legs. A partial flow rate decrease was also observed in the A loop cold legs. These flow perturbations resulted in a hot leg temperature increase of 11 K (20*F) due t. the essentially constant primary heat sources (9 MW reactor core decay heat and contributions from the primary coolant piping) and the - 116
decreased heat transfer to the steam generator secondaries. The initiation of core flood tank injection resulted in a temperature dec'rease of 8 K (14'F). The net effect was a temporary temperature gain of 3 K (5'F), which was subsequently reversed by the steam generator secondaries. It is
. therefore concluded that the inadvertent depressurization of the primary system and the subsequent core flood tank injection dit not significantly affect the final results.
The steam generator low level conditicas were calculated using level tap locations corresponding to the startup level taps for the 0.91 m (3 ft) setpoint, and the operating taps for the 6.1 m (20 ft) setpoint. It was later determined that the startup taps are always used for determination of steam generator low level conditions. Additionally, the startup level calculation incorporated a density head compensation term (which is not done in the prototype generators); and modeled the lower startup tap location upstream of the downcomer orifice plate. The combined effect of the above three model discrepancies was a 300 s delay in tripoing off the emergency feedwater sources. This corresponds to an additional 10,200 kg (22,500 lbm) of liquid inventory in each steam generator. The excess heat transfer capability resulted in an estimated 15 K (27.5 F) underprediction of the final vessel downcomer temperature. 6.4 Conclusions The calculation was carried out to 2 h of transient time, the entire period of interest for pressurized thermal shock. The minimum calculated - reactor vessel downccaer fluid temperature was 387 K (237'F) near the end of the transient. The maximum calculated reactor vessel downcomer pressure was 16.99 MPa (2465 psia), the opening setpoint pressure for the power operated relief valve. The uncertainty in the calculated downcomer fluid temperature due to startup valve operation, code-calculated condensation effects, and low level limit determination method was assessed. The uncertainty was found to be minor and the conclusion is that the calculated temperature represents the lower temperature limit of the uncertainty band.
. 117
- 7. PRESSURIZER SURGE LINE SMALL BREAK TRANSIENT The foll'owing subsections contain the trar. stent scenario description, modeling changes effected to perform this calculation, detailed analysis of the transient results, and conclusions drawn from the analysis.
7.1 Transient Scenario Oescription 4 x A description of the sequence analyzed is presented in Table 18. This sequence was developed at Oak Ridge National Laboratory. The transient was initiated by a 0.0508 m (2 in.) diameter break in the horizontal section of the pressurizer surge line with the reactor at full power conditions. Operator action is assumed to trip off power to the reactor coolant pumps (RCPs) 30 s after initiation of high pressure injection (HPI). This scenario is similar to that discussed in Section 3.3 except the break is larger, has been relocated from the PORV to the surge line, and the operator action to trip RCPs is assumed. 7.2 Model Changes ,/ The basic RELAPS model used to perform the pressurizer surge line small break transient is described in Section 2. - The thermal-hydraulic model used for this transient was the same with one exception. The transient description specified the break should be midway between the hot leg riser and the pressurizer. Therefore pipe - Component 600 (see Figure 3), which represented the pressurizer surge line, was split and the break was inserted between the two sections. The break had an area of 0.00203 m2 (0.02182 ft2 ), which corresponds to a diameter of 0.0508 m (0.167 ft). The transient was initiated from normal steady state conditions with the core at full power. Reactor power was set to trip off when the primary system pressure dropped below the pressure determined by the following relationship: 118
TABLE 18. PRESSURIZER SURGE LINE BREAK TRANSIENT SCENARIO Initial Conditions:
- 1. Full reactor power
- 2. Nominal temperatures / pressures
- 3. PZR spray /htrs operate as designed
- 4. Decay heat = 1.0 ANS standard Sequence of Events:
- 1. 2-in. diam hole midway between pressurizer and the riser portion of the candy cane of the "A" steam generator
- 2. Reactor trips, turbine tr;ps, TSVs close
- 3. HPI actuates at setpoint
- 4. Operator trips RCPs 30 s after HPI actuation
- 5. TBVs/SRVs in secondary fui.ction as designed
- 6. MFW and EFW systems funct:an as designed
- 7. Core flood tanks dump; LPI system actuates a
- 8. No further operator action:, i.e., no throttling of HPI D Failures Assumed:
- 1. SBLOCA
- 2. Violation of procedures (Event number 8)
Basis for Transient Selection: Determine cooldown experienced during " stagnant" SBLOCA. RCS pressure will probably stabilize below PORV/SRV setpoints. Addresses the importance of I vent valve flow in the B&W design. l [ . a. Event may or may not occur, depending on phenomena encountered.
- b. Calculation restart point: on attainment of 75*F subcooling in least-subcooled coolant loop, operator throttles HPI sufficiently to i
maintain 75 25 F subcooling. Note: In calculating subcooling with RCPs i off, pressure for each loop is taken from top of hot leg riser, and temperature is taken from core outlet. 119
PTRIP.= 13.26 TH0T
- 5989.
where P TRIP
= pressure at which scram occurs, psig T = h t leg temperature, L p A, F.
HOT This scram' relationship was obtained from the Oconee-1 FSAR and, on cooldown transients, typically is encountered before the low pressure scram setpoint is reached. After scram, the decay heat was assumed to be the ANS standard level. The reactor coolant pumps were set to run at a constant speed of 1226 rpm until they were tripped, 30 s after HPI injection began. The secondary system was controlled as follows. The steam flow control valves were set to close at reactor scram to simulate the closing of the turbine stop valves. The heater drains into the feedwater system were set to ramp closed in 5 s after reactor scram. The ICS was allowed to control the feedwater to both steam generators throughout the transient. ,
/
When the reactor coolant pumps (RCP) trip, feed flow is diverted from the main feedline into the emergency feedwater header. Also at the time of RCP trip, the low level limit in the steam generators changes t.-om 0.61 m (2.0 ft) to 6.10 m (20.0 ft). The minimum speed of the main feedwater pumps, prior to pump trip, was set to 4270 rpm. The hotwell and booster pumps were set to run at constant speeds of 1180= rpm and 3581 rpm, respectively. - 7.3 Transient Results This section chronologically describes +.he events. and their causes, that occurred during the transient calculation. A sequence of calculated events is also presented in Table 19. The calculation covered 6200 s of 120
l TABLE 19. PRESSURIZER SURGE LINE BREAK TRANSIENT SEQUENCE OF EVENTS lire Event (s)
. Open break in pressurizer surge line 0.0 Reactor scrams on P versus T relationship. Turbine stcc valves 45.2 close. Heater drains to feed train begin to close. MFW pumps begin to run back.
TBV open on high steam generator pressure 47.0 SRV open on high steam generator pressure 50.0 Pressurizer heaters turn off on low pressurizer liquid level 55.3 SRV close as steam generator pressure drops 69.0 MFW pumps trip on high pump discharge pressure 70.0 HPIS i ection begins on low primary pressure 78.5 Upper head begins to void 86.0 RCP trip 30 s after HPI initiation. Realign main feedwater to 108.5 emergency feed headers. Emergency feec begins on low steam generator level. TBV close as steam generator pressure drops 117.0 Upper head volume completely drained of liquid. Top volume of ~300.0 downcomer begins to void. Feed to Steam Generator B stopped as low liquid level limit 500.0 is exceeded Emergency feed to Steam Generator A stopped as low liquid level 503.0 . limit is exceeded Vent valves open 554.0 Bubble forms in tubes of Steam Generator A. Bubble starts to move 768.0 to top of candy cane Flow stagnates in Loop A as bubble fills top of candy cane 815.0 Circular flow between cold legs of Loop A 872.0 121
TABLE 19. (continued) Time Event (s) Bubble formed in tubes of steam generator B. Bubble starts to 892.0 move to top of candy cane Flow stagnates in Loop B as bubble fills top of candy cane 1020.0 Bubble in top volume of downcomer collapses 1383.0 Partial collapse of bubble in upper head. Liquid drawn up is 1400.0 heated and vaporized by hot metal, primary pressure momentarily r11ses Bubble in upper head collapses, volume refills with liquid. Large 1488.0 flows throughout primary system as liquid moves toward upper head. Core flood tank begins injection on low primary pressure 2215.0 Circular flow between cold legs of Loop B starts 2378.0 Circular flow between cold legs of Loop B reverses direction 2443.0 Pressurizer begins to refill with liquid 5064.0
'J LPIS injection begins on low primary pressure 5124.0 Calculation terminated 6200.0 4
122 e
~. _ __ - _ _ _ _ _ _ _ _ _
transient time. The responses of both steam generators were essentially the same throughout the transient, therefore, any reference to the steam generators will apply equally to both steam generators.
. The transient was initiated by opening the break in the pressurizer surge line. The mass flow rate out the break is shown in Figure 86. The "rimary system pressure and the liquid lavel in the pressurizer dropped immediately as mass was lost from the primary system (see Figure 87).
Almost all other conditions in the plant, both primary and secondary, remained essentially unchanged until 45 s, at which time the primary system pressure had dropped to the point where reactor power was tripped. The steam flow valves on both steam generators closed, thereby increasing secondary pressure and temperature. Figure 88 shows secondary system pressures. At 46 s, the turbine bypass valves opened, and at 47 s the secondary relief valves opened, and both generators started to dump large quantities of steam. Figure 89 shows the SG liquid levels drop as the mass inventory in the generators was depleted. This resulted in a rapid i cooldown of the fluid in the steam generators and, as shown in Figures 90 through 93, a cooldown of the primary system fluid. As the primary fluid cooled, it shrank, decreasing the pressurizer level even faster, and therefore increasing the rate of depressurization. As reactor power dropped after scram, the ICS calculated a reduced feedwater demand and ran back the main feedwater pumps to their minimum speed and began to close the main feedwater (MCW) and startup valves, sharply dropping feedwater flow to the generators. The MFW valu s close - faster than the startup valves, and were closed by 50 s. The startup valves began to close at about 67 s. As the startup valves began to close, the pressure downstream of the MFW pumps increased. At 70 s, the outlet pressure of the MFW pumps excaeded the high discharge pressure trip set point, and the MFW pumps were tripped. After the pump trip, the ICS began to reopen the startup valves and, by 74 s, they we*e full open again. By 69 s, the energy removal rate from the steam generators had dropped the secondary fluid temperatures to the point where the saturation pressure in the steam generators was below the relief valve setpoint, and these 123
400 . 4 i 4 . . . 800 m 300 - 7
- 400 \
.m v
$M g:
t l 400 y o tj ll '
.i i
i
- k. i g..I.. e! . .e
. \, g-pf **
O O O 900 2000 3000 40C0 $000 6C00 7000 Time (s) Figure SG. Pressuriser surge 49 brook. breck rrose flow r a t e. 20 , , , , . . - 2500 m O O
$ 15 r -
[ 2000 v
=>
I a Ng< . tS00 $
$ E A Q.
y -
.g
.J S -
> 500 $
f f f f I i 0 1000 2000 30C0 4000 5000 6000 7000 Time (s) Figure 87. Pressurlaer surge line break, hot leg pressure. . 9 v e 124
3 , , i i i > SG A
-- SG B 100C
^
O 9 A J g . . m v w a 300
- e 6 w 3 3 E
e 4 -.
. 600 $e b
b a c. e . 400 , E *- 3 E
-O 2 - -
.s
> o 200 >
0 O O 1000 2000 30C0 40C0 5000 6C00 7000 Time (s) Figure 88. Pressurtzer surge fine brook. S.C. secondary pressures. 10 . . 4 i . . 350 SG A
... $g g 300
^ 8 - - ^ '
E --
- ,2 .m . :
, v 1 i,s y. %.,..- . r -- w
.a w
J r* '$. 250 _J w b "G I - 200 c c 3 O 3 - O
.35 0 -
- 4 . . a 10 0 2-O 1000 2000 3000 4CCC 5000 6000 7000
. Time (s)
Figure 89. Pressurizer surge Ene brook. S.G. stettup levels. 125
600 , , , ,
^ ^
M 600 v & e e
' 550 - - w 3 3 g
e500 g u - E 500
- I p j - $
l ,4 E
,. s 4C0 f.
-- 450 - -
3 3 T T
- 300 =
{400 3 {3 350 O ?000 2000 30C0 4CCC 5000 6000 70C0 Tim. (s) Figure 90. PressurIrer surge 'Ine breau, R.V. fluid downccreer tempera t ur e. m M SOO
- l ^
" - LOOP A 600 *b
--- LOOP 9 e e b b 3 3 550 -
O
- 550 -
c
- 6 e o O. c.
E 500 E 3
.e.
l
~, ,
3 T 500 450 3cr -
*: =
3 e E 3 400 E 3 O C 450 ' ' O 900 2000 30C0 4CCC 5000 EC00 70C0 Time (s) Figure 91. Pressurizer surge line brook, t'ot leg fluid t emper a t ur es. . 126
l 600 , , i
^ 600 ^
$ - LEG A-1 LEG A-2
{ e 33o - . e
$ 500 O
t O
- 500 -
lfi '
- 6 I
I 400
- 450 -
L J
.f.
" 'hf*'*%%**%
"% s.,,, .,
300 2 _5400 l* ^..g.,, - h
,Q,. r. , -
' 2" l350 -
T l' 3 0 o
> t f f f f f 10 0 >
0 1000 2000 30C0 4000 5000 6C00 7000 Time (s) Figure 92. Pressurizer surge time brect. loop A cold leg fluid tempe r at u r es. 600 , , , , , , n 600 ^ M - LEG 8-1 b
... LEG 8-2 e no - -
e 5 560 $
-O * -
O g 500 - - g g { 400 { 2 450 -
%s -
2 '
%.,4 300 2 5400 - ' - -
5
=
! =
. , %.,. 200 .
. h350 - E 3
0 o
' ' ' ' ' 10 0 >
300 0 400 2000 3000 4000 5000 6000 7000 Time (s) Figure 93. Pr essurizer surge line br ook, loop a cold leg field tempera tur es. 127
valves closed. This stopped the cooldown in the secondary system. The turbine bypass valves continued to release just enough steam'to hold the secondpry pressure near the turbine bypass system opening pressure setpoint. The fluid in the upper head of the reactor vessel as modeled is outside of the normal reactor vessel circulation ficw path and remains fairly stagnant. Therefore, as the rest of the primary fluid cooled, this fluid stayed at near the initial, steady state hot leg temperature. 'At about 86 s, primary system pressure had dropped to the saturation pressure of this fluid, and it began to flash, as shown in Figure 94. This flashing slowed the primary depressurization rate significantly. Primary system pressure reached the HPI setpoint of 10.45 MPa (1515 psia) at 78 s, and the HPI system began injectir] cold water into all four cold legs. At 108 s. 30 s after HPI initiation, the primary coolant pumps were tripped. At that time, the main feedwater train was realigned to the emergency feedwater header. Also, because the steam generator liquid levels were below the new low level limit of 6.10 m (20.0 ft), emergency feedwater was initiated to both steam generators. Total j feedwater flow to the generators is shown in Figure 95 This cold flow into the top of the steam generators depressurized them somewhat, which in turn closed the turbine bypass valves. The liquid injected into the top of the generators fell down to the bottom and began to fill the generators from the bottom up. After the primary coolant pumps coasted down, natural circulation kept " some flow moving around both loops. Natural circulation was maintained because significant energy was removed from the primary system into the steam generators to warm the cold emergency feedwater being injected into the steam generators. This ener;y removal from the primary system also caused a smootn cooldown in primary system fluid temperatures between 109 , and 500 s, as seen in Figure 90. At about 300 5, the upper head of the reacto? vessel had completely drained of liquid and, as a result, primary system pressure began to drop more rapidly. However, another pocket of stagnant, relatively hot fluid was present in the top volume of the downcomer. The primary pressure very
~
128
I -
- 1 i . . . .
,,7 7 ,
1 C 08 * *
.2 8
o.s - -
.io 0.4 - -
6 C k 0.2 - -
, il 21s i 0
O MOO 2000 3000 40C0 SC00 6C00 7000 Time (s) Figure 94. Pressurf ter surge line brook, void fraction in R.V. opper j hood.
\
800 , , , , , .
- SG A SG 8 1500
%Q * -
m n a N
) 20 E
$ soo - -
9 v A B 200 ~ - , 500 $ O 3 2 0- As -'O i . ,
,yg , , ,
0 2 00 2000 30C0 4CCO 5000 6c00 7000 Time (s) Figure 95. Pressurizer surge line brook, total feedwater flow rates. 129
soon reached this fluid's saturation pressure, and as snown in Figure 96, it began to flash and slow the primary depressurization. Th,e effect of this flashing and draining an be seen as ripples in the primary system pressure shown in Figure 87. Even though the MFW pumps had been tripped earlier, the condensate booster pump in the MFW train was still running against a closed check valve and delivered no flow. At 416 s, the pressure in the'stea- generator - secondaries had dropped to the point where the head produced b', e condensate booster pump exceeded the pressure in the steam generators, and feedwater began to flow from the MFW train into the steam generators via the emergency feedwater headers. By 500 s, the combination of main and emergency feedwater flow had filled the steam generators to above the low level limit, and the emergency feedwater flow was shut off. The ICS also began to close the startup feedwater valves, ana by 621 s, main feed to the generators had also been terminated. The natural circulation flow rates dropped as the emergency feedwater was terminated and the temperature of the fluid in the steam generators , j increased and approached the temperature of the ,,rimary hot' leg fluid. At 554 s the natural circulation flow dropped to tt ? point whera the reactor vessel vent valves opened, allowing flow between the upper plenum and the top of the downcomer in the reactor vessel. This mixeh a significant amount of warm fluid from the upper plenum with the fluid in the downcomer, and slowed the cooldown rate in the downcomer, as shown in Figure 90. By about 675 s, the temperature of the primary system hot leg fluid dropped below the temperature of the fluid in the steam generator secondaries, and the steam generators becarne heat sources. The primary system pressure had been continu111y dropping throughout this cooldown period, and by 768 s had dropped to the saturation pressure of the fluid in the primary side of the tubes of Steam Generator A. This primary fluid then began to flash and create voids inside the. tubes. Figure 97 gives the mass flow rates in both cold legs of Loop A, which show a large surge of flow moving in the reverse direction starting at 77G s. At that time, the bubble in the primary side of Steam Ganerator A had grown j 130 I i
1 ,, i i i i c 0.8 - - TC
& 0.6 - -
2 o 0.4 - - o it 0.2 - -
' ' I '
0 O *000 2000 30C0 40C0 S000 6000 7000 Time (s) Figure 96. P r e ssuri zer surge fire breck, void fractionin tcp of R.V. [ downcomer annulus.
\.
1000 , , , , , ,
- LEG A-1 2C00 800 , --- LEG A-2 .
1500 g a 600 - - g l 6
.x . 9000 .0
- 400 - ,-,
- l. le
,l e o
)
0-- - ' e -0 o
? -
,, - ,, mu n - -
- 200 ,- t r (, r M I '-500
?
!l O W 2003 2000 4000 5000 6000 70C0 Time (s)
Figure 97. Pressurizer surge line break, loop A cold leg flow rates. 131 w-- - _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . . _ _ _ _ _ _ _ _ m ____ _
to the point where the buoyant force on the bubble exceeded the momentum force of the fluid trying to push the bubble in the positive, direction around the loop, and the bubble began to rise up the tubes of the generator. This resulted in the large reverse flow all through the A loop as the bubble pushed liquid aheadqof it down the hot leg, and pulled liquid
~behind it though both cold legs. By 815 s, the bubble had collected at the top of the candy cane and flow through the A loop stopped.
The flows encountered when the bubble moved to the top of the candy cane also served to redistribute mass around the primary system. Temperatures in the reactor vessel downcomer rose when these reverse flows pulled fluid from the core region into the downcomer. The primary system depressurization was virtually stopped when the fluid in the hot leg of Loop A began to flash. Bubbles began to form in the tubes of Steam Generator B at about 890 s, and as shown in Figure 98, by 1020 s enough voids had collected at the top of the Loop B candy cane to cut off loop flow. The primary system then began a smooth, gradual depressurization as the energy removed out the break exceeded the energy added from decay heat. From about 1000 s to the end of the transient, the fluid in the reactor vessel also experienced a gradual but irregular cooldown. Two-primary factors control the temperature of the fluid in the downcomer. They are the mass flow rates and temperatures of the fluids flowing through che vent valves and from the four cold legs. Three phenomena occur in this portion of the transient which prevent the cold leg flows from . stagnating completely, and therefore, mix the cold HPI water with the warm primary
~
fluid somewhat before it flows into the downcomer. The first of these phenomena was a manometer-type oscillation between - liquid levels in the upflow and downflow legs of the candy canes. After the bubble had formed at the top of each candy cane, there was distinct liquid level in the riser portion, which was balanced hydrostatically by the liquid level inside the steam generator tubes. As shown in Figures 92 and 93, when the loop flow stagnated, the cold HpI fluid injected into the cold legs quickly dropped the temperature of the fluid moving from the cold legs to the downcomer. As a result, the downcomer fluid became a little 132
O O 5 4 LEG B-1 2000
--- LEG B-2 1500
) 500 -
N 6 ; SN 3
, i i .
O O<- Ik e o . o .. ;;, , . -- O (, , . . 2 i ;4 h *La 44 Lied, f .i g , }.f, y f e vr ef' ; }j g f 2
-500 4 ,.
e - l
' ' ' , , -10C0
-500 ,
0 1000 2000 30C0 40C0 SC00 6000 i 70C0 Time (s) Figure 98. Pressurizer surge time breau loop B cold leg flow rates. l l o S i e 133
-em ---e-,-- - ~ - ,- ,~ +-w- n--- _ . . . - a .-..-g ,.,np, ,,e, 7 g y- ww- ,.e,w-w, ww-y v- --yrw,-w-,ymy- --w w =
cooler. Since the fluid was cooler, and therefore more dense, it exerted a slightly higher hydrostatic pressure. To balance this higher pressure on the cold leg side, a little fluid flowed up the hot leg and increased the liquid level in the riser portion of the candy cane. However, as the fluid in the downcomer became cooler, this also increased the differential pressure across the vent valves, and tended to open them more. The increased flow of warm fluid through the vent valves warmed up the fluid in the downtomer, and therefore, decreased the hydrostatic head imposed by the ' downcomer liquid. Fluid then flowed in the reverse direction down the hot leg to drop the liquid level in the riser, and again balance the fluid in the downcomer and inside the tubes of the steam generator. Warming the fluid in the downcomer also decreased the differential pressure across the vent valves, and the valves closed somewhat, decreasing the vent valve flow. The downcomer started to cool again, and the whole cycle was repeated. The overall result of this phenomena was that the hot and cold leg flows oscillated slowly back and forth, therefore mixing the fluid injected from the HPI with warmer primary fluid before it flowed to the downcomer. j Start,ing at 872 s in the A-loop and at 2378 s in the B loop, another mixing phenomena occurred which completely overshadowed the mixing encountered from the above manometer oscillation effect. Figure 97 shows that at 872 s, flow began to move internally within the two cold legs of the A loop. The flow moved in the positive direction from the steam generator lower plenum down Cold Leg A-2 to the downcomer of the reactor vessel, and then flowed in the reverse direction up Cold Leg A-1 back to . the steam generator lower plenum. In the A loop this inturnal circulation lasted until 952 s and then stopped, but then started attain at 1230 s and continued until the end of the transient. Figure 98 shows that in the B loop, the circulation flow started at 2378 s, reversed directions at 2443 s, and then continued to the end of the transient. The driving force for this flow is a difference in the temperatures between the fluids in the oipes connecting the steam generator outlet plenum to tha two primary coolant pumps on the same loop. There is an elevation change of about 5.2 m (17 ft) in this piping. When the 134
circulating flow is moving, the cold HPI water injected in,to one of the cold legs is swept up the pipe toward the steam generator and down the vertical section of pipe in the pump suction. The flow then moves through the outlet plenum'of the- steam generator and up the other cold leg.
. However, the steam generator outlet plenum is a large. volume, fi" ed with relatively warm fluid, so when the flow moves through it and is mixed with the fluid in the outlet plenum, it was warmed 8.3 K (15'F). As a result, the colder, more dense fluid in one of the pump suction pipes exerted a greater static pressure than the warmer column of fluid in the other pump section pipe. This difference in static pressure exerted by the two columns of fluid pushed fluid around the loop formed Dy the two cold legs.
A hand calculation was performed which verified that the differential pre >>ure due to the static heads of the two columns of fluid.was sufficient to force the fluid around the cold legs at the velocities encountered in the calculation, about 0.3 m/s (1 f t/s). It is not understood, however, what initiates this flow. From a geometric standpoint, both cold legs in a ( given loop are identical. One theory is that the mixing of the HPI flow in the cold legs by the oscillating manometer effect discussed earlier moves some cf the cold HPI water backward up the cold legs anc into the pump suction piping, and that due to stochastic differences in the oscillating flows, the fluid in one pump suction leg becomes a little colder than the other, and starts the internal flow. The third phenomena which affects the downcomer fluid temperature is the rapid ccndensation of steam bubbles in the primary system. Figures 94 and 96 show the void fractions in the upper head of the reactor vessel and the top volume of the downcomer. At 1400 s the bubble in the upper head began to collapse, drawing liquid into the upper head. This liquid contacted the hot metal structures in the upper plenun and immediately began to boil and produce vapor, forcing the* remaining liquid out of the volume, and slightly increasing the primary pressure. A few seconds later, at 1483 s, another surge of water entered the upper head. This liquid did not boil to steam, but began to condense the steam already there. As the steam was condensed, its volume shrank, drawing even more liquid in from 135
.___._._____i__ ______.__._ - _ _ . _ _ . _ _ _ _ . _ _ _ . . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - . _ _ . - _ _ - _ _ . _ - _ _ . - - - __
the upper plenum. The additional liquid condensed even more steam, and the condensation est.ntially fed itself. The end result was that the upper head steam bubble completely collapsed and filled with liquid in about 3 s. This unrealistically high condensation rate caused very large flow rate surges throughout the primary system as liquid rushed toward the upper head to fill the volume left by the condensing bubble. These large flow rates redistributed warmer fluid into the region of the downcomer, and temporarily raised fluid temperatures there. The large flow rates due to ~ the condensing steam bubbles also warmed the fluid in the downcomer to approximately the same temperature as fluid in the core. As a result, the differential pressure at the elevation of the vent valves was not sufficient to keep the valves open. With the vent valves closed, the colder water in the cold legs flowed into the downcomer and dropped temceratures there. The cooler temperatures in the downcomer caused the dif f aential pressure at the vent valves to increase again, and the valves reopened. Because of the relatively large flows encountered when the vent vahes open, this cooler fluid was soon flushed out of the downcomer, and temperatures returned to the point they were prior to the bubble collapse. This effect is seen as sharp spikes in the downcomer temperature between 1200 and 1700 s, as seen in Figure 90. After 1700 s, smaller bubbles are formed, then collapsed many times both in the upper head and in the top of the dcwncomer, however, the flow rates induced when these bub'o les collapse are not large enough to greatly perturb the downcomer temperature. Between 1700 and 5100 s, the primary system gradually cooled down and depressurized, as more energy was removed out the break then was added by decay heat. The mass flow out the break was only slightly greater than that input from the HPI, so the primary system was only gradually losing mass. At 2215 s, primary pressure dropped below the core flood tank setpoint pressure and some flow was delivered by the core flood tank. The mass flow rate from the core flood tank is shown in Figure 99. The , additional mass add,ed by the core flood tank shifted he mass balance so that slightly more mass was being added to the primary system than was leaving out the break. However, the core flood tank flow rates were never large enough to significantly affect either primary system pressure, or the downcomer temperature. 136
- . . . . _ _=_ _ _ . - . . . _ . . _ _ _ _ _ _ . - . __ _. . _ . _ _
e i i , f , e f J 4 I 300 . . , , , , 400 130 - . 7 300 7 N N
- E i w 50 -
. o
, y 2C0 O
{ 2 f 1 So-- 8 L Il I 3 I t ,
,oo
- 2 k) 1 0 i kV 6 o
.g- , , , , , , -10 0 O 9 00 2000 3000 40C0 $000 6C00 7000 Figure 99.
Time (s) Pr essurt zer surge Ree break, core nood f ew Inlaction ra t e. 4 1 137
At 5124 s, the primary system pressure dropped below 1.34 MPa (195 psia), and the LPI system began injecting fluid to the r,eactor vessel downcomer. The mass flow delivered by the LPI is shown in Figure 100. The initial LPI flow rate was sufficient to cause a more rapid cooldown in the downcomer fluid, but the large. flow rates also increased the primary pressure, which in turn reduced the LPI flow rate. By the end of the transient, at 6200 s, a quasi-steady state had been achieved where just enough flow was obtained from the LPI to keep the primary pressure very - close to the shutoff head for the LPI, 1.48 MPa (214 psia), but not enough fluid was added to cause a cooldown of the downcomer fluid greater than that seen just prior to LPI initiation. Based on the cooldown rates experienced just prior to LPI initiation, downcomer temperatures would drop to 355 to 361 K (180 to 190*F) if extrapolated to 7200 s. Because of the action of the LPI system, primary system pressure would stay close to the LPI shutoff head of 1.48 MPa (214 psia) out to 7200 s. The temperature of the fluid in the downcomer of the reactor vessel is , dependent on the cold leg mass flow rates and fluid temperatures, and the vent valve mass flow rate and fluid temperatures. The following paragraphs will identify and assign uncertainties to those phenomena predicted by RELAPS which af fect the downcomer liquid temperature and which could possibly have been insufficiently characterized in the calculation. Two such phenomena have been identified, the unrealistically high condensation rates in the reactor vessel upper head and drwncomer, and the internal cold . leg flows which started at 872 and 2378 s in the A and B loop cold legs, respectively. Whenever the large condensation rates collapsed the steam bubble in the reactor vessel head, a spike occurred in the downcomer temperature for reasons discussed earlier. If the condensation had occurred at a more reasonable rate, the large flow rates calculated around the primary system would not have occurred and the vent valve operation would not have been as severely perturbed. As a result, the spikes in the downcomer temperature 138
i 150 4 i i , , 300 7 250 7 No 00 . N E
$ ~200 D
, c o
E . 15 0
, I 0 a 50 -
/,
C /. 1
= A
\ 2 ' 80 0 a O
f 2 0 1 50 0 ' ' ' ' l 0 0 1000 2000 30C0 40C0 5000 6000 7000 Time (s) Figure 100. Pressurizer surge fine creck. Iow pressure f eiection r a t e, f t l l l l 139 I
between 1200 and 1700 s would not have occurred. However, the absence of the spikes would not have affected the temperature of the flyid in the downcomer af ter the condensation induced effects were over, af ter 1700 s. The downcomer fluid temperatures would continue the gradual cooldown, with no significant change from the calculated result. Thus condensation effects in the reactor vessel upper head have little effect on the final downcomer fluid temperature. The internal cold leg flows caused mixing of the fluid within the cold legs, instead of allowing them to remain stagnant. Because the cold legs were mixed, warmer fluid was delivered to the downcomer volume, thereby maintaining higher downcomer fluid temperatures. A sensitivity calculation was performed to assess how cold the downcomer fluid would have become if no mixing at all was allowed between the fluid injected by HPI and the fluid in the cold legs. In this sensitivity calculation, the fluid injected by the HPI system was mixed directly with the fluid flowing into the downcomer from the vent valves, and a mixture temperature determined. The calculated mixture temperature is shown in Figure 101 along with the RELAPS generated downcomer temperature. The downcomer did not become j significantly colder in the se'isitivity calculation because the temperature of the downcomer fluid is controlled mainly by the temperature of the fluid passing through the vent valves. Tnis was becaase the vent valve flow rate shown in Figure 102 was much greater than the flow injected by the HPI system shown in Figure 103. However, although the temperature of the downcomer did not drop significantly when no mixing was assumed, the temperature of the fluid in the cold legs flowing into the downcomer would , drop to the temperature of the fluid being injected by the HPI system, 283 K (50*F). , Figures 104, 105, and 106 show the downcomer pressure, fluid temperature, and heat transfer coefficient extrapolated to 7200 s. As shown in Figures 104 and 105 uncertainty bars have been added to i the RELAPS-calculated reactor vessel downcomer fluid pressure and I temperature curves. A comprehensive study of the uncertainties in the i calculation is beyond the scope of this work. However, an estimate of 140 i
I I I I 4 6
^
M 600
* - NO LOOP rL0w e
--- wiTH LOOP FLOW
' 550 ,
s . U . 500 ^ e e ' gpvI , 4 500 - ' '
- =
E ' e i s'"e' 3
~
,\ .,&* '400
- a
- 450 -
l .
\g -
h , 's. E
= s. 300 =
l400 -
/* - -
1 '*'.. y 350 ' ' ' ' '
/' 200 O 1000 2000 30C0 40C0 5C03 6000 70C0 Time (s) figure 101. Pressurizer surge line breck, R.V. downcomer fluid temperature with and without cold leg circulation.
4 \ 4000 , , , , , , 8000 3000 . n 6000 7
\
en E
$ 2000 -
3 4000 v o I G: 0 1000,- -
, 2000 r
n w
~
i
, y)/%[W/S:g;:y p44 _ ,
i l
.g< > i i i i i
-2000 0 MOO 2000 3000 40C0 5000 6000 70C0 Time (s)
Figure 102. Pressurizer surge line br eak, vent volve mass flow r a t e. 141
- , - - , - - - - , - , - - , a-, - - , . ,, . - - - - - - - - - - , - . . - - - -
80 , , , 15 0 m 60 - - Q .
\
m E 6 10 0 e. w A ~ E
= 2 l n n a o 0 a
2 20' - 0 O O 1000 2000 30C0 4000 SC00 6000 7000 Figure 103. Pressuriter surge line brook, total FPI rmess flow r o t e. s 20 , , , , , , , j'1
- CALCULATED
- - -. EXTRAPOLATED 2500 m ^
O O
$2 2000 v i
W = U $ N-w _ e00 : oc w Q- o. W 1000 o 3 a s - - E 2 O o
> 1 500 >
L' ' ' 0 0 0 1000 2000 3000 4000 50C0 6000 7000 8000 Time (s) Figure 104. Pressurizer surge Gne breck, entropolated R.V. downcomer , pressure. 142
l l ( l
- S00 , , , , , ,
^ 600 ^
M - CALCULATED e
--- EXTR APOLATED e e
- - 6 2' 550 ..
2 300 3 0
- n.
1 500 E -
.N'- n [ --
" 9 ..
~400 m ,
- 450 -
.m_
b \ h
= N 300 =
{400
' ~~
G -
> 200 > 0 3g i i 0 1000 2000 3000 4000 5000 60C0 7000 SOCO o ressurizer Time (s)
I 'gur e 105. surge line break, en tropolated il v. wowncorrer
, fluid imperature.
i
\
30 ' ' i i , , ,
>5000
- CALCULAYE0
* "- EXTR APOLATED .
c e e 4000 %
.2 =Y io g ** e o1 un- 3000 u"
- i. C 6 g eu
.e5 $ *A
- ht 2000 E o)_
0 10 - - g 28
==
0
- WOC $
. $ .1l ,
& 1
, i , , , , , - - ,- -
0 0 0 1000 2000 3000 4000 5000 6000 7000 80C0 Time (s) Figure 106. Pressurtzer wrge Sne break entropolated R.V. downcomer InsIce surf ace hoot tronsf er coefficient. 143 . l
-------~.-.x _ _ _ _ _ _ _ _ _ _ " - - - - - - - - _ _ _ _ _ . _. _
l l l I l l c uncertainty in the calculation was required by the analysts performing fracture mechanics calculations who will use the o wncomer p,ressure and temperature-time histories as boundary conditions for further analyses. Therefore a limited uncertainty estimate has been made based primarily on i insights gained from the comparison of counterpart TRAC and RELAP5 calculations presented in Appendix C. l i The uncertainty bars were constructed by accounting for three
- effects: (a) break energy removal differences, (b) cold leg circulation uncertainty, and (c) main feedwater pump trip uncertainty. Each of these will now be addressed separately.
As discussed in Appendix C, a difference between the TRAC and RELAPS break energy removal was observed. This difference was due to more liquid being available at the break with RELAP5 than with TRAC. As a result the break energy r moval was higher with TRAC resulting in a higher primary system depressurization with TRAC. Since this difference was unresolved, it represents an uncertainty in the primary system depressurization and cooldown rates. The lower end of the uncertainty bars were constructed by ,/ assuming the higher TRAC depressurization and cooldown rates'. The effect of stopping the RELAPS-calculated cold leg circulation on downcomer fluid temperature was presented in Figure 101. An average -20 K (-36'F) shift in reactor vessel downcomer fluid temperature was used in constructing the lower end of the uncertainty bars. It is estimated this uncertainty does not significantly affect the primary system pressure. . The effect of uncertainty in main feedwater pump trip, discussed in Apperdix C, was estimated to be a +40 K (+72'F) change in reactor vessel - downcomer fluid temperature. This estimate was made using Figure C-2 as a , guide. It is estimated this uncertainty does not significantly affect the primary system pressure. 144
7.4 Conclusions . The minimum downcomer fluid temperature calculated was 380 K (225*F) and occurred at the end of the calculation, 6200 s. By extrapolating . trends evident near the end of the calculation, the downcomer temperature would drop to 355 K (180*F) by 7200 s. The minimum cold leg temperature calculated (adja _9t to the reactor vessel) was 333 K (140'F), which occurred at 1130 s when the primary system pressure was 6.1 MPa (740 psia). At the end of the calculation, the primary system pressure was at the LPI shutoff head,1.48 Mra (214 psia). Since LPI injection was holding primary pressure at this point, it would be expected to stay very close to that pressure through 7200 s. 6 9 145
- 8. REACTOR COOLANT PUMP SUCTION SMALL BREAK TRANSIENT The following subsections contain the transient . scenario description, modeling changes effected to perform this calculation, detailed analysis of the transient results, and conclusions drawn from the analysis.
8.1 Transient Scenario Description
. The transient is initiated from full power steady state (nominal temperature and pressure). The pressurizar heaters and spray ooerate as designed. The transient is begun by a break downstream of a valve in the letdown line connected to the bottom of the A-1 pump suction leg.
Following reactor scram the decay heat is assumed to be the ANS standard. The turbine stop valves close and it is assumed the integrated control system operates as designed. After the core flood tanks empty the break is isolated and high pressure injection flow is throttled when 28 K (50*F) subcooling is attained, but only if the primary system has regressurized to the power operated reliaf valve (PORV) setpoint. A transient scenario is provided in Table 20. ,, 4 8.2 Model Changes s The primary change made to the steady state model, presented in Section 2, was adding a pipe and valve component to the A-1 pump suction , leg. As stated above the break was in the letdown line which is a 0.0635 m (2-1/2 in.) diameter line. The break was assumed to occur downstream of a , motor valve that could be isolated. The distance from the primary piping l to the valve in the line was approximately 15.24 m (50 ft). This line was modeled with three volumes. The valve throat was assumed to have the same - inside diameter as the pipe. 8.3 Transient Results This section presents the results of the RELAPS pump suction leg small break transient. A sequence of events of the calculation is presented in l ! Table 21. l 146 l-
,i TABLE 20. PUMP SUCTION BREAK TRANSIENT SCENARIO
- 1. Break in letdown line connected to the A-1 pump suction leg occurs downstream of an isolatable valve.
. 2. Reactor scrams, turbine trips, turbine stop valves close, heater drains close, and the integrated control system operates as designed.
- 3. Turbine bypass valves and safety relief valves on steam generator secondary operate as designed.
- 4. Main and emergency feedwater systems function as designed.
- 5. High pressure injection initiates at setpoint.
- 6. Operator trips reactor coolant pumps 30 s after high pressure injection actuation and main feedwater is realigned to EFW headers.
- 7. Core flood tanks dump; low pressure injection system actuates.
- 8. Break isolated af ter core flood tanks empty.
- 9. High pressure injection throttled when 50*F subcooling occurs but after the primary system repressurizes to the power operated relief valve setpoint.a
- 10. Maintain 50*F subcooling.
- a. In calculating subcooling with the reactor coolant pumps off, pressure for each loop is taken from the top of the hot leg riser, and temperature is taken from the core outlet.
_G 147
TABLE 21. PUMP SUCTION BREAK TRANSIENT SEQUENCE OF EVENTS Time Event (s) Break in Loop A-1 pump suction leg occurs 0.0 Reactor scrams (P, T relationship), turbine stop valves close, 48.60 heater drains begin to close, turbine bypass and steam generator secondary safety valves operate as designed. . Pressurizer heaters latched off due to low level in pressurizer. 60.92 Main feedwater pumps trip on high discharge pressure. 75.88 Primary pressure dropped below high pressure injection setpoint 94.32 1 and injection began. Upper head began to void. 100.0 Reactor coolant pumps trip 30 s after high pressure injection 124.32 i initiation. Main feedwater flow realigned to the EFW headers. Motor and turbine driven emergency feedwater systems initiated due 124.36 to low level limit in the steam generator secondaries. Steam Generator A and B turbine bypass valves close. 135.0 ,
)
i Upper head completely voided. 424.0 Steam Generator B secondary pressure dropped below the booster 500.0 pump head in the main feedline and main feedwater commenced to
- flow into B generator secondary via the startup valves and l
Crossover, i Steam Generator A secondary pressure dropped below the booster 545.0 ' l pump head in the main feedline- and main feedwater commenced to l flow into A generator secondary. via the startup valves and . I Crossover. l Steam Generator B emergency feedwater flow stopped due to liquid 573.71 5 level above the low level signal. , High level limit signal ir. Steam Generator B reached due to flow 596.97 ! through startup valves. Loop natural circulation established. 600.0 l i - 148 i a
TABLE 21. (continued) Time Event (s)
. Steam Generator A emergency feedwater flow stopped due to liquid 631.90 level above the low level signal.
High pressure injection volumetric flow rate exceeds the break 650.0 volumetric flow rate and primary system depressurization essentially stopped. High level limit signal in Steam Generator A reached due to flow 660.80 thr3 ugh startup valve. ICS ramped B startup valve closed, isolating Steam Generator B 675.0 secondary. Vent valse flow commenced. 680.0 Level in pressurizer began to recover. Void in upper head began 700.0 to collapse. ICS ramped A startup valve closed, isolating Steam Generator A 731.0 secondary. Primary mass lost out the break regained by high pressure 900.0 injection mass. Upper head became liquid solid. Primary system began to 1350.0 repressurize Natural circulation lost in Loop B. 2000.0 Drimary pressure turned over. Volumetric break flow exceeded 2200.0 volumetric HPI flow. Flow in loop B reversed. 2380.0 Flow reversal in Loop B recovered. 2800.0 Series of flow reversals in Loop B mixed cold HPI water with warm 2800.0 primary liquid in the Loop B cold legs. to 4900.0 Transient terminated. Pattern established in transient to enable 4900.0 extrapolation of downcomer pressure and temperature to 7200 s. Downcomer pressure and temperature at termination of transient were 6.0 MPa (870 psia) and ^70 4 (386 F), respectively. Extrapolated downcomer r-v>; ' a .c fluid temperature are 5.1 MPa 7200.0 (740 psia) and 446 K J. 7 aspectively.
. .~ -
The initiating event was a break in the letdown line downstream of an isolatable valve. The letdown line is connected to the A-1 pump suction leg. The primary system rapidly depressurized, as shown in Figure 107, resulting in a reactor scram at 48.6 s due to viol,ation of the pressure / temperature relationship scram criteria of the Oconee plant. At the time of reactor scram the turbine stop valves closed, which produced a rapid increase in the steam generator secondary pressure, shown in Figure 108. The turbine bypass and safety valves operated as designed to
- arrest the secondary pressure increase. Also, the feedwater heater drains were closed over a 5 s period starting at the time of scram. At 60.9 s primary mass inventory had decreased enough that the pressurizer heaters were turned off due to low liquid level in the pressurizer.
At 75.9 s the main feedwater pumps were tripped due to high discharge pressure. At reactor scram, the integrated control system (ICS) ran back the main feedwater (MFW) pumps to their minimum value of 447 rad /s (4270 rpm). The MFW and start-up valves, also controlled by the ICS, began to close. The closing response time for the MFW valves was faster than the 3 start-up valves and these valves were closed at 62 s as shown in _, Figure 109. It was not until 74 s that the ICS began closing the start-up valves as shown in Figure 110. When the start-up valve began to close, the pressure upstream of the valves began to increase due to the head produced by the MFW pumps. By 75.88 s, the pressure upstream of the valves increased above the high discharge setpoint and the MFW pumps tripped. When the pumps tripped, the ICS began to open the MFW and start-up valves. The opening response time for the start-up valves was faster than the MFW valves, therefore, the start-up valves opened up almost immediately after the pump trip, as shown in Figure 110, while the MFW valves remained closed i because a stronger signal from the ICS was needed to open them. - The primary system depressurization rate increased due to volumetric liquid shrinkage. The primary pressure dropped below the high pressure injection (HPI) setpvint at 94 s. At 100 s the primary pressure dropped below the saturation temperature in the reactor vessel upper head, liquid flashed to steam, and the upper head began to void as shown in Figure 111. 150
If
, ~
F yOaD1w AmwMM$y V3L7 F i
>O.s33w AawMMDEW h 7 gi g g m 1 2 M W u 4 8 8 u 8 r O ,- r 0 '
e - e - . . 1 1 0 5 0 5 . 8 0 7 0 _
. 0 I
, 0 t P
1
- P 1
0 0 u u 0 0 m 0 I , m 0 t p ' p
' 1 1
s 5
- s 5 u 0 - u 0 ct C I t
n 1 t i oi 2 o 2 n 0 b a n r e 0 0 0 2 I
, b r
e a 0 0 2 t [ . T05 5 1 ,k i0T0 .
,k i0 _ .
5 I , t 1 S. m 3 .N' hm e e0 3
.G e 00 t 0
m [H, ge(0 l s(0 t i
)s 3 I -
es c)3 o 5 p 5 n 0 r 0 c 0 SS . e 0 t _ or TT s EE . 4 s 4 y C A A u C p C MW r C O I , e C t r . GG s e
}H.
s s 4 5 BE EE 4 5 0 u 0 . RR 0 . r 0 I . , t LL AA e
- s. 5
.,. TT 5 CO OO 0 . OO 0 PD .
C . RR 0 0 I . 0 1 BA SA 5 5 5 5 0 0 - - - 0 . _- - - - 0 - . .. - 4 6 8 1 0 1 0 g 2 0 0 0 0 0 0 o 0 0 0 0 0 0 o 0 yo _5 , cseemg vC 2
. L T >o- 3Eo Cweaa36e L a* 7
, y
~
3 $ B .
- l WAIN FEED VALVE A -
i--- WAIN FEED VALVE B w 0.8 - - e . 4 w J 0.8 - - c w N 0.4 - - J 2 /.
-m O 0.2 - -
2 0 O 20 40 83 80 10 0 t Time (s) Figure 109. Pump suction brook, normalized mcIn feedwater volve cr eo. 6
/
1 . i i i 4 . w : m *
< l w *e
>a >
0.3 . . c w N a ~ E START-CP VALVE A
--- START-UP VALVE 9 l.
O Z 0 0 20 40 60 80 10 0 Time (s) Figure 110. Pump suction break, normalized startup volve creo. . I 152
I I I & I l 3 I I 3 C o U Q M 2 0.5 - 1 0 6 0 1 O O O $00 0 00 1500 2000 2500 3000 3500 4000 4500 50C0 5500 Time (S) Figure 111. Pump suction br eak, reactor vessel upper head void froc11on. G 153 9 _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ - - . _
The production of steam in the upper head momentarily stopped the depressurization and then resulted in a decrease in the deprassurization rate as shown in Figure 107 as the upper head voided. Also contributing to the decrease in the depressurization rate was the trip of the reactor coolsnt pumps, 30 s after HPI initiation. Coincidental with the reactor coolant pump trip at 124 s the main feedwater was realigned to the emergency feedwater headers and the steam generator secondary low level signal was increased from 0.6096 m (24 in.) to 6.096 m (240 in.). At this time the level in both generators was below the new low level signal and emergency feedwater was initiated to both , steam generators as shown in Figure 112. As a result of injecting cold emergency feedwater into the secondary, the secondary depressurization rate increased as shown in Figure 108. At 135 s, the turbine bypass valves closed as shown in Figure 113. At 424 s the upper head was completely voided .:nd the depressurization rate increased until the pressure reached the saturttion condition of the fluid in the top of the reactor vessel downcomer. 'he fluid in this volume , flashed, slowing the depressurization rate. By 550 3 this' volume was essentially voided and the depressurization rate ao.in increased until the pressure reached the saturation condition of the flu'd in the next lower volume. At 500 and 545 s, respectively, the pressure in Steam Generators B and A r.econdaries dropped below the pressure in the main feedline which was ' primarily determined by the condensate booster pump head. Main feedline < fluid commenced flowing into the steam generator seccndaries via the emergency feedwater headers fed tnrough the start-up valves, as shown in Figure 112. The secondary liquid level rose above tne low level setpoint, terminating emergency feedwater, at 573 and 631 s for Steam Generators B and A, respectively. At termination, the temperature of the fluid entering the secondary via the emergency feedwater connection increased, thus decreasing the subcooling cf the liquid entering secondaries, and resulting in an increase in secondary pressure, shown in Figure 108. Flow from the main feedline to the steam generators continued, however. -t a decreased 154
~ __ _____ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
-200 , , , . . . . . . i
- STEAW GENERATOR A 400 j --- STEAM GENERATOR 8
. 15 0 ;:l n
- 300 m
( o, !. g
\
E 6 100 -
! i, t 200 v S
D ! !
$ 50 ,-
[!; 1 j - 10 0
- k. . $
3 o
- l_ ..li), s 0-- -
0
.g. > > , e i i . e i i '00 0 500 1000 1500 2000 2500 3000 3500 4CCC 4500 5000 5S00 Time (s)
Figure 112. Pump suction brook, emergency f eedwater flow ro t es. 300 . , . . . .
- STEAM CENERATOR A 600
--- STEAM GENERATOR 8
. 500 7 N
N, 200 - - g 6 , 400 v3 R o % g 300 o m - n 10 0 e o -200 a 0 2 2 Lf fg' . sgo C ' ' ' ' ' 0 O O 10 0 200 300 400 500 600 700 Time (s) Flgure 113. Pump suction break, turbine bypass flow roles. 155
rate because of the repressurization of the secondary systems. At 597 s the high level limit in the B steam generator was reached, sending a signal to the ICS to close the B start-up valve. At 675 s the B startup valve was closed, isolating the B steam generator. Similarly, the high level limit in the A steam generator secondary was reached by 661 s and the A start-up - valve closed by 731 s, isolating the A steam generator. Shortly after the isolation of the secondaries'the steam generators became heat sources to the primary, and the secondary side pressures declined as snown in Figure 108, due to cooling of the secondary as heat was removed from the secondary to the primary. Natural circulation was essentially established in the prima *j loops by 600 s as shown in Figures.114 and.115. After the steam generators became heat sources to the primary fluid, the temperature of the primary fluid exiting the steam generators was warmer than the fluid entering the generators as shown in Figures 116 and 117. A density gradient was established in the loops such that the fluid density entering the steam generator and in the cold legs was higher than the fluid density leaving s the steam generator (see Figures 118 and 119). The low c3nsity fluid ,) between the two high density fluids tended to slow the loop mass flow rate as shcwn in Figures 114 and 115. At 650 s the volumetric HPI flow rate exceeded the volumetric break flow rate as shown in Figure 120, and the pressurizer level began to increase as shown in Figure 121 and the primary depressurization was stopped as shown in Figure 107. Also, liquid began entering the upper , head, resulting in condensation effects which resulted in the primary pressure oscillations shown in Figure 107, pressurizer liquid level oscillations shown in Figure 121 and the upper head void fraction oscillations shown in Figure 111. By 1350 s the void in the upper head had collapsed, the pressurizer level rapidly increased and the primary system . repressurized. At approximately this time, the mais lost out of the system through the break was exceeded by the mass injected into the system by the HPI as shown in Figure 122. 156
4 000 , , , , , , , , ,
- HOT LEC 20M0 f{
COLD LEG (4-2) l X COLD LEG (A-1) 8000 m a 850C0
) , o
$ 5000 -
O 3r 4000 _
- O M
y 2 2 5000 2000~ -
-- t m.
0 500 1000 1500 2000 2500 3000 35C0 4C00 4500 5000 5500 Time (s) [ Figure 114. Pump suction breck, loop A hot and cold leg moss flow
\ ref oe.
M
- a e i i i , , ,
HOT LEG 200C0
't ---
COLD LEG (B-2) 8000 - X COLD LEG (B-1) e m 15000 a N
\", 6000 - _
e
- .a v
10000 O j g goo . , ' = C - n - 5000 , g 2000 - _ 3 E 2 l 0.. '< g- 4: .g ,o ( -2000 ' ' l
~
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Time (s) Figur e 115. Pump su:llon brook, loop B hot and cold leg mass flow roles. l I i ( 157 F l
\.
I a 600 , , , , , , , , , ,
^ ^
M INLET &a-
..- CUTLET e 580 w
- e w .
2 3 O o
' S40 - .,
- 550 %
- e Q. '. a E '.. E o
- 540 -
.o 7 '... 'E g
.&520
....., - a
= =
. - .~. . .% - - .....' ..,0 e E E 3 500 3
0 -0
> , , i i i i i g i i i 0 500 1000 1500 2000 2500 3000 3500 40C0 4500 5000 5500 Time (s)
I"Igure 116. Pt.rro suc tlon brook, Ioop A S.G. primory inIet and outiet fluid temperatures.
}
J 600 , , , , , , , , , ,
^
n .. 600 M 1. - INLET W
--- CUTLET e e w 330 . u 3 % s
- * . . . . . . . . . . . . . . . . . 500 -
o p o
. j . k E
k300 : i - m
- y
- 4M - -
3 .,*
,. .. y -
y y
= f I,. ' *, ,300 =
*[ ( '.,,. g l400
-3 .m O 200 > 0
> , i . . .
g i i , i i 0 500 1000 1500 2000 2500 3000 35C0 4C00 4500 5000 5500 Time (s) Fi gur e 117. Pump suction brook, loop 8 S.G, prirrory inlet and outlet fluid temperatures.
. 158
m 950 , , , , , , , ,
- STEAW CETRATOR IN
--- STEAW CENERATOR 007 p 800 -
X COLO LEG - O E 55 : N , M \ E m aso - _ 6 - 0
~~' ._ " " . ... ......
5 a00 -- ,... ,,,,..... 50 y e . . e .
- c D
- e
'O O
7503 ' - o 45 .O -
- 7% .
O 650 O 500 1000 1500 2000 2S00 3000 3500 4000 4500 50C0 $500 Time (s) Figure 118. Pur9p suc tion break, loop A S.O. prireary Inlet and cutlet f, fluid denstiles. N . . . i . , , , , . STEAM CENr.RATOR IN 65
^
pg . --- X STEAMCOLD LEG CE!.ERATOR QUT l E N : * *- s0 N E O , e ., ,
.* , ,, i j .o 300 -
l - , A ,,,, - . j 55 c h w 4 ** - 50 5 m _U M - C
=
[ 700 .
]
w 40 600 O 500 1000 1500 2000 1500 3000 3500 4000 4500 5000 5500 Time (s) Figure 119. Purnp suc tion break, loop B S.G. primor y inlet and outlet fluid densities. 159
0.2 , , , . , T m -- ~ r i
- VOL FLOW 162
--- TCT HPI VOL FLO 7 7 .
.N.
N m v E l v 5 ) o 4 o 0.1 'I - l i
.o.
l f l! ' l 0 d ............................... e E ,- W <
,, .,- E 2
r . . C e O
> l >
0.0 O O 500 1000 1500 2000 2500 3C00 3500 4C00 4500 5000 5500 Time (s) FlyJre 120. Ptmp suction break, breck ord HPl volumetric flow rates.
)
/
a , , , , . . . . . . 25 .
, 6-- - 20 m 15 E v v e 13 e>
24 - 2 m u 5 ** 5cr .
.e. -
J 2 5 O L ' !' ' ' ' ' ' ' ' ' O O 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Time (s) - Figure 121. Pump suction break, pressurizer colicpsed liquid level. 160 .
400000 , , , , , , , , , ,
^ INT WFLOWWJ102 n m --- TOT INT HPI FLOW 8.0 0*10' E
.at h v
g 300000 -
- w o I C ,
8.0 0 *10' 3
.- ~
m =* a n E 200000 - a o
*,- 4 00+10 E 2
, ...- =
0 *
$ g)oooo -
n
,, - w
,~ 2.0 0 *10, m
.e .- o c **
_C
~
0 ' O.00 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Time (s) Figure 122. Pump suction break, break and HPl Integrcled mass flow rates. e G 161
The oscillations observed in the break flow (see Figure 120) were i caused by unsteady flashing phenomena occurring in the pipe,between the pump suction and the break. During the primary system depressurization and i repressurization, the fluid exiting the break oscillated between single phase and two phase due to the fluid temperature in the break pipe being near or at saturation. This oscillation resulted in oscillations in the mixture density, which is used to calculate the break flow rate. As the primary system repressurized the fluid exiting the break became single phase, subcooled and the oscillations damped out. At 22u0 s the break mass flow rate steadied out at a value near the HPI mass flow rate, as shown in Figure 123. The primary pressure, however, turned over because the volumetric break flow exceeded the volumetric HPI flow as shown in Figure 120, and the primary system slowly depressurized. By 1800 s the primary system was water filled except for the top of , the pressurizer. Natural circulation continued in Loop A due to the break and HPI acting as heat sinks to remove decay heat. In the 8 loop, however, natural circulation was essentially stopped due to the steam generator changing from a heat sink to a heat source. Loop flow oscillations, which ,) may have been established due to asymmetric flows and density gradients between the loops and the vent valve flow, further perturbed the B loop flow rate as shown in Figure 124. Cold fluid from the HPI began to pool in the B loop cold legs, further slowing the loop flow rate until the HPI inflow exceeded the cold leg flow into the reactor vessel. This excess injected HPI back flowed in the cold leg toward the reactor coolant pump. l This back flow occurred at 2250 s and resulted in an increase in the fluid density of the volumes upstream of the HPI connection point. The duration of the back flow was 20 s as observed in Figure 124 Another similar back flow occurring at 2320 s and lasting about 30 s further increased the density of the fluid in the cold leg upstream of the HPI connection. 1 Finally at about 2390 s a third reversal resulted in a density gradient between the pump suction leg and the steam generator tubes (see Figure 125)
-such that a complete flow reversal in the 8 loop occurred as shown in Figure 18. As a result of the B loop flow reversal, the A loop mass flow rate increased as shown in Figure 114. The flow reversal lasted 445 s, during which time the cold fluid in the B loop cold leg traveled through 162 o
e 10 0 , , -- T - - r t-- t- 6 7-a - BPEAK FLOW l .gon
--- TOTAL HPl FLOW i 80 -
m n I a
.* ~~ - - * - A +g ;,4. . . . ................_... = %
- c. L
(
! i, , i
.tso E
,f,, 60 , ; i E
e %.# 3 l I I O g g l 1 W ,s i 1C0 o a $. e n . o . w 0 2 l 50 2 20 .-l l l 0 - O O 500 1000 1500 2000 2500 5000 35C0 40C0 4500 50C0 5500 Time (s) Figure 123. Pump suction brook brook and HPl mass flow rates. 60 . . . 4
- IPsLET FLOW
--- OUTLET FLCW 10 0 40 -
X Ifl FLOW - n . .....' ;. n e
--. I'. 30 N
(, 20 . s:x _
.....,L..- / '. L '. ml i.
E
.w ' '
.< ... ., 0 v
.. v
$ 0-- ,
0 3
= -
2 m . g-20,-
- _g a 2
- O 2
B
-40 - -
'! .t00
-60 1600 1800 2000 2200 2400 2600 Time (s)
Figure 124. Pump suction break, cold leg B-2 moss flow rates in tre region of the HPI Injection site. 163
110 0 , , , , PWP SUCTION LEG b- COLD LEG 6S , m 3 TEAM GENERATOR OUT ,,, a . E 1o00 - - N si:n .***-
- if . / '. E v
m **....,,... .... ,. . i 60 m-v
= ,no , .
-t-c -
55 m e 'f c w a t _.O -
~ 900;- ^f} j -
50 ~C o , m s_
> o
-l' 45 700 1800 1800 2000 2200 2400 2600 Time (s)
Figure 125. Pump suction brook. fluid densifles in loop B cold leg pump sueilon and S.G. outiet rogions. 4 164 a .A---- - - _ _ - _ - - - - - - - - - . _- - - - - _ _ . _ - . _ - ---
the pump and pump suction leg and into the steam generator tubes, while warm fluid from the reactor vessel downcomer entered the cold legs. At 3165 s the density difference between the fluid in the steam generator tubes and the cold legs was such that the flow reversal stoppeo and normal
. loop flow was reestablished. Similar flow reversals occurred throughout the remainder of the transient. The magnitude of the reversals decreased with time as shown in Figure 115. '
The transient was terminated at 4900 s after a pattern in the system pressure and temperature had been established. The primary system was experiencing a continuous cooldown which was expected to continue. The depressurization of the primary system was not sufficient to reach the core flood tanks injection pressure. Thus core flood tanks flow was not established as per the scenario and the break was not isolated. After HPI initiation the HPI volumetric flow exceeded the volumetric break flow and the primary system repressurized. At 2200 s an equilibrium between the creak mass flow rate and the HPI mass flow rate was established, however, the volumetric break flow rate exceeded the volumetric HPI flow rate and the primary system slowly depressurized. Had the calculation been continued, an equilibrium between the volumetric break and HPI flows would have been established and the primary system pressure would have stabilized. The flow reversals in the B loop which began at 2390 s effectively , increased the downcomer fluid temperature, as shown in Figure 126, by l mixing the cold HPI fluid with the warm downcomer fluid. If the flow reversals continued, the effective temperature at 7200 s was extrapolated ~ to be 446 K (343'F). The corresponding vessel downcomer pressure, shown in Figure 127, at 7200 s would be 5.1 MPa (740 psia). Had the flow reversals in the 8 loop not occurred, cold HPI liquid from the B loop would have
- entered the vessel downcomer at the rate of the 8 loop HPI injection rate.
. This cold liquid would have mixed with the vent valve flow and the flow from the A loop cold leg. The overall result would have made the vessel l downcomer temperature colder than calculated with RELAP5. Uncertainty bars l hn e been incluaed in Figure 126 to represent the uncertainty in the liquid temaerature history if the flow reversals had not occurred. As shown the colcest downcomer fluid temperature that would have been achieved, had l
l 165 l l
Volume liquid temperature (K)
, VOLUME PRESSURE (WPa) ,
8 8 5 Y $ $ $ $ I= o" . , ;a , , G 5 n. i
.,8 -
2 I n g E 3 5$ S - - 2E $ -
=
E !; 8 w 3 8 l 8 - -
.i 3
- Sw 71 $r g '
-8 f!8 - _
j!$ id n
'v8
* ?,,,u ,
8 8
,'d
- o - r . .
e :* s *
- - 8 *
. s
- 8 o n 8 .8 00 gr -
g o - F.g .
;
- g I gr g , ' >Q ~ , 28
= g $>
>M g
3 gA o f t ,a no 8 O
= . a :
E g E E . :
" o . '. .i .
E 8 5 h $ Volume presssrt (psfo) V lume iquid temperature (T) 9 l w
the flow reversals not occurred, was 390 K (242'F). The primary system pressure history would not be affected by the presence or absence of the ficw reversals. To aid in future fracture mechanics calculations, the reactor vessel wall inside surface heat transfer coefficient is shown in
. Figure 128.
8.4 Conclusions For the reactor coolant pump suction break sequence, the downcomer fluid temperature was 470 K (386'F) at the end of the calculation (4900 s). The transient was extrapolated tc 7200 s (2 h) and, at that time, the downcomer temperature is estimated to be 446 K (343*F) at a pressure of 5.1 MPa (740 psta). p The major uncertainty in the calculation is related to the existence of loop flow oscillations. Had these oscillations .at been p' resent in the calculation, the downcomer fluid temperature at 2 h is estimated to be 390 K (242'F). The sequence requirement to isolate the break (by closing the motor block valve in the letdown) was not encountered since core flood tank injection did not occur. Had this isolation occurre; the subsequent maximum primary pressure would have been ~16.99 MPa (2465 psia), the opening pressure setpoint of the power operated relief valve. l 167 r ,, y , _ -., - . - , -
e t 30 , ,
- CALCULA*ED
\ --- EXTR APOLATED y,
. W c -
e
- 4000 Om 6 u 6
-0 20 -
4o
*X m
dd ud s000 o= mE m w 'i
- 6 2
- M\
"6 2000 23
-$ g -
<~
g CD w >v o 900 & S E t
. - - =. .. .......
Y O O 2000 4000 6000 8000 Tim. (3) I'Igure 128. Pump suctiom broeir. reactor vesset downcomer insio, surf ace heet t*enef er coefficient. 2 168
- 9. STEAM GENERATOR TUBE RUPTURE TRANSIENT 1
The following subsections contain the transient scenario description, modeling changes effected to perform this calculation, detailed analysis of
. the transient results, and conclusions drawn from the analysis.
9.1 Transient Scenario Description The transient is initiated from full power steady state conditions (nominal temperature and pressure). The pressurizer heaters and spray operati as designed. The transient begins with a rupture of a single steam generator tube (double ended break) in the A loop steam generator. It is assumed the ICS operates as designed. The rea.: tor scrams and the turbine i stop valves close. Emergency feedwater starts on low level limit signal and secondary safety relief and turbine bypass valves operate as designed. At 20 minutes, the operator uses the PORV to reduce the reactor coolant system pressure to 7.17 MPa (1039 psia). The operator also isolates the af fected steam generator based on either a time or hot leg temperature criteria. After steam generator isolation, the operator throttles the HPI to maintain 90 K (50*F) subcooling and starts one reactor coolant pump per loop. A transient scenario is provided in Table 22. 9.2 Model Changes
, dhanges were made to the steady state model to incorporate the tube rupture as follows. The rupture was assumed to be a double ended break
- occurring at the top tube sheet in the A steam generator. The tube side of the break was modeled using a four volume pipe component with the broken end of the tube connected to the upper volume of the secondary (Cell 32501 in Figure 5) witn a junction. The tube sheet side of the break was connected to the same secondary volume with a junction. Two time dependen; volumes and junctions were incorporated to the primary system to represent the letdown and makeup system. The makeup system flow was controlled by the pressurizer liquid level such that. if the pressurizer level dropped below 5.588 m (18.333 in.), the makeup flow increases by approximately three times. The letdown flow remained constant until the time of HPI initiation, when both the letdown and makeup systems were isolated.
169
N TABLE 22. STEAM GENERATOR TUBE RUPTURE TRANSIENT SCENARIO !
- 1. Steam Generator A tube ruptures (dnuble-ended break).
- 2. Reactor trips, turbine trips, and the turbine stop valves close.
Heater drains close.
- 3. HPI actuates on low primary pressure.
4 Operator trips the reactor coolant pumps 30 s after HPI initiation. .
- 5. ICS controls main feedwater based on steam generator level signals.
- 6. Main feedwater pump protection system operates as designed.
- 7. Main feedwater pumps trip on high Steam Generator A level.a
- 8. Er..ergency feedwater pumps start on low main feedwater pump discharge a
pressure.
- 9. Emergency feedwater flow controlled based on steam generator level.*
- 10. Secondary safety relief and turbine bypass valves function as designed.
- 11. At 20 min, operator uses PORV to reduce primary pressure to 1025 osig (if necessary). j
- 12. Operator isolates affected steam generator at 20 min or when hot leg temperature <545'F, whicnever is later.
- 13. Following isolation of affected steam generator, operator throttles HPI to maintain 50 F subcooling and restarts one reactor coolant pump per loop.
'a. May be phenomenologically dependent.
170
9.3 Transient Results This section presents the result of the steam generator tube rupture transient. A sequence of events of the transient is presented in Table 23. The initiating event was the tube rupture which occurred in the A steam generator. The secondary pressure in the A steam generator increased as a result of primary fluid flashing as it entered the secondary side. The pressure increase propagated up the main feed line to the A main feed valve, reducing the differential pressure across the valve, and therefore reducing the flow into the steam generator. The feed flow mismatch controller opened the valve to maintain the flow and the main feedwater pump speed increased to maintain the differential pressure across the valve. At the same time, the main feed valve for the B steam generator closed because of the higher flow from the increased main feedwater pump speed. The valves oscillated until a new steady state valve was determined by the ICS. ( Primary pressure decreased as a result of the tube rupture as shown in Figure 129. Pressurizer heaters were turned on as the primary system pressure decreased. By 180 s all the heaters were on and the primary pressure momentarily increased, resulting in a slight increase in the break mass flow rate through the tube side of the break as shown in Figure 130. The increase in break mass flow rate was enough to reverse the primary system repressurization and at 220 s primary depressurization commenced dgain.
- At 319 s the reactor tripped due to violation of the pressure /
temperature relationship resulting in a turbine trip, closure of the turbine stop valves and closure of the feedwater heater drains. Also, the
. ICS ran back the main feedwater pump to its minimum speed and closed the main feedwater valves. Upon closure of the turbine stop valves the secondary pressure, shown in Figure 131, increased to the turbine bypass 171
w l 18 . , , , i . 1 2600 .
- LOOP A
--- LCOP B q @ ----- 2400 9 q -
% is - _
- S w 2200 ,
e w 3 3 m
- my y , . -
2000 m 9 a: 6 A 4 w e 3 teoc E 2 - 3 a 12 - o o 1600 g , , i , , 0 500 1000 1500 2000 2500 3C00 3500 Time (s) Figure 129. 5.C. tube rup ture, hot leg pressures. 4 . , , , , , _ . , 5 m 3 % - 7
$m 6 \
E
.x n
(
=
$2 - -
*s 3
E E - O 3 - " o 1 - 2 3 0 0 0 500 *J00 1500 2000 250c 3000 3500 Time (s) Figure t30. 5.G. tube r6pture. tute side breow it ass flow rate 172
. _. . . _ - -. . . ...-__- . _.. __ . _ ___ . _ -- . . . . . _ . . . , . _ . . . > - - . ~ . - . _ . . . - __ .._ -m._ _ , _ _. __m i
4 4 i N 4 8 . i i 4
- STEAM CENERATOR A l --- STEAM GENERATOR B m 4
n- t - o o 7,- -
.1000 -
a c1 v w w
- ce g .
- u 2 3 M 'E b
=
m e
. a A$ .
( e W 3 3 . --.. ...... , .. - E 3 i a C
,* .600 o
'l -
>4'-
j 1 J I *000 1500 20C0 2500 3C00 3300 O S00 ' time (s) Figure 131. S.G. t @ ri.pture, st eam generator secoadory pressures. i 6 l e
. f-e I
I l . 173 i
. ym y9,,, , ._,w..r. y ,3.w,p.y_wm,7T,,,...-,,.9.-, , , , , ..-,.y..,,-w,y
* + - -
- Tey e.g '-&6e', e-.,-%.,,,, -,.-ww7 ., _a,__g-9qgw.m pp, g--y- ,,y3-.y.qq-q,,, g 9, p 9 9yp--_ygypm
TABLE 23. STEAM GENERATOR TUBE RUPTURE TRANSIENT SE00ENCE OF EVENTS Time Event _ _ _ . _ _ _ _ _(sL Steam generator tube rupture in Steam Generator A occurs. ICS 0.0 adjusts MFW valves to compensate for perturbation in Steam Generator A secondary system. Reactor scram (p/T relationship) turbine trip, turbine stop 319.2 valve closes, heater drains close over 5 s period. ICS runs back MFW pumps and MFW valves. Pressurizer heaters come on due to low primary pressure. 321.6 Pressurizer level drops below 220 in. and makeup flow increases. 325.4 Pressurizer level drops below 127.6 in., cutting off the 336.3 pressurizer heater power. Main feedwater pumps trip on high discharge pressure due to ICS 349.0 running back the startup valves. Primary pressure increased due to reduction in primary to 650.0 secondary heat transfer. ' Level in Steam Generator B secondary oscillates around low level 775.0 setpoint, allowing periodic flow of emergency feedwater into secondary system. Periodic flow of emergency feedwater enhances primary to secondary heat transfer and primary pressure turns over Primary pressure dropped below HPI setpoint, HPI flow initiated, 942.4 makeup and letdown flow stopped. I RCPs trip 30 s after HPI initiation. Steam generator low level 972.4 setpoint increased and motor and turbine driven emergency . l feedwater systems turned on. Increased primary to secondary heat transfer enhances primary fluid cooldown. Liquid in the main feedline commenced to flow into Steam 1280.0 - Generator 8 through the start-up valve and crossover due to the secondary pressure dropping below the booster pump head. Liquid in the main feedline commenced to flow into Steam 1300.0 Generator A through the start-up valve and crossover due to the secondary pressure dropping below the booster pump head. Liquid level in Steam Generator A rose above the low level limit 1396.6 setpoint and EFW into the secondary was terminated. 174
TABLE 23. (continued) . Time Transient Event (s) . Liquid level in Steam Generator B rose above the low level limit 141,5.5 setpoint and EFW into the secondary was terminated. Liquid level in Steam Generator A rose above the high level 1426.0 limit setpoint and ICS closed the startup valve, irolating Steam Generator A. Liquid level in Steam Generator B rose above the high level 1437.6 limit setpoint and ICS closed startup valve, isolating Steam Generator B Steam generators become heat sources. Primary temperatures begin 1600.0 to increase. Primary pressure reachea PORV setpoint. The valves oscillate 2200.0 around the setpoint maintaining primary pressure. Pressurizer becomes liqu d solid. i 2450.0 Transient terminated. Vessel downcomer liquid temperature at 2800.0 517 K (471 F) and increasing. Pressure at the PORV setpoint. e 4 S 175
~
setpoint' and, momentarily, to the secondary safety valve setpoint shown in Figures 132 and 133, respectively. Primary system pressure 1,ncreased as a result of the increase in secondary pressure and a decrease in primary to secondary heat transfer, however, due to incre. sed steam cooling when the turbine bypass and safety valves opened, the trimary system pressure rapidly decreased as shown in Figure 129. As the primary system pressure cecayed, the break mass flow from the tube side also decayed as shown in
- Figure 130. The break mass flow from the tube sheet side, however, -
increased as shown in Figure 134 The mechanism to increase the flow appears to be related to ficw choking and a decrease in the primary liquid temperature, resulting in a higher liquid density. At 400 s the tube sheet side break mass flow oscillated as the conditions at the break oscillated between choked and unchoked conditions. This behavior persisted for 600 s. Occasionally it has been observed that the RELAp5 code has difficulty in calculating break flows and in determining choked or unchoked conditions. It should be noted the effect o,f the flow choking and unchoking was minor on the overall results of the transient, as will be explained later.
,1 J
As a result of primary depressurization the level in th'e pressurizer decreased due to system shrinkage as shown in Figure 135. At 325 s the level in the pressuri:er dropped below the setpoint levei for increasing system makeup flow (220 in.) and the makeup flow increased from 1.4 kg/s (3.1 lbm/s) to 7.6 kg/s (16.8 lbm/s). At 336 s the pressurizer level dropped below 127 in. and the pressurizer heaters were turned off. The secondary liquid level rapidly decreased due to the opening of the turbine bypass and safety valves at reactor scram as shown in Figures 136 anc 137. As a result of the decrease in liquid level, the heat transfer between the primary and secondary decreased, as shown in Figure 138, and the primary depressurization rate decreased as shown in Figure 129. Also affecting the reduction in primary to secondary heat transfer was the tripping of main feedwater pumps due to high discharge pressure when the ICS began closing the startup valves at 349 s. After the pumps were tripped the startup valves cpened because of a feed-flow mismatch controller. At 650 s the primary to secondary heat transfer further
- 176
O 200 , , , , , ) STEAW CENERATOR A 400
--- STEAM CENERATCR B n tSO - -
7 (m 300 \ E
.x 4 w
g { @ - - 200 0 3 E
- o
- So ,- ,
1C3 2 l :
' ' 1 " ' ' ' '
O O O 500 1000 1500 20C0 2$00 3000 3500 Time (s) Figure 132. $7.. tube ibpture, turbine bypass mass flow r a t es. 500 , , , , , , STEAM GENERATOR A 1C00 STEAW GENERATOR 8
. 400 - -
a n # 300 N
# N m E 6 300 - -
f V 600 D o D C 2
~
200 - - m
'l 400 ,
O # 2 ,! O i 2 10 0 ,- - 200 0 0 0 500 1000 1500 2CCO 2500 3000 3500 Time (s) Figure 133. 5.C. t6be rup tur e, S.C. relie f volve mass ficw rates. 177
- 9. - __, _ _ _
. . _ . . _ - _ . _ - - - - - - . . . -- .- - ~ . . . _ _ , .. - . - - _ - . - - _ . - .
B 3 l l 5 I 30 l e 6 '20 S s
) to -
E I
$ \ 20 we 3 i
'i 3 .M 3
~
f 13 o J 3: h 3, - ig
- 2 {
i 5 0 O 0 SCO WO 1500 20C0 2500 3000 3500 Time (s) Figure 134. 5.G. t une r up t ur e, t ube snee t s ide br e* mese ti on r a t e.
- IS
)
j 40 m a F C C io . .v d
> 30 >d
! N d 9 -
*20 9 i -3 s - -
8 *
- a a e
. to
) . i i , '
~0 0 500 :000 1500 2000 2500 3 Coo 3200 Time (s)
Figure 135. 5.0. tube r6plure, pressurizer collapsed Equid level. 3.- l I . . -' ! 178
s = i
,4
. 4...
l
- l
^ s'- ! - 20 m E : -
~
J g ..te ~ ~; i
- w
. , t3 e w 4 . l .
a -
.e
. O ; V
. ,,o -
- s . e 3
O ; l - T 3 2 - l ~ "
, - STEAW GENERATOR A l l ---
STEAW GENERATOR B S I. 0 0 0 SOO 1000 1500 2000 2500 3C00 3500 Time (s) Figure 136. S.O. tube rupture, steem generator operating levels. 10 , , , , , ,
- STEAM GENERATOR A 30
--- STEAM GENERATOR S n : 25 m v
E -- l 0 w J s.. : - 20 - e w e
=
e a :
. . ts c
- m 4
y .;p : l 5 . . . . . .
- i. .,0 -8
_J ; e J e
- 2 -
< ,/
l 5 0 O O 500 1000 15C0 20C0 2500 3000 3500 Time (s) Figur e 137. 5.G. t ube r up t ur e, s t eam gener a t o r s t art up l evel s. 179 l
b S 7 , i
- STEAw CENERATOR A
--- 0.CO*'O.
STEAW GENERATOR B T ^ c-E E-- _ - 0.00 ( a g -10.00**0' O- ~ u E -s - 2e o -2.0C'fC, i C . u w c '
. s
= -m e l
-f -3.CC*t0* -
c J e
- 4.0 0'10' , 93 t i , i i - 5.0C '10' O S00 1000 1300 2CCO 2.500 3000 JSCO Time (s)
Fi gur e 138. S C. tube rupture, steam generator hea t r etroval re tes. e 180
decreased due to a further decrease in secondary liquid level and resulted in an increase in the primary system pressure as shown in Figure 129. The primary pressure continued to increase until 775 s when the level in the B steam generator decreased below the 24 in. low level set point and the
. emergency feedwater system for the B steam generator was started to maintain the 24 in. level as shown in Figure 139. The introduction of cold emergency feedwater flow into the steam generator secondary enhanced the primary-to-secondary heat transfer such that the primary pressure began to decrease once again.
The primary system pressure continued to decrease until 942 s when the pressure reached the HPI setpoint, initiating HPI flow. As shown in Figure 140, the HPI flow overwhelmed the total break flow, and with the primary system being liquid full (the upper head had not voided), the HPI inflow began filling the pressurizer, as shown in Figure 135. This repressurized the primary system, as shown in Figure 129, which resulted in an increase in break mass flow rate shown in Figure 140. Thirty seconds
, af ter HPI initiation the reactor coolant pumps were tripped and loop flow decreased to natural circulation flow, shown in Figures 141 and 142. Wnen the reactor coolant pumps tripped, the secondary side low level setpoint signal increased to 240 in, and full emergency feedwater flow was initiated to each steam ger.erator as shown in Figure 139. The cold emergency feedwater entering the secondary system resulted in a rapid depressurization of the secondary, shown in Figure 131, and termination of turbine bypass flow shown in Figure 132. At 1280 and 1300 s the pressure in the B and A steam generators, respectively, decreased below the concensate booster pump head and flow from the main feed line commenced to flow into the steam generators via the startup valves to the emergency feedwater headers. At 1396 and 1415 s the level in Steam Generators A and B, respectively, increased to tne low level set point and emergency feed water was terminated. Flow from the main feedline continued, however, until 1426 and 1437 s for Steam Generators A and B, respectively, at which time the level had increased to the high level setpoint and the ICS closed the startup valves, isolating the steam generator secondaries. With the cooling mechanism removed from the secor.dary side the pressure began to increase as shown in Figure 131.
181
'mLu 4-o I
2SO i . . . i
- $TEAW GENERATOR A SCO
--- STEAM GENERATOR 8 200 -
,t -
n
<00 n m
N" N
- 50 - ,. I. -
E f
'. i
'O.
300 o O
~
O ~ 200 e l** , a 50.- i ' - co 2 ~~ L? o l 2 he l - 0 0 a t t t t t t =l$$ 0 SCO 1000 1500 20C0 2500 5000 3500 Time (s) Figure 139. S.G. tube rupture. EFW Neoter mass flow rates (sum of EFw ond WFW deDwered .29 the Erw hooder). ' i 60 . . . . . .
,a ......,* - TOT BREAK FLOW 12 0
; --- TOT tel FLOW
* ^
n l . 10 0
- w ., *
, N N 40 -
l '. E f V so a O l
- m : Ah ,
- l 60 o
~
l c 3 0 20 - !*
-'40 g a 0 20 0 O O $00 1000 1500 2CCO 2500 3000 3500 Time (s)
Figure 140. S.G.
- ube r up t u r e. t o t a l b r eck cnd t o t al *l mass flow ro t es.
w 182
I B i J 4 5 HOT LEG 20000
% --- CCLD LEO 4-2
~~ ~'
8000 - - m n n
$a 15000 %
E c s000 - - e v 3 o 5 g -, . , . . , _ . -. . . 10000 o EM - , 1 -
- O i g 2 C 2000-i .1 5000 2 i.
0 500 1000 1500 2000 2500 3C00 3500 Time (s) Figure 141. S.C. tube rupture, loop A hot and cold tog mass f ow r a t es. 10000 , , , , , ,
- - HOT LE3 23000
% - - COLD LEG 9-2 8000 - - COLD LEG 8-1 a
m n (. 15000 \ s 6 s000 - - w e I
- D
*" _ , ,,_ ..8 10000 o 4000 -
l - C
$ l ! e o '
I a 2 U
*I 5000 2 2000'- j -
'. t n
0 ., 500 '000 1500 2CCO 2500 3000 3500
. Jime (s)
Figure 142. S.C. f ube rupture loop B hot and cold leg tross fice rates. 183 e
'El '
- - - - - - , - - - , - , - - . . _ - _ _ . - _ - - ----.c- - - - -
Primary system coolant temperatures decreased upon initiation of HPI as shown in Figures 143 and 144. At 1600 s the secondary 1.iquid temperature became warmer than the primary temperature, such that the overall primary-to-secondary heat transfer reversed and the steam generators became heat sources to the primary system. Secondary pressures decreased as a result of the heat given up by the secondaries as shown in Figure 131, and the primary system liquid temperatures increased as shown in Figures 143 and 144. The lowest primary system liquid temperatures were calculated at this time. At 2200 s the primary pressure reached the PORV setpoint and oscillated around the setpoint as shown in Figure 129. Throughout the duration of the transient the PORV held the pressure.at its openir.g setpoint. Also, the primary liquid tenperature had increased enough that the net prim ry-to-secondary heat transfer was nearly zero, with the generators being a slight heat sink. At 2450 s the pressurizer became liquid full. The transient was terminated at 2800 s when trends in the pressure and *.emperatures in the primary system had been established allowing ext spolations to 7200 s. 9.4 Conclusions At the termination of the transient, the primary system pressure was oscillating around the PORV setpoint pressure of 1.7 MPa (2465 psia). The reactor vessel downcomer liquid temperature, which is of PTS concern, was at 517 K (470*F) and increasing slightly. A time history of the vessel , downtomer pressure, liquid temperatures, and heat transfer coefficient are shown in Figures 145, 146, and 147, respectively, with extrapolations to 7200 s. As shown, the extrapolated values of pressure and temperature out to 7200 s are 1.7 MPa (2465 psia) and 544 K (520 F), respectively. The lowest vessel downcomer liquid temperature was 505 K (450*F) which oc:urred at 1700 s. The heat transfer coefficient is provided for use in future fracture mechanics calculations. 184
. j j 5 .# - Hot LEc
-600 w E
--- CCLD LEG A-2 e -- CCLD LEG A-2 e
' 580 - - w s s
- a-O w
, o
.*s---' -'.*
o ',.
'" \. ..550 E g 540 -- E e .
.t. .o.
s \
', a y
.'O. Sec - -
- p 4 T
- s
'. *\ -50C e
= -
o - g . e E s20 . s ', \, ........ - . ._. E o . \ ,.'
> '..' . . . . ,h,* .- .45o j_
3oo 0 500 1000 1500 2000 2500 3000 35C0 Time (s) Figure 143. 3.0. tune rupture, toop A hot cr'd cold leg fluid temper a t ur es. 600 a i i . , , m 6 / - HOT LEG
-600 v h
CCLD LEG 9-2
-- COLD LEG g-1 e
' 540 - - w s
=. g s
*=
i O w e ,:." , . J. O u 560 <- ,; " . , . ' ' * * *.5 .55o [ s e
- '. \
a 5 t 'O .
- 540 -
'. '\ .
s . s
=
w ','
'\ -500 7
o e
. \ .-.
g E,520 ., g-
.......y.._.
s ( o
.....b' a
... s -
t
-450 $
i 500 0 500 1000 1500 2000 2500 3000 3500 Time (s) Figure 144. S.G. t utse rup t. re. loop B ho t and cold leg fluid t empera t ur es. l i i i se l 185 w , y - -p-- w 7.--.c a m ,, -p,
9 18 - . . . . > i i 2600 O 2W O
- 5
-I 16 [ 3 w 2200 ,
a 3 $, m , b 84 '
-,2000 U m w Q- c.
W - CALCULATED
- 1300 E hs 12 -
-- - - EX TRAPOL A TED -
2 O o 1600 i , g i 0 1000 2000 3000 4000 5000 6000 7000 50C0 Time (s) Figure 145. 5.C. *ube rupture, entrepolated 4.V. downcomer fluid p r essur e. 1 580 , , , , , . .
^
-)
n M - CALCULATED b
- - - EXTR APOLATED e
6 j e 6 3 560'- . 550 3 O O 6 L e o Q. c. E E
.e- S40 -
.e-T 500 -7 3 3 -
7 0"
- = =
,520 -
E E 2 2 0 450 >0 500 O 1000 2000 3000 4000 50C0 6000 7000 80C0 Time (s) Figure 146. 5.C. tube r6pture, entropolated R.V. downcomer fluid t emper a tur e. 186
- -- u ,
,- p .,
e a-30 , , , ,
- CALCUL4TED #D c
--- EXTRAPOLATED $
a.a e - g 40 A U n Om we 20 - g ax v) o ,.' e 3000 00 uE = i.k U wsx{ 86 m 2000 z3
. M 26 w
a 1000 Q E E 0 ' 0 0 1000 3600 5400 7200 Time (s) Figure 147. S.G. tabe rupture, ex trapolots: R.V. downcomer inside surf ace hoof trarsf er coefficioet. 9 e 187 hmi maism
- 10. OVERVIEW AND CONCLUSIONS This report has presented analyses of the results of ten RELAPS calculations pertinent to the study of pressurized thermal shock (PTS) in the Oconee-1 pressurized water reactor (PWR). One of these calculations simulated a turbine-trip transient which occurred in the Oconee-3 PWR. A comparison of code-calculated and measured data indicated generally good agreement thus providing an informal and limited, but useful, qualification of the model. The remaining nine calculations cimulated hypothetical rapid cooldown sequences with potential for primary system repressurization.
The details of these sequences were defined at Oak Ridge National Laboratory, integrator of the multi-laboratory pressurized thermal shock study. Table 24 presents a summary tabulation of minimum fluid temperatures and maximum subsequent fluid pressucc in the reactor vessel downcomer for eacn of tha nine sequences. The pressures and temperatures were calculated at the elevation of the first reactor vessel circumferential weld below the cold leg nozzles. Note the pressures and temperatures shown are generally , nct coincident. The temperatures shown represent the lowest calculated or, in the event of calculations terminated before the end of the PTS 2h period of interest, the lowest extrapolated temperatures within the 2h period. Uncertainty analyses were performed for calculations with the most severe outcomes and the effect of these uncertainties on the minimum downcomer fluid temperatures are also shown in the table. With respect to the PTS concern, the most severe of the sequences investigated were found to be, in order of severity:
- 1. Failure open of four turbine bypass valves with the reactor at hot standby .
- 2. Two and one-half inch diameter reactor coolant pump suction break
- 3. Main steam line break with reactor coolant pumps restarted at the time subcooling margin is obtained.
s-183 ,
TABLE 24. SUP9WtY TABULATION OF OCONEE-1 FTS RELAPS CALCULATION RESULTS Maximum Subsequent Minimum RV Downcomer RV Downcomer Fluid Temperature Fluid Pressure Reference Section
'F MPa psia Sequence in this Report L flain steam line break RC pumps restarted 10 min 3.1 481 407 16.99 2465' after subcooling attained 431 b 3g7 b g7,34a 2515 RC pumps restarted at 4.0 494 429 16.99 2465 287c g7,34a 2515a time subcooling attained 415C
,$ team generator overfeed MFW pumps tripped on 3.2 505 450 16.99 2465 Som suction pressure 17.34a 2515' Mastmum sustainable without 5.0 500 440 16.99 2465 -
tripping MFW pumps 17.34a 2515a Iallure open of 4 TBP valves 6.0 387 237 16.99 2465 p M eactor hot standby [mallbreakLOCA
' Stuck-open PORV. RC pumps 3.3 54 5 521 11.38 1650 not tripped Two. inch diameter pressurtrer 1.0 355 180 1.48 214 surge line break 305d god f
Two and one-half inch 8.0 446' 5.17 f 740 390 h 343'h 242 17.34a 25159 diameter RC pump suction break Steam generator tube rupture ~~ 9.0 505 450 16.99 2465 17.34a 2515a
- u. Calculation was entrapolated to 2 h, this pressure assumes code safety valve is demanded. ,
ie . Includes maximum effects of uncertainty presented in Reference 1.
- c. Includes maximum effects of uncertainty sumnarized in Section 4.4.
ii . Includes maximum effects of uncertainty described in Section 1.4.
- c. Minimum temperature entrapolated to occur at 2 h see Section 8.4
. 1. Extrapolated pressure at 2 h. see Section 8.4 4 Includes uncertainty due to operator closing letoown line block valve,
- h. Inclurfes uncertainty due to finw oscillations. See Section 8.4
. 189 mi uni umim u in muu . .
O
r The 2-inch diameter pressurizer surge line break was found to produce the coldest reactor vessel downecmer fluid temperatures. However, in this sequence there is no mechanism to repressurize the arimary system.
)
e 190
- 11. REFERENCES
- 1. M. A. Bolander et al., RELAPS Analysis of Oconee-1 PWR Transients for
, Pressurized Thermal Shock Integration Study (Proprietary), EG&G Idaho Report EGG-NTAP-6190, March 1983.
- 2. Internals Vent Valve Evaluation (Prop-ietary), Topical Report BAW-10005, Babcock and Wilcox Company, July 1969.
- 3. "First Pass Feedwater Train Model for Oconee 1" letter f rom R. A.
Hedrick (SAI-0ak Ridge) to R. C. Kryter (CRNL), May 11, 1982.
- 4. "Oconee 1 Control System Model," letter from J. M. Keeton (SAI-Oak Ridge) to R. C. Kryter (ORNL), October 13, 1982.
- 5. Instruction Book 620-0003, Duke Power Company, Oconee Nuclear Plant, Unit No.1. Bailey (Babcock and Wilcox) Company, March 15, 1977.
- 3. " Estimation of Uncertainty in a RELAPS Calet.'atier. Of a Main Steam Line Break in Oconee-1 PWR," EG&G Idaho Let er JAD-114-82 from J. A.
Dearien to J. E. Solecki (00E-ID),
Novemcer 4, 1982.
- 7. B. Bassett et al., " TRAC Analyses of Sever: Overcooling Transients for the Oconee-1 P'aR," Los Alamos National Labo. atory, Draf t Report, March 1983.
S. N. G. Trikourous et al., "RETRAN Analysis of Rapid Cooldown Transient, Three Mile Island Unit 2, GPU Service Corp., August 1978. 191
4 m O APPENDIX A COMPUTER RUN TIME STATISTICS i f 4 f 4 A-1 t f
APPENDIX A COMPUTER RUN TIME STATISTICS Table A-1 presents a timing survey of the final six RELAPS Oconee PTS
- calculations containing geometric and timing information. Figures A-1 through A-6 show differentiated CPU time versus time for the six
,- calculations. The computer used to perform the calculations was the CDC 176 at the Idaho National Engineering Laboratory. The calculations were performed using the RELAP5/M001.5 computer code. Timing statistics for the initial four calculations appear in Appendix A of Reference 1.
e 9 A-2 1
. - - . .- ~-
50 ' i 7 (u 40 - , s O 30 - , O N
^ - ~
$_ 20 l
3 a a 10 - - v i 4 y k f sk 0 ' t 0 200 2000 3000 Time (s) 1 Figure A-L Diff erentiated coteputer run time, revised WSl B. I I 30 , , , 7 ' w N u I e ' O 20 - . O N C 2 -
- m - _
>=
3 I g ,_ ,t , O O
' ' t 0
0 500 1000 1500 2000 Time (s) Figure A-2. D!fferentiated computer run time, maximum sustainable overfeed. t l A-3 1 L.
- 30 i i i . . . .
T N
~
0 20 - s o W m - 3 J b , E
- 111 J t _ ,
_ 1_i__ . , _ 1 _ _[1 _i__ , 0 1000 2000 3000 4000 SOCO 6000 70C0 8000 Time (s) Figure A-3. Otfferentiated computer run time. T8e f ailure at Pot standby. 60 i i i i . . i T s S O 40 - F-O w 3 20 t-
~
3 a lJ $ dpy', 0 0 1000 2000 3000 4000 5000 6000 7000 8000 Time (s) Figur e A-4. Dif f erentiat ed computer run time, pressarizer surge line break. A-4
N 60 , , , , , m V vi N N o 40 - s-c h
' W
--- 20 -
" /
D ll l l h O W ' 1 e O 1000 2000 3000 4000 S000 Time (s) Figure A-5. Olfferentiated computer run time, purrp suction break. . j. so . 9 N S O 40 - 6-O N C - S to D
) -
8 l e g 0 10C0 2000 3000 - Time (s) Figure A-6. Olff erentiated comput er run time, S.G. tube rupture. A-5
TABLE A-l. TIMING STATI$11CS Revised Pressurlier Pot Standby Surge Line RC Pump Suction Steam Generator Calculation Main Steam Maximum Sustainable Small greak Tube Rupture
,. Parameter Line Break Steam Generator Overfeed TBP valve failure Small Break 220 222 226 Total num er 220 218 218 volumes (#C) 208 208 211 Total number 208 208 208 heat structures 6200 4900 2800 m Transient time 2697 1695 7200 ?.econds (RT) 40971 25180 14889 .
Total CPU 19305 11435 14008 f.econds used 142765 80597 48316 Total number of 65442 35934 54859 tline steps (f0T) 1.95 6.61 5.14 5.32 EPU/real time 7.16 6.75 0.893 3.00 2.31 - 2.35 CPU n 10 2 3.25 3.09 IWBC . 0.163 0.210 0.287 0.487 CPU u 106 0.497 0.tl61 El a #CTfGT 1.17 1.30 1.41 l.36 (PU x 103 1.34 1.46 FCTFDT . 0
APPENDIX B MAIN STEAM LINE BREAK REVISED TRANSIENT, COMPARISON OF COUNTERPART TRAC AND RELAP5 CALCULATIONS e 9 e B-1 1 _ - _ _ . . _ - . . _ _ _ . _ . - -. _ - . . _ _ _ . _ . . - . _ _ _ - . _ _ _ _ - - - - - _ . _ . _ - - _ _ _ - - _ _ - - _ _ - _ >
APPENDIX B MAIN STEAM LINE BREAK REVISED TRANSIENT,' COMPARISON OF COUNTERPART TRAC AND RELAPS CALCULATIONS i
- This Appendix presents a comparison between counterpart calculations ,
of the revised main steam line break sequence performed at Los Alamos National Laboratory (LANL) using the TRAC-PF1 computer code and Idaho National Engineering Laboratory (INEL) using the RELAPS/M001.5 computer code. A detailed description of the sequence and analysis of the results of the RELAP5 calculation appears in Section 4. A detailed analysis of the TRAC calculation appears in Reference 7. A comparison of initial, steady-state conditions indicates no significant differences between the starting points for the TRAC and RELAPS calculations. For the TRAC calculation these conditions may be found in Tables II.0-I and II.D-II of Reference 7. For the RELAP5 calculation these conditions are shown in Table 1 of this report. i Figure B-1 shows a comparison of hot leg pressures for the two calculations. TRAC output data was available only for the first 900 s of the calculation and comparisons are shown only for that period. This is f not a significant limitation since most events of interest occur during the initial portions of the sequence in both calculations. As shown in i Figure B-1 the primary system initially depressurized faster in the RELAP5 calculation but the depressurization progressed further in the TRAC calculation. Figure B-2 shows the reactor vessel downcomer fluid - temperature was lower in the TRAC calculation. The primary purpose of this Appendix is to explain the differences in primary pressure and downcomer l , fluid temperature between the two calculations. i . Table B-1 compares the timing of events between the calculations and
.significant differences are noted. The following discussions examine these differences. For purposes of discussion the sequence is separated into
! four phases: (a) transient initiation to time of reactor coolant pump B-2
20 , , , l'
- RELAPS
--- TRAC n h2500 ^
O 0 l
$ 13 5
t TRAC /RELAPS
':000 "
I S. RELAP5 FW TERMINATEC E
$, m .- i RCP RESTART ,,330 R g
w a a 9 {' ., 1000 O s -
.......~~~~~'i,,,,.....---~~ -
Q 3 500 > T AC RCP RESTART 0 o
-250 0 250 SCO 750 1000 Time (s)
Figure 9-1. TRAC and 4ELAP5 comparison, revised MSLB. not leg pressures.
. ,600 t600 E - RELAP5
"- mAC
, TRAC = LEVEL 5 CELL 10 ,
w w a a o
- RELAPS TRAC /RELAPS o
[ kRCPRESTART FWl TERMINATED[E* E
- 500 -
\ , - ~a '.. o '; N, TRAC RCP RESTART 400 ; .
v e
- = ,,
. .... =
e ' o' e E -
; ,. ' E
- 2. ' . ,T 300 3
0 .' O
> i i i ,
4g
-250 0 250 500 750 1000 Time (s)
Figure 9-2. TR AC and RELAPS comparison, revised MSLS, reac tor vessel downcorrer fluid temperatures. B-3
TABLE B 1. COMPARISON OF TRAC AND RELAP5 SEQUENCES OF EVENTS, MAIN STEM LINE BREAK REVISED TRAN51ENT Time of Event (s) Notes TRAC RFLAPS (T = TRAC. R = REL.APS) Event I;reak opens 0.0 0.0 11FW pumps tripped on high level 7-trip occurs later Indication in SGA Not calculated 0.3 (47.8 s) on low suction pressure 0.09 R-one timestep plus Peactor trip 0.5 0.05 S Turbine trip 0.5 0.5 f eedwater heater power and drain flows tripped 1.0 1.0 tose condenser feed from turbine 1.5 Not calculated 5.0 2.0 T. low pressure safety
%B turbine bypass opened not challenged TSP Not calculated 3.0 cycled twice LG3 safety opened Not Calculated 12.0 R-safety and TBP
%8 safety closed challenged and cycled 39.9 13.0 once each
'.GB turbine bypass closed iiPI initiated on low hot leg pressure 21.2 5.3 29.4 4.4 IFW initiated to SGA 47.8 Not calculated R-trip occurred at 0.35 ftFW pumps tripped on low suction pressure on high SGA level indic ation 48.7 4.4 tFW initiated to SGB FCPS trip, MFW realigned *u Eid header 51.2 35.3 Enndensate bnoster pump trip on low section pressure 53.9 Not calculated 346.7 370 LG8 level recovered EFW throttled 42K subcoolin9 attained in hot legs,
. A-1 and B-1 RCPs restarted and HPl t5rottled 526 300 ikstor driven EFW terminated to both SG Not calculated 5.3 Lnre flood tanks inject 530.9 - 537.9 Not calculated All EFW, MFW, and TBP isolated to both $G 600 600
%8 TEP an.1 EFW restored 900 900 194V setpoint pressure reached 4678 2432 Calculation terminated 7200 2697 R-data for 2697 -
7200 s was estrapolated 3 B-4 i ___ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ . _ _ _ _ _ _ _ . _ _ _ . . . _ _ _ _ _ _ . . _ _ _ _ . _ _ . _ _ . _ _ _ . _ _ _ _ . . _ _ _ _ _ _]
trip (b) time of RCP trip to time of RCP restart. (c) time of RCP restart to time of feedwater termination, and (d) time after feedwater termination.
- Table B-2 shows how the timing cf the four phases of the sequence differed between the two calculations. Specifically, the times dividing
~
Phases 1 from 2, and 2 from 3 are different. The differences between the timing of the sequence phases will be discussed first, followed by a discussion of differences between phenomena occurring during each of the phases. Phase 1 ends and Phase 2 begins at the time of RCP trip. This trip is scheduled 30 s after HPI initiation which occurs when the hot leg pressure is reduced to 10.44 MPa (1515 psia). The RCP trip occurred at 35.3 s with RELAP5 and 51.2 s with TRAC. As shown in Figure B-1, the primary pressure declined quicker to the HPI initiaticn setpoint with RELAP5 than with TRAC. This differeace appears to be caused by a greater affected steam generator heat removal rate with RELAPS, as shown in Figure B-3, which in turn was caused by the higr.- RELAPS break mass flow rate shown in Figure B-4. A complete explanation of the differences between TRAC and , ) RELAPS initial secondar.y blowdown behavior would require a detailed evaluation of mass distribution, heat transfer processes, and break flow during the blowdown and is beyond the scope of this comparison. The differences in heat removal rate shown in Figure B-3, however, explain why sequence Phase 1 ended sooner in the RELAPS calculation than in the TRAC calculation. Phase 2 ends and Phase 3 begins at the time of RCP restart. One RCP per loop is to be restarted only when 42 K (75*F) subcooling is attained in all hot legs and cold legs. This criteria was met at 300 s in the RELAPS - calculation and 526 s in the TRAC calculation. The reason for this difference is that in the RELAP5 caiculation loop natural circulation flow continued in both loops while in the TRAC calculation the unaffected loop stagnated as shown in Figures B-5 and B-6 during the time period from about 100 s to 300 s. Loop B flow was terminated in the TRAC calculation when the fluid at the top of the Loop B hot leg flashed and formed a large steam bubble as shown in Figure B-7. This flashing occurred in the TRAC B-5
a 6000 ' ' RELAPS
... TRAC 4000 -
m 3 2 w ;. ' g 2000 - l ' i y b . .................. 3 o Q. 0 -
' t
-2000 60
- to 40 Time N Figure 3 3. TRAC and MLAPS compaMom rev sed 69. oftected S.G.
Mot removel rates. 6000 * ' MLAPS
.-- TRAC n 10000
# e n
Na 4000 - N
.,6 V E A
O 0 7, 5000 6 2000 - 3 l' **. O
* = .
i . .. 3 . g e -O 2 n 0
- 0 2
' 1 2000
. 20 40 60
-20 0 -
Time (s) Figure 3-4 TRAC and RELAPS come?!SC"- vised Wst.B. break mo23 flow rates. B-6
1 l l i 1$000 , , , , , RELAPS 3oooo !
--- TRAC j
^ o N 25000 O - 3 N w 10000 - E I e '
, ; 20000 -Q o : O w : y y l 15000
~
o e o I Sooo . t. n
.10000 e o
m . g
- 2 2 *
. . 5000 0 O
-250 0 250 SCO 750 1000 Time (s)
Figure 9-5. 7%C and RELAPS comport son. revised WSLB, loop A hot tog mass flow rates. j
)'
10000 , , , i RELAPS 20000
--- TRAC m
IS000 (, !, m 2 I -
\
- N/ * -
30000- 6 0 e - O (*: .
- c
' '. l M= 3
. o I C .
.f ,
0-- O a n m . o a 2 5 -5000
-i==
_g . , , ,
-250 0 250 SCO 750 1000 Time (s)
Figure B-6. TRAC and RELAPS comparison, revised WSLS, loop B hol leg mass flow r a t es. B-7 , w <--w--r w w =r
e d i h k 4 i-I i 1
; '. i RELAPS v ---
TRAC l c l l i o . . 4
*4 l .
u . . 0 :
- 6 .
3 0.5 - l l _ c . . u .
- o. l- l a
O 4 . 0 ' ' ' i. i
-250 0 250 500 150 10C0 4
Figure 8-7. Time (s) TRAC and RELAPS comparison, revised WSLB. void froc tions of t op o f loop 9 he t f ees. 1 i 4 e 9 e B-8
TABLE B-2. TRAC AND RELAPS SEQUENCE PHASE TIMING TRAC RELAP5 (s) (s) Sequence Phase Number From To From To__ 1-initiation to RCT trip 0 51.2 0 35.3 2-RCP trip to RCP restart 51.2 526 35.3 300 3-RCP restart to FW termination 526 600 300 600 4-after FW termination 600 7200 600 7200 m B-9 e _ _ . _ _ - _ _ _ _ _ - - _ _ _ . _ - - - _ _ _ _ _ _ _ ___ __w
calculation, but not in the RELAP5 calculation, because the Loop B hot leg fluid temperature was higher with TRAC than with RELAPS as shown in Figure B-8. Figure B-9 shows the opposite was true in Loop A. When Figures B-6 and B-9 are compared, one observes the RELAPS Loop A and B hot
. leg fluid temperatures are identical while the TRAC Loop A temperatures are below, and the Loop B temperatures are above, the corresponding RELAPS temperatures. These differences are caused by the one-dimensional nature of RELAP5 reactor vessel model as comoared to the three-dimensional nature of the TRAC reactor vessel model. The two hot leg temperatures are identical with RELAPS because the same reactor vessel cell feeds Doth hot legs. The temperatures are different with TRAC because different reactor vessel cells feed the hot legs and these cells contain different temperature fluids. The high TRAC Loop B hot leg temperature caused it to flash and cut off the natural circulation in Loop B. With loop flow stopped the subcooling margin necessary for restarting the RCPs was not attained until 526 s in the TRAC calculation, wnen HPI volume addition had repressurized the primary system and sufficiently raised the saturation temperature. Because of the lower RELAPS Loop B hot leg temperature, Loop B hot leg flashing was not calculated, the loop flow continued, and the subcooling margin was attained much sooner at 300 s. As shown in Figure B-2 the RCP restart had a significant effect on the reactor vessel downcomer temperatures in both calculations. TRAC and RELAPS calculated different RCP restart times primarily due to multi-dimensional behavior within the TRAC reactor vessel model which the RELAPS model was not capable of calculating. Specifically, in the TRAC calculation the fluid entering the affected loop hot leg was colder than that entering the unaffected loop '
hot leg because the affected cold legs were colder than the unaffected loop cold legs. The asymmetric fluid temperature distribution around the reactor vessel downcomer inlet caused the upper plenum fluid temperature distribution to also be asymmetric. The one-dimensional RELAPS reactor vessel model, however, perfectly mixed the flow from all cold legs and donored the same temperature fluid to each hot leg. The extent of this asymmetry in the prototype plant will be a function of the geometry of the flow paths from the downcomer inlet annulus to the upper plenum and the velocities of fluid through the paths. For a fast flow and complicated B-10
4 4 4
^
R
^
M
- 1 RELAPS 600 *>
... TRAC e , e w .
u 3 3 550 - O A.* c
\ *.
- w
'e 550 . e a '. c.
E ., .-------- --------- 500 E e . . , ,
.e
~ .
7 I t
~
- 450 3 5500 -
w
= ; =
o : e E : 4c0 E 3 .. _3 o :, . ....... o 450
-250 0 230 SCO 750 N000 Time (s)
Figure 9-8. TRAC and REl.APS comparison, revised MSLB, loop B Not leg liquid temperatures. 4 5 4 5
^
n 4: ] RELAPS
- b*
TRAC e e b b 5 SSO - O O 500
'e .
'e k '*
, ,,*#.'., Q.
c .. . E e ' e
- 500 - . -
T , y
- . 4oo ,
3 3 v '.g ~.., = w
= ; ,... '*.........-
e 450
.. e E ,' E 3 3
- 300 -
O o 40
-250 0 250 500 750 1000 Time (s) .
Figur e B-9. TRAC ord RELAP5 comparison revised WSLB, loop A hot leg liquid terrperatures.
/
B-11
geometry, symmetric behavior could be expected. The true extent of asymmetry expected for the Oconee-1 plant at the flow rate's calculated by the computer codes is not known. Existing test data from a subscale facility with different reactor design indicates that a degree of. asymmetry
- is to be expected. Due to proprietary considerations, this data is not referenced here. No attempt has been made to compare the magnitude of the TRAC-calculated asymmetric with the test data. It is likely, however, that for the main steam line break sequence, which includes hignly asymmetric steam generator secondary behavior, the TRAC-calculated asymmetric hot leg conditions are more reasonable than the RELAPS-calculated symmetric hot leg conditions. Thus the RCP restart (and the end of sequence Phase 2) calculated with TRAC at 526 s is likely more reasonable than that calculated with RELAPS at 300 s.
Differences between the TRAC and RELAP5 calculations during sequence Phase 1 will be discussed next. Phase 1 extends from transient initiation to time of RCP trip; 51.2 s with TRAC and 35.3 s with RELAP5. 1 As discussed in Section 2.3, in the RELAPS calculation the secondary level indication was determined in a prototypical manner using differential pressures and a reference density. The TRAC calculation used a nonprototypical downcomer collapsed level indication. As discussed in Section 4.3, the main feedwater (MFW) pumps were tripped at 0.3 s in the RELAPS calculation as a result of high level indication in the affected steam generator. This trip did not occur until 47.8 s in the TRAC calculation when the MFW pump trip occurred oue to low suction pressure. - The early MFW pump trip calculated with RELAPS better represents the behavior expected during a main steam line break than the late MFW pump trip calculated with TRAC.
. The first ef fect of the discrepancy in MFW pump trip time is a difference in the initiation of emergency feedwater (EFW). With TRAC, EFW was initiated to Steam Generator A at 29.4 s and to Steam Generator B at 48.7 s. With RELAPS, EFW was initiated at 4.4 s to both steam generators.
As discussed in Section 4.3 with RELAPS EFW initiation should have been delayed by 4 s for Steam Generator A and 11 s for Steam Generator B. Since initiation of EFW requires a low MFW pump discharge signal and a low B-12
secondary level signal and since continued MFW pump operation delays receipt of both of these signals, EFW initiation occurred.later with TRAC than with RELAP5. The break mass flow discrepancy from about 10 to 35 s in Figure B-4 is due to EFW injection with RELAP5 but not with TRAC during this period. A. comparison of affected steam generator secondary pressures is shown in Figure B-10. The figure shows excellent agreement between TRAC and RELAPS until 5 s when EFW is initiated with RELAP5 but not with TRAC. After 5 s the pressures diverge with the RELAPS pressure about 0.6 MPa (87 psia) below that of TRAC. The RELAPS pressure is lower due to the depressurizing effect of injecting cold EFW liquid. With a colder affected steam generator secondary, heat removal 'is enhanced thus explaining the slightly higher affected steam generator heat removal rate with RELAPS shown in figure B-3 from about 10 to 35 s. A second effect of the discrepancy in MFW pump trip time is the different affected steam generator MFW injection benavior shown in Figure B-11. The flow rates shown in this figure diverge over j approximately the first 10 s due to opposite MFW valve response induced by the different secondary level indications. In the TRAC calculation the MFW valve initially opened wider due to the low level indication while in the RELAPS calculation the valve closed due to the high level indication. After about 10 s the RELAPS indicated level also fell to the low level setpoint and the MFW valves opened. Between 10 and 35 s compensating differences are noted between the calculations. During this period with . TRAC the MFW pumps were running but with RELAP5 they were not. However, as just discussed, during this period with RELAP5 the secondary pressure was significantly lower than with TRAC resulting in a higher RELAP5 affected steam generator MFW flow. The final difference between the calculations observed during sequence Phase 1~ concerns reactor vessel upper head voiding. Figure B-12 compares the RELAP5 upper head void fraction with that of a representative cell in the TRAC model. As shown, about 65% of the RELAP5 upper head voided during the first few seconds of the transient while only very minor voids were B-13 u_.--,--__--_---- ---- - . _ .-,.---- - - - - - - - - , - . - - - . - - - - _ . - - - - - - - _ _ , - - _ _ _ - _ - . - -
=
i 8 , , RELAPS
--- TRAC 1000 m ^
O O
}g
. 00 -
- e b
'E 3 g 4'. 600 a'
- u a
400 e
& 's ' E
.' 2 -
-. - =
* -........ ' -- ,gnn f 0 ' ' '
O
-20 0 20 40 60 Figure 9-10.
rim. (3) TRAC and RELAPS comparisca, revised MSLS off ected s.c. .condory pr...ur... 1500 . . . RELAPS 3000
--- TRAC 7
)u iOOO MFW REALIGNED- O v ,.. 2000 N
. ,..,....' ' E e
g..
.c
=
u ., / ' - - ' Soo . ge i 1000 3 g , o O C - M
# e o
g 0-- 0 2 2
' ' ' -1000
-500 ~
-20 0 20 40 60 Time (s)
Figure B-11. TRAC and RELAPS comparison, revised WSLB, offected S.G. main f eed header mass flow rates. l I ? e B-14 f
1 . I 0.8 . . i i R(LAPS
--- TRAC e
.o . o.s -
l TRAC - LEVEL.8 CELL 3 - o l 0-5
- 2 0.4 -
l
~ . O b
o
$ 01 I
~
')
J w 0.0 ==
-250 0 250 500 750 1000 . Time (s)
Figure 3-12. TRA0 and RELAPS comoorison r evised WSL9. reoctor vessel upper hood void fractions. e
./
B-15 m i
calculated using TRAC. The reason for this difference is related to modeling in this part of the vessel. The TRAC model considers the entire upper head region as being in a flow path between the core outlet and the hot legs. The RELAP5 model considers the uppermost 1.74 m (5.7 ft) of the upper head to be a stagnant region removed from-this flow path. As a result of this modeling difference the RELAPS upper head fluid remained at its initial temperature and flashed considerably as the primary system pressure depressurized while the TRAC upper head fluid was cooled by flow during the depressurization and did not flash significantly. This difference is important in that either the upper head er a hot leg may be expected to flash first and the location of flashing was different in the two calculations. As explained earlier the hot leg flashing calculated by TRAC delayed the RCP restart and was caused by asymmetric hot leg fluid temperatures. The upper head behavior represents an additional reason why the hot leg flashed with TRAC but not with RELAP5. The two models represent the extremes of flashing potential with the TRAC model assuring virtually no flashing and the RELAP5 model assuring significant flashing. Reference 8 investigated a cooldown transient wnich occurred in the Three Mile Island, Unit 2 plant during 30*." power operation. The conclusion of this report was that for a primary system cooldowh rate of 0.37 K/s the plant experienced partial flashing of upper head fluid. The amount of fluid flashed is sensitive to the cooldown rate and the main steam line break cooldown rate is about 5 times that for the plant transient examined in the reference. Therefore, significant upper head flashing is likely during the main steam line break sequence. ! Differences between the TRAC and RELAP5 calculations during sequence Phase 2 will be discussed next. Phase 2 extends from the time of RCP trip (51.2 s with TRAC, 35.3 s with RELAPS) to time of RCP restart (526 s with TRAC, 300 s with RELAPS). l During Phase 2 of the sequence main feedwater (MFW) is realigned to the emergency feedwater (EFW) header. Flows from the MFW and EFW systems are mixed before injection near the top of the steam generator boiler region. To flow downward into the lower portions of the boiler the B-16
feedwater must penetrate tubs support plates which restrict the flow area to 43% of the area in the tube bundie itself. Countercurrent flow limiting phenomena at the uppermost tube support plate has been discussed in Reference 7 for the TRAC calculation and in Section 4.3 for the RELAP5 calculation. The first difference between the calculations during sequence Phase 2 is the tripping of the condensate booster pump. This pump is to be tripped if its suction pressure falls below (0.21 MPa) 30 psia for a period longer than 5 s. The pump was tripped off with TRAC but not with RELAP5 and this difference is likely caused by the earlier MFW realignment to the ERJ header with RELAP5. After realignment MFW is delisered only through the
.,small flow area of the startup control-valves, thus limiting further depressurization of the upstream MFW train including the condensate booster pump suction region. With the condensate booster pump powered more flow is delivered through the MFW train to the EFW header of the affected steam generator. This explains the higher RELAPS EFW header flow as shown in Figure B-13. s
/
During the injection of EFW and MFW through the EFW header, the f fbw enters the boiler section and is flashed due to the much lower pressure in the boiler section than in the EFW line. Figure B-14 compares the void fractions calculated at the top of the boiler in the affected steam generator. During the period from about 50 to 350 s the void fractions calculated by the two codes are in excellent agreement. Note that at the low pressures involved here, a 98% void fraction carresponds to about a . 10% mixture quality. The divergence which began at 350 s was caused when the TRAC affected steam generator EFW flow increased because of unaffected steam generator EFW throttling. After entering the affected steam generator boiler section most of the . EFW flow was bypassed to the break in both calculations. A portion of liquid, however, penetrated downward into the lower boiler section and was vaporized there which produced a vapor upflow that limited further penetration of liquid. Figure B-15 compares the liquid volumetric flow B-17
l 600 ' * . RELAP5
--- TRAC
^ f000 M
Nc4 4on . L
" _L,._
Q N v s E D 2 I 500 o E' '. y, - y a- 200 -
.f .t"'*.'
.-py o y ,v.
;(, w',%7.7A5 h, '.,,;J . =
o a C I. .
- e
- O m .s 0 2 m 0 O
2
-200 0 250 500 750 10C0
-250 Time (s)
Figure 9-13. TRAC and RELAPS comparison, revised 1451.9, offected S.C. emergency 1eedricter heoder ftow r a f es. 1
' s r
. RELAPS l
; --- TRAC
# 4 N.;
2 .: . , o : ..- >
)
e
=
- 'l ,
o : ..
.. t g : . $3
'A : I 9 l :
o
'I: :
o . e x '. i
!.. .. : I.
o a 9
'il I::
0.9
~
-250 0 250 500 750 200 Time (s)
Figur e B-14. TRAC and RELAPS comporsion, revised MSLB, void fractions at top of off ected S.G. boff er. G-18
LIOUl0 VOLUMETRIC FLOWRATE (m3/s) g 9 e nP P O C l e 8 -* m 0 1 o23
- o > o .
2an
-oo ggg .........
W 3
- e .- O. m .
a
.G u :
lIa - an ..1 8 - l
- y a __
o33 ...... . . ss . - -
,e ge m =_.
.._...........g s
. . . . . . . . . . . .-5.,33333333_up3v ..
8 _^ ..... :::::.::::
..... ..... u...s-m m_ : . ~ ~- ~.
(4,Og
- ;?
E - ;.2.y].G..6.6- . .j:j.j:j:::j ,2,2::::2::22~,'.,"d'~~-
,7 ,e =.:.:<.:.:d, _.-. -__..-..
. _. o.f:m:..ss
- <: -mss
- - =- - - u-sw. .~ .-z m s,
- e. --.. .
,e -
' =1 e D E
~
u r-I O g' c ~ 1 r- -m 25 m D pr
- gO C OFy
~-
. g u.
j .e 4
'. o a
e l l 1 0 6 l l
l l rates at the upper tube support plate of the affected steam generator. The rates compare favorably with the TRAC liquid penetration rate generally 10% more than that of RELAPS.
. Figure B-16 compares the affected steam generator heat removal rates of the two calculations. For the period from about 50 to 300 s the heat removal rate with TRAC is shown to be much higher than with RELAP5. Only a small part of this difference is accounted for by the slightly higher TRAC liquid penetration just discussed. Another small part was found to be caused by a RELAPS calculational problem discovered as a result of this comparison. Figures B-17 and B-18, respectively, compare the liqu'.d and vapor temperatures at the top of the affected steam generator secondary.
As shown, the RELAPS liquid temperatures are below and the RELAP5 vapor temperatures are above the corresponding TRAC temperatures. The rate of heat removal from the primary is proportional to the differential between the primary liquid and secondary vapor and liquid temperatures and the vapor superheat calculated with RELAP5 from 50 to about 500 s was excessive. The problem is caused by an understatement of the interphase mass transfer area and the f'act that the liquid phase must remain at the saturation temperature. The interphase mass transfer is based on the Saha correlation and the difficulty is encountered only in situations combining low pressure and low quality. The RELAPS calculational problem was not, nowever, determined to contribute significantly to the differences between the TRAC and RELAP5 heat removal rates shown in Figure B-16. This was
~
determined by comparing the original main steam line break calculation, which was not affected by the vapor superheat problem, with the revised
- calculation. Had the superheat problem not been encountered the RELAP5 heat removal rate would have only increased by 10 MW. The primary dif ference between the TRAC- and RELAPS-calculated heat removal rates during the period from 50 to 300 s, is believed to be the steam generator
. tube outside surface heat transfer coefficients in the top calculational cells of the steam generator secondaries. These coefficients are determined by the code heat transfer packages and are based on the flow regimes and wall temperatures present. With TRAC this coefficient is believed to be larger than with RELAPS thus causing a higher TRAC heat removal rate. This finding is supported by observing the different heat B-20
/
~ t 600 ' ' ,
. i l RELAPS .
l --- TRAC m 400 - : - 3 ,! 2 v m l ll. . *... sa . R l' ll .7;; o C- 200 -
- ,,' * ;;;tn
- g;
. ...o.. -
t%.A*.. v. , '; *y ( : l 0 ' i . -_.
-250 0 250 500 750 1000 T,me i (s)
Figure B-16. TRAC and RELAPS con *porison, revised WSLB. off ected S.G. hoot removal rate. 0-t00 s. r n M 600 *
- u ,
600 ^
')
" RELAP5 v
e e --- TRAC w
- s g 500 $
g 500 - , __ E u e 400 a. e . g
~
; e
= ;
.' - ---.. 300 a
d = e '" E 200 e E 2 s o g
' , 10 0 >
300 + ,
-250 0 - 250 500 750 1000 Figure B-T7.
Time (s) TRAC and RELApS comparison, revised MSLS liquid t empe r a t ur e o f t op o f o f f ec t ed S.G. bo il e r. B-21
600 , , , n 600 m M
- e h- RELAP5 TRAC
$ 500 y O O 9,7- -
w
'e 500'- l l -- l [i ,
{. - i i 400 a g i p 2 N " ' -
, , , , w ... ,,E,,..J' 300 $a o ,.
O' 400 ' '"**1 - o
=
- 3 4 200
- E E 3 3 O O ~
' ' 10 0 >
J0. ' '
- 250 0 250 SCO 750 1000 Time (s)
Figure 3-18. TR AC and RELAPS comparisen, rev! sed MSLB, vcpor t errpo r o t ur e a t t op o f o f f ec t ed S.C. bo il e r. 9 9 4 B-22
removal response when the RCPs were restarted (526 s with TRAC and 300 s with RELAP in Figure B-16). The TRAC heat removal dramatically increased when RCPs were restarted but with RELAPS it did not. This in31 cates that heat removal is being limited in TRAC by the tube inside surface coefficient and in RELAPS by the outside surface coefficient. A direct comparison of heat transfer coefficients' between the two calculations was not possible because noncomparcble heat transfer data is stored on the calcuIation output tapes. A further consideration in determining the heat removal rate above the upper tube support is the size of the upper calculational cell in the models of the secondary boiler sections. This is important for, even if either code perfectly modeled the heat transfer processes on the outside of the tubes, the size of the cell determines the effective heat transfer area.over which it-applies. In both the TRAC and RELAPS models this area is overstated because the uppermost calculational cell is taller than the length from the uppermost tube support plate to the bottom of the upper tubesheet. With the TRAC model the heat transfer area apportioned to the region between the plate and tube sheet i s overstated by about SC% and with RELAPS by about 100%; both represent a conservatism in i that heat removal is overstated. The affected steam generator heat removal ' rate difference shown in Figure B-16 caused the different cooldown rates shown in Figure B-2. It is uncertain, however, whether the TRAC or RELAPS calculated heat removal better represents the phenomena. Finally, differences between the TRAC and RELAP5 calculations during sequence Phase 3 will be discussed. Phase 3 extends from the time of RCP restart (526 s with TRAC, 300 s with RELAPS) to 600 s when all feedwater is . terminated by definition of the sequence. When the RCPs were restarted a droo in primary pressure was calculated with TRAC. This pressure drop caused a momentary core flood tank injection which was not of significance to the reactor vessel downcomer temperature as shown in Figure B-2. The rapid drop in primary pressure with TRAC was due to the spike in h9at removal to the secondaries as indicated on Figure B-16. With RELAP5 the pressure decline was much less severe due to the lower affected steam generator heat removal rate. This difference was B-23 l
'M caused by the heat removal rate to the affected secondary being limited by the steam generator tube inside surface heat transfer coef'ficient with TRAC but by the outside surface heat transfer coefficient with RELAP5. Thus, when RCPs restarted, the TRAC heat transfer rate responded dramatically to
. the increased heat transfer coefficient on the inside surface of the steam generator tubes, but the effect on the RELAPS rate was much smaller. With both codes, restarting RCPs is shown in Figures 8-9 and B-2 to bs a time when the primary system cooldown was reversed or moderated. Therefore, the differences between TRAC and RELAPS noted for sequence Phase 3 are not signficant to the overall results of the study. For the sequence Phase 4 which extends from 10 min to 2 h no differences between the TRAC and RELAP5 calculations were observed which are significant to the overall results of the study. In summary, a comparison of the TRAC and RELAPS calculations has been performed and significant differences have been examined. These differences are (a) a difference in primary system heat removal rate during the first 10 s of the transient, (b) different main feedwater pump trip behavior, (c) different reactor coolant pump restart times resulting from as,ymmetric hot leg behavior, (d) different flashing benavior in the reactor vessel upper head, and (e) different affected steam generator heat removal rates during periods of emergency feedwater header injection. The insights gained while performing this comparison were used in the evaluation of uncertainty in the RELAPS calculation presented in
~
Section 4.3. O B-24 N .. . . .
. _ _ _ _ _ _ _ _ _ _ _ _ _ _ - - - . - - - _ - . - _ - - - - _-i-
APPENDIX C PRESSURIZER SURGE LINE BREAK, COMPARISON OF COUNTERPART TRAC AND RELAP5 CALCULATIONS O C-1 p,- , --m --- ,, c- e
APPENDIX C PRESSURIZER SURGE LINE BREAK, COMPARISO'N OF COUNTERPART TRAC AND RELAPS CALCULATIONS
. This appendix presents a comparison between counterpart calculations of the revised main steam line break sequence performed at t.os Alamos National Laboratory (LANL) using the TRAC-PF1 computer code and Idaho National Engineering Laboratory (INEL) using the RELAPS/M001.5 computer code. A detailed description of the sequence and analysis of the results of the RELAPS calculation appears in Section 7. A detailed analysis of the TRAC calculation appears in Reference 7.
A comparison of initial, steady-state conditions indicates no significant differences between the starting points for the TRAC and RELAP5 calculations. For the TRAC calculation these conditions may be found in Tables II.0-1 and II.0-II of Reference 7. For the RELAP5 calculation these conditions are shown in Table 1 of this report. Taole C-1 presents a comparison of sequence of events timing between the two calculations. A difference in time of reactor trip is noted., In the TRAC calculation reactor trip was assumed to occur a+. 0.5 s while in the RELAPS calculation the reactor trip did not occur until a reactor protection trip was actually encountered. In the RELAPS calculation, violation of the pressure-temperature reactor trip was encountered at 45 s. While the RELAP5 reactor trip model is more prototypical of actual plant behavior, the difference in time of reactor trio did not materially ~ affect the overall results of the calculation. The difference did however cause an offset in the timing of primary system pressure and temperature changes during the first 120 s of the transient as shown in Figures C-1 and C-2. The main feedwater (MFW) pumps were tripped in both calculations due to high MFW pump discharge pressure. This occurred at 70 s with RELAP5 but not until 788 s with TRAC. MFW pump discharge pressure is sensitive to MFW pump speed and MFW control valve flow area both of which are controlled by C-2
15 i
- k '
1 RELAP5 2000 TRAC n 0-0
't 2
2 v us o. 10 1500 v e - w 1 o 3 u w i 3
$ J , e g
N ' 1C00 o-E 3 - - o 3 E
- 3 o 500 -
o
, e '
O O O 2000 4000 WO E rime (s) Figure C-t. TRAC and RELAPS comparison, pressurizer surge fine brook, hot leg pressures.
- j 600 _/
6 - '
- g .
600 ^ ss RELAPS
--- - TRAC C e e w
3 u
- % 500 3 o /%g *-
O w 500 -
,1I.M u 400
.e.
3 300 3 3 3 7 A00 - - 7
= =
3 200 e E E 3 3 o o
' ' 10 0 >
300 ' O 2000 4000 6000 8000 Time (s) rigure C-2. TRAC one RELAPS comparison, pressurizsr surge line brecir, reactor vessel dowrcomer fluid temperature. C-3
==
l TA8tE C-l. COMPARISON OF TRAC AND RELAP5 SEQUENCES OF EVENTS. PRESSURIZER SURGE LINE BREAK IRANSIENT Time of Event (s) Notes Event TRAC RELAPS (T = TRAC. R = RELAPS) I:reak opens 0.0 0.0 l'eactor and turbine trips 0.5 45 T-scram at 0.5 5 transient time R-scram on violation of P-T relationship Turbine bypass opens (both loops) 4.2 47 n . i iPI initiation 42.8 18.5 Turbine bypass closed (both loops) 93 117 l'C pump trip. MFW realigned 72.8 108.5 R. T-by sequence definition 3C s after RCP trip IFW tripped on 72.9 108.5 Vent valves open 100 554
" Candy canes
- volded 600 815 (toop A) -
1020 (Loop 8) flain FW ptmp trip 788 70 R. T trip on MFW pump high discharge P Core flood tank flow initiated Not calculated 2215 IPI initiation Not calculted 5124 Calculatton terminated 1800 6200
the integrated control system (ICS). MFW pump speed during the initial portion of this sequence is expected to be run back to the minimum speed as a result of input received from the BTU-limit calculator and MFW control valve differential pre:sure indication. Main and startuo centrol valves are expected to close during the initial portion of this sequence due to the BTU-limit calculator. Both of these effects increase MFW pump
. discharge pressure which causes a MFW oump trip. It has not been resolved why the MFW pump was tripped so much earlier with RELAP5 than with TRAC.
It is suspected that the two calculations used different minimum allowed main feedwater pump speeds or that the startup or main feedwater control valves were closed at different times. The effect of the MFW pump continuing to run with TRAC but not with RELAPS was that the strim generators were fed with a mixture of MFW and emergency feedwater (EFW) in the TRAC calculation but only with the colder EFW in the RELAPS calculation. This caused the RELAPS steam generator secondary pressures t: be far below those with TRAC as indicated by the Steam Generator A pressures shown in Figure C-3.
)
l The primary system depressurization shown in Figure C-1 continued with RELAP5 until it was stabilized by flashing of the reactor vessel upper head and hot legs, as shown ir Figures C-4 and C-5. The upper head flashing was more complete with RELAPS and this caused more liquid to generally be present at the break with RELAP5 than with TRAC as shown in Figure C-6. The generally higher TRAC primary system pressure from about 500 to 1500 s caused the TRAC break mass flow rate to exceed that with RELAPS as shown in . Figure C-7. The depressurization response comparison is thus related to many effects: pressurizer level at time of reactor trip, hot leg and upper head flashing behavior, and critical flow models. A discussion of upper head flashing appears as part of the main steam line biaak comparison in Appendix B. Loop natural circulation flow continued longer with RELAPS than with TRAC as shown in Figure C-8. This may have been caused by the different MFW pump trip behavior discussed earlier. With TRAC, the secondary feed C-5
. W , , i RELAPS
-- TRAC
^
. o a r -
T., 0- u s 1000 O e e g6 a
- '~ } s g
e s e b w a 4 .
- a
- 500 e E E 2 2 o 0
>2 - -
0 O O 2000 4000 6000 5000 Time (s) Figure C-3. TRAC and RELAPS comparison, pressurizer surge line break, steem generator A secondary pressures. 1 k - RELAPS y -- - TRAC 8 , TRAC LEVEL 8 CELL 9
= l i o I e >
E i l TRAC LEVEL 8 CELL 3 i 1.
' l ll 2 0.5 -
'e il -
I' 7f a t J g- o p. l' RELAPS
. . 'f,*% t (
i1 1 % 1 Ifl' 0 b ! udlI kl N _ $ IL !J i l j a. , 0 2000 4000 6000 8000 Time (s) Figure C-4 TRAC cnd RELAPS ccmparison, pressurizer surge li re break, roccior vessel upper head volc froctiors. C-6
e I l I iF RELAP5 f i
-- - TRAC I
b l
= .
C l !' c h t II f 4 i ~ 2 0.5 - i I
?
s I _ O l 1
, 11' lY "I '
O' 8000 C 2000 4000 6000 Time (s) Figur e C-5. TRAC and RELAP5 comparison, pressurizer surge line break, vold fractions at top of loop A hot leg. Y 1 i i 3 . RE' AP5
; --- TRAC
- l. y
! l 8 i I n
u I 4 y C ! 1 u g
- i >
T 0.S ,, I Id o
, l O g ,
J,$
\\ '
bl !
' ll
@ l -
g
> li d l l hj,a r hidsil,n 0 2000 4000 6000 8000 Time (s)
Figure C-6. TRAC and RELAP5 compcrison. pressurizer sarge li ne brecx, void fractions in coils upstrecm of break. C-7 1
400 i , i RELAPS . goo
-- - TR AC g
7 a c " 6 600 e o C 6 200 - - 3 g l 400 o, S I' hh ;
- r. I e
- l. il
. e I 11 ;- ,
y e I l I 0 O O 2000 4000 60C0 5000 Time (s) Figure C-7. TPAC and RELAP5 comparison. pressurizer surge line break. break mcss flow r o t es. 6000 . . . RELAP5
-- - TRAC
- g 3 10000
\ 4000 i -
7 2 i E
.o I O
O < 1 5000 O
' 2000 1 3
3 S 2 j 0-(
-- h n u :2: i -0 $
a
. 2
-2000 O 2000 4000 60C0 8000 Time (s)
Fi gur e C-8. TRAC ond RELAPS comparison pressurizer surge line break. cold leg A-2 mass flow rates o? racetor vessel C-8
i was performed with a comoination of EFW and MFW while with RELAP5 the feed was only with EFW and was colder. The colder feed may have <aused the extended RELAPS natural circulatien. The earlier termination of natural circulation with TRAC caused the . cold leg temperatures at the vessel to decline sharply as shown in Figure C-9 for the A-2 cold leg. With RELAP5 the decline occurred later.
~
The Loop A hot leg liquid temperatures, shown in Figure C-10 indicate the TRAC hot leg fluid temperature exceeds that of RELAP5 up to about 1500 s. The TRAC hot leg temperature was higher even though the cold leg temperature was lower because of a higher TRAC total vent valve mass flow rate, as shown in Figure C-11. The higher TRAC vent valve flow rate is
-consistent with the-lower TRAC. loop: flow rato.
The TRAC calculation was carried out to 1800 s and the RELAPS calculation was carried out to 6200 s. The figures showing comparisons between the two calculations used an abetssa of S000 s primarily to highlight the similarities between the calculations. While different TRAC and RELAPS short term behavior has been discussed, the plant conditions and .,/ trends at 1800 s, the end of the TRAC calculation, are similar. The primary system pressures, shown in Figure C-1, are in good agreement and the trends are converging with the TRAC depressurization rate slowing and that of RELAPS increasing. The same similarity is present in the secondary system pressures shown in Figure C-3. Loop flow has been essentially stagnated in both calculations at 1800 s as shown in Figure C-8. However, different minor flows have been established. With TRAC a manometer-type , oscillation is observed between hot leg and steam generator tubes but with RELAP5 a steady circulation is observed between cold legs on the same loop. Details of the RELAP5 circulation appear in Section 7.3. At 1800 s, hot and cold leg temperatures and trends of the two calculations are approximately the same as shown in Figures C-9 and C-10. The TRAC reactor . vessel downcomer fluid temperature, shown in Figure C-2 was about 25 K (45"F) higher than of RELAPS at 1800 s because of the higher vent valve flow shown in Figure C-11. A difference is noted in the break void fraction comparison at 1300 s as shown in Figure C-6. The TRAC void C-9 .
l 600 , , ,
^ 600 ^
M
!. RELAP5 b e h-- TRAC e =
3 g6 , 500 $- 0 ,l 4 AL 0
' 500 g
- 6 e l, l 1,9 I .e.
T 'd g II , 300 ."
=
h400 I - h li .. e 200 e h o o
> , , , 10 0 >
0 2000 4000 60C0 80C0 Tim. (3) Figure C-9. TRAC and RELAPS comparison, pressurizar surge line breck. cold leg A-2 liquid terrporoture at reactorvessel 600 , , , m ^ x -i RELAP5 600 *b
... TRAC e 1 e
6 6 5 % 550 1 o O
' 550 - 's% - 6 8 \ e 4 g Cl.
y 500 E V T ~ 30-500 450 % 6-
= =
e e
. E -4c0 E 3 3 o o 450 O 2000 4000 60C0 8000 Time (s)
Figure C-10. TRAC cnd RELAP5 comparison, pressurizer surge line breck, loop A hot leg liquid teraperatures. C-10 4 4
i 1500 . , , RELAP5 3o00
- - T9AC 7
) jv 2500 7 5 1000 - -
e i
\ 2000 *e o i I
b f y
, f I 1500 o o i
~ ~
E l .
} ] I 1000 I
- I g !
[ O O O 2000 4000 60C0 50C0 Time (s) Figure C-11. TRAC and RELAPS comparison, pressurtzer surge une brecic, tolol vent volve mess ficw ro t es.
\
l C-ll w--..-_ __ ._ __ _ - . _ _ . _ _ - -
fraction remains above that with RELAPS indicating more liquid at the break with RELAP5 than with TRAC. The nodalizations of the two models were compared and modeling of the break location was found to be consistent. The difference in break void fracticn apnears to be caused by the manometer oscillation in the TRAC calculation allowing vapor to enter the pressure surge line while with RELAPS the hot leg surge line nozzle remained essentially liquid covered. The higher. break void fraction at 1800 s with TRAC is the cause of the slightli higher TRAC depressurization rate at that time. In summary the two calculations have inconsistencies during the first 1800 s of the transient, the time period for which the TRAC calculation was run. Plant conditions and trends at 1500 s, however, are similar indicating that, had the TRAC calculatic been carried further, TRAC and RELAPS results would have compared favorably over an extended period. There is an uncertainty, however, in the depressurization ra*.e to be expected from about 1800 to 7200 s. The TRAC depressurization rate during this period is extrapolated to be higher than that calculated with RELAP5 which is due to a higher average break said fraction with TRAC. This app 6ars to be caused by minor differencet in hot leg flow behavior. C-12
'0"" . U.S NUCLE A A AEGUL ATOAV COMMISSION BIBLIOGRAPHIC DATA SHEET EGG-NSMD-6343 4 TITLE AND SUBTITLE 2 fteneescal RELAPS THERMAL-HYDRAULIC ANALYSIS OF PRESSURIZED THERMAL SH0CK SEQUENCES FOR THE OCONEE-1 PRESSURIZED 2 RECIPIENT s AccESSitw No WATER REACTOR 1 Ats f HOH(Si 6 OATE REPORT COMPLE TED C. D. Fletcher, M. A. Bolander, B. D. Stitt, M. E. Waternun " "'" July "'"1983 t) M HF OHMING ORGANI/ A flON N AME AND M AILING ADDRESS Itactus I.a Cool DATE REPORT ISSUEO wontM l vt aa July 1983 EG&G Idaho, Inc. , , , , , , , , ,
Icabo Falls, ID 83415 e ite- twel 17 SPONSOHING OHGANt/ AflOrd N AME AND M AIL'NG ADD RE SS (tacivem I.a Cowl p . Division of Accident Evaluation Office of Nuclear Regulatory Research ,, nN No U.S. Nuclear Regulatory Commission Washington, DC 20555 A6047 13 T Y PE OF HE POH T PE RIOD Cov E mio trec#ve,,e asest Technical l'a SUPPt i VE N I A H V N O TE S 34 (t, , es; 16 AHST H AC T (100 wman o< ecus Thermal-hydraulic analyses of pressurized thermal Shock sequences for the Oconee-1 pressurized water reactor were performed at the Idaho National Engineering Laboratory (INEL) using the RELAP5 computer code. This final report summarizes the results of previously reported calculations and presents the results of recently corapleted calculations. Comparisons of the two counterpart calculations performed using the RELAPS code at INEL and the TRAC code at Los Alamos National Laboratory, are included as appendices. 0 11 KE 4 WOHiri ANO DOCUME N t AN A L YS'S 1 74 OE SC hip T O MS 1in IDtN TIF IE R$ ope'd E NDE D TE RVS 18 AV AIL A81LITY bf A TE VENT 10 SE Cvqi f y CL Ass f raavoons 21 NO OF P AGE S Unclassified Unlimited :o u cuaiTv clas5 tra...,m :: PaicE Unclassified s mac s oav 2n . .. .}}