ML17255A450
| ML17255A450 | |
| Person / Time | |
|---|---|
| Site: | Ginna  |
| Issue date: | 10/14/1983 |
| From: | Dick G Office of Nuclear Reactor Regulation |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8310190029 | |
| Download: ML17255A450 (16) | |
Text
Docket Ho. 50-244 October 14, 1983 DISTRIBUTION, CDocket ~
NRC PDR Local PDR ORB ¹5 Reading DCrutchfield GDick OELD EJordan JTaylor ACRS (10)
NRC Participants NSIC LICENSEE:
ROCHESTER GAS AND ELECTRIC CORPORATIOft FACILITY:
R. E. Ginna Nuclear Power Plant
SUBJECT:
SUt1t1ARY OF OCTOBER 5, 1983 MEETING TO DISCUSS THE IHPO SIt'lULATIOH OF THE GIHHA STEAtl GENERATOR TUBE RUPTURE EVENT On October 5, 1983 members of the tlRC staff met with representatives of Rochester Gas and Electric Corporation (RG5E), Institute of Nuclear Power Operations (IHPO), and Argonne t<ational Laboratories (AttL) for the subject discussion.
Using steam generator tube rupture (SGTR) i nfor-mation obtained from RGSE, IHPO modeled the event using the RETRAN-02 computer code.
IHPO will provide the program to the NRC and AHL. It will be used to study the thermal-hydraulic response of the system during the SGTR and the implications regarding pressurized thermal shock.
The attendance list (Enclosure
- 1) and agenda (Enclosure
- 2) are enclosed.
IHPO provided an overview of how the SGTR event was modeled and the analysis of the results, (Enclosure
- 3) summarizes the general in-formation provided.
The answers to specific technical questions were provided by IHPO.
AHL will be provided a copy of the program tape and the data necessary to run the program within 3 weeks.
Adapting the program to another
- computer, running it, and analyzing the results is expected to take approximately 10 months.
Original signed by George F. Dick, Jr., Project t1anager Operating Reactors Branch ¹5 Division of Licensing OFFICEIN SURNAMEP DATEf
~
¹
~
( ~
DCru ield
- l4i ifi9.3',:
¹5
~
~ ~
~
~ ~ ~ ~ ~ ~
ick:jc Jd!P.!s 8310190029 831014 PDR ADOCK 05000244, P
PDR gg>f ~(i Qgy gEE'g(/4)
~ ~ ~ (
(
~
~
~\\( ~ ~ ~ ( ~ ~ ~ ~
~
~ ~ ~ ( ~ ~ ~
~ ~ ~
~ ~ ~
(
NRG FORM 318 (10.80) NRCM 0240 OFFICIAL RECORD COPY USQPO: 1981-339.960
p,g AKg(
~cw C
/y qO
++**+
0 UNITED STATES NUCLEAR REGULATORY COMMISSION LVASHIIVGTON,D. C. 20555 October 14, 1983 Docket Ho.
50-244
'I LICENSEE:
ROCHESTER GAS AND ELECTRIC CORPORATIOH-FACILITY:
R.
E. Ginna Nuclear Power Plant
SUBJECT:
SUMMARY
OF OCTOBER 5, 1983 MEETING TO DISCUSS THE INPO SIMULATION OF THE GINHA STEAM GENERATOR TUBE RUPTURE EVENT On October 5, 1983 members of the HRC staff met with representatives of Rochester Gas and Electric Corporation (RGSE), Institute of Nuclear Power Operations (INPO),
and Argonne National Laboratories (ANL) for the subject discussion.
Using steam generator tube rupture (SGTR) infor-mation obtained from RGIEE, INPO modeled the event using the RETRAH-02 computer code.
IHPO will provide the program to the HRC and,ANL.
It will be used to study the thermal-hydraulic response of the system during the SGTR and the implications regarding pressurized thermal shock.
The attendance list (Enclosure
- 1) and agenda (Enclosure
- 2) are enclosed.
INPO provided an overview of how the SGTR event..was modeled and the analysis of the results, (Enclosure
- 3) summarizes the general in-formation provided.
The answers to specific technical questions
- were, provided by INPO.
ANL will be provided a copy of the program tape and the data necessary to run the program within 3 weeks.
Adapting the program to another
- computer, runni ng it, and analyzing the results is expected to take approximately 1 0 months.
Geor F. Dick, J Project Manager Operating Reactor Branch g5 Division of Licensing
ATTENDANCE LIST OCTOBER 5, 1983 MEETING t)NE Robert Eliasz Robert tlecredy Roger Myrick Gary Fader William Brown Edward Minkler Jack Tessier Tom Hie Jack Guttmann George Dick ORGANIZATION RGKE RGSE IHPO IHPO INPO IHPO AHL ANL HRC HRC Enclosure 1
AGENDA for Meet'ng with INPO,
- RGRE, NRC, and ANL on I
BZTRAN Analysis of Ginna Steam Generator Tube
. Rupture Event 8: 30 Welcome 8.45 9:00 9:30 12:00 Lunch Introduction NRC Analysis Goals
& Objectives INPO/RG&E RETRAN Model Objectives of INPO Analysis Observations and Results Important Physical Phenomena Modeling Assumptions,and Limitations Conclusions and Suggestions G.
Fad er R. Mecredy J.
Guttman R.
Wyrick 1.00 3:00 Additional Technical Discussions Adjourn All All Enclosure 2
THERMAL-HYDRAULICANALYSIS OF THE GINNA STEAN GENERATOR TUBE RUPTURE EVENT USiNG RETRAN-02 R. K.
WYRXCK) E.
N.
- WlNKLER, AND W.
W.
BROWN Institute of Nuclear Power Operations Atlanta, Georgia INTRODUCTION A detailed analysis of the thexmal-hydraulic response of the R.
E. Ginna nuclear power plant during a major steam generator tube rupture event of January 25,
- 1982, was performed.
The event illustrated a nu ber of thermal-hydraulic phenomena and operational problems that can occur durir~ tube rupture transients.
For this
- analysis, a computer simulation of the G'na event was performed using the RETRMl-02 computer code (Ref.
1) and the results com-pared to the available plant data and sequence of events.
The objectives of this analysis included the following:
o a thorough understanding of the thermal-hydraulic response of a nuclear power plant during a steam generator tube rupture transient o
better understanding of the sequence of events and physical phenomena tha" occurred during the Ginna event than was available rom the plant data o
better understanding of various procedural actions and their corresponding affect on the physical response of a nuclear power plant during steam generator tube rupture transients.
GONNA EVENT OVERV le The P,. E. G'nna nuclear plant, is a Westinghouse-designed two-loof pressurized water reactor (Ph'R) that has been in operation since 1970.
The plant full-power output is 1520 Hwt.
On January 25,
- 1982, a single tube ruptured in the "B".steam generator (SG) at Ginna, initiating a complex plant transient, Encl osure, 3
lasting approximately 33 hours3.819444e-4 days <br />0.00917 hours <br />5.456349e-5 weeks <br />1.25565e-5 months <br />, until cold shutdown conditions were reached.
Salient occurrences and phenomena during the event included the following:
I o
rapid reactor coolant system (RCS) depressurization, reactor trip on low pressure, and initiation of safety injection (SZ) o natural'circulation cooling of the RCS due to procedural requirements to stop reactor coolant pumps o
formation of a steam region in the reactor vessel upper head dur'ng RCS depressurization by opening of a pressurizer PORV o
overfilling of the "B" SG and flooding of the steam line with water o
pressurization of the damaged "B" SG and steam line, causing opening of an SG safety valve o
liquid releases through the "B" SG safety valve, as indicated by radiological analysis of site samples o
higher temperatures in the "B" SG and steam line than in the RCS during plant cooldowm.
ANALYSIS The thermal-hydraulic response of the Ginna steam generator tube rupture event was analyzed by simulating the transient with the RETRAN-02 computer code and compar ing the calculated results with the available plant data obtained from the plant computer and operator logs.
The RETRAN-02 model (see Figure
- 1) consisted of 0N fluid volumes, 61 junctions, and 20 heat conductors.
Initial conditions from the event were used first to develop a steady s ate case to establish the numerical stability of the model.
The transient was initiated by causing a break flow area of 0.0066 fti 2 in a "B" SG tube.
The tube rupture flow rate was calculated using the extended Henry-Fauske critical flow correlat'n (Ref. 2) available in the R=-.RAN code.
A discharge coefficient for the break of 0.075 was de ermined by a series of parametric studies in which the discharge coefficient was varied until the calculated pressurizer and RCS pressures reasonably matched the available pressure data.
Twelve flo" boundary condi"ions were. used during the transient calculation.
Eight were RETRAN filljunction boundary conditions for "A" and "B" steam flow to the turbine/
condenser,'A".and "B" main feedwater flow, "A" and "B" auxiliary
These flows were determined from a combination o
known valve and pump actions obta'ned from the plant computer and operator logs, and iterative variation of the flow ra"e within design limits until calculated plant transient results were in reasonable agreement with the available plant data.
The plant data included RCS and steam generator pressures, pressurizer and steam generator water levels, reactor vessel upper head and core exit fluid temperatures, and RCS cold leg temperatures.
The other four boundary conditions were safety injection pump flow, pressurizer PORV flow, "A" SG atmospheric steam dump valve flow, and "B" SG safety valve flow.
Safety injection flow was calculated using the RETRAN control logic from pump performance data for flow versus RCS pressure.
The pressurizer PORV, the "A" SG atmospheric steam dump valve, and the "B" SG safety valve were modeled by connecting he downstr earn side of the valves to a large volume at atmospheric pressure.
Flow through the valves was calculated.using the extended Henry-Pauske (Ref. 2) and Moody (Ref. 3) critical flow correlations available in the RETRAN code.
Reactor Vessel Upper Head Modeling Flashing of fluid in the reactor vessel upper head region was suspected to have occurred during the event.
This was based on temperature
'ndications rom thermocouples located 'n the lower port'on of the upper head and pressurizer level data'hich showed a rapid increase during RCS depressurization, when a pressurizer POPV was opened.
The calculation of steam formation, the size of the steam bubble, and its rate of collapse required a model capable of a reasonable prediction of the temperature distribution in the upper head fluid.
The actual reactor vessel upper head geometry and flow distribution is complex, involving flow paths connecting with the outlet plenum region through 33 control rod guide tubes, and with the downcomer region through flow holes in the upper support plate.
A relatively simple RETRAN model was used to allow a two region axial temperature distribution in the upper head.
(See Figure I)
The final model was developed by a series of transient calculations in which the flow rate (i.e. the junction loss coefficients) entering the upper head and the relative sizes of he two upper head volumes were varied to determine an upper head enthalpy distribution required to predict the available upper head empe. ature and pressurizer level data.
The RETRAN non-equilibrium thermodynamic "pressurizer " option was used to facilitate modeling water 1'evel and the subsequent collapse of the steam bubble in the upper head.
"B" Steam Line Modeling During the event the "B" SG water level increased to the top of the SG and overflowed into the "B" steam line.
The "B" steam line eventually filled completely with water.
The SG and steam line model included the use of the non-equilibrium thermodynamic "pressurizer " volume option in the SG and 'n the steam line to improve the modeling of non-equilibrium effects during the filling of the steam line.
The spray junction option and the steam-water heat transfer coefficient of the "pr essurizer" model in the steam line were varied in a series of transient calculations to provide condensation and compression rates necessary to match the available "B" SG pressure data.
Once the pressure data was matched the model provided a reasonable estimate of the water level trans'nt in the steam line.
RESULTS Selected calculated transient results from the analysis and comparison to plant data, where available, are presented in Figures 2-5.
Figure 2 shows calculated flow rates for the tube rupture flow, safety injection pump flow, and charging pump flow.
Figure 3 shows a comparison of calculated RCS and steam generator pressures w'h the ava'able plant data.
'Figure 0 shows calculated temperatures in the reactor vessel upper head and at the core exit.
The upper head temperatures include the liouid temperature in the bottom region of the upper head (RETRAN Volume 20),
and the liquid and steam temperatures in the top region of the upper head (RE'iRAN Volume 19).
Available data from thermocouples located. in the bottom region of the upper head (near the interface of RETRAN Volumes 19 and 20) and at the core exit are also shown in Figure N.
Figure 5 shows the calculated reactor vessel upper head steam volume.
Figure 6 shows the calculated RCS cold leg temperatures and their comparison to the available cold leg temperature data.
OBSERVATiONS AND CONCLUSIONS Principal observations and conclusions of the analysis are briefly summarized below for one hour and twenty minutes of the transient, during which time most of the significant thermal-hydraulic phenomena occurred.
Fur the. details regarding the Ginna event and i's thermal-hydraulic response can be found in References 0, 5, 6, and 7.
Elapsed event time is expressed as hours:minutes:seconds.
The maximum primary-to-secondary flov rate through the ruptured tube occurred shortly after the rupture and vas calculated to be 632 gpm (62 ibm/sec.).
2 ~ After the reactor coolant pumps were stopped at 0:04:10, flow into the
. eactor vessel upper head was calculated to decrease to about 1 percent of the normal flow rate.
As a'esult of this low. reduction, the vessel upper head'luid cooled more slovly during plant cooldown than the remaining RCS fluid and became the hottest region in the RCS.
This condition lasted approximately 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, until a reactor coolant pump was restarted at 1:57:00.
3 0 When a pressurizer PORV was opened at 0:42:30, the hottest fluid region in the RCS was at the top of the reactor vessel upper head.
The fluid temperature there vas calcula ed to be about 20 oF hotter han the fluid 'n the bottom region of the upper head (vhere three thermocouples were located) and about 100 F hotter than the core exit fluid.
The calculated subcoo3.ing marg'n in the hottest fluid region was only 0
F at the time the POHV was" opened.
Formation of steam ws calculated to have occurred in the reactor vessel upper head region dur'ng the initial pressure decrease follow'ng the tube rupture and during the depressur-ization assoc'ated with the pressurizer PORV openings.
The calculation indicated that 7.5 ft~ of steam formed during the nitial pressure decrease and approximately 200 ft of steam formed dur ing the pressurizer PORV openings.
5-Based on this analysis steam did not form in the RCS hot legs or steam generator U-tubes at any time during the trans'ient.
- 6. Complete isolation of steam flov from the affected "B" steam generator appears "o have occurred about 6 minutes after closure of the "B" Y>SZV at 0:15:00.
Et is likely that flow from the "B" steam gener ator to the turbine-driven auxiliary feedwater pump was not isolated until the pump was stopped at 0:2'1:00.
7
~ Based on the calculation, flov through the ruptured tube caused the water level to increase in the "B" steam generator, causing compression and superheating of the steam regions at the top of the "B" steam generator and in the "B" steam line (after the steam ne was completely isolated).
- 8. After 0:02:30, when steam formed in the reactor vessel
- head, the system pressure was influenced by three effective "pressurizers."
These included the real pressurizer, the steam/water interaction in the reactor vessel upper head, and
the steam/water interaction in the "B" steam line.
9
~
10.
The calculation indicated that the "B" steam generator began overflowing into the "B" steam line'a about 0:43:10.
The SG overfill caused a slight reduction in "B" SG pressure due to the condensing effect of subcooled fluid from the "B" SG on the superheated steam in the "B" steam line.
Based on this analysis, the "B" steam line header, where the SG safety valves are located, filled with water at about 0:47:30, prior to the first safety valve opening.
Therefore, water releases occurred during all openings of the "B" SG safety valve during the event.
The calculation indicated that the entire "B" steam line filled completely with water up to the HSEV at 0:58:28.
One of the four "B" SG safety valves opened several times during the transient.
During all of the valve openings, the valve flow area was calculated to be less than the full-open flow area.
This calculated result was consistent with the post-event examination of the valve, which also indicated that the valve never reached a full lift position.
12.
Based on the calculation, the "B" safety valve opened and closed at least once before the "B" steam line f'lied solid with wa er.
After the "B" steam line became water-solid at 0'.58:38, the analysis'indicates that the safety valve flow area incr eased and decreased corresponding to pressure changes in the s
earn 1'ne, but likely did, not, close completely during the two-hour and four-minute interval from 0:58:00 until 3:02:00.
The inability of the valve to close can be explained by the continued flow through the ruptured tube (due to RCS safety injection and/or charging) into the water-solid "B" SG and steam line.
13 ~ Based on the plant.data, the stopping of safety injection at 1:12:00 caused RCS pressure to decrease from about 1370 psig to about 930 psig at 1:16:00.
The "B" SG pressure correspond-ingly decreased from about 1035 psig to 840 psig.
The RCS and "B" SG pressures decreased after stopping safety injection but did not equalize.
"B" SG pressure remained 50-100 psi less than RCS pressure until 3:02:00.
This was most unlikely due to continued leakage through the "B" SG safety valve during this time period.
14.
Based on this analysis the RCS and "B" SG pressures would have eoualized after safety injection had been stopped if charging flow had been reduced to the value of the letdown flow rate.
- 15. With reactor coolant pumps off and natural c'rculation flow nearly stoppeo 'n the "B" pr'mary coolant loop, decreased thermal mixing of safety injection fluid in the cold leg significantly recuced the "B" cold leg flui'd emperature.
The analys's 'dicated that the net flow direction in the nearly stagnant.
"B" cold leg was directed toward the reactor vessel rather than toward the ruptured SG tube.
The reactor vessel wall directly beneath the "B" inlet nozzle likely experienced much cooler water than other reactor vessel
- regions, creating a
potential for thermal shock of the vessel s~ll.
(Fracture analysis performed by Westinghouse indicated that no flaw propagation occur. ed during the Ginna event.')
i6. Continued operation of safety injection and charging caused RCS presssure to remain higher than the safety valve opening setpoint of the damaged steam generator during most of the time between 0:00:00 and 1:12:00.
Continued safety injection and charging caused the "B" steam generator to overfill and to flood the steam line, and caused "B" steam generator pressur e to increase and a steam generator safety valve to open several times.
REF"RENCES "RETR&-02, "A Program for Transient Thermal-Hydraulic Analysis of Complex Fluid Flow Systems,"
EPRI NP-1850-CCM,'ay, 1981.
2 0 Henry, R.E.
and Fauske, H. K. "The Two-Phase Critical Flow of One-Component Mixtures in l<ozzles, Orifices, and Short Tubes, "J. Heat Transfer, 93, 179-187, 1971 3 ~ Moody, F. J.,
"Maximum Flow Rate of a Single Component, Two-Phase Mixture, J.
Heat Transfer, 87, 130-102, 1965 "Analysis of Steam Generator Tube Rupture Events at Oconee and Ginna,"
INPO Report 82-030, 1<ovember, 1982 5.
6.
"Thermal-Hydraulic Analysis oT Ginna Steam Generator Tube Rupture Event",
INFO Report 83-,
1983 "Incident Evaluation, Ginna Steam Generator Tube Failure
- Incident, January 25,
- 1982, R.
E. Ginna Nuclear Power Plant,"
Rochester Gas and Electric Company, April 12, 1982 7.. "t<RC Repor t on the January 25,
- 1982, Steam Generator Tube Rupture at R.
E. Ginna Nuclear Power Plant,"
NUREG-0909, April 1982.
J gE I I
%W k L 0
%W'
~
~
es
~
II 1 ~
~
~
II
IS00 e
1 I4 1000
~
~ ~
~
~ ~ ~ ~ ~ ~ ~
~ ~ r~ ~ ~ ~
~ ao carou>>>>d Rcs ortaaure IRETRArtvolumt 6'I) 4>>u>>red A SraamLne freaaure IRETRAkVaume 66)
CaIOIaltd 8 SraamLnt Oreaaure IRETRAIIVdrumtooI
~
Rcs arcature oara ICrom'A'vuR Leo)
A A'ream Vr>> arcature Oaia
~
% Srtam Lne oreuure OaIa
~ ~ r ~
~
~
~ ~ ~ ~ ~ ~ ~
ro, e
uu1 A.
2, 0 0 0
ELAPSEO TIME.MINUTES Figure 3.
HCS and Steam Line Pressures vs.
Time ur C0:
I aSG
~ ~ ~ ~ ~ ~
Card>>ared Lieu>> TtmoeraIae n VdammtaOIRETRArt Vaumt 191 4>>urartd Core EIo 9tmoeralure IRE,RANva'ume 131
~
C>>La o rom Abrade O'ue Cat Era Tr>>rmooaoree I
I I
'I
~~~~ a
~ ~ ~ r ~ ~
~ ~ ~ ~ ~ ~ ~ ~
~ ~ a<a
'iu'Pkt
~
~ ~ ~ ~ ~ ~ ~ 4>>Iaarto Seam Temotrarure uI voder tread IRET RArtvruume 191 L, ~ ~
L,o,, r<~,
Ca>>urared Voad Temotrarrte n
~
Oav crom Vooe r>>to Tr>>,mov4drr Vg
~ ~ ~ ~, ~
~
~
Voarr WadrRETRArr Vaumt ECI Looaredrerr>>r>>
AOI>>eVo~
Ln
~'a '
~
SeOOn O>>re
%~1,~ /
0
'IC 0
ao 50 ELAPSEO TIME. MINUTES iC Figure 0.
Reactor Vessel Upper Head and Core Exit Temperatures vs.
Time
~
t OORV 6VJC ~ VJVCCIOICO(04$001 I
'Ilo 110 0
140 Io 110 I
0$
~'00
(
)
(o I
Ly CLI C
PORv Ooesb (boe StJn CJ IOIJJt4Vs (R ET4At(VOIVCC 101 10 00 40
$0 10 ELAPSEO TIME.MINUTES b
Figure 5.
Reactor Vessel Upper Head Steam Volume vs. Tme
~
~ bb ~
~ ~
4
~
~~
~ ~ ~\\ ~ ~ e ~
~ ~ tb ~ e ~
~ ~ eeet
~
bb I
I I
~
~ ~
st ~
~
~
~
I
~
I
~
~
s
~
~
~
~
~
~
~
~
bl
~s I,
~
beb I
I
~'t II I ~IIII I I I I
~
I
~
Iv ts lb II I~
~ ~
~
4 (4bt I{ sl I
V I
~ b
~
~ I ~ ~
ISIt
~b 4
II I
~
4
~ I
~ I
~
I I I I 4
Ii
(
I ~ I I ~ I(
I ~
~ Lr (4 sst ss s ~ ssl Its',
ss I I I,I Is
~ II
~
) I bbl I ~
I I I
~ I tl b4 I
~
~ ~
II 4
4)
IIII~
II
~ ~
swirr
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~
~ ~ ~ ~ ~
CJbovtsteoLooo
- 'C(IOLeoyt(tbotsetbrcl461RAI(v(bbrtst6$ (
Cesevteteo Looo 6'(vo Leo Ttvbot(et(re tRETRAN V((br(be $41 Looo A Coso Lco 0414 I~I(ters of Sl Itsy I(btcotsoss Iso(l tet
~
Lao.e m LtQOJIJ(vxt(44~of $4'kty Ie~~NoD<i 10 20 XI
$0 10 ELAPSED TIME. MINUTES Figur e 6.
RCS 'Cold Leg Temperatures vs.
Time
l'gpss
~,
4