ML20217C508

Forwards Rept on AP600 Matrix Test 12 Performed in Facility in Italy.Purpose of Test Program to Acquire Data for Development & Validation of W Accident Analysis Computer Codes

ML20217C508
Person / Time
Site:	05200003
Issue date:	10/25/1994
From:	Levin A NRC (Affiliation Not Assigned)
To:	Mcpherson G NRC (Affiliation Not Assigned)
Shared Package
ML20217C479	List: ML20217C474 ML20217C501 ML20217C508 ML20217C514 ML20217C522 ML20217C544 ML20217C550 ML20217C556
References
NUDOCS 9803270044
Download: ML20217C508 (4)

Text

p KICg p

Iog UNITED STATES

((y g NUCLEAR REGULATORY COMMISSION WASHINGTON, D. C. 20$55 g E

,/ October 25, 1994 MEMORANDUM T0: G. Donald McPherson, Senior Thermal Hydraulics and Testing Expert Division of Systems Safety and Analysis THRU: Robert C. Jones, Chie Reactor Systems Branch Division of Systems Safet and Analysis  ;

FROM: Alan E. Levin, Senior Reactor Engineer Advanced Reactor Systems Section Reactor Systems Branch Division of Systems Safety and Analysis '

SUBJECT:

REPORT ON AP600 MATRIX TEST N0. 12 IN THE SPES-2 FACILITY Attached is a report on the Matrix Test #12 performed in the SPES-2 facility at SIET Laboratories, Piacenza, Italy. This was the final test in the original SPES-2 test matrix, and simulated a large main steam line break (MSLB) at hot, zero power conditions. It was also a " blind" test, for the purposes of Westinghouse's computer code validation and qualification activities.

The purpose of this test program is to acquire data for the development and validation of Westinghouse's accident analysis computer codes, including MCOBRA/ TRAC, NOTRUMP, and LOFTRAN/LOFTTR2. SPES-2 is a 1:395 volume-scale, full-height, high-pressure integral test facility, which includes representation of all AP600 safety systems and key non-safety systems, plus actuation logic for these systems based on the AP600 control systems. The 1 procedure for Test #12 called for safety systems only to respond to the MSLB.

Consistent with Westinghouse's analysis of this event in the AP600 SSAR, no single active failure was simulated in the test. -

Tw) unsuccessful attempts were made to perform the test on October 5 and 7, 1994; the test was completed successfully on October 11. I monitored test activities from October 5-11. I was joined by Mr. Gabriele Del Nero and Mr.

Giuseppe Marella of ANPA on October 5-7. On October 11, members of the SPES-2 quality assurance (QA) inspection team also observed the test.

My report on SPES-2 Test #3 contains details of SPES-2 and Westinghouse personnel and facility operation responsibilities. Since those personnel and responsibilities are essentially the same for Test #12, that information will not be repeated here. In addition, since considerable time during the earlier visit was spent reviewing facility operating procedures, QA, and other related  ;

,E1 test activities, it was felt to be unnecessary to repeat those reviews for I this test. Since the test was " blind," there was also no detailed post-test F data review. Therefore, this report concentrates almost exclusively on day- E of-test observations of the two unsuccessful attempts and of the final, ' -

successful test run.

l Any questions on this report should be directed to Alan Levin, at 504-2890.

DO 00003 A ,

PDR ,

ATTACHMENT 1 AP600 SPES-2 TEST PROGRAM - MATRIX TEST #12 TESTING REPORT Introduction This test was originally the final test in the SPES-2 high pressure, full-height integral systems test facility at SIET Laboratories in Piacenza, Italy.

Westinghouse will, however, perform at least two additional tests in SPES-2 as part of the AP600 testing program. The test simulated a large main steam line break (MSLB) in the AP600 at full system pressure and temperature, but with no power to the heater rods after accident initiation (hot, zero power conditions), without a single active failure, consistent with Westinghouse's SSAR analysis of this event. In addition, the SPES-2 facility was modified slightly to permit the use of all three of the passive residual heat removal (PRHR) heat exchanger tubes, rather than the single tube used for all previous tests, to maximize facility cooldown, which is also consistent with the SSAR analysis. It is also important to note that the heat loss compensation employed for previous tests was not used in this test.

Two unsuccessful attempts were made to perform the test on October 5 and 7.

The test was run to completion on October 11, with preliminary indications that it was successful in achieving its objectives. All three attempts are described in detail below.

A description of the test and procedures is provided in Attachment 2. The testing on October 5 and 7 was monitored by Alan Levin, SRXB lead AP600 test reviewer. The NRC monitor was accompanied by Mr. Giuseppe Marella and Mr.

Gabriele Del Nero, ANPA, as part of the cooperative agreement for test monitoring between the NRC and ANPA. On October 11, the test was observed by Dr. Levin and by all of the members of the quality assurance (QA) inspection team at SIET for the QA inspection of the SPES-2 Program. The remainder of this report focuses on test performance, with some general observations on test conduct.

Test Performance The staff at SIET has remained the same throughout the SPES-2 program. O.

Vescovi is responsible for control room operations, with Dr. C. Medich having chief oversight responsibilities. As has been the case in previous tests, the control room staff operated smoothly throughout test operations, even when things went wrong. The professionalism and competence of the SIET staff was clear throughout the program.

The test was run under conditions similar to those assumed in Westinghouse's l SSAR analysis. The facility was brought to full operating pressure and temperature using relatively low power, in contrast to previous tests in which full scaled power was required for test performance. The power was brought l immediately to zero at test initiation, to simulate the SSAR assumptions of hot, zero power reactor conditions. No heat loss compensation was used to J make up the estimated facility environmental heat loss of 150 kW. The other significant step taken to maximize facility cooldown rate and resultant primary coolant shrinkage was simulation of full passive residual heat removal (PRHR) capability, by use of all 3 heat exchanger tubes rather than the single tube employed in previous tests. In addition, because SPES does not exactly simulate the AP600 control logic, the reactor trip ("R") and safety injection

("S") signals were triggered manually within one second after the break was opened. This is close to SSAR conditions, which essentially assume that full i

PRHR capacity comes on-line at t=0. Two aspects of the SPES-2 test did not replicate SSAR conditions: availability of start-up feedwater (which helps increase cooldown rate) was not simulated; and the break represented was not equivalent to the full double-ended guillotine rupture calculated in the SSAR.

As a consequence, while the SSAR calculation of this event shows primary pressure dropping well below accumulator initiation pressure (approximately 4.9 MPa) before recovering, the pre-test prediction for this test, performed by Ansaldo using RELAP5/ MOD 3 and carried out to 2000 seconds, shows primary pressure decreasing to about the intact steam generator (SG) pressure  ;

(approximately 6.9 MPa for this test), and leveling off, with flashing of the l primary fluid in the SG tubes helping to maintain primary pressure at that j level. Pressure in both the primary and secondary sides then drops slowly to just over 5 MPa by 2000 seconds due to facility heat losses and PRHR cooling.

This sequence of events was onsidered to be " conservative" by Westinghouse, since it maximizes the reliance on the core makeup tanks to provide make-up inventory, without accumulator injection.

The first attempt to run the test, on October 5, was aborted before test l initiation conditions were reached, due to a leak in one of the primary I coolant pumps. The leaking seal was repaired on October 6, and the second I attempt to run the test began satisfactorily on October 7. However, l approximately 10 minutes after test initiation, a leak developed in at least  !

one of the simulated fuel rods in the loop, with water entering the heater and flashing out the bottom of the rod near the electrical connection. This effectively created a combined MSLB and primary-side small-break LOCA. Since the test was at zero power, the main result of the heater leak was to help decrease primary pressure very slowly, until it dropped below the intact SG pressure and, eventually, to the accumulator pressure. The accumulators were allowed to inject water into the system for a brief period, and the test was then terminated. The CMTs recirculated throughout the test, and did not drain. While the test did not reveal any unexpected potential AP600 safety problems in such an event, it did not, of course, meet Westinghouse's test acceptance criteria, and a third attempt was required. SIET plugged the bottom of the leaking heater rod, and in the process of completing repairs, found two additional leaking rods. These were also plugged. SIET personnel believe that the leaks were caused by cyclic stress (due to multiple tests) on the welds in the heaters. The loop could still be brought to test initiation conditions for both the MSLB and simulated full power tests, but test personnel were concerned that if the heaters were damaged sufficiently to cause the tubes to become severed, the unrestrained ends might contact other rods, causing short circuits. However, this apparently did not occur in subsequent tests.

The third attempt to perform the test occurred on October 11. The SPES-2 QA inspection began on that date, and, in addition to Alan Levin, the rest of the inspection team, including NRR representatives Richard McIntyre, Uldis Potapovs, Robert Latta, David Tang, and RES representative David Bessette, observed the test.

Test results were consistent with those from the second attempt (prior to heater leakage), and also tracked very close to Ansaldo's pre-test prediction.

Pressure coasted down slowly, with CMTs recirculating throughout the test. No drop in CMT level was evident (although the indicated level decreased by about 0.2 m due to heat-up and resultant decrease in fluid density). By about 2000 seconds, primary and secondary pressures were almost equal at about 5 MPa.

The test was carried on to more than 2500 seconds, with the primary pressure reaching that of the accumulator at about 2300 seconds. The accumulator began

injecting very slowly at that point, but did not cut off CMT recirculation.

The inventory from the accumulators increased primary vessel level to the point where the pressurizer refilled to approximately 0.7 m. The test was terminated at that time.

Since this test was one of the " blind" SPES-2 tests, performed to test the -

ability of Westinghouse's codes to model thermal-hydraulic behavior (in this case, the code is LOFTRAN), there was not a detailed post-test data review.

SIET verified that pre-test-initiation conditions were within acceptance limits for virtually all parameters, and no obvious events occurred during the test that would appear to compromise its acceptability. On that basis, the I test appears to have achieved its objectives. The increase in PRHR cooling l was evident during the test, but no condensation events or water hammer were noted in the upper plenum / upper head region. Since ADS actuation appears to be a trigger of such events, and no ADS actuation occurred in the test, this result is not unexpected. Some condensation-related flow oscillations were expected in the PRHR tubes, however, once flashing of primary fluid occurred.

These types of oscillations have been noted in previous SPES-2 tests (e.g.,

steam generator tube rupture), as well. More detailed review of the data will be necessary to determine if other significant thermal-hydraulic phenomena, such as cold-leg thermal stratification, can be discerned.

Other Observations (

My previous test monitoring report, for Matrix Test #3 in February, detailed specific observations about the performance of the SPES-2 operating staff. As previously noted, the staff has, in general, conducted the SPES-2 program with a high degree of professionalism and attention to detail. It is interesting to note, however, that control room housekeeping has not improved significantly since the first test, despite specific observations, which were transmitted to Westinghouse, regarding the operating staff placing objects (including partly-full beverage cans) on the facility control panel, although there are signs on the control panel prohibiting this behavior. Apparently, this manner of operating is firmly entrenched at SIET, and is difficult to change. While a mishap involving these objects on the control panel could threaten conduct of the test, this has apparently not occurred.

The QA inspection continued through the week.at SIET, and concluded on Friday, October 14. The results of the inspection will be documented in a separate report.

Conclusions Matrix Test #12 in the AP600 SPES-2 program appears to have been performed successfully, and to have achieved its objectives. Further observations regarding the detailed thermal-hydraulic behavior occurring during the test can only be made after detailed review of the test data.