ML20023B018
ML20023B018 | |
Person / Time | |
---|---|
Site: | Three Mile Island ![]() |
Issue date: | 09/30/1982 |
From: | Berglund G, Dimenna R, Shimeck D EG&G, INC. |
To: | Landry R NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
Shared Package | |
ML20023B009 | List: |
References | |
EGG-SEMI-6022, EGG-SEMI-6022-DRFT, NUDOCS 8211010476 | |
Download: ML20023B018 (130) | |
Text
{{#Wiki_filter:~
-- + - , , *
- 7..
, g -i \: .x EGG-SEMI-6022 September 1982 -
ANALYSIS OF PRIMARY FEED AND BLEED COOLING IN PWR SYSTEMS D. J. Shimeck G. R. Berglund . R. A. Dimenna C. B. Davis J. P. Adams Idaho National Engineering Laboratory Operated by the U.S. Department of Energy
, s * }7 , . $ .h. '
Manguisanne mermag ,
.y r 4" M-_ . _ Mi m musumet ammans samme um 2.'F f
- s-me
- ; 1 ~ % y 4 ,-
wm. -- sem/ V g-2 _n
=7 .._p t
1
.$ % '. ' ~ % pa _T_ _~J __ y- - -3.h - , ~- W ; L ~' , ; , .p - t l
v - q, - sN .
' ~ ~
c -% - %h -
% . . .w . , -
Myg. b ,
-- 1 '
_- _ _ahM;: This is an informal report intended for use as a preliminary or working document Prepared for the U.S. NUCLEAR REGil.ATORY CCrtilSSION under DOE Contract No. OE-AC07-761001570 0 ggg 8211010476 82102b PDR ADOCK 05000289 P PDR ,
-er -- r-v -
e env-- --w------ - - - - - - - ' - - , + y -www
.E3rtS w EartMB !NTERIM REPCRT Access:en No.
Rooort No. EGG-SEMI-6022 Contract Program or Project
Title:
Semiscale 3ualect of : hie Documents Analysis of Primary Feed and Bleed Cooling in PWR Systems Type of Document Topical Author (s): D. J. Shir..eck C. 3. Davis G. R. Serglund J. P. Adams R. A. 01menna - Date of Cocument September 1982 -- i Resoonsable NRC Individual and NRC Office or Division: R. R. Landry, Reactor Safety Research . i This cocument was precared primanly for prelimina.'/ or internat use. it has not receivec l ' full review and acproval. Since there may be substantive cnanges this cocument shoutc l not ce considered final. ! EG&G icano. Inc. Idano 8 alls. Idaho 83415 Prepared for the U.S. Nuclear Regulatory Commission Wasnington, D.C. '- Under DOE Contract No. DE AC07 751001570 NRC FIN No. A6038 INTERIM REPCRT t ( ~ 4
. ... , .- e.. .: . . s .
- 1 9 ANALYSIS OF PRIMARY FEED AND BLEED COOLING IN PWR SYSTEMS 94 Approved: , . .
P. North, Manager Water Reactor Research Test Facilities Division Approved: % U). _e-- - G. W.dJohn'sen,Nanager WRRTF Experiment Planning and Analysis Branch 11 m
. q_.u _ _ m _... . _ . . . . . . _ . _ _ . . . . . _ ' '
- 1
. L e o .
ABSTRACT , Primary feea and bleec cooling as it pertains to a pressurized water ' reactor, describes an operation whereby reactor core cooling is maintainec oy injecting liould witn pumpaa emergency core cooling systems and removing the heated / vaporized fluid via the pressurizar power operstec relief . valve. Tnis report presents a systematic analysis of tne capanilities anc . limitations of primary feed ana bleeu. First, tne system parameters tnat 90vern sne uitimate capanility of feed and sleed are examinec along with . tne innerent assumptions. Data from Semiscale experiments is analyzec. l Phenomena tndt influence the results are identified anc analyzeG witn regara to typicality. Calculations witn the AELAPS ccmputer coce.are then . presentaa that verify tne ability of tne code to precict tne pnenomena ! ooserved in Semiscale. Finally, results from RELAPS are used to predict j full-scale plant response curing selectec feea anc oleed scenarios. [ F I 3 o
'L I
L f l i iii t b
~ *>QM Ne .h y. *MS . . . .r. . .: - . . w: ; >.a. .. . . . . .s . A ,
l SUPMARY This report presents results from an experimental and analytical study
- of primary feed and bleed cooling in pressurized water reactor (PWR) type systems. Primary coolant. feed and bleed cooling as it pertains to a
.. pressurized water reactor, denotes an operation whereby reactor core cooling is maintained by injecting coolant with pumped emergency core cooling systems and removing the heated / vaporized fluid via the pressurizer power operated relief valve (s) (PORVs). m ile a number of scenarios may be .
hypothesized in which feed and bleed cooling might be called for, this study is concerned with conditions in which a complete loss of secondary heat sink occurs. Other assumed conditions include:- the reactor has scrammed (decay heat levels), all or some pumped injection systems and the power operated relief valve (s) are operative, the pressurizer heaters are off, and the primary recirculation pumps are tripped. Although limited in scope, the study systematically examines relevant thermal-hydraulic
~
pnenomena attendant to feed. and bleed. The scope of this study encompasses , a simplified analysis of the mass and energy balances associated with feto and bleed, examination of experimental data from the Semiscale and LOFT systems, and RELAPS computer code analyses of both Semiscale and a full-scale Westinghouse plant. The parameters that govern the ultimate feasibility of feed and bleec were first examined, and the variables and uncertainties associated with those parameters identified. Examination of primary feed and bleed identifies four key parameters: core decay heat, cooling water injection capacity, PORV energy removal rate, and PORV mass removal rate. Other than core power, the remaining parameters are functions of system pressure. Simple operating maos may be drawn which define a steady-state feed ano bleed operating band for selected sets of these four parameters. A lower bound pressure for steady-state operation is defined by an energy balance between the core power and the PORV energy removal rate. An upper bound is defined by a mass balance between the injection capacity and the PORV mass ' j removal rate. Steady-state operation within this band is possible by cycling the PORV to maintain system pressure, and throttling the injection rate to maintain a net mass balance. iv e
----, - w -
. . . .i . .. 5 . -t r , , .
j
, 'l .
Of the above parameters which define a primary feea and bleed , operating pressure band, the PORV mass flow rate, and therefore energy removal rate are subject to the most uncertainty since they are strongly *l influenced by pressurizer coolant conditions. Variaticas may also occur in f the injection capacity (such as assumed failure of injection trains, etc.), and in the core decay heat power, which continually decreases with time. . Although limited in their ability to predict transient phenomena, the ! operating scos provide a method for quickly calculating the influence of ! variations or uncertainties on the ultimate capability of feed and bleed cooling. ; Two primary feed and bleed experiments were conducted in the Semiscale f Mod-2A facility to examine the transient thermal-hydraulic bt.havior [ associated with feed and bleed and to provide data for comouter code l assessment. The first of these, Test 5-SR-1, experienced excessive uncontrolled primary leakage and was merely used as a data base for
. examining selected phenomena. The second test used boundary conditions ;
simulating a decay heat level of 2% full power. The high pressure pumped ! injection capacity was scaled from Zion nuclear plant high pressure l injection system flow rates, but disallowed charging pump flow. The PORY
. relief capacity was approximately representative of PWR-scaled values f
(actually 20% larger than that of Zion). The aggregate result of the l chosen bouncary conditions was an operating map with a steady-state band f between 7 and 8.2 MPa when saturated steam PORV discharge is assumed. ; 1 Results from the experiments showed that when PORV was latched open the pressurizer filled and the PORV discharge appeared to be directly affected by hot leg coolant quality. In the second of the two Semiscale experiments the resulting high mass discharge rate was far in excess of the I injection rate and resulted in a rapid system mass depletion. Once ! sufficient mass inventory had been lost such that the hot leg had voiced, the PORY flow closely reflected the predicted saturated steam flow rate. i At that point the core was still adequately covered, but a small deficit - still existed in the inflow / outflow rates and eventually the core uncovered ! , and heated up. The deficit in the mass balance was small enougn that it was within the range of experimental uncertainties, and therefore makes no i > l l
.-a . ..r. l
. . . . . - . . . - - . - . . - -- . - . . . . . ~.- T .- ~ . ~. . : . a . . . ~~. *~. ... .,r s . ; =
i i^
~
direct sLatement as to the viability of primary feed and bleed. However, the excessive PORY discharge prior to hot leg voiding and the phenomena which caused.it are well outside the range of experimental uncertainty. Irrespective of experimental uncertainties, the mass inventory in the
- l . Semiscale experiment. could.not have been maintained with the PORY latched * ; open, until the hot leg voided.. ] Two factors were identified which directly influence mass flow out the
{ [ PORV. The first involves differences observed when the PORV was cycled open and shut, versus when it was latched open. Periodic closing of the l
} PORV in Semiscale promoted phase separation and resulted in an average PORV flow rate (when- the PORY was open) close to the predicted saturated steam flow rate. Secondly, analysis of Semiscale and LOFT data and analysis of code calculations-indicated that the orientation of the surge line connection to the hot- leg influences phase separation at the hot leg.
Suostantial hot. leg voiding was necessary to reduce the PORV flow in the i Semiscale experiments with a horizontal centerline connection, while minimal voiding was required in a LOFT test with a vertical top connection. Both orientations, and some intermediate angles, are used in current PWR designs. Posttest calculations were performed with the RELAPS computer code to predict the Semiscale experimental results. The RELAPS calculations correctly predicted the filling of the pressurizer and showed excellent quantitative agreement with the two-phase PORV mass discharge rate. The mass loss rate of the system was' closely m'atched. However the calculation l used the specified HPIS injection rate (as opposed to the actual rate injected during the test which was lower than specified). The results showed that once the hot leg had been voided the mass depletion trend toward core uncovery may have been arrested in the experiment had the specified injection rate been achieved. Until sufficient mass inventary had been last to void the hot leg, however, the PORV discharge rate still l far exceeded the calculated HPIS injection rate. The RELAPS code was next used to perform a calculation for a full-scale plant. The Westinghouse RESA0 plant design was selected. A vi
. t. .
i . complete loss of feedwater transient was chosen as a convenient scenario
. leading to primary feed and bleed cooling. The boundary conditions for this calculation were best estimate and included an ANS decay heat curve
- and full capacity HPIS and charging pump injection rates. Consistent with the low core power levels at the time feed and bleed was initiated (more than I he after shutdown), and the full . injection capability, the injection rate far exceeded the PORV discharge rate and the system was returned to a subcooled condition. .This result was consistent with a simple analysis of the operating map for the RESAR plant design, which indicated a wide .
pressure band within which primary feed and bleed was theoretically possible. Moreover, the RELAPS analysis of the RESAR plant gained significant credibility by virtue of the success of the Semiscale calculation,.which employed the same modeling techniques. While limited in its overall scope the present study, consisting of both experimental and analytical investigations, has shed considerable
- light on the subject of primary feed and bleed.' A simple analytical ,
approach to determining the feasibility of feed and bleed has been . developed and corroborated by experimental data and computer cada calculations. The Semiscale experiments have identified the factors i influencing PORV discharge, which is the most variable of the boundary {, ccnoitions influencing feed and bleed. The RELAPS computer code has been j shown capable of predicting the Semiscale experiments, and when applied to a full-scale plant has indicated that primary feed and bleed is a viable cooling mechanism. , , Based on the results of this study it seems safe to assume that primary feed and bleed would be a successful recovery procedure in the Westingnouse-type plant designs examined (i.e., RESAR, IION), assuming undegraded injection capability. Further. analysis appears warranted to predict the probable response in Combustion Engineering and Babcock and Wilcox plant designs. In addition, the current NSSS emergency operating procedures relevant to complete loss of secondary heat sink should be '- reviewed to determine if they adequately reflect an understanding of anticipated plant behavior in light of the results reported herein. vii e e amme w ese ,=es - qu> 66-uD-e*--
- ~ - ' ,,,.,,,,,c,.,,,.-,g, . ... - . . t e ,.. x. -c. . 4 , .. s . , . ,,,---l .s ,
FORE 4 0 l At tne reouest of tne Nuclear Regulatory Consission the Semiscale . Program conducted experiments designed to investigate the feasibility of primary coolant system (PCS) feed ano aleed as a means of rejecting decay heat in tne absence of steam generator heat removal. The results anc preliminary analysis of the experiments suggested that a reasonanle uncertainty may exist in the ability to effect stacle PCS feed and Dieed. Since current pressurizac water reactor emergency operating guidelines call for primary feed and nieed under certain abnormal conoitions, it was consicered of some importance that the general suoject of feed and bleed be studiec in some-depth anc that tne Semiscale results be carefully analyzed so tnat they might De interpreted in the proper perspective. To this end, tne Semiscale Program nas conducted an analysis effort involving Dotn experimental results (i.e., Semiscale and LCFT) and full-scale plants.
'4estingnouse design plants were chosen for the stuoy ou e to tne availaoility or infomation and existing computer cecxs at the'DiEL. The purpose of this report is to present the results of tne analysis of feed anc oleed, incluaing the recent Semiscale results.
- e viii e -o4,~*s- p ---w --v , , - - - 1 u -
. .a . . _ . . .. . . . - .. -.. .. - - . . . _ . . . .. .
, a ~
Q t
- ACKMOWLEDG4ENTS The authors wish to extend their thanks to L. J. Martinez for his assistance in analyzing the Semiscale data, to G. W. Johnsen for his direc-tion and review,.and to J. Berrey and T. Demitropoulos for the preparation of the report.
, .1 . - . ~ . .
h o t w 9 E e e i e ix i s'
~ . . . . - v:m - - . .:, ~~ ..s. . . -:v-:.u.- ^
s.. . .' . . . . r- r- a- '^=x .c.O
. s e t
- i. .
i CONTENTS
-t A85 TRACT ..................................................... iii
SUMMARY
...................................................... . iv FOREWORD ..................................................... viii L
ACKNOWLEDGMENTS .............................................. ix i 1. INTRODUCTION ........................................... 1
- 2. PRINCIPLES OF FEED AND BLEED OPERATION ................. 3 i
2.1 Theoretical Feed and Bl eed Operation . . . . . . . . . . . . . . 3 2.2 Uncertainties Associated With Steady-State Operating Pressure Sand ........................... 5 2.3 Factors Affecting PORV Discharge .................. 10 2.3.1 Pressurizer Coolant Conditions ana Primary Inventory .....'............. ............... 13 2.3.2 Pressurizer / Surge Line Geometry ............ 13
, 2.3.3 Surge Line Orientation ..................... la i
2.4 Summary Observations .............................. 15
- 3. RESULTS FROM SEMISCALE EXPERIMENTS . . . . . . . . . . . . . . . . . . . . . 16 3.1 System Configuration .............................. 16 3.2 Test Procedures and Conditions .................... 18 i
3.2.1 Pre-Feed and Bleed Operation ............... 18 3.3 Test Results ...................................... 20 3.3.1 Test S-SR-1 ................................ 20 3.3.1.1 S-SR-1 Predicted Response and Objectives ........................ 20 3.3.1.2 Test S-SR-1 Results ............... 22 3.3.2 Test 5-SR-2 ................................ 31 3.3.2.1 S-SR-2 Predicted Response and Objectives ........................ 31 l 3.3.2.2 Test 5-SR-2 Results ............... 34 3.3.3 Conclusions from Semiscale Experiments ..... 54
- 4. TYPICALITY OF SEMISCALE RESULTS ........................ 57 4.1 Experimental Uncertainties ........................ 57 X
4
i- . . .
. l, -
4.2 Surge Line Flooding ............................... 60 - 4.3 Surge Line Orie.ntation ............................ 60 4.4 Supporting Analysis, Applicable LOFT Data ......... 65
- 5. RELAPS ANALYSIS OF SEMISCALE TEST S-SR-2 ............... 72 5.1 Model Description ................................. 72 5.2 RELAPS Analysis of Test S-SR-2 .................... 74 5.2.1 Case 1 - Baseline Calculation .............. 75 5.2.2 Case 2 - Steam Generator SecondarygHeat Loss 81 5.3 Sen s i ti v i ty S tudi e s ... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 5.4 Conclusions from the RELAPS Analysis .............. 90
- 6. FULL-SCALE PLANT FEED AND BLEED CALCULATIONS . . . . . . . . . . . 91 6.1 Mode 1~ Description ................................. 93 6.2 Best Estimate Calculation Results ................. 93 6.3 Loss of Secondary Heat Sink With no ECC ........... 103
- 7. fnNCLUSIONS .........................~................... 110
- 8. REFERENCES ............................................. 113 FIGURES
- 1. Typical primary feed and bleed operating map . . . . . . . . . . . 4 j
- 2. Zion primary feed and bleed operating map for 2% core power .................................................. 7
- 3. Zion primary feed and bleed operating map for 2% core power without charging pumps ........................... 8
- 4. Zion primary feed and bleed operating map for 1.5%
core power without charging pumps ...................... 9
- 5. Effect of fluid quality on PORY mass flow and ener removal at 10 MPa ................................gy ...... 11 l 6. Zion primary feed and bleed operating map for 1.55
- core power, no cnarging flow, and 75% quality PORV i
upstream conditions .................................... 12 -
- 7. Semiscale system configuration for primary feed and
! bleed experiments ...................................... 17 l xi i
, , - - . - -. ----m ,- ,, r. ,- e_ -,s.
. ._c ,i.- . ,. .. . . 1 1: ... . . .. e
- 8. Semiscale Mod-2A primary feed and bleed operating map for Test S-SR-1 (high head HPIS) ....................... 23 -
- 9. Test 5-SR-1 pressurizer pressure and collapsed liquid -
level .................................................. 25
- 10. System pressure compared with' selected system fluid temperatures for Test S-SR-1 ........................... 27
- 11. Measured PORY and HPIS flow for Test 5-SR-1 ............ 28
- 12. Primary system mass bal ance for Test S-SR-1 . . . . . . . . . . . . 29
- 13. PORY flow rate compared tar predicted values and hot leg density. Test S-SR-2 .................................. 30
- 14. System pressure and PORV setpoints for Test S-SR-1 i sint 1 ................................................ 32
- 15. PAV flow rate compared to predicted value and hot leg density ................................................. 33
- 16. Predicted primary feed and bleed operatin map for Test S-SR-2 .............................g ............... 35
- 17. Pressurizer level and system pressure for Test S-SR-2 point 1 ................................................ 37
- 18. PORV and HPIS flow and primacy mass inventor for Test 5-SR-2 point 1 ........................y ............ 39
- 19. Measured and predicted (assuming 100% steam) PORV flow comparison with pressurizer collapsed liquid level for Test S-SR-2 point 1 .................................... 40
- 20. Collapsed liquid levels 1n steam generator tubes for Test S-SR-2 point 1 .................................... 41
- 21. Pressurizer collapsed liquid level and system pressure for Test S-SR-2 point 2 ................................ 42
- 22. PORV and HPIS flow and primary mass inventory for Test S-SR-2 point 2 ......................................... 43
- 23. Steam generator tube collapsed liquid levels and primary system pressure. Test S-SR-2 point 2 .................. 45
- 24. Measured and predicted (100% quality steam) PORV flow ccmparison to hot leg density for Test S-SR-2 point 2 .. 46
- 25. Pressurizer pressure and collapsed liquid level during Test S-SR-2 point 3 .................................... 47 xii e G e
.. i
- 26. . Measured and predicted (100% quality steam) PORY flow comparison to hot leg density for Test S-SR-2 point 3 .. 49
- 27. PORY and HPIS flow and primary mass inventory for Test '
S-SR-Z point 3 ......................................... 50 .
- 28. Steam generator tube collapsed liquid levels ........... 51
- 29. Pump suction levels. Test S-SR-2 point 3............... 52
- 30. Vessel collapsed liquid level. Test S-SR-2 point 3 .... 53
- 31. Core liquid level compared with representative temperatures during Test S-SR-2 point 3 ................ 55
- 32. Semiscale Mod-2A primary feed and bleed operating map for 21 core power with uncertainties ................... 59 33.- Comparison of flooding correlations with surge line steam velocities for Semiscale and ran of.PWR geometries .........-..................ge . . . .. . . . . . . . . .. . . . 61
- 34. Semiscale Mod-2A hot. leg / surge line qcnfiguration for feed and bleed tests. (Not to scale.) .................. 63
- 35. Liquid entrainment predictions for side and top entry surge lines ............................................ 64 .
- 36. Axonometri c projection of LOFT system . . . . . . . . . . . . . . . . . . 66
- 37. LOFT Test L9-1 PORY mass flow and intact 1000 hot leg density ............................................ 69
- 38. Pressurizer liquid level respense ...................... 70
- 39. RELAPS nodalization of the Seuiscale Mod-2A system ..... 73
- 40. Comparison of measured PORY flow with RELAPS predicted PORY flow. Test S-SR-2 point 3 ........................ 76
- 41. Comparison of measured and RELAPS predicted upper plenum pressures. Test S-SR-2 point 3 ........................ 77
- 42. Ccmparison of measured and RELAPS predicted pressurizer liquid level. Test S-SR-2 point 3 ..................... 78
- 43. Ccmparison of measured and RELAPS predicteo vessel liquid l 1evels. Test S-SR-2 point 3 ........................... 80 -
- 44. Comparison of measured and RELAPS predicted HPIS flow.
Test S-SR-2 point 3 .................................... 82 i xiii-l l
,-n ,,,---w
..,. . . . . _ ~ , ~ . + .s 9
i
;_ 45. . RELAPS predicted PORY and HPIS flow rates. Test - -f f
S-SR-2 point 3 ......................................... 83
- 46. Comparison of measured and RELAPS predicted PORY mass flow rato. Test S-SR-2 point 3 ........................ 85
- 47. Comparison of measured and RELAPS predicted system pressure. Test 5-SR-2 point 3 ......................... 86 l 48. Comparison of measured and RELAPS predicted pressurizer a : liquid level. Test S-SR-2 point.3 ..................... 87
- 49. Comparison of measured and RELAPS predicted core liquid levels. Test S-SR-2 point 3 ........................... 88
- 50. Primary feed and bleed operating map for RESAR 4 -
calculation ............................................ 92
- 51. RELAPS model of RESAR .................................. 94
- 52. Core power'for the RESAR feed and bleed galculation .... 95
- 53. Calculated pressurizer pressure for RESAR feed and bleed .................................................. 99
- 54. Calculated pressurizer liquid level for RESAR feed and bleed .............................................. 100 t
- 55. A comparison of PORV and ECC flow rates during RESAR feed and bleed ......................................... 102 4
- 56. The effect of ECC on pressurizer pressure (RESAR)....... 105 i
.i 57. The effect of ECC on pressurizer level (RESAR) ......... 106 i
- 58. The effect of ECC on PCRV mass flow rate (RESAR) ....... 107 i 59. PORV flow rate and hot leg density for the RESAR calculation without ECC ................................ 109 TABLES
- 1. External Heater Power Levels ........................... 19 1
i
- 2. Initial Test Conditions ................................ 21
- 3. Sequence of Events for Test S-SR-1 ..................... 24
- 4. Sequence of Events for Test S-SR-2 ..................... 35 l
ivx l
.e , .
- 5. Semiscale Experimental Uncertainties ............. ..... 58 ,
- 6. Initial conditions for the RESAR Calculation ........ . 96
- 7. Sequence of Events in the RESA'R Calculation ............ 97
- 8. Sequence of Events in the RESAR Calculation Without ECC .................................................... 104 e
e e { I
. i VX I. 1 4 , , ,,,--- r- -, y- - % ..-,w , , , . . . - . ,w ._ - ,,n-,--
. - . - .. __ - . - - - - - ~. _ .. - . . . . . .g . .5 . ~'. .
ANALYSIS OF PRIMARY FEED ANO
- SLEED COOL!tiG IN PWR SYSTEMS
- 1. INTRODUCTION Cartain transient scenarios may be postulated for pressurizec water reactor plants wherein the capability for delivering water to the seconaary of tne steam generators is lost. Once tne remaining water stored in the seconaaries is cepleted as a result of oeing coilec off by tne cecay heat
, generated in the core, the loss of heat sink will result in pressurization of the primary system. Shoula tnis occur, one metnoa tnat une operator nas available to maintain adeauata core cooling ano to control primary coolant system pressure is.to open tne power operated relief valve (PORV) on top of Ine pressurizer and use nign pressure pumped emergency core cooling (ECC) injection to maintain inventury. Inis procecure is commonly referred to as primary feed and oleed. Feea and oleea ccamences wnen the PORV(s) are openea (oleed) and hign pressure injectio'n oegins (feea). The passage of steam out tne PORV(s) provices for tne rejection of ascay heat wnile tne ,
introcuttion of ECCS' coolant provices makeup for the resultant coolant loss. I A numoer of concerns arise when examining the feasioility of primary feed ano oleed. There is the general question as to whicn parameters ultimataly govern the auility of a given plant to maintain a steacy-stata j feed and bleed operation; from what range of initial conditions is it possiole to depressurize,a system while retaining sufficient mass inventary to xeep the core cooled?; what effects ao geometry ano, in tne case of experimentai systems scale, inauce on integral systen cenavior? While a multituce of variaoles ano scenarios exist that coula leaa to tnis situation, tne focus of tsis present analysis is on primary feea ana bleec cooling in a multi-loop pressurizea water reacter system typical of a a-loop Westingnouse cesign. It will examine the feasioility of acnieving a favortole coolant ano energy balance under conaitions in wnicn:
- 1. The reactor nas scrammeo
- 2. The steam generator secencartes are c:mpletely ceoletec of ecolant 1
O
- - .--...w.- . . , .. . -. _ , . . , ._....x . 1 .
- 3. The nign pressure injection system (ECCS) is operative
- 4. The pressurizer heaters are inactive ,
~
- 5. The pressure-operated relief valve (s) (PORVs) are operative
- 6. Primary recirculation pumps are off.
Tne next sectinn of tnis report will examine tne tnearetical feasioility of maintaining steady-state feec anc bleed cooling by examining the parameters inich govern it. ,The variaoles and uncertainties that affect feec and bleec operation are icentified and briefly discussec. Tne follow'ing section will present an analysis of experiments conducted in the Semiscale Moo-2A experimental facility which involvec feed ano bleec witn scaled PORV and ECC flows. The typicality of the results will then be ciscussac by reviewing the experimental uncertainties anc variaoles. The last two sections will present results from RELAPS coce precictions of the Semiscale experiments ano a oest estimate calculatten of a scanario involving primary feed and oleed in a full-scale PWR. Finally, conclusions will oe summarizeo from tne analysis as a -nole. l l l l l 1 1 2 6
-m_. -,,,. -- - - - . _ .- _ . , _ , _
. . _. . ;.e J. , ,
4
- 2. PRINCIPLES OF FEED AND BLEED OPERATION ,
2.1 Tneoretical Feea ana Sleea Coeration The ocjective of primary feed and bleed is to remove core cecay heat in the aosence of heat transfer in the steam generators wnile maintaining a favoracle c.colant inventory. Figure i snows tne important parameters for determining the feasibility of primary feed and bleea operation anc incicates tne ,,ossibility of a steacy-state operating bana. The governing parameters wnich cetermine this operating band are cecay neat level, HPIS flow rate, and PORV flow rate (ana enthalpy flow rate). Except for tne core cecay heat level the remaining parameters are functions of primary system pressure. Tne lower bouna of the operating bana represents tne ininimum pressure at which tne PCRV can pass enougn steama (witn tne coolant replacea oy amoient temperature water) to remove sufficient energy from the system. Steacy-state operation oe, low tnis pressure witnout acoitional energy removal patns is not possiole. Operation at a pressure aoove the lower bouno may me tneoretically accomplisnea by cycling tne PURV cpen anc closea witnin a cesired pressure cand. Tne upper pressure cound to the steacy-state coerating bana is cefinea .I sy a talance bet..een tne PCRV average coolant removal rate ano tne HPIS lf coolant injecticn rate. The average PCRV coolant removal rate is sicoly !, defined as the core power divided by tne cifference between inlet anc ~ outlet entnalpies: ,
-h AVG ' Ocore/I"out in)
(Tnis relationsnip assumes tnat tne coolant eemovec :nrougn tne PORV is replacec witn amoient temperature water at :ne same flow rate. Actually a coolant ceficit exists at pressures higner tnan ene uoper bound anc a steady-state condition cannot exist aue to a continual loss of system .
- a. It is assumec nere tnat lucy. cuality steam is aiscnargea enrougn tne PORV. The effect of reaucea cuality is examinea later.
3
i 5 i .
! j r '
4j -
+ .r s '
N ~ s o
~
Core Power
- , g .
i
,/ --- --- 2 ~
l iM,/ j ; o i E T-o\ D p o . j (a g e a 1 i c n I I I l l 3 , 0 2 4 6 8 10 12 14 16 System pressure (MPa) , .
- ' Figure 1.
Typical primary f eed and bleed operating map %wun : ll l .- 8 .- 0 O egg , e
< \'
)) , ( I
- 6 * +
y~ " 5 0 5 0 S 0 7 7 6 6 S 5
. 0 j .
1 s a 9 P
- l M i _
_ a 0 - v 1 o m
= e
- e r
* : 8 r
y-g u r s e s n e e p r . d m - n e i 7 a t s w y S % l f s s 6 a V m e y R t t O a i P
- r l a f l
a u o i v i 5 Q y
% e o
m d i i u t l r l a F u
* % y g
r q i 4 d e i i n u e l f V . R f a O oP P N e 3 t c0 M e1 N' f ft Ea S
, N a 2
e r u N i f g
, \ t 1 \ *D 0 0 0 0 0 o 6 4 3 2 1 ^ d' E ?"
- E C
l !
i'. ; I 4
- I I
i p. 200 45 ,
\. +
40 17 5 -
. h . m .-
35 \" ' i 15 0 - N4'.s G .' l
.M 30 v n
3 2 12 5 - \ 4 e l-.. [ v f PoltV Moss llemoval - 25 0 ! '- L N x oi 10 0 - 7, - - , , L 8
- e 20 o C -
ad 75 -
, ' , - .I . 15 i ' N '
50 - Core Power
'# -~ -
s O
, _ g ,
25 - S
' ~
e6' 9'193ffS ' ' ' ' l ' ' 0 --- O O 2 4 6 8 10 12 14 16 . Syslem pra:;siare (MPa) 5i waitr u-a
, {-l Figure 6. Zion prima,y feed and bleed operating snap for 1.51 core power, no charging flow, and 75% quality PORV upstream conditions. l ]' . ? .. 1 t
v f
- g. .
I .
. o- .=
200 45
\
17 5 - 40
- e. _
Vo', 35 # ! 15 0 - N N ! s ta j h 1 12 5 -
\,
e v 25 o
- o. 10 0 -
y _ A _ _t9(v uase_p g a! ~ - 20 $ - Es 75 - g C
,/' ~
15 a 50 - g,- y , y 25 e j[ 5 t 0 1 8 8 ' ' ' 1 - 0 0 2 4 6 8 10 12 14 16 System pressuru (MPa) w aaras-e figure 4. Zion primary feed and bleed operating map for 1.51 core power willicHL charging ptaips. l f f i 1 ! . l . ,
a -. - - ... . .- ._. -- .- -
, -.. . . . .. ..s. ..L.
l l The above curves are cased upon.the assumotion tnat 1007. cuality steam exists at the PORV. Figure 5 snows tne sensitivity of tne PORV energy removal curve to lower cualities as determinea with the HEM flow mocel. , Since the energy removal per unit mass decreases wnile tne mass flow rate increases the energy removal rate initially aecreases witn decreasing quality. However, since the mass flow rata increases substantially with decreasing quality the. energy removal rate eventually increases. The effect on tne lower operating bound pressure is not large; however, tne large increase in PORY mass flow witn increasea quality rapialy lowers tne upper eno of the bano. As an example, for tne conditions usea in Figure 4 tne operating Dand coes not exist at qualities below approximately 75% (see _. Figure.6). ,, The foregoing analysis is useful, in that it provices a basis for examining the feasioility of feed ana bleed and for quantitatively assessing the effects of uncertainties or variations in tne councing parameters. However, it coes not address transient behavior tnat may have an important bearing on the ultimate viaoility of primary feed anc Dieea. In particular, it snould be evicent tnat tnere exists some uncertainty regarding the aDility to safely bring the primary coolant system to witnin the " feasible" operating pressure band without sustaining unacceptable coolant loss in the process. Factors wnicn bear on tnis transient process incluce tne primary coolant system state at tne initiation of an attemot to feea ana oleea, ano the nature of tne coolant discnargea enrougn tne PORV(s) in depressurizing tne system to witnin tne operating cana. These questions can only ce acoress'ea tnrougn" experimentation ana One use of computer coce analyses. 2.3 Factors Affecting PORV Disenarge Of the factors previously discussed the largest uncertainty affecting tne feea and bleea operating bana arises from the influence of two-pnase PORV flow. The mass flow througn tne PORV is depencent on upstream fluid - conditions at tne top of the pressurizer. Several factors contribute to 10
I V i
}'
200 - 12 0 , i N ,
'Qq 17 5 -
_ g30 m .,
% M l' 15 0 - \ ! ? En ^ .3f 80 v
{ y 12 5 -
\ '
u
' f N 100 - -
60 I' t C /,\ , - Ed 75 - Core Power ,/ N-y ' 4g , ; 50 -
- 4 '16/ 3 !
PORV Mass Runevol_ _' N h ,p,- I vf - 20 25 -
',4-g / i ' ' I ='
0 i 8 0 , 0 2 4 6 8 10 12 14 16 System pressure (MPa) we.aw is-i figure 2. Zion primary feed and bleed operating map for 2% core power.
}
d j. j . 2 e %
.~ ;1 '
i
..!. i ti 45 \ . -
4 us -
~
n ..; 15 0 - 35 g" ,i I tm n x .. 30 s-3 12 5 - ' 3 ronV Avara --- -
- 7--- - Disn~ fI*_*9*_ _ -
25 D x ,
- 0. 10 0 -
/ ,g y ~
ia 75 v . ;.
-_ _ _ _ Cum _lh m -
2 -:- - 5 m I
/ / .
5 50 - 1/ 2 10 j ,3 25
)
5 i i e i .i i 1 'l - t 0 2 4 6 8 10 12 !4 16 Sysiem prassure (MPa) Figure 3. Zion primary feed and bleed operating map for 21 core power .! without charging pasaps. :] N
m_. _ . _ . . _ . . . _ _ _ . -_ . j _- . Coolant inventory.)- Below the upper bound tne. system mass inventory can . theoretically be ..saintained *11 thin'a desired operating range'oy either j tnrottling tne HPIS or cycling it on and off.
- I l
,i A numoer of studies have been conductea to predict PhR system behavior under feed and bleed operation following loss of secondary heat sink (e.g.,
References 1 and 2). 'However, tnese studies nave made use of large j thermal-nyoraulic computer codes to examine limited numbers of scenarios. ! The simplifieo approacn presenteo here, altnough lacking in its ability to preoict tne influence of transient phenomena, allows examination of tne factors that determine the ultimate capanility of feed ano oleec for given plant parameters. It presents a starting point.for examining tne sensitivity of any variations or uncertainties with minimal expenditure of time. Sucn a stuoy is presented oelow. 2.2 Uncertainties Associated witn 5teaoy-State coerating Pressure Band In practice tne curves discussed above are not well defined oue to several uncertainties. Suoject to the greatest uncertainty are tne PORV mass removal and energy removal curves. The mass flow througn the PORV is cepencent on the fluid conditions at tne top of tne pressurizer. If the pressurizer is nearly . liquid full.the flow tnrougn the PORV will be a mixture of liquia and vapor. At a given system pressure tnis results in greater mass flow tnan for saturateo steam flow tnrougn the PCRV. The result of naving two-phase flow tnrougn the PORV is therefore to lower the upper bouno pressure. Uncertainties also arise in the PORV energy removal curve due to two-onase flow. With decreasing quality tne energy removal per unit mass decreases wnile tne mass discnarge rate increases. Depenoing upon the quality tne energy removal rate at a given pressure may be less tnan or greater tnan that for saturated steam. The lower bound of tne operating band will vary accordingly. - Another significant variaole that affects tne wiatn of the operating bano is tne actual heat load that must ce rejecteo tnrougn tne PORV. The 5 4
.e.,-_ n.__-,.---e --~..--r- -. - , - - - e--=-~-
= :
energy that must be rejecteo by tne PORV is reduced as core decay neat decreases witn time after shutoown and also if additional neat. sinus exist. These additional heat sinns may be sucn tnings as environmental , heat' loss or residual water in tne steam generator seconaaries. The result of adottional heat removal paths and/or lower core power is to lowe: the bottom eno of the operating cano. Coinc.ioent witn tnis is the reouction of PORV average mass flow which raises.the upper bound pressure. Another factor whicn affects the operating band width is the HPIS injection flow rate. As Figure 1 inoicates, tne lower the HPIS injection capacity the lower tne upper bound of the operating band will be. The quantitative effects of the uncertainties and/or variances discussed above are illustrated in Figures 2 througn 6. For tnese examples the curves were generaten using availaole cata ootained from tne Zion I nuclear generating plant,3 a 3411 MW(t) pressurizea water reactor. Figure 2 snows a primaFy feed and oleea map for a 27. cecay heat power l level. A steady-state operating band is seen to exist between 7.5 ano 14 MPa. A decay heat level of 27. of full power is typical of tne time period from about 10 min to 20 min after snutdown'. F.igure 3 is a similar curve, but nere no makeup pump injection is assumeo; only the HPIS pumps were assumed to oe operating. The HPIS pumps are snown to deachead at about 10.3 MPa. For this case no steady-state operating cano exists, since at the minimum pressure wnere the PORV can remove the energy there is a mass ceficit between the PORV coolant removal and the HPIS injecticn capacity. ~ Figure 4 snows the primary feed and bleed map for 1-1/2Y. full power, a decay neat level typical of the period from 1/2 to 1 nr after shutdown, and for only HPIS injection. Comparison to Figure 3 snows that tne reduction in core power and corresponoing PORV average mass flow born act to estaolisn a steacy-state operating bano. [ l l
- a. Assuming end-of-life reactor fuel conditions 6
l
~* '
estaolisning pressurizer flufo conditions. The ones discusseo nere , are: pressurizer coolant ccnditions, primary coolant system concitions, pressurizer / surge line geometry, ano surge line orientation.
- 2.3.1 Pressurizer Coolant Conditions ano Primary Inventory If feed and Dised is not initiatec soon after losing tne secondary heat sint and pressurizer heaters, tne crimary licuia swell will fill tne
.pressurizar anc collgse tne steam bubbla. Several conditions may form or sustain a vaoor bubble at tne top of tne pressurizer. A vapor aucale can be produced by loss of pressurizer licuia inventory, neating of tne fluid to saturation, and/or aepressurization. In tne present study tne ~ ~
pressurizer heaters are assumed to be n'onoperational and direct heating is therefore precluceo. In a transient cearessurization, licuid flashing in tne pressurizer will tend to create a nign cuality regicn near tne too as long as tne fluia in tne pressurizer is tne nottest in tne system. However, the liquid swell that accompanies' bulk flasning will tena to decrease tne quality at the top of the pressurizer. For either a cuasi-steady-state situation, or in a transient once the original pressurizer inventory has oeen replaced with coolant from the' hot leg, the PORV fluid canaitions are cepenaent upon tne conditions in tne not leg. If i tne coolant lost tnrough the PORV is replaceo by low auality fluid tne mass - discnarge out the FORV will remain fairly nign. This will' occur until the primary system inventory is reduced enough to cause significant voicing in tne hot leg. Once significant hot leg voiding occurs pressurizer / surge line geometry and orientatio,' become important as cescribec celow. 2.3.2 Pressurizer /Surce Line Geometry
, For a given vapor volume a pressurizer witn a large length-to-aiameter ratio would nave a " tall" voia height relative to a pressurizer with a smaller ratio, in acottion to also naving a smaller cross-section. A steam Ducole of greater neignt would tend to ennance separation from the vaoor of .
licuia droplets createa by cuboles breaking througn the licuia surface, due to'tne greater wall surface area ana recuced potential for croplets being thrown upwara into tne nign vapor velocity area near tne PURV line 13 i 1 I l
. - . - . . . .2 .
l
~ . . f. s. .. u .. 3c .. . c . ~ :~ . . .. . %., .. .s .
E -entrance." However, since vapor must Dy necessity pass througn the pressurizar liquid from the surge line to the PORV, a large L/0 would tena to promote liquid swell and droplet entrainment due'to tne smaller , cross-sectional area. In any case,'tne' influence of pressurizer geometry is prooably 4 oversnadowed by the. preclusion of counter-current flow in the surge line. Even if a lieuto/ vapor separation mechanism dia exist in tne pressurizer, typical surge line velocities are well acove floocing limits.D Therefore, pressurizer liquia could noE arain cack to tne loop anc woulo continue to be storedL in the pressurizer until tne PORV aisenarge quality self-adjusted to accommodate removal of tne mass. It tnerefore appears ._ . necessary to.have nign quality steam supplied from the not leg in order to have_nign Quality PORV disenarge. ! 2.3.3 Surce Line Orientation 1 If hot leg voicing aces occur, the orientation of the surge line woulu influence the primary system inventory at wnicn nign cuality steam entered tne pressurizar. Surge line to hot leg connections of various orientations, from horizontal side entrance to vertical top entrance, are used in current PWRs. With tne top entrance line, and.cutescent hot leg conoitions, minimal hot leg voicing is necessary to allow nigh quality surge line flow. With a sice entrance line tne not leg pipe liquia level [ must arop much lower before hign quaTity flow begins. In eitner case the j surge line flow may stili ce'variea si'gnificantly if nonouiescent l conditions exist that aisrupt stratifiea flow, sucn as wnen primary recirculation pumps are turned on, or a transient cepressurization is occurring.
- a. For typical PWR pressurizer dimensions, the vapor velocity ~
open PORV) in a vapor fillea cross-section is on ene order of 1(due ft/s to wnicnan presents little cnance of croplet entrainment.
- b. Otscussed in Section 4.2 14
. - . _ e sa ...._.2 m _u.u .
[. . 2.4 .Sumary 00servations , Basea on tne foregoing discussion it is concluaed tnat a simplifiec ' approach to determining tne feasioility of primary feed ano oleed in a pressurized water reactor lies in the mapping of energy and mass flows. Moreover, tnis tecnnique can De used to quantitatively assess the sensitivity of tne operating pressurt cana to variations in the boundary conditions of ECCS flow, PORY flow, ano decay heat. The operating map represents an ultimate statement as to wnetner feed and bleed is possible,. and is tne starting point for examining specific design features tnat oear on the operating bounds. It is evident tnat plausiole variations and , uncertainties in these parameters can lead to tne elimination of a steaoy-state operating pressure range. Principal among tnese uncertainties is the coolant discharge tnrough the PORV. The predictaoility of this single parameter is suoject to mucn greater uncertainty tnan eitner cecay neat or ECCS flow. Irrespective of the existance of a tnearetically feasible operating pressure cana, there remains tne question as to wnetner tne reactor system can ce safely maneuvereo into tnis pressure range. In Inis regara it is clear that a cepenaence must be placed on computer coce analyses (witn suitable verification) and acequate supporting experimental cata. Sucn analyses and/or experiments snoulo examine tne plausiele scenarios wnicn leaa tne operator to comence primary feea and bleea, since tne initial condition of the primary coolant system (particularly inventory) will nave a significant effect on the outcome. . e 15
~
.u m <. :,a c.. .. . :: . .. . . . . , . ., . '3. RESULTS FROM SEMISCALE EXPERI!ENTS Experiments were conducted in the Semiscale M00-2A facility to -
- evaluate system behavior curing primary feed and bleea operations. The
; Douncary conoitions of' HPIS injection rate were scaleo from Westingnouse plants with either "hign head" (injection capacity up to the safety valve setpoint) or " low head" (pump deadneaa at typically 10.3 MPa) HPIS pumps.
1 The PORV discharge capacity was scaled close to the value for a full-size plant, but was sligntly larger. This was consistent with tne use of core power levels that were on tne high eno of typical oecay neat values, so as
'to allow more- positive ooservation of system performance witn less aistortion from environmental neat losses. Consistent with tne analysis of Section 2 regaraing' tne ultimate feasiality of primary feea anc bleed cooling, tne councary conaitions selected for the Semiscale experiments provioed for a steady-state operating bana. As descrioso celow, tne experimental results therefore allowea for examining tne influence of the assumptions implicit in tnose simplifiec analyses on the true feasilibity of feed and bleea cooling.
3.1 System Configuration For Semiscale Mod-2A Tests S-SR-1 and 2, the Mod-2A system was configured as shown in Figure 7. The major comoonents of the system were the vessel witn electrically neated core and external cowncomer, intact and broken loop steam generators,. broken lopp recirculation pumo, and loop piping. The vessel core consists of a 5 x 5 array of internally neatea electric rods, 23 of wnicn were powered. The rods are geometrically similar to nuclear rods witn a heated lengtn of 3.6e m ano an outsioe aiameter of 2.072 cm. The power was distributed sucn tnat tne center nine , roos were powereo at 1.25 times tne average rod power. The primary system also incorporatea tne use of external neaters on loop piping anc on tne pressure vessel to mitigate tne effects of neat loss to tne environment. A more cetailed description of tne Mod-2A system may oe found in Reference 5. Since tne recirculation pumos were not useo during unis test, tne ! intact loop recirculation pumo was replaced with a section of pipe to i i 16 i e
- - ., .. . - , ,w,. -
e a
+
i ej i l i i el . l. M ! l
@ t i
1 l'l ! l l ,i
! ! ti esenen i i ii o'"""'**
s J I
/
t I 3asem generosor Stamm g.nasesor
- f. W '
I s 'l W g .ea, w tsstarzta f
)uhe n
j ngsy ;\ vg . . _ _ _ . _ _
,., .%,. me a wr: . __
g[$ . Qi - f
. ,n. - C -M ! i, j l$ k* .ns ts.sa.n G- !!! '
ewcz wt > s.sa.n } e f a
$%f an,a,.=,,,,n,a,,s,,,m, i.. ,3, ,
ji s - jlI d;4 /'-l
/ M '
c1 4
- aeomw' 'a# i 1 5! l l q t.
!. I i =i l M
v i i! !
. 1 m : mr
- 1 Figure 7. Semiscale system configuration ?0r primary feed and bleed exper1 merits.
17 l
- l
- -, .- , , , , - - _ , - _ . - . - n --- ,- - -,,
m.. . . . , .,_ . . . . .-. . . precluoe a'relatively large primary leax. Tnis section of pipe was orificea to represent the scalea nyaraulic resistance of a pressurizea water reactor primary pump ir. a locxeo rotor (stoppea) configuration. . Other important configuration details were as follows:
- 1. Tne HPIS fluic enterac tne primary system tnrougn octn tne intact and broken loop cola legs auring Test S-SR-1. During Test S-SR-2 .
all of tne HPIS entered tnrougn tne intact loco cold leg. Tne broken loop HPIS was usad for leakage makeup only.
- 2. A square edged orifice of 0.1549 cm 10 was placea upstream of a remotely controlled valve to simulate two PORVs. The critical flow rate proviceo by this orifice was typical of the scaled relief rate of PORVs in a full-size PWR, Relative to tne puolishea Zion plant PORV capacities the orifice used in tne Semiscale experiments had a flow area 20ie greater tnan the correctly scalea value.
- 3. The outlet line from tne pressurizer PORV was cunnected to tne concensing and measurement system. This system was usea to concense PORY effluent ana measure it so an accurate aetermination of PORV mass flow rate could be made.
4 For these tests tne seconcary sices of tne intact ana broxen loop steam generators were crained ana isolated. 3.2 Test Proceaures ana Cenditions 3.2.1 Pre-Feed ano 31eea coeration Activities Prior to initiating each experiment the primary was brougnt to the l desirec equilibrium temperature (minimal temperature gradients around the . primary). The external neaters were powerec at the levels inaicatea in ! Table 1. Once the desired temperature was obtained the core power was l adjusted to maintain steady-state. Thus, tne core power ccmaensatea fcr 18 L .
,, _.,., - _ - , _ _ - -w
table l'. EXTERNAL HEATER POWER LEVELS Power , Location (kW) Vessel 20 Hot' legs 7.1 Cold legs 3.3 I.L. pump suction 8.5 8.L. pump suction 4.2 Total 42.1 e e e 19
~
e
-. ,- %, . , _ . . , , , - _ - - . _ _ _ _ _ . - . _ . , _ , , - , ,m-. - , .. - , . _ , _ . , , .- , , . , . , _ _ _ __.,, ,y-, -
3- - - .. - ---
;sn :i.. .w . . . . . , ,-:.. .a . : , u ... f.:... . . , = ~
the remaining primary environmental neat losses ana -also neat transfer to 4 the seconaary sfoe if the steam geInerators. Leax rate checks were also mace. . For Test 5-SR-1 the coolant loss was uncompensatea. For Test 5-SR-2 , the leakage was mace 'up ny cold water injection-into tne oraken loop cold leg. Results of the lean rate tests are given in Taole 2. t 3.3 Test Result Test results are presented in this section for Tests S-SR-1 ano 2. Eacn of these tests was performed with a constant net core power of 40 kW." For Semiscale this represents 2% of full power. This level of heat is representative of decay heat approximately 10 to 20 minutes after a scram and is high enougn above Semiscale environmental neat losses to have a measureable effect on system response. The orifice used to simulate tne pressurizar PORVs in a full-size PWR l wassizactoprovicearepresentativesteamreiiefcapacity. The ECC systems of importance for primary feea anc bleed are those capaole of injecting water at relatively high pressure. The injection rates usea for these experiments were typical of the HPI systems in PWRs. As aiscussea below, two different injection rate versus system pressure curves were usea in tne tests performed. The aggregate result of the selectea bouncary conditions on tne theoretical primary feea and bleed behavior addressed in Section 2 is aiscussea for each experiment. An analysis of actual system behavior is then presented. 3.3.1 Test 5-SR-1 3.3.1.1 5-SR-1 Predicted Response ana Oojectives. Test 5-SR-1 was performec using a "nign neaa" pumpea HPIS injection capacity. The HPIS flow was powered scalea basea on HPIS flow information for the Nortn Anna plant.6 The feed and oleed operating map for Semiscale casea on tnis
- a. Core power nas augmented to compensate for measured environmental neat i losses and heat transfer to tne seconcaries.
l 20 l l
g------ - _ _ _ _ ..._.._= TABLE 2. INITIAL TEST CONDITIONS i l Test Test Test - Test 5-SR-2 5-5R-c S-5R-2 Parameter S-SR-1 Point i Point 2 Point a < Pressurtzer pressure 12.43 MPa 8.16 HPa 6.30 MPa 15.2 MPa I Pressurizer level 80 cm 125 cm 143 cm 108 cm Cora power 78 kW 78 kW 73 kW 78 xW het core power 40 kW Ju KW 40 ta 40 xA (eeyono that neecec for environmental neat loss) Radial power peaxing 1.32 1.33 1.33 1.33 Cold leg fluid temperature Intact loop 575 K 520 K 540 K 547 K Broken loop. 5 51 .< 502 X 500 X 517 % Time of initiation 1300 s 3,750 s 8700 s 14,940 s ECC injection HPIS type nign neac low heaa low nead low neaa Temperature amoient amoient amoient amoien t
- Leakage a 0.00G kg/s 0.007 kg/s ' O.007' ag/s 0.007 xg/s Environmental neat loss 72 kW 72 kW 72 kW 72 ti
- a. Total primary coolant system inventory was approximately 150 kg at typical initial conditions.
I e 21 m
-- .. - ,_m .,,--%. , - -
~~
- 2. . c- ...m. . . .. . . . .
scalec HPIS flow capacity and 40 kW core power is snown in Figure e. Figure 8 indicates tnat steady-state feed and cleea is tneoretically possible between 7.6 and 15.4 MPa assuming 100% oualf ty steam flow tnrougn tne PORV. Tne specific objective of 5-SR-1 was to determine tne feasiDility of operating witnin the predicted operating band and to examine the tnermal-hydraulic effects accompanying rapid depressurizations to Icwer operating pressures wnen significant primary mass cepletion had occurrea. As will be shown Delow, operational proDiems with uncontrolled coolant leakage from the system prectuaed tne use of results from Test S-SR-1 for airect interpretation as to the viaDility of feed anc bleea cooling. Ratner, tne test furnisheo cata tnat is useful for identifying ano examining phenomena that influence feed and cleea operation. 3.3.1.2 Test 5-SR-1 Results. The transient was initiated at I 1300 seconos into the test. (Tne first 1300 , seconds were used to cetermine s primary leauage and environmental neat losses.) Table 3 inaicates the sequence of major events for unis test. Tne initial conoitions of important parameters prior to feed and bleed initiation are given in Table 2. The first part of the transient simulatea a'"nanos off" situation wnere tne system was allowea to establish conditions at the safety relie.f
. pressure in response to a loss of feed water and loss of offsite power, Figure 9 shows the response.cf the pressurizer pressure and collapsed liquia level. At the initiation of the transient (1300 seconds) the l pressurizer licuid level is seen to rise. Thisfresulted from terminating pressurizer heater power with the resultant cooling of the pressurizer a j causing collapse of tne vapor space. The vap'ar buDDie in the pressurizer continues to collapse until tne pressurizer is full cf licuta.
- a. Tne pressurizer anvironmental heat loss was determinea to be approximately 4.5 kW at 535 !<.
l 22 4
/ ~
h, ' ,
i1 ' i 1 l
,e '
12 0 i r i i i i , 30 l . < ,, p 10 0 - fy l -- 25 -h
- ?
ec ff% . NP IS 80
%e 20 N a
e - - M
$ 3 a v v . >- PORY A [991
- i m 60 -
IJass ow
\- 15 +-
0 . m s_
%^a g , L C '
Ld s' ) !
---- *E *-- *E * '-- 'I---
- 40 -- - - - --
10
/
9V, M 4 M U 20 - - 5 m t.. d" i 0 ' ' ' ' ' ' ' '- 0 j 0 2 4 6 8 10 12 14 16 ; Sys t em pressure (MPa) , I i 9m e !!. Semiscale Hod-2A primary leest ental fileeni oper.nting m.yi l'ir lest
! S-Sit-l (iiigli licoil til*l';).
I I' l
- i l t
- I n' , . s >
.1 . \
TA8LE 3. SEQUEhCE OF EVENTh FOR TEST S- SR-1 Time (seconds)
- 0-700 Determine environmental heat loss and leanage Pressurizer isolated from primary
. Steam generator secondaries empty 700-1300 Pressurizer and primary equilibrated anc brought to initial conditions 1300 Transient initiated Pressurizer neaters off Core power increased by 40 kW Pressure = 17 HPa Leakage makeup ter.ninatec 3450 Primary system pressure reacnes 15.17 MPa--setpoint of PCRV 3700 HPIS enabled 6050 Depressurization to 12.41 MPa 6300 Depressurization to 11.03 MPa 6550 Termination of test i
l i l l i e 24 4
l l l CH33 "!3A17 E3IIE0353Ed l
- e. e. e. e. e. e. e e e 3 e. e. e.
m
.v 2 2 .e e e . - - . - I as & w -
i o - a j g .gi;g . i g i j i l4 3 - b
" ca 3 8- - / ~n - e e ?r - .s * -.3 , - aw . - u .,. -. . .e - e. . - e ,a.. - e - - .. _. 6 g . 4 e - . e a u .e
- n e -e .c. 1
, -f - e o 7 r e '
e l)lb i
~%
a= b
- I, e 2 s, * -3 2 - s .z e a.
e . > - g e -' e a u
- ,t e g 2 ., -. . - - e u .E a d , =. = , e .e-i m
C . H 22
, u -
e
= $ e ~ ~ e - . .e n x e
l 3 .
- * . m. %.,- e . m - . . , u .
e
, c. .. u. ~ - e u - ~,~ - e ~ - . ~
m auted 2 sat Jo uotsetstuI s s,- y a; f o
~ -
g u
~.
s
/
e o - [ [ l " " b
= - 1 - - M = -.
t 0
"'f i t f f f f f f t t t t t t f e t ttttttete it? _sttt!aeestrestfettttestefsteettte g
e m , n m cn
.e e
(UdH] 3 J ft S 5 3 J d 25
, e e - . - - - , . . . - - . - , , - . , , - - , , , - - . y- . . , , . ..,_
7 *
- a l .
Figure 9 also snows tne system pressure initially dropping at tne onset of the test. This is also a result of the vapor ouocle in the pressurizer collapsing due to the termination of power to tne pressurizer . heaters. As Figure.10 snows, the system pressure begins to rise wnen the
; upper need temperature reacnes saturation. Tne system pressure continues to rise to the ' set pressure of the PORY (15.17 MPa). At this time - (3450 seconds).the.nleed portion of.tne operation begins. Several minutes later (at 3700 seconds) HPIS was enab*ed wnich began the feed portion of the operation. /
Figure 11 shows the PORV and totsl HPIS mass flow rates. The PORV flow curve is not smooth due to the cycling of tne PORV to maintain tne system pressure at 15.17 MPa. Figure 12 snows tne integrated PORV, HPIS and estimated leakage flow and the cet effect on mass inventory. Beginning at 1300 seconos tne mass inventory is reaucing cue to recistribution of mass into the pressurizer" (vapor ouoble collapsing) and from leanage. At 3450 seconos tne rate of change of inventcry loss Can De seen to increase due to latcning open tne PORV. At 3700 seconds the HPIS was enanled and the mass inventory hela relatively constant, until tne PORV mass flow increased. The PORV mass flaw increaseo due to cnanging fluid conditions in the pressurizar. This conoition continued until tne collapsec 11ouid level in tne core was just above the neated lengtn. At tnat time (6050 seconds) the PORV setpoint was readjusted and the system pressure was reduced to 12.7 MPa in an attempt to recover primary mass inventory by obtaining larger HPIS flows. Tne analysis of Test S-SR-2 described in ene following section will snow a distinct relationship between tne PORV flow rate and the hot leg density (near the surge line entrance) for a situation in which tne PORV is latched open. In Test S-SR-1 the PORV was cycleo (witn a 70 kPa nysteresis between tne opening and closing pressures) to maintain selecteo pressures. Figure 13 shows the measureo PORV flow, predicted 100%-ouality steam flow
- a. For the mass calculation shown, coolant wnien entereo tne pressurizer was subtracted from tne " primarydsystem mass.
26 i S
i i .
* .l .g i
i . g 1 l l ! .) l 700 yi;i;egil ligi;i 16 gigi;i; ;il gig:l ;i;ig ,. 15 SY5 tem Pressure ; 675 _-- 8 14
<t . '.e. ~
o 650 -- - 13 - Ei.
-"1
- i.
se 12 3 ion
- 625 - ** -
" 7- llot leg fluid "
a.: N N - Upper head saturation temperature, II E ,'
'~' ~
600 --- temperature 1~ - ^ ,>.<. [-- # + #,,# ,c o #~10 - x f . ,- 7, cold leg fluid - 0 l in temperature g m
.. ,--,,..- (
575 - . . r - . l
'_ f f-..Hpperheadfluidtemperature -
8 d
., v >
550 f I 'I I' T ' I ' I ' I ' I 'I 'I-'l 'I 'I ' I ' I ' I ' I ' I ' l ' I *- 7 O 400 800 1200 1600 2000 2400 2800 3200 3600 4000 Tlas Isl Figure 10. System pressure compared with selected system fluid teinperaturen for lest S-Sit-1. ; 4
. 0
e . o e l
. . e tl i - l l l l e
l a e 9. a - e s e m 4 t . - m 7 m i - W
- m I - - - e u J - *
- C ,i - e u
-s e ,' L 1
3 d. - 5 3
- E "N C,- [ e e g 5 E - E -
e
- m -
_a
- b. 3u e *
- n. e '.E
~ .a ,, -4 w - ? I * ~4 4- 5 =
2 e > e
= h 2 2 - : x . 1 -. -d . * - . e
- 1- e v
, d o ===. * .=.=4 3
- n. =,
g -
- e '- - I e - , - . m - m n m..
I , l , l , I . I i ! , I i ,n $ o m e e e e e e e e e e e e e o e e e e v e ao e v eu o
- n.a
. cu 1 . . . e. e. o. e. e. e. .-
e e e e o e e o e gs,!M: .s e [ J s 9 e'A 8 28 1 l l l l
. . _ . . . . . . - . . . . . . . _ . _ . . . . _ . - _ . ~ _ . . . . _. ,
SM SSYN H21SAS eee ~
- e. a. =v=$ .=.eeeeeee$m=e
. . $. .e e. = a os r- e m e ee - - - - . g . . - .i.i.i.i.i.i.e.g.i.i.i.i.i.i. ,
g 4 = S b l
~ .
2 f g g 6 N M
" % wt %g w " e.S W
g g e g @ *'*
% y o - @
S u m o
)% % 9 g
4
.ns W % . s 4 . # \ ~
W e 3
- . I m 3 . ?
a i% P . g
- e 9 1 % =
t I s, = E
\' y 2 $>
2 %
= ] we g '. >= w > a % 1 h a
5 a v g l. 6 e 6 n m g e 5 N N % g O 6
- t %
y .
$ %I N - e e u .O, -
- E 9 @
E I m C
" 6 6 =
5 .$ J 5 h a " x e o () m a e N "
- c -
(
.1fttttttt!,ttttttt*!ttittttt!lteteettptls,etet'88lt'eit** ;4tt - O N 4 in v (7 ru =
tSM) ssch 29 l l e
- wy----
b .
.- t i
i s t
- e.se g
; . g i g , ; ,
g i seo e.es y_ ,, ,
" s~.> -
see e.es i g- . ,, .. i-,i ,, 5, , s,, 1,',
, r, , o,#4 ,,g ,
llot leg density is','n
- f t,8,S'st sg
V # # ,# "sig 8-g E ,
e.07 g, p8 s i, g, - g j . 4ee x II e.es 2- Heasured PORV flow >- i / '.~ x 5 Calculat'ed PORV / **
.., .3 e.es f steam flow l$
300 [a i o si. 2 A
- l e.04 5- PORV steam flow a l
- i. . .
_ _ _' 6!9'J.ted_by sycle,J.M , _ _ , , , ._ j as X e.e3 i- e~ l f < E
! 6.ea 5- I 4
0.es iI-5
-v% _.
1
- d. .
f: l 9 t; - see e.ee i ' I ' I ' -A LAL - I I i' ;
\
e 348e asse assa 4ees 4ase 44ee 4see 4see sees saae s40e i Tlas tsi . Figure 13. PORV flow rate compared to predicted values and hot leg - density. Test S-SR-2. . 3. ;t
., +
rate, and the steam flow rate weignteo by tne valve cycle time required to , remove the core neat, compared to the hot leg density. It is seen here
, tnat even with a low void fraction in the hot leg the flow out the PORV '
reflected a high quality oischarge. Referring to the pressurizer liquid level of Figure 9, it appears tnat a vapor space existed in tne pressurizer that would allow this.
~
At- 6050 s tne PORV setpoint was lowereo from 15.7 to 12.41 MPa, ano further lowereo to 11.03 MPa at 6300 s (Figure 14). Immeciately following tnese enanges the PORV was effectively latened open until tne new pressure was reacned. Referring to Figures 9 and 15 at these times it is observed tnat the mass flow rate increased even though tnere was little cnange in pressurizer level. Also, referring to the hot leg density curve of Figure 15, the PORV flow rate was strongly dependent on the not leg density, as is apparent by the difference in flow rates from 6100 to 6400 s. It is surmiseo from this benavior that closing tne PORV for periods of time in the cycling process allowed a phase separation mecnanism to occur in the pressurizer whicn maintained a steam flow disenarge. At 12.7 Pfa the mass balance was still unfavorable so the pressure was reduced to 11.1 MPa in anotner attempt to recover primary mass inventory by increasing the HPIS injection rate. Altnough HPIS injection was large enough to begin to recover mass inventory, core uncovery was too extensive to prevent excessive roo temperatures. Tne test was terminated at 6550 seconds. 3.3.2 Test 5-SR-2 3.3.2.1 S-SR-2 Predicted Rescanse ano 00jectives. Test S-SR-2 was performeo using a ". low head" pumped HPIS injection capacity. The injection rate was powered-scaled based on HPIS flow information from the Zion plant.7 Only the combined injection capacity of tne HPIS pumps, which l oeadhead at 10.34 MPa, was considered. No contribution was assumeo from . l tne charging pumps, which are capable of injecting up to the safety relief pressure. The feed and bleed operating map for Semiscale based on this
, 31 e -- , .yy . -_ ..-, --.w. -
s
.a e e , og 9
o S e C ' um i i . e i i . tu
, * = =
- m. - T e
= = c - 3 E - s3. -
E E Y e (
" ~
- m a
e.mm. W L
.".' W = *J = >== = m .
- E we . 9 M b t
@ =s = w m .-em e =. > ' ~ W W g . Ett t G,
M w
.- s == 0 E f = = ,
M
-s =
W w g
?- '. 'a M w e E ' 2 a z . ~ ~J - ' >== c r3 = = ,
C
, W ""=9 .O u 0 "E 6
- N wt
@ 2 . u s
1,; e 3 M W u 3 we 6 ws @ 3. E $ I u t @ 7 $ E. C. c !! 4J a t
' I =
O O-CJ
]" eMe w
W h
== Q j' - J. .. = " > e g ., =' @
3ll C " -e
? @ =
kI s L ~' M m y u
==
- 3
==o " am ~ ^~ '" t ? ?
o 9 ' t O e G M N l . gefM} sansssJd 1 32 l
1 ! ( CW/0041 AilSN30 037 10H -
\
3 m 3 3 8 3
,- n tu - .
g... ,,g, h . e - gC t 5 ( o -
+
- g g 4
i
@ z ,r G% % "
u y
*1 ~ . . ' ' . .h8% $
g yc. (; & fg - s
, p<, w % 8 ~ ~
r
$eC " =
T I e 5 h # 43 u 1 s 5
- ' 9, s . 4 ' , 4s : : . , i i g -
g d i e g a s\ - rE s S e o
.. - \ - .5 8 C g m ## s 1 1.t3 < s -s ~~
s (, e >; 8 a O 3 3 T co
= %) . . s m 0 'C L , ,
y D e- E o $
~ "s =s e o r, , , , , , m , a G
O CO U3 Y UI v W
= === w fu . Oe CO Cu o l tu. e. o.
e e e
. . . . o. o. o.
o e o e o e e e . l
$ 8s 3MI Acid SSPM l
I i 33 l l l
,. ., . _ _ _ . , . ,w. . _. ,.- , _ , , , _ - . . . _ _ _ , . _ . ,.,.__,_,_,._,,,_,.y,,,,,,,.,m.,,.w,. ,.-.._,.,,wy,_,_..
s' 1 scaledHPIhflowcapacityand40kWcorepowerissnowninFigure16. Figure 16 inoicates that steady-state feea ana bleec is eneoretically possible between 7.0 ana 8.2 MPa assuming 100A Quality steam flow tnrougn ,. the PORV. The specific oojectives of S-SR-2 were to determine tne feasicility of operating within the predicted operating band using a representative low head pumped HPIS anc to cetermine the feasibility of initiating a
~
steady-state feed and bleed operation by depressurizing from a representative operating pressure. L 3.3.2.2 Test S-SR-2 Results. A significant operational change for Test 5-SR-2 was that the primary leakage was made up with cold water injection. This resultea in minimizing the effects of leaxing mass from the primary. Test S-SR-2 consisted of tnree separate feed and bleed operations. The first attempt was to try ana find the upper steady-state operating - limit for favoracle energy removal ano coolant inventory. Tne secono operating point was below tne preaicted operating oand for PORV steam l disCnarge as snown Dy Figure 16. The tnied operation was a , cepressurization from a representative operating pressure into the precicted operating band. This operation was performed by enabling the [ HPIS ana latching open tne PORV. This maneuver is representative of feed , ano bleed emergency procedures specified for PWRs. Taole 4 inoicates the ( sequence of major events and tne times at which tne three feed and bleed ! operations were conducted. I L The first test point consistec of operating at several pressures in an f attempt to cotain a constant mass inventory. Pressure control was [ accc:nplisnea oy setting tne PORV to open and close at selecteo pressures. l The HPIS was allowed to inject at a rate governea by the system pressure. Figure 17 indicates the system pressures and pressurizer licuid level during the first test point. As incicated, liquid level in the pressurizer [ L l 34 l P i L _
E' 7
. i I
L a e i 30 h ^ s ; 10 0 - 25 52 , j, x ;
^ 80 -
20 N 3: tn - d5 d5 on 60 -
's\ -
VA rpga as ow 15 + m ' ..
- o l-sa - s, .. ,i 40 ' ) \ \
10 c 20 -
/ p @ .'ef -
5 '
)
I I t L i I i 0 2 4 6 8 10 12 14 16 1 System pressure (MPa) i
. Figure 16. Predicted primary feed and bleed operating map for Test S-SR-2.
h
; i , s
. . . . .... ~ .. TA8i.E 4. ' SEQUENCE OF EVENTS FOR TEST S-SR-2 0-2750 seconds Determined environmental neat loss from '
pressurizer and primary. Determineo primary leak rate. Steam generator secondaries empty. 2750-4190 seconos First test point
-2750-3330 PORV open at 8.70. PORV closec at 8.13 3330-3750 - - - - PORV open at 8.01. -PORV closed at 7.94 3750-4000 PORV open at 7.93. PORV closeo at 7.86 4000-4190 PCRV open at 8.19. PORV closed at 7.34 4190-8700 seconos Reinitialize system for secono test point 8700-9390 seconos Secono test point 8700-8930 PORV open at 6.38. PORV closed at o.03 8930-9390 PORY open at 6.55. , PORV closed at 6.21 ~
9390-15,000 seconos Reinitialize system for decressurization test point 15,000-17,500 seconos Oepressurization from representative operating pressure PORV lattned open j HPIS enabled l 36 ? r
{ l[ J's .J 1
- i. ' a . . ..[; ' . - .' ', * -
4h 1 t 0 9 , n 0 9 , i o
, p 9 8 7 0 .
0 2
, - , t e - - - . , 24 R ,. S - j S - s 0 t - ' 0 s
_ 0 e _ T
'\ e %.3 4 *p e r r fi
_ r u o u s f s s 8 s e 0 e _d e r r 8 r s p p u s
- e . s, . n s t. e l
ep o r p e l 0 v - o c . 0 m e - V R V R
- s. .
6 3 e l d
- O O .
t s P P l y i
- a. e s u
q T- i_
" . 0 t i
l
' s t
0 4 3 e m d n a r ,1- i
- l e .
L T e z - v _ i r -. L
. 0 0 l e
u ; 2 s r s ; e . 3 e e r z P r ; u s L.. i r 3
- s e
0 0 u s
; r i. 0 s p e ~ ; .. 3 r m P 3 ; t e t.. ; s y .. 0 .
j . . 0 7 S J. 0 2 1 1',g, . e j r
. J. u g - - ~ - 0 -
- _ ~~ ~ - . - :.
0 i F 6 0 0 0 0 2 0 0 0 0 , 6 4 2 0 8 6 4 1 1 1 1 I
.n. .
- j E a .. J.
e 5 i<3 !l, ,: l;' I I
remained full and did not change substantially during the test. Figure 18 shows PORV and HPIS flow anc mass inventory. As Figure 18 indicates, initially PORV flow was approximately equal to HPIS flow resulting in no . mass inventory reduction'. During this time the PORV flow was approximately equal to calculated steam flows througn tne PORV as shown in Figure 19. The upstream conditions of the PORV then changed to a lower quality fluia as evidenced by increasec mass flow tnrough the PORV. Tne result of tnis increased PORV mass flow was to decrease tne mass inventory in the system. Once this decrease was toentified, the system pressure was reduced at 3325 seconds to 8.0 MPa by resetting the PORV. This pressure reauction is snown in Figure 17. The reauction in system pressure naa very little effect on mass inventory loss as seen in Figure 18. The system pressure was then further reduceo to 7.58 MPa at 3750 seconds. Mass inventory was still continually lost. The PORV set point was then increaseo to 8.10 MPa, but as evidenced by Figure 17, the system pressure began to rise beyond this set point. As discussed in Section 2, tnis is an inoication that insufficient energy is being removea from tne system. Figure 17 indicates that the liquid level in tne pressurizer.did not cnange suostantially during this time and Figure 19 inoicates the PORV mass flow rate also remained unchanged. Figure 20 shows the licuid levels in tne steam generator tubes were falling. It was surmised that this reduced the primary heat loss 1.o the steam generator resulting in the pressure increase. The secono test point das an attempt to establisn a steady feea anc bleed operation at pressures lower tnan those attemptea in test point 1. Figure 21 snows the system pressure and liquid level in tne pressurizer during the test point. As indicated, the collapsed licuid level in the l pressurizer remainea full during tne test. Figure 22 shows 90xV and HPIS flows and corresponding mass inventory. The PORV flows are much nigner than HPIS flows wnich results in the continuous cepletion of coolant inventory. As indicated in Figure 21 several PORV pressures were attemptea witn continuea loss of mass inventory. The last pressure range attempted ! was 6.21 to 6.55 MPa. Figure 21 indicates the pressure never cropped below l 6.4 MPa at wnich time tne pressure began to rise sligntly. This is an 38 G
- a. . . -
. 4 t1N3383d1 AkB1N3ANI SSYN 3 S e . e - e n e 3 . 3 .. __ - ......,1,,s., ,I,s.......I,,,,,,sussIs ssess, - a,s e u g C - e -
i - e x
- -3 -
8 e g
- _ e > - - e c g -
n -
,s - m ,- m -
e 2
, E -
- w
= g ' , 't - L, - - 3e t ,.
e - - a i t - 3 y
.o..
n -
= 2. .
e r g c ( - e g. _=
-o e s . . 3 m m w -
M
- . - - g N. - x > =
am
- 8 e %A 3
n m.
== - O Q, - c. - .. g .
g 2
. t m 2, - u i , I , l ,I ,! ,~ C I , I . I I, 3 =
m e w m e e e e m o
. . . . e. e. e
- e. e. m.
e e e e e . e e s .- (s/5M no ld sse'; f 39 m
= -a - _ - -- v- w- v v,r-,- - - ---,w y y,ww yea y..w--w, yv -
we.gp----w- 4
W3 73A37 Q10CI7 H3 IHASS3Hd
? 9 9 9 . .
I 2 5-3
. 8.
A o.iiiii p - n .. . ....... q ,.......,,,,,,,,,,,, g nog i e
- s I _ T=
s , c y.- _.---:_----_ ______l
~~ --
i I e Ex
-- m.,V ,__ - - .1.--
y
==- -- s . , l _
su
%=. M 4 C ~~ ~ \
l g - - v~ i, I l a=s. m M _s 2 "3 s ,- r 3, -
. 35 s l v. - ,/ , ;; .
uu ., s8 3 V q y t, l ,=6x n .- 2,s a s ~ p o-w
> g . l -u "o - - s _- u su , m - s( n -. =~
t ,
, 3 ~ \ , y . am u < . 8 - ,a . u b u N =--. g ) vs n -b=
_ u g ,
= e. =,a -
9* r1 EQ % - u Em _ .a - u - _
" i
t s g s s -
- m I. e -
IM f
~ - I , .I , I . I'. i, g m !7 $
7 7 9 9 I 2 --
. 4 4 4 g (5/3M) A c i ,4 sst3 ,
40
it. , l ; , < [ - . l '
,~ ,' ' ,.: .
T , ' * ' t - * ' . ;. a * . e t i n . e l e r h s t e n e e b c t u g f l e o d p l o o f c 0 11 - 6 ; l. ;, 1 1
- 2 t 1
4 s 4 e y j 4 g g g 5
^
6 1 4 T r
. s g-g 0 o p ) I 6 f s 6 s . - 4 e - 6 n ' 6 di~
9 t
% - 0 3 r 1 8 t o
8 a 3 r
~ - '
6 0 e n 7 e
, a 3 g m 8 m
- i
- d s
6 6 m a i 3 e l i : e 6 t t e d s d 1" 6) y/ ei
~
rds ei rds 6s n _ oi - 3( i oi t sw 8 _ t sw a o a o l 6 s rwl - 4 e l T rwl eof eof nl 3 0i m e v nl ef n e ef n g 0T po w li 1 l m _ g w 3 i po mud "- 3 d . l i mud a 6 y . a ess ess t ee b"a 8 i1 u qt m -
- t ee sbb 2 in m - sbb uu l
a 3 l i i uu ptt t 0 o g .- o ptt o oyy l s e 0 dp e . oyy l rr n n 1 3 e s . 9._ f
.* _ ^
l rr irr nmm eii OPP aa t mm cii arr tPP n aa l i e m u i u 8 6 6 3 0 lS p2 a - l R o-s r I h 0 C5 u i m i-l m 9 2 s . T - 8 y - m - l b0 m 8 0 2 e i 2 u q _ bm 0 8 r u s_ g a _ 7 i m i p - m 0 2 'l l i 0 6 2 0 0 6 0 0 0 0 e 6 0 0 0 0 6 0 4 2 6 8 6 4 2 1 1 1
. 33 * .C s. 'l. :s ar 2 ~
u- -
< i , - < .ll - l , ,l
.. .. . . .-.s..._ .. ,. . . . . . . ...--.-
t YeH : 3W.*.SS2Wd XIISAS
- e. e. e.
> e e m u . . .. gi g . . . . i i . ,e - ~,. 7 a e.
L s
- e N -
s s om
-u I
a a usw a 5
- >W W e \s ge -
e x e sa -
- N E .s
- > w u - a 3=
- \ ,s,, - 8 w g 1.. -
- a. a e 4 e
, /- u a
o e., e
/ 8 - #
e e - - c.w -e a s e :
- 5
'N * -
3 m 8
~ o No -
m .o 4 m
~
Ns ~ _ $ &-oc. - s e o- .- e a c.
- a u . . m -
e 6 U N. W N2
. *M w "
g b : t i -- e =*
- - = .s ==
o o
== W eW Q - A >=
I e . e -
- (s -
N
=
N N s h
- = .........!.......,, , , ,s , . . . .- 3 t w e e a ,
l @ @ L W v m
** we "" .ru. .
s - l e (III l I **. Il h l O I I ", O I b
** s m
H i U l w r l 1 l 42 '
(1N33M3dl AB01N3ANI SSYN S e e
- R e - 3 e
e
~ - - . . . . - . . . jrvinissi6agianisi6:4 .46ii....
p u g, -
/f , 1 -
e o - is m 3
- /
i
- e g - i o e - / / 3 w -
o e 5 c I f 3 .m
- a m ~ / % -
i 3 a l m
- 8. , D
- e a y
t - - Tc. i -
- t - e a e -l e = m - a , , t - ~ =.
o - wN
/ u a e ^ 2 o e e -*
g a -
- i - a se i
o - a C
- i g -
e =~
-, x e - >x J - x xo m, - a p - LW a _
e e 4
~ / 5 g i - e u =
T -
=
F, l . I , l .l . I , !. l . . w m e a v cu e m e v cu o eu
= . . =. =. e. o. e. e. e. e.
e e e e e e e o e e . i (5/8M1 acid sseg 43 c
;~ ~ [ :- . .. , . . . _ - _ ~ - -
Indication tnat tne energy being removed from tne primary system is being reaucto. Since PORV mass flow ano pressurizer 11 auto level remain constant auring the same period it is assumed that tne energy removal rate of tne PORV is also unenangec. Anotner possiole energy patn is to the environment. Figure 23 snows tnat licuid level in the intact loop steam generator is falling at the same time tne pressure increases even witn an open PORV valve. Tne falling liauld level in the steam generator tubes may have caused reduceo neat transfer to tne empty seconaaries causing tne system pressure to rise. Figure 24 snows actual and calculated PORV flows (assuming 100% auality steam) and hot leg coolant density in the intact loop. The reduction in hot leg coolant density indicates minor voiding occurring in the hot leg due to loss of mass in the primary. The minor voiding had little effect on PORV mass flow rate as snown in Figure 22. Tne cojective of the last test point was to cetermine the feasibility of obtaining a steady state feec and oleea operation (using a icw neaa pumped HPIS) initiated from a representative operating pressure. Tne proceaure for tnis test point censisted of cepressurizing the primary by latcning open the PORV. Imeciately prior to tne cepressurization tne pressurizer neaters were turned off and the HPIS enaolea. Figure 25 inaicates the pressurizer pressure ana collapsea liquia level response during tnis test point. As can be seen, the pressurizer level initially dropped and then recovered to a level inoicating a licuid full pressurizer. Tne initial level reduction is a result of flasning of the pressurizer liquid. Initially the pressurizer is the nottest volume in the primary aro this fluid flasnes first wnen tne PORV is opened. As tne steam bubble is vented througn the PORV the pressurizer inventory cepletes, ano cooler not leg liould flows into the pressurizer filling tne pressurizer. After the initial drop and rise the indicated level remainea l constant at a near full value. The system oressure dropped rapidly from - the initial pressure of (15.17 MPa; to 8.0 MPa wnich was approximately the saturation pressure of the not leg fluid. Tne pressure then slowly oecreasea to 6.8 MPa. 44 e
i . l , l l . '. S l l l 1400 7.80 iiiigsisigvisigsi,igisiigisingssi,geisigivingsisigvisi;iisigivingissigivirgiis . N s.90 1 t,00 c _
~
Intact generator tube 2 6.00 ' level
- roken generator 4 tube level l g 1000 -
-! 6.70 I - PMV opn 4-
- , 6,68 i pressure 7 e 's ,/ s
- i 800 - i
,i 's s a / - ~
6.50 i i
=
a
= 600 -- / / Ne --
6.40
; ,N y - / g J. _ System pressure : / ~i 400 - / \ -j 6.30 e' \ ,v' : # opTof tube sfieet LPORY close ~~
0 T' ! 200 pressure E !: i, - I 6.10 . hulu u l u u lHiilrm inul an li m i n u l u n i n n i n n i n n i n n l u n i n J ! o 6.00 i 8600 8700 8800 8900 9000 9100 9200 9300 9400 .! i ! Time tel t ! Figure 23. Steam generator tube collapsed liquid levels and primary system j pressure. Test S-SR-2, point 2. , l . 1,
. .,, . o c.i ::
ai 1 .; ; aj a a
~,
4 0.16 , a i , , i 1000 Densitometer beam orientation f
- - 900 0.14 -
t 1- Hot leg density I 3-n 0.12 -
! 800 *E ^ " st j- N . 700 2m -
Nm0.10 - i v ' 5 Measured POR f., . . . . . . . . .. f- 600 2x ! 0.08 - fIOW i
- t. !'.....,, .
i - ;*' E ,.. ..~ ........: ; g
- s. O :
- \ts - 500 ,c i-j
* . t e 0.06 - .e t. o .
t 3 l \ ' '1.
\ l-400 m j 0.04 - ! k ! If .
l
.I t \
i I.* 300 ' 0.02 - e l i ' O r ___ _._--J_'._.___._____._._._._.i / 200
\ .
L Predicted PORV 0.00 - - - - - - - - - -.! g saturated steam flow
-0.02 ' ' ' ' ' ' '
0 i 8600 8700 8800 8900 9000 9100 9200 9300 9400 Time (s) . i waanso-a Figure 24. Measured and predicted (l001 quality steam) PORV flow comparison to liot leg density for Test S-SR-2. point 2. i
~
h S'
i . -.
! . .- i 1 .i i
200 --
, , , , ; ; ; ; ; ; ; ; ; 16.0- --~_- ---
gg,,
. Pressurizer level i --
14.0 ' 150 - ' - i _ _/ - 13.0 $
" Pressurize %-_ _ _= i -
run - - i _- - - =
---.+ 12.0 p: .
r . 3 100 - - 11.0 El .
,e . - i System pressu're -
10.0 x .. i 3 -
.i g 9.0 3 50 --
i - m 8.0 . s,.
,s___~' '
_ -- 7.0 ; 1 .t o luduul u u l i m l o uluub tuhtal o n l i n i l i n i l o n l i n i l o n g,e ? 14000 14500 15000 15500 16000 16500 17000 17500 Time ts) { 3 Figure 25. Pressurizer pressure and collapsed liquid level during Test i
! S-SR-2, point 3. .I ,. f g= r , 4
n _._ aa. . ~. _ : . 1-
-~ -
o . . I . . I a Figure 26 shows botn the actual and tne calculatec (assuming 100% auality steam) PORY flow compared to tne hot leg density. An interesting point observec here is that the PORY mass flow rate remains high (indicating low quality fluid upstream of the PORV) until tne not leg [ ; substantially voios. After tne not leg had voided the PORV flow rate ' inoicated that high quality steam was being discharged. Referring oack to Figure 25 it is seen tnat tne pressurizer still remainec nearly full of liquia at tnat time. Flooaing calculations presented in Section 4.2 precict that the steam velocities in tne surge line were nign enougn to prevent countercurrent craining of pressurizer liquia into tne voiceo hot leg." It is apparent from these benaviors tnat tnere is a close coupling between the conditions in the hot leg and the quality of fluia tnac enters the PORV, irrespective of pressurizer inventory. Due to tnis coup 1ing, until there was substantial voiding of tne not leg the PORV mass discnarge rate remained much nigner tnan values typical of 100% quality steam flow. Since the HPIS injection capacity was on tne order of tne PORV steam discnarge rate, tne nigner discnarge resulted in a net oeficit in the inflow / outflow mass calance. Figure 27 compares tne measured PORV flow rate to tne HPIS injection rate and snows the net influence on system mass inventory. The primary system fluio distribution resulting from the mass cepletion was generally characterized by a craining of fluia from the upper elevations. Figures 28 and 29 snow the collapsed licuta levels in ene loop component;: the steam generator tubes and the pump suction piping. Figure 30 snows tne collapsea licuid levels for cifferent regions of tne vessel. The upper heaa and upper plenum regions of the vessel voided rapialy. Drainage of tne steam generator tuoes exhioiteo a ceiay enat was most likely caused by the residual heat transfer ciscusseo earlier. This acted to promote concensation in tne tubes ana kept voius from forming. Once the inventory nas decreasea to aoout 60 to 50% the intact loco steam ( a. Section 4 will acdress particular hydraulic pnenomena tnat bear on tne , l typicality of tne Moc-2A pressurizer benavior. l l ( .18 - i-
f * , 1
. ~ ~
6 0.16 , i : , 1000
- 0.14 - ~
M llot leg density n 0.12 -
+ ^ 800 "E n -
I- Measured PORV flow N 700 m
) 0.10 -
l i Densitometer beam d t 3 ilg .
., orientation 600 x
- 0.08 ,il -'y! $. %
, o !! / m e ::: :: : 1 -
500 c ; 0.06 : #
- / g f. , V .
a a !! t t ht - 400 0 :: ! , ' 2 0.04 :: i ___.g! l t, - 300 ,
- 0.02 -
'C . I.-._.___._.
A !; . 0. S
.(.. . _ .'. .'_ _ _u ._ 4l _. _. _. _ .r_ _.<. ._.. _ 200 j Predicted PORV '
s O.00 ------* 4 saturated steam flow _l - 10 0
-0.02 ' ' ' ' ' ' ' '
O 14800 15100 15400 15700 16000 16300 16600 16900 17200 17500 4 i Time (s) t teswie.s Figure 26. Measured and predicted (1001 quality steam) PORV flow i comparison to hot leg density for test S-SR-2, point 3.
- l
! i 1 i ; J , 1 1 ; - e- j. . . .
_ = _ _ __. . _ _ . _. ..
\
t , k
/ + ~- \
i s s r 3 (IN3383d1 AUDIN3ANI SSYH r s - e e e e e e e e e a- a es r- e m v m 4 aaaajieiigiiiig.i g . i a i gi i g i i i sy
- 4 e r
- w s - 's M L g -
r= o
-
- u s
- I - c I
s - -
, s, .
1, _
., e - f -- o a m
w
- c - 8 - e., a I -
m a
~- - $ - u m - t s . e I
M e g 6
- 5 - . _- m c.m
- b i s . % I
- = " *e sa c;; =
s> q , - ~ m-
> 1 , e o . . cc - ' - 3 c. . o o a r e , - .
s w 6 e a % n, e m a:
,, o , ~ -
a $ e > -w s _ h , ( 2 IA a e zu e s
- om - ao s
8
,,e ,e e / .
e > s I . 4 .- p e ."g L q - z , m oo e%
- e s \ -t ,' .. ~
n ~ s 3 --
,e ..
e 3
\ ,s / - . m $ h 5 . u.
2,0.I . I , I , i , i , iN.:
> a v
2 m e m e e m o m
. . ==.= a. e. . e. e. e. o. e, r e e e o e e -e e o e a i .
(S/8M) Acid sseH s
- ) so i > s'
.i V'
s
>,' Y /;- s , -
i
. . .s. . . . m. . ..
l l - l 9
.. .. --.-=
s [ 4 g g e l p, g 4 i, e I e
- s -
n
- o, c 3 t. - e , s' e- - 2a ! - * - s. - -.
- i *
- aCei g -
- o
,,,, =Ew s.
e-s I i a e en o -
- cw s, -
e g ,] , i i s 2
= - Eae eoo us ! ,e -
a3a -
- m s I . -
O e L- a - e ( - -
$ 3 -Sxx ~
eu,s. ,
~ , -, e. _
5
- c.a o6L -- a --_--- ! \
o w w a U c., e -, I. a s I 7 e e S
, cII '% , J e a 4 ' - - -- ; ., 4 e c
s i L 2 -
- 's s- - /
I
]'2 e g
5
- s =
e m
- s , -
- e. _
- 3 - g - - 1 I &l s - - - t - o a w ._- m re - . ed C CJ ' d 4 4 -
s o
~o ~
o e d
- is o u x- x-s = -
e .y
- s -cuw-a m u 1 ee m @
O u
- a =
1 uea3u3 a, ou-w-
- s. - -
- e - o e c w a. w =u -
w
- j -o e
I l t e f i i l . . , l . o a T e e e o e e o e e e e e o e o T
==
fu m e 03 Q T fu ( *J J ) l 3 A 3 "; 3[.iDi"
- c. . .. . . . ..~...:,.. .
. . .. ~ .. . - . .. ..,c. = = d i =
e_ _ _-
= - = <
e - a e
=
sv = I 1 : , . . . . . _ i _l 7 1 C _, n-
~
1 1 1 1 5 I I I
,al l I ,
l l C C
- I # - e i / N I r " - r I 8 -
H f I
/
[ C c3 *
- L ;; f i f --- n a \ = ,
l
- N
. u \ - -
L / t
= i G I < = , C "54 C = t g <3 - C , -a z
w ( %
)~ 3 ~ # v5 eo =
l " w vt Q. C.
= 1 5 * -
Y f O 8"* C7
" \
g l e [ i.;;
=
C ~ C n >= oz um A p S
- ~b / ,,,,s f
i e _. W q p C >> r / - 8 o r 'l -
- n. M s' -
=
I s e e
- ifI g
C n T "w I -
- ,1 -
t t is) I l lI f l I l I I I C C v C C n C C v C C M CD C2 N C C C C C
==*
cl3 C N C C M C C v a:D CD n 1 I I I l ('A3 ) "'4P.AM CICCI 1 53 G 4
-a . _- 3 ;- _ .- 3 *
. . . . .. i.-
generator tubes naa emptied. as seen in Figure 2s tnere appears to nave Deen sufficient neat loss to tne broken loop generator to cause a slow, large amplituae fill and aump benavior out to 1630u s. Temperature fluctuations from tnis phenomena were ooserveo around tne system. Tne overall liquid level in tne system aroppeo to aoout colo leg
. elevation. Due-to heat losses in the loop piping and lack of a oriving mecnanism to force water from the loops, a stagnant volume of succooled water remained in tne pump suctions. Furtner cepletion of fluid was at the expense of tne saturatec liquia in the vessel. As seen by examining Figure 27, even after tne hot leg had voiced tnere was a small deficit in tne PORV/HPIS mass balance. This resultea in a slow depletion of vessel inventory and eventual cryout of the core. Incipient core dryout for this feed and bleed transient occurred at a system mass inventory of 55;. as opposed to inventories of typically 35% for small colo leg break experiments. The reason for this nigner value is that'a substantial cuantity of water remained in the loop pump suction piping, and also levitated in the pressurizer, wnere it aid not contricute to core cooling.
Comparison of the core cryout benavior versus vessel inventory response founa that ne inventory at whicn incipient dryout occurrea was consistent witn tne two-pnase level swell eenavior reportec for small cola leg break 1.0CA ' s .8 Figure 31 snows tne response of selected fluia temperatures relative to a roc temperature and tne collapsea liquid level in tne core. Note tnat tne temperatures remain near saturation until the core begins to uncover. Tnese fluid temperatures do therefore not provide an accurate reflection of vessel liquia levels other than indicating superneated vapor following dryout of the core. 3.3.3 Conclusions from Semiscale Exceriments Tne inability of tne Moo-2A experiments to attain steaoy-state feea - ana bleed operation once not leg voiding had occurred is subject to 54
; i1 f ,I: j :: !J!!' ,
- i,f1 i 1 ,,l 2 :@f .c: . a72n - it 6 6 0 0 0 0 0 0 0 0 9 8 7 6 5 0
s 0 II g - - _ - 1
- _.i - - - h 5 s gi a 7 e al - _
l u l 1 -i s i p s ua n ~ gl f t i m i 0 t
'rea e
i s ' u p g l g s 0 s i. h s l i 0 e a g t i 7 t n s l l u 1 i s e
; i i p
i e l u v l n s - ul i t i l s t s s Q ul 0 t a p s m 0 n i l a 5 e l u n n 6 s l a - u 1 e r
'i p
i l f l t u l p e l i - l i r l t s e i g t '
' u 0 l s
l g s s , ^ l a 0 l l i l s l 0 ) n i s ~ l u 6 s iw. 3 k i
*8 i t l
t 1 i d s j ' u et l i l l e rn - m i. ' l i m ai po . i s s e s u 0 i np e r - t i o i g e y r u - l u 0 T c , l e i t l 5 l i t e t a .. u 5 2 n r r a r s u l - u e r e 1 eR i g , t - e p ni vS i g a d p m l n le-n , c i ni e i S n a d f i e u e t l l 0 dt r _. e er f l t u is i j , sl vu d d i l t l 6 ue m pe it a i 0 gl t a m e ui u u b e n
, av l e ar mr l l
f n i t 5 1 i l g g _ ll tnpe eu f n n o ei p! n . l t g
. cd em pa g e u rr u , l se r e t ou u
l eu et re l l t u 0 Cd a rg r ep d l p o1 pd pu t l u 6 l i eo o o 5 m n i c1 Rr Ukl p I C l l i lu 4 1 1 3 l e i p n o g A D y luis l r i i u n l ig i 0 i m.~- .:- : -{ -
; n
_ . ~- i 0 'l 6 1 0 0 - 0 0 0 0 0 0 0 0 1 0 5 - 5 0 5 0 5 0 5 0 1 - 1 1 2 2 3 3 4 s . j -
. *. -- . .:*_ u ;,_ a.
e gi n el l r e dt ,.. l n ,u oe cc
- I ! , l- ;l ; 4 { :{:l !1 3
- numerous experimental uncertainties. Tne results will ce examined in terms of system typicality'in tne following section. The conclusions tnat can be cirectly orawn from these experiments are as follows:
- 1. PORV discharge rates dominate primary feed anc bleed capaDility, and are more variaole tnan the ability to select either a given core power or representative HPIS injection capacity.
- 2. PORV flow rates are very dependent on upstream fluio conditions which in turn may be directly affected by the cunditfora in the not leg, particularly in a situation where the PORV is continuously latched open for an extenced period. However,.
cycling of tne PORV to maintain a constant system pressure appearea to promote pnase separation and allow nign cuality steam flow even witn low quality hot ley conoitions.
- 3. Temperature response in the not leg, upper plenum, and upper nead do not appear to be good inoicators for cetermining liquio level in the vessel. As evidenced by test data, tne not leg, upper plenum, and upper head temperatures dia not respond to local liquid levels.
J 56 G
*L, - '
w .
- 4. TYPICALITY OF SEMISCALE RESULTS .
In and of themselves,' the results from the Semiscale experiments do , not inoicate the existence of a proolem regarding primary feed and oleeo.
-The importance of the results from the Semiscale experiments lies in-oemonstrating tne cominance of the PORV discharge rate on primary feeo ano bleea capability and the dependence of the PORV aiscnarge on nut leg conditions and consequently system coolant inventory.
4.1 Exoerimental Uncertainties The inability to maintain system inventory in Semiscale during periods when the PORV flow was in near agreement with the predicted steam flow rate is subject to experimental uncertainties. Uncertainties exist in the actual PORV orifice characteristics, HPIS injection rate and measureme.it thereof, system neat losses, and fluia leakage. Each of the parameters that create the operating maps of Figures a ano 16 are subject to experimental uncertainties. Taking the conditions of Test S-SR-2 (Figure 16) as an example, cue to the narrow steacy-state operating oand,'tne uncertainty in almost any indiviaual parameter can eliminate tne operating band. (Or, likewise, expand tne band.) The calculatec or estimated uncertainty in each parameter is listed in Table 5 ano plottec in Figure 32. It is seen oy examing tne figure that the uncertaintles in eitner the net core power or HPIS injection rate coula have acted to eliminate the steady-s' tate operating band. However, the observec PORV mass discharge relationsnip to not leg conditions during tne transient depressurization, witn a factor of 5 increase aoove tne predicted steam flow rate, lies outside the effects of the uncertainties mentioned aDove. Given that all the other parameters coula nave been accurately imposed, tnen, this pnenomena woula still nave acted to eliminate the steady-state band until such time as tne hot leg nad - voiced. It is therefore concluded that witn the PORV latened open a steady-state mass inventory could not have Deen established until after sufficient primary coolant inventory hao been lost so as to void the hot i 57 ___e -e-v,- -
--+y- - - - - - - , =.-----w--g ,.-,,,,,,------.y,---w,-,- , ~ , - - , - - - - - - , - , - , - - . - , - - -
TABLE 5e SEMISCALE EXPERIMENTAL UNCERTAINTIES Parameter Uncertainty Basis for Uncertainty Net Core Power 225% -Component neat losses from Reference 6, structural neat
, transfer, and interpretation of neat loss cnecks prior to testing.
Average PORV Mass Flow Rate =25% Based on uncertainty in net core power. Predicted PORV Mass Flow Rate 23% Uncertainties in i flow area, oisenarge j coefficient and
- compressioility factors.
! Predicteo PORV Energy Removal Rate 23% Sased on uncertainty l in predicted PORV ! mass flow rate. l l HPIS Injection Rate 2.008 kg/sec Reportec accuracy of turbine meter flow mesurement. t I 58 O
1 . s i .. I f
/ ^
10 0 - 25
^ 80 - / -
l 60 - ass o - 15 i
% ANNN 5-xxwwmmm 2 s c . ~ h.c.M.,\ \ " g ; . ^ w v\ A \ \ \ \ \ \ \ g "a
- = 40 - - -
20 - M Wu'v\\\\\\\\ _ l
' ' i i , ,
4 i
; o 2 4 6 8 10 12 14 to System pressure (MPa) i w m.:
i figure 32. Semiscale Mod-2A primary feed and bleed operating map for 21 cora power with uncertainties. ! 4 4 6 l
..-. . . . . - . - - n. . . :.
leg, given tne imposed HPIS bouncary concitions. Once tne not leg naa votoed the ability, or inability, to maintain coolant inventory is nignly subject to uncertainti.es in tne actual HPIS injection rate anc PGRV discnarge cnaracteristics. Witn the relatively narrow operating bana oefined for Test S-SR-2 small changes in core power, PURV flow, or HPIS injection rate coula influence the ability to obtain a steacy-state-condition. The questions remaining as to the typicality of the Semiscale results aeal witn geometry and scaling and are addressed in tne following sections. 4.2 Surge Line Flocaino As aiscussed in Section 2.2.2, altnough on coula postulate some liquia/ vapor separation mecnanisms that woula be influencec by tne pressurizer tank geometry, the net effect coulo be negated if tne vapor velocity in the surge line resulted in flocaing conoitions. Any water tnus separateo in the pressurizer tank would tnerefore remain in tne tank until eventually aiscnarged out the PORV. Figure 33 snows tne velocity for 1007. cuality steam flow througn tne Semiscale surge line (ID=0.94 cm) as a function of time for tne transient oepressurization of Test 5-SR-2. con' pared to tne calculatea floocing velocity. Also snown is tne scaled flooding velocity for a range of typical PWR surge line sizes.' As seen from these curves the. velocities expected in ene surge line are significantly above the floocing limit. From these calculations, supplemented by the fact tnat such apparent floooing benavior has been ooserved in actual reactor transients 10 it would appear that this pnenomena was act significantly distorted oy any Semiscale geometrical atypicalities. 4.3 Surge Line Orientaticn As oiscussec in Section 2.2.3 tne orientation of the surge line connection to the hot leg may nave an influence on tne PORV aisenarge by :.
- a. The floocing curves were caiculated with .tne correlation from Reference 9 wnicn unifies tne Wallis and Kutatelacze correlations and appears very useful for scaling tne effect of pipe size.
60 e
1 i i j
.= 'J !
I 9 -s s l s s l s s l s s l s s j s s l 8 slssl :lsslssls l sg l 5 PWR upper bound 5 8 .:l
, b.- ____------- -- _ , - -
7 fi- / l l
/ / / -- _
l lsi3,r.ge line velociyesl/ l '
/b -g = . / : /-/ /
semiscale i
/. / . / f ; 6 5- /
s !
/ ,
s \ j a i s ; i / , s' E
'. u : - - / :
s .. g _
/ ,./ ' % ! l l
- o l
*5 a, .". =. / / / / / / / i - i / < i -
s i
,/ ' -_. _ _ . _ _ _ _ _ _ _ _ _ - _ _ _ - _ _ __.. _ q , ;
3 F- PWR lower bound : ,
; 5 :.
2 I-
-5 g j_ Zero penetration velocities _g j,_____-.____m_________-_--,..,,.,,.,,?
O IlIlIIIlIlllIE 14930 15300 15700 16100 16500 16900 17300 Time (s) l Figure 33. Comparison of flooding correlations with surge line steam l , velocities for Semiscale and range of PWR geometries. J
" promoting phase separation at the hot leg. The Semiscale surge line-to-hot les connection geometry is snown in Figures 7 and 34 As snown, the surge line connects to the side of the pipe on the horizontal centerline.
Results from Semiscale natural circulation experiments have shown that minimal loss of primary inventory results in suostantial voiding of the hot leg. As an example from Reference 11, a loss of only aoout 5 to 107, coolant inventory results in a void fraction of 40% in the hot leg. (This result must De tempered oy the fact snat tnere was a seconaary heat sink that incuced natural circulation loop flow.) Due to the horizontal side connection of the existing Mod-2A surge line, substantial voicing of tne not leg is necessary to uncover tne surge line entrance." This degree of hot leg voicing must be preceeced by voiding of tne steam generator tubes and the upper regions of the vessel. Results from the experiments showed that a coolant inventory loss on the order of 30 to 40% was required to uncover the surge line." Since substantial loss of primary inventory was found to De necessary to void tne hot leg to the extent that high quality steam entered tne surge line in the feed and bleed experiments, it appears that surge line orientation can dramatically affect the inventory and timing at which steady-state feea anc bleea becomes feasible. A worst case for mass loss from the system when stratifiec flow exists in the not leg is a two-pnase flow until the surge line uncovers and then continued entrainment of liquid by the vapor flow. If sufficient entrainment potential existec at typical cischarge rates it would reduce the importance of surge line orientation. Figure 35 snows the steam flow velocity (assuming 100a, quality) into the surge line as a function of tne not leg stratified liquid level measured with gamma densitometersc for
- a. A void fraction of 61% in stratified conditions will void tne Semiscale hot leg to the bottom of the surge line.
D. The ratio of surge line diameter to not leg pipe diameter for Semiscale Mod-2A is 0.17 wnereas that for PWR's ranges from approximately 0.3 to 0.1 Surge line uncovery in a PWR would therefore require even more voiding. _
- c. The flow regime for typical conoitions during feed and bleea is stratified flow as cetermined with correlations of Reference 13. Durino portions of the experiment however, the natural circulation flow rates '
inouced by environmental neat losses can snift conoitions close to the intermittent (slug) flow regime. 62
,. [:
i
- 1; I
1 5.40 cm I.D. , {- N
\ x Surge line
s
} "MA .,,,, sis, /.94 cm I.D.-
l To pressurizer
\ 4 <
1 u Y // f) m w s
\ s ,
llot leg piping .
\ 1.79 cm I.D. l i-4 6
I Figure 34. Semiscale Hod-2A hot leg / surge line configuration for Feed and Bleed tests. (not to scale.) i i i
/
i l .i IL Vold Fraction (%) ! 100 90 88 78 60 58 48 36 29 le e i 36 0
/ I I l l l VI I I l. . - ,/
27.5 /
- Side Top ;
25.0
? dntrance Entrance" l _ i, 22.5 h -
_~ 3 E; 8
- i!-
- 20.0 h u i !! = ~
- ro i :
17.5
" E!
- 8 o : .
o. s. D 15.0 h. u g R- -[ a : -; i' 12.5 . y gg,g --
..~ / 5eE1"sc'aIe~~ -
7.5 :- /
' surge line ' --
5.0 5.... /
' ' ~ ~
l1ocation s \ ; 7 Saturated stean ~~"~~~~~"~T~~~~~~~~'~~~~ :
- velocity in surge i e :
2.5 -line A'N + N -
. s .- ,
0.0 3 111- 11 lLLLA l LLLA l1 ' AJ Lidi'LA 1 1 LI ' l ' ' ' ' l ' ' ' ' ! '- 2, 0.000 1.966 2.069 _ top of pipe (2-1/8" I.D.) 3.ees 4.eet 5.000 Liquid level from bottom of hot leg (cm) Figure 35. Liquid entrainment predictions for side and top entry surge ' lines.
- Correlation band denotes pressure range of transient mg -
..l .h I
Test 5-SR-2, point 3. Also snown are the velocity Doundaries for
,j -entrainment for both a side and top exit (Reference 12). (Ranges indicated )
are for pressures occurring during the test.) It is evident tnat while - there is.some chance for entrainment with both geometries tnere is a very Droao difference. between tne-not leg liquid level (voto fraction) at wnien - phase' separation is plausible for a top versus sioe entrance. Also,
.results presented in the next section show that separation apparently occurred during a feed and bleen operation in the LOFT facility wnich has a top entrance surge line. ,?
4.4 Supporting Analysis, Applicable LOFT Oata _ . . _ . _ _ _ _ . To examine tne effects of scale un these various phenomena and to ontain independent data to check Semiscale results, applicaole data generated in the Loss-of-Fluto Test (LOFT) pressurized water reactor were also examined. In April 1981, a Loss-of-Feeowater Accident (LOFW)l4,15,16 simulation as well as simulations of two LOFW recovery procedures were conoucted in LOFT, a 50 HW(t) integral nuclear experimental facility. The LOFT configuration for these simulations' is shown in Figure 36 and detailed descriptions of LOFT ano its scaling basis are available in References 17 and 18. Of particular interest to tnis report witn respect to scaling are tne following:
- a. The LOFT pressurizer is of cylindrical geometry witn internal dimensions 0.85 m diameter and 2.02 m neignt. Tne L/D of 2.38 compares with 7.'13 for thi Zion pressurizer. Thus, the LOFT pressurizer nas a larger cross-sectional area (relative to height) than Zion. This will affect, for example, the velocity of. steam in tne pressurizer and, consequently, the amount of carryover into the PORV line.
- a. Test L9-1/L3-3.
l 65 L t I
4 ,s,=,. y4. Wirr r= - go.-*-#--.m P" 8 "N* 6*8 * *""***' ' D M-E M' &' 4 -W
%p4 s 4 men ++.i e u seM4W Md-we eh-1 *@ .O ~~
p'.
- O F p e
I y%m -.. , -g
=* 1 ~ - . ~ . . / ~l 1 i g -_-= " 9 g i \l/ = *' 3 5 u -
5 Q W d s 2
$ E .5 %.L 2 3g $$ r $ $_4 5 - -. s s, - u s ! a i : =w a F2 3 >
Es 44 2 j 2
- T ~
ji ES
/' ! \ "'
FM/
= ? .
s. m nin! s 3ea a
\N \ , ~
y q( l[y. 1-N
'\ \ ?
si an- s = 5$ t.p
\
U5 34 ~k$h s q
\ N s, e a w, 9= , $sab! \ 's " ,-.22 . .3 .; Y.'\ \ 'J 2 i' 5,{ u.a s r?. n'\u2 $,
o ,
- i 3 3=-.. 3{ <
n.33 v ~v,r u, , a -h-y.
=
i :g- = w NI! . w (T , Y.h a i
< A :
ibl N) 5;1 )M1.;N jlk
}ll a
3=, -
\
S ,,S i'
=
i i ss-
,> -! e 9 [-: 2a 2 -a S *[
( 512* / _, b 9
= .
gi j/', i k_j2
=-
m ?= b $ Y9 2 66 6
l
,Y ' .. 1
- b. The LOFT pressurizer surge line is a 2 in. Scneaule 160 pipe wr.ich leaos vertically upward from the norizontal hat leg which is a 14 in. Scnedule 160 pipe. . .
- c. The power operated relief valve (PORV) installed for the LOFW simulation nas a relief capacity of 0.66 k 3/s saturateo steam at the relief setpoint of 16.2 MPa. This corresponds to a Westingnouse plant type with minimum PORV relief capacity (1.32 kg/s *NW( t) . ,
The LuFW simulation was initiated witn tne reactor at full power (50 MW)-by stopping all seconaary feedwater flow. Reactor scram was celayed to 65 s to maximize tne depletion of secondary water inventory. The plant was allowea to react to tne LOFW for 3270 s (54.5 min) with no , operator intervention. During this time, tne primary system neat imoalance caused tne coolant to swell into tne pressurizer and collapse tne steam bubble. The primary system was liquid solla oy approximately 1250 s (20.8 min) and remainea solid throughout this pnase with tne PORV automatically cycling tcr control pressure. At 3270 s-(54.5 min), tne PORV was manually latened open to initiate tne first LOFW recovery simulation, a primary feed and bleed operation with severely degraded primary to secondary neat transfer. The decay heat level was approximately 0.53 MW or 1.17. of initial power. The PORV remained open for 1580 s (26.3 min), oropping the primary system pressure rapidly to saturation and tnen continuing to c'epressurize tna system. In order to cotain maximum primary system voiding for the second LCFW recovery simulation, no primary feea was initiated. However, based on tne measureo PORV mass flow and Known nign pressure injection capacity, it is estimated tnat a steacy state primary system heat and mass calance coulo theoretically nave oeen achieved. When tne PORV was latened open, the mass flow out tne valve initially - transitioned from low quality to a hign cuality flow as the succooleo primary system was unaole to expano fast enougn to keep uo with tne increased volumetric flow. When the pressure reducea to saturation, voids l Gl l l l l .
. _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~ _ _ -_ _ . . _
' ~ ]_, ~.__ :__._< _ ~ . :x.: .j -c.t .,,.? ' started to form outside tne pressurizer ano this increaseo tne surge rate into the pressurizer. The 'PORV mass flow transitioned oack to a low quality flow wnien persistaa for a wnile'ana tnen continuea to gracually decreased to even-lower quality. Figure 37 snows the PCRV mass flow ouring tnis time. . Also shown in tne figure are PORY mass flows for saturated liquid and saturatea steam (fluid ouality = 0.0 and 1.0, respectively) calculated using tne homogeneous Equilibrium Model critical flow moce1 I9 as well as fluid density in tne upper part of tne hot leg. As snown, the fluid in the hot leg started to stratify after the pumps were stopped and -primary system voiding occurred. Since tne' pressurizer surge line is connected vertically to the top of the hot leg, as soon as voics started to occur there, the surge line flow quickly transitioned to a higner Quality and the PORV mass flow followea suit. This is similar to the PORV mass flow transition in Semiscale altnough tnat transition was celayed until more hot leg voiding occurrea. This supports the conclusion that the surge line/not-leg connection geometry has a strong influence on pnase separation and PORV flow rate.
Figure 38 snows the response of tne LOFT pressurizer liquic level. The pressurizer was liquid full prior to latching open the PORV. The level tnen decreasec into tne indicating range and remainea approximately
-constant while tne PORV was open. The rapid drop in level in 5100 s
- (85 min) was a response to restoration of steam generator auxillary feedwater injection.
An as yet unresolved question about the PORV mass flow concerns the flow oscillations after 4000 s (67 min). Apparently, there was a rapid fluctuation of tne fluid density upstream of tne PORV with tne density oscillating between saturated licuid and saturated steam densities. The cause of tnis phenomena is as yet not clear. To suninarize, the LOFT data exhibit a correlation between tne PORV mass flow and not leg density which is cualitatively similar to that - measured in Semiscale. Tne pressurizer volume itself coes not appear to significantly affect this correlation thougn ene correlation is very 68
.;! . j .0-ector ~ ' pipe l 500 1500 2500 3500 4500 5500 ,
d*"5'* i 0 1000 2000 3000 4000 5000 6000 "***;y I'0 2 1 I I I I l~ l l I I l O.9 - PORV l cycling l
- 0.8 - .
_m -s , m] s 07 -J l h '*N!"$ated 8 4
=
g 0.e 0.5 l f\ l g, 7 i
-f -
[ L 7 1 - 0.4 - I uld l 3 IIEM flow ,
- 0.3 -
y,,,, ; L 0.2 \ \ 1 Y I I PORV open f 0' k I B P! MS _ gg l l iu-m =-- - - = - - l l l l l 0 0 1000 2000 3000 4000 5000 6000 500 1500 2500 3500 4500 5500
- TIME (S). ,
l 1 i figure 37. LOFT Test L9-1 PORV mass flow and intact loop hot leg denstt'. y I ! 1
i OL i l LIQUID LEVEL (m) u o w9*9e9a9 P .- e r ? m
- e a N o !
a s n . . s s a 2 o
'i2 n so m =
w o- 9 F ( m s
?
n y , F 7" begins cycling qp
- oi
-os - = to o 1
_s
~ .i I c l
e c
, z 2
mul mof an o
?_ col z i . F c -
g
- m PORY latched coen m
$ A
_ -o - mo
'm
-
two eh l 0 **8
, 6 . .i;glg) i g*!,l!
._; -. fig,- __ t . _2 _
casa c= *!! -
1 .
p l ~casil?.~j- ii :.; cui cua cui ,
dd ; LEIS]*{aa (c m d. - cui ex> cui
.t
; l.!:
i'/g$_.J._rted21mJ'C'O'l(]A
.piir 1 ,
cr4 ,1
.i _. l -
s . . ..[ -
i ! N U'!.
_ .y l tasa 8888 'I A .'. ,a r'-- 8 88 i tHe
%%8 58* h $ b--l-- M
-- d., .1_
-h.
cs
_ _ _ _ ___ _ . ._ . . .. 35 ! " l 58 "
l * *3stac -
i I
lE295 !!!ifMIf.'." ""I figure 39. REl.APS nodalization of the Semiscale Mod-2A system.
' l i; i
' i i
(
- ! j. !!'!
- i ;t
- 5. A DreaK junction representing the PORV was modeled on the top of tne pressurizer. The piping leading to tne flow limiting orifice used to simulate the PORV was not modelec, out rather was represented with a two-pnase cisenarge coefficient of 0.84 on tne PORY juncticn. Tne selection of tnis value is ciscussea later with the sensitivity studies.
- 6. Pressurizer wall neat loss was modelea mecnanistically. A separate pressurizer mocal was used to modify tne tnermal conductivity of tne insulation material until the calculaceo heat
- 7. Hign pressure injection system flow was set to EOS specified values, ratner than actual values delivereo curing tne test. Tne
- smoother function with respect to pressure in the EOS specification made the results easier to generalize.
- a. Pressurizer neat loss, as determinea in sensitivity stucies witn a secarate pressurizer model, was incluced in all three stages.
- a. 14975 s experiment time correspanos to O s in tne RELAPS calculation.
- =
- c. s e 2 a
- 8 AM 4
- C I
- N
- 1 ase E 5e
- 1. - ..
- C e *
- 4 C C*
- b
- test. Analysis of the. vessel 1fquid level comparison is confounded ,
- a. As statec previously, HPIS flow was modeleo accorcing to tne EOS '
- b. A minor correction was mace to an internal pressurizer junction.
- r'mB+ .-.a m # 6 6- -* cdi mm' M9' 4e*1 Og ec h ' .6 64Ng@m $
- 7. - @
- 3 J
- e. .m s
- b s,'**.. .., ,-
- a. Suosequent calculations showeo tnat 10 kW would nave seen a cetter estimate of steam generator. heat loss.
- a. m 2 e D
- a. _- e w .o.
- m 66
- e e e s .7 -
- L S* -
- g
- s er
- W 5
- a. = g U
- g *- ed
- 1g
- e e e I
- i e t t 6 O O S S O 9 @ C C
- 1. Pressurizer noding nad a small Dut noticeable effect on disenarge mass flow and on pressurizer quality.. Calculations using
- 2. Pressurizer wall heat transfer resultaa in a PORV mass flow rate aDout 207. greater than tnat calculated witn an' aciabatic boundary condition on the outer wall surface. The acianatic wall calculation was noiser than tnat' nitn neat transfer and gave worse agreement with data.
- 3. Calculated PORY mass flow rates were comparea witn data for steam flow at 8 MPa to cetermine a value for the oiscnarge coefficient in tne RELAPS mcoel. The calculated ficws were a linear function of tne di:cnarge coefficient and showed 0.82 to give the best comparison with data.b 4 Doucling and,haIving the pressurizer surge line resistance, R',
- a. Tne 20-node mooel was oevelopec Dy halving the top four noces in :ne 16-node model. .
- b. The system analysis used CO2 = 0.84 instead of 0.82. However, the l estimated value of 0.84 was so close to the derivec value of 0.32 :nat no system calculatians were repeateo.
- l. RELAPS correctly calculate <. tne overall system response ooservea in Test S-SR-2. In addition, it gave a good representation of botn the magnitude and timing of tne P0kV flow, system pressure, and pressurizer and core liquid level transients.
- 2. RELAPS showed agreement with the test result tnat estanlisnment of a steady-state feed and bleed operating point is sensitive to the exact PORV and HPIS flow enaracteristics,
- 3. Modeling sensitivity studies showed that pressurizer wall neat transfer and node size upstream of the disenarge junction have small
- l. ,
- 6. FULL-SCALE PLANT FEED AND SLEED CALCULATIONS ,
- Pro WIS a MU.prietor[on Inlect curv*nd -
- leg orean calculations. Tne mooel represents tne primary coolant system. [
- E -
- 5 I I
- f A p S e S
- 8 g
- S
- a. Incluoes tne liquio in the feecwater line,
- b. Includes tne liquid in the surge line.
- .~~, -
- 1 Event 0.0 Luss-of-offsite power, reactor coolant pumps tripped 0.5 Control .roa drop initiated 1.0 Turoine stop valves closec 3.1 Control rocs fully insertec -
- aecrease racialy until 4150 s wnen fluid in the stagnant portions of tne upper plenum ana upper neaa began flashing. Flasning in tne vessel reducea the aepressurization rate directly througn tne effect of steam generation.
- to
- 3. 2 = 7
- U D U
- p '
- I i
- I at %1 I g iss b8 I '
- s 25 -
- r. =e \
- C- -
- r -- I ) C '
- t.
- as 100 - F -
- s.
- 2.~
- f I
- v.
- e. - -
- f. w x x. : m , : ..n. m q. ,... .x.r..~ .:: - m.:.w ,:x.. x... . . .,_ n . . ..>..- % :a.- 9 s
- 7. CONC 1.USIONS Primary feed. and bleed cooling in -a pressurized water reactor system i with no secondary heat sink has.been analyzed through a study of the basic ,j parameters that govern feed and bleed) the interpretation of experimental i data,cand. finally by verification of computer codes and extrapolation of j the: identified phenomena"to full! scale plan 1!sIAn or'derly examination of the results draws- the following conclusions: (
- by the boundary conditions used, the energy and mass balances were highly subject to experimental uncertainties. A small reduction in core power, PORV mass flow, or an increase in HPIS injection rate would probably have
- resulted in steady-state cooling after the hot leg had voided. -
- bleed cooling becomes viable.
- W.,
- c,
- e,. *
- 1. W. Tauche, Loss of Feenwater induceo toss of coolant accia.nt Analysis .
- 2. N. S. OeMutn, D. Dooranich, R. J..Henninger,1oss-of-Feedwater -
- 3. " Zion Station Final Safety Analvtit Recort", Comonwealth Edisort _. __ - __ -_
- 5. System Desien Descriotion for Moa-2A Semiscale Svstem. Addendum I,
- 6. G. R. Berglund, " Experiment Ocerating Specification for Semiscate Mod-2A Primary Feed and 31eed Experiment 5-SR-1," EG6G Idano Inc.,
- 7. G. R. Sergluna and D. J. Shimecx, " Experiment Operating Specification
- 8. D. J. Shimecx, M. T. Leonard, Results from Semicale Mod-2A Upper Heaa Injection Test Series, Transactions of American Nuclear Society 1981 Winter Meeting., Volume 39, Novemoer-Decemcer 1981.
- 9. H. J. Richter, " Flooding in Tubes and Annuli", Int. Journal of Multignase Flow, Volume 7, No. 6, pp 647-658,1981
- 10. T. K. Larson, G. G. Loomis, R. W. Shumway, Simulation of Three Mile Islana Transient in Semiscale, SEMI-TR-010, EG&G Idano Inc., July 1979.
- 11. G. G. Loomis, K. Soda, C. P. Fineman, Ouick look Recort for Semiscale Mod-2A Test S-NC-2 EGG-SEMI-5507, EG% Idano Inc., July 1981.
- 12. N. Zuber, Decolamt in Modalinc of Cmall Reaak fOCA, NUREG-0724, Octooer 1980.
- 13. Y. Taitel, A. E. Dukler, "A Model for Predicting Flow Regime Transitions in Horizontal and Near Horizontal Gas-Liouid Flow," AICnE Journal, (Vol. 22, No.1), January 1976, 14 James P. Adams Ouick-Look Recort on LCFT Nuclear Exoeriment ~
- 15. Mary L. McCormick-Barger and Janice M. Divine, Exceriment Data Recort for LOFT Anticioateo Transient witn Multicle Failures Exoeriment L9-1 ano small oreaK Experiment L3-1, WREG/Cn-2119, EFr-2101, June tydl.
- 18. . Larry-J.--Ybarrondo,' et al. I" Examination of ' LOFT Scaling", 3
- 19. DougTas 'G. ' Hall ana Lfnda'S'i Czapary,:Tablei of Homogeneous Eauilerium 9
- 22. G. G. Loomis. Sunnary Report Semiscale Mod-2A Heat Loss ., ,