ML20217C550
| ML20217C550 | |
| Person / Time | |
|---|---|
| Site: | 05200003 |
| Issue date: | 07/21/1994 |
| From: | Levin A NRC (Affiliation Not Assigned) |
| To: | Mcpherson G NRC (Affiliation Not Assigned) |
| Shared Package | |
| ML20217C479 | List: |
| References | |
| NUDOCS 9803270056 | |
| Download: ML20217C550 (6) | |
Text
1
. p nog
!f-h E-UNITED STATES i
NUCLEAR REGULATORY COMMISSION-
/
waswisoros, o.c. rosssaooi e...+
July 21, 1994 MEMORANDUM FOR:
G. Donald McPherson, Senior Thermal Hydraulics and Testing Expert Division of Systems Safety' and Analysis THRU:
Mark P. Rubin, Acting Chief Reactor Systems Branch Division of Systems Safety and Analysis i
FROM:
Alan E. Levin,' Sr. Reactor Engineer i
Advanced Reactor Systems Section i;
Reactor Systems Branch Division of Systems Safety and Analysis
SUBJECT:
REPORT ON AP600 MATRIX TEST SB19 AT THE APEX FACILITY AT OREGON STATE UNIVERSITY Matrix test $819 in the Advanced Plant Experiment (APEX) Facility at Oregon State University (050) was performed on July 14, 1994. The test simulated a 2" cold leg small-break loss-of-coolant accident (SBLOCA). The test was similar to test SBl in the APEX loop, with the exception of the imposition of a varying simulated containment backpressure on the loop components that would be directly affected by containment pressure: the sump (simulated by a large tank at an appropriate scaled elevation); the in-containment refueling water storage tank (IRWST); and q
the break and accident measurement system (BAMS), into which the effluent from both the break and the fourth stage of the automatic depressurization system (ADS) exhaust, rather than flowing into the containment as in the actual plant.
The containment backpressure was programmed to vary as a function of time during the simulated accident, based on an AP600 calculation performed with the WGOTHIC containment code.
APEL is a one-quarter height, reduced-pressure integral test facility.
The rspresentation of the AP600 primary and secondary systems and associated safety systems is as close to prototypic as is practical. No containment is included as. part of the facility, but the capability does exist to vary the effective
. backpressure on. system components.
The test was monitored by Alan Levin of NRR. Following a short pre-test briefing describing the conditions and. expected behavior for the test (based on the results of 581), SB19 was performed over a period of approximately 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, including 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> of sump recirculatory (long-term) cooling. Results of the test were largely as expected. The timing of. the early stages of ADS were almost the same as in SBl as would be expected due to the existence of critical' flow from both~the break and the ADS. No' core uncovery or heatup was experienced during
~the test. Systems. interactions were noted, however including flow oscillations
'between the IRWST and the core makeup tanks (CMTs),that the OSU staff indicated J had no_t been seen in SBl.
The testing staff conducted the experiment in a L competent and confident, manner, demonstrating complete familiarity with.the facility operations. -The smoothness of facility operation was quite impressive. - contains: detailed observations of the. test:and a brief ~ review of
- preliminary, data.
b k b 3-
A.
PDR.
a
G. Donald McPherson July 21, 1994 Because the NRR observer was present for the OSU test readiness review in May 1994 and was familiar with test procedures and basic QA practices, these items were not reviewed on this trip.
This test was also witnessed by R. T. Fernandez of the Electric Power Research Institute.
Michael. Carter was the test supervisor, with John Haytas and Dan Holland operating the APEX facility. Dr. Jose Reyes, OSU Principal Investigator for the APEX program, and Moshe Mahlab, Westinghouse's APEX Program Manager, also attended the test.
Please refer questions to the undersigned at 504-2890.
h Ala E. Levin, Sr. Reactor Engineer Advanced Reactor Systems Section Reactor Systems Branch Division of Systems Safety and Analysis cc:
M. Virgilio R. Jones
4 AP600 APEX TEST PROGRAM - MATRIX TEST SB19 OBSERVATION REPORT Introduction Test SB19 was performed on' July 14, 1994 in the APEX Facility at Oregon State University, a one-quarter-height, reduced pressure, integral test loop.
The i
conditions for this test simulated a 2" cold leg small-break loss-of-coolant J.
accident (SBLOCA) at the bottom of a pipe on the "CMT" side of the plant. The conditions were similar to those for the first APEX SBLOCA test, SB1, with the addition of a varying simulated containment pressure.
The test continued out into the long-term cooling, sump recirculation mode, for a period of about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />. While the test facility represents the primary and secondary loops of the AP600, including all safety-related and key non-safety systems, in a prototypic configuration, it does not include a containment structure. However, the effect of containment pressure on reactor system components can be simulated by varying the backpressure on tanks representing the containment sump and' the in-
' containment refueling water storage tank (IRWST), and on the break and accident measurement system (BAMS), which collects the effluent from the break and automatic depressurization system (ADS) to permit measurement for the purposes of calculating system mass and energy balances. The variation in backpressure was based on a calculation for this accident using the MG0THIC computer code,
~
which is designed to calculate AP600 containment conditions during an accident.
The target pressure curve ramped quickly to approximately 15 psig, with a gradual monotonic decay to about 4 psig.
The test was monitored by Alan Levin, SRXB lead AP600 test reviewer. It was also observed by R. T. Fernandez of the Electric Power Research Institute, on behalf of EPRI's Analysis and Test Review Team (A&TRT). Other observers included Jose Reyes, OSU APEX Principal Investigator, and Moshe Mahlab, Westinghouse APEX Program Manager.
Preliminary indications based on observation of the test and post-test review of test data are that the test was successful. Early facility response matched that of SBl very closely, as would be expected, since backpressure should have a negligible effect on early blowdown behavior (critical flow). Long-term behavior varied somewhat from SBl (see discussion below), but the core remained covered and cooled at all times.
Detailed observations are included below.
Pre-test Briefino and Test Preparation The NRC observer has visited the facility previously and has reviewed elements of Westinghouse's QA program as implemented at OSU; consequently, no tour was scheduled nor was a pre-test QA or instrumentation review conducted. A copy of the test procedure was made available for use at the facility, but was not provided for inclusion in this report.
The test was scheduled to begin at approximately 6 a.m. on July 14. This observer arrived at OSU at about 5 a.m.
and ~ reviewed briefly the test procedure.
The facility was almost at test initiation conditions', with the exception of warming the core makeup tank (CMT) pressure balance lines, which was accomplished shortly befure the test began.
'In addition, Michael Carter, the test supervisor, gave a short pre-test briefing on procedures and. expected system behavior, such as timing of ADS' operation, based on results 'from test SBl.
The test procedures are similar to those followed at the SPES integral test facility, with some small differences. The facility is brought to test pressure (approximately 370 psig) and temperature
l (approximately 400*F), and the test is then initiated by pressing the appropriate
-button on the facility control panel. Two minutes of pre-test data are acquired by the data acquisition system, and the break valve is then open to initiate the test. Instead of waiting for the system to generate an "S" signal based on AP600 control logic (e.g., low pressurizer pressure), an "S" signal is automatically generated approximately 0.5 seconds after test initiation. The "S" signal scrams the - facility, opens the CMT-isolation valves and the passive residual heat removal system outlet valves, and trips the primary coolant pumps.
All non-safety systems are also tripped. The transient is then allowed to proceed based on simulated AP600_ control logic.
. The other significant difference between APEX and SPES is the presence of air-operated valves in the IRWST discharge lines to the direct vessel injection (DVI) lines.
These' valves do not exist either in SPES or in the AP600.
They were
' installed in APEX due to concerns about back-leakage through the check valves
' that are nominally supposed to isolate the IRWST from RCS pressure. These valves are opened when the facility pressure reaches about 40 psi, which is considered
' adequate to allow the-system to settle out before RCS pressure is reduced to IRWST_ injection pressures.
The facility was at steady-state pre-test conditions by about 5:55 a.m., and the test was initiated shortly thereafter.
Test Conduct and Data Recordina The~ control room staff operated very smoothly and competently throughout the test.
It is clear that the experience gained over the past several months of operation has allowed the staff to anticipate system behavior. Michael Carter oversaw test operations. John Haytas was-the facility operator, and Dan Holland operated the data acquisition system (DAS). The DAS uses three PCs for recording the data on disk, and each PC monitor can be set up to permit real-time observation of up to 4 instruments, which are plotted as " scrolling" graphs.
The " test" button was pressed at approximately 5:55 a.m.
Following the two-minute steady-state data collection period, the test itself began at about 5:57 a.m.,
with the "S"
signal following LOCA initiation per procedure.
CMT recirculation followed almost immediately for a brief period, with draining occurring thereafter.
First stage ADS initiated at about 9 minutes into the test, followed by ADS 2 at 10 minutes and ADS 3 at 12 minutes.
During the initial stages of the test, significant " banging" was heard from the test facility. ~ This was deduced to have come from the BAMS system. Additional noise associated with less violent boiling / condensation behavior was noted as testing
- proceeded.
'The test continued without incident.
ADS stage 4 actuation occurred approximately 15 minutes into the test.
As the IRWST drained, the pressure vessel: inventory rose to the point that the CMTs began to refill. BAMS pressure reached its peak value of about 11 psig at around this time; it did not reach the target value of 15 psig, most likely as a result of insufficient steam generation to pressurize _the appropriate tanks. IRWST injection continued, with injection from the CMTs as well. A series of slow flow oscillations was observed on the
.DAS monitors, lasting for.several. tens of minutes. This was clearly due to an L
+
n
.r
. ' nteraction between the IRWST and CMT-injection: CMT flow would increase slowly, i
accompanied by a decrease in _IRWST flow.
Overall flow into the DVI line generally increased slightly during.these oscillations. Backpressure decreased during this period, undershooting the 4 psig setpoint and falling to about 1.5 psig.
It then rose gradually to the 4 psig setpoint and remained there for the duration of the test. -The CMTs were finally emptied of liquid at about 3-1/2 hours into the test. CMT temperatures were observed to increase to saturation temperature as the CMTs filled with steam.
Sump injection initiated _ at approximate 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> (9:57 a.m.), triggered by IRWST low level. When the change-over to sump injection occurred, both DVI flow and the level in the pressure vessel began to fall, due to the reduced driving head available from the sump, which is at a lower elevation than the IRWST.
However, the level stayed well above the top of the simulated fuel rods. Sump injection was generally steady, but the injection flow was extremely asymmetric, with the flow through one sump line of the order of 5-10 times that of the other. The reason for. the asymmetry Lwas explained as being due to differences in piping for the two lines.
It Lappeared as. if flow was going from one of the sump lines back to the IRWST, and then from the IRWST to the pressure vessel.
Sump recirculation (long-term cooling) continued for slightly more than 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br />, and the test was terminated at approximately 12:10 p.m.
No instrumentation nor data' acquisition difficulties were noted. The initial impression of the test was that it was successful; however, OSU personnel stated that some of the thermal-hydraulic response in SB19, especially the flow oscillations between the CMTs and the IRWST, had not been observed in test SBl..and would require some additional evaluation to see how this might be related to the elevated backpressure in SB19.
Post-test examination of the data was conducted the day after the test, and is discussed below.
Post-test operations include loop cooldown and a zero check of all instruments.
This was accomplished during the afternoon, following the test.
Data Processina and Quick-Look Reports The OSU testing staff began data processing immediately following the test. The entire process requires approximately one day, as the data are transferred from the DAS PC's to CDs, and are then plotted. With approximately 800 instruments being recorded for more than 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />, this translates to a very large amount of data processing. Plots are generally ready for examination on the afternoon of the day after that of the test.
Quick-look reports are planned for issuance approximately 30 days after test completion.
Post-Test Data Examination
-On July 15, Lthe NRC monitor returned to OSU for a brief post-test data review.
The plotted data were - available --early in the afternoon.
The data plots essentially confirmed the observations during the test. A pressure spike early in the test was quite evident on a number of the plots, resulting from the water hammer event, f Aside from the momentary disruption of these readings, however, no significant additional-effects'were noted. The flow oscillations between the
'CMT.and IRWST were also very apparent. One parameter that was_ checked was the n
level in-the pressure. vessel.
DuringqlRWST. injection, the level clearly rose l
- t: y.
l over'the elevation of the cold legs, allowing the CMTs to refill partially. As described previously, the level dropped several inches when the changeover to
. sump injection occurred. Based on thermocouple readings in the primary loop, the
-mixture level was around that of the hot legs (which in the AP600 and in OSU are below the elevation of the cold ~1egs), but the cold legs were uncovered, allowing steam to vent through the break. This also allowed steam to fill the CMTs, as observed during the test.
{
The number of plots made a comprehensive evaluation of the test data impossible in the time available.
However, the plotted data did not show significant evidence of instrumentation problems; a few instruments did show signs of zero shifts, but no gross malfunctions were noted. A post-test acceptance rev.iew will be-performed by OSU,.and the data will then be forwarded to Westinghouse for a more in-depth evaluation.
General Observations "Overall, the impression received of the test operations was one of confidence in facility response and competence in performing the test.
It is clear why the test program has proceeded with relatively little time between tests.
The facility gives a favorable physical impression, as well; the test area is clean, and control room decorum is appropriate.
While observers are permitted to observe test operations and monitor data on the DAS, test operation personnel are not to be disturbed.
.It is not clear at this time what the results of Westinghouse's post-test review will.be.
The inability of the pressure controller to match the. programmed backpressure curve for the BAMS, IRWST and sump will have to be evaluated, and it is conceivable that Westinghouse will ask for a repeat of the test with an improved pressure program. It is clear, however, that the elevated backpressure does have an observable effect on the loop thermal-hydraulic response, which should be evaluated further when the data from SBl and SB19 are made available to the NRC, to help determine whether Westinghouse's assumption of an atmospheric pressure containment during this type of event for licensing basis analyses is, in fact, conservative.
Based on these observations, the OSU APEX program appears to be a valuable adjunct to the SPES-2 high-pressure integral systems test program, and should provide high-quality thermal-hydraulic data to assist in the analysis of AP600 accident response.
f I
1
.e.,m.,,.
.., i,,.
..i
,.,y