ML19281A966
| ML19281A966 | |
| Person / Time | |
|---|---|
| Issue date: | 03/28/1979 |
| From: | Murley T NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Mattson R Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7904170340 | |
| Download: ML19281A966 (6) | |
Text
..
~
~
1 P00R quAUTY PAGES 3 L
MAR 2 81979 Distributica Subj Chron Circ Branch R/F TEMurley MEMORANDUM FOR: Roger J. Mattson, Director LSTong WDLanning Division of Systems Safety CEJohnson WDLanning R/F Office of Nuclear Reactor Regulation VStello
-PE LRubenstein ACRS FRCH:
Thomas E. Murley, Director ZRRosztoczy VStello Division of Reactor Safety Research RLTedesco Office of Nuclear Regulatory Research
SUBJECT:
ANOMALIES BETWEEN SEMISCALE TEST S-07-6 RESULTS AND ANALYTICAL PREDICTIONS The purpose of tus memorandun is to sumarize the Semiscale system behavior based on the results of Test S-07-6 and discuss the differences between the RELAP4 prediction'and the data. Results of additional tests and post-test calculatiJns are also presented which have contributed to further understanding of tiie system behavior and explaining the l
anomalies between the experimental data and analytical pretest calculations.
l
Background
Semiscale Test S-07-6 was the~ sixth test conducted in the Semiscale MOD-3 baseline test series and was the first integral blowdown reflood test conducted in the H0D-3 system.? The MOD-3 configuration is a one-dimensional non-nuclear representation of the thermal hydraulic behavior of a pressurized water reactor. The nuclear core was simulated by 25 twolve-foot long electrically heated rods. The downcomer annulus was simulated by a single pipe ext'ernal to the vessel. The primary differences between the present MOD-3 configuration and the previous MOD-1 configuration j
are the external downcomer, full length core sumulator UHI vessel internals and active steam generator and coolant pump in the broken loop.
The external downcomer design was chosen to permit extensive density measurements in the do'wncomer and vessel. The other changes were made to convert from the.' LOFT configuration to a representation of
~
a PWR with Uit! amergency core, cooling system.
Although the facility was desfgrad to simulate ECCS/UHI performance during a loss-of-coolant experiment. Test S-07-6 was intended for a system check-out test with ECC injection only into the cold leg. This test was to provide referener' data which would peinit an evaluation of the integral blowdowr..e1@d behavior of the MOD-3 system during a 7904y7O p
Contact:
Wayne D. Lanning, RSR/SEB 42-74260 6,me, save >
/
s.
R. J. Mattson double-ended cold leg break loss-of-coolant experiment. The test results would also be compared to Test S-04-6 data to provide an indication of the effects of differences between the MOD-3 and M00-1 system configuration on the resulting system thennal hydraulic behavior.
Prior to completing the integral test, separate effects tests of blowdown /
refill and reflood were completed. The system thennal hydraulic behavior was as expected and reasonable comparisons were obtained with the pretest calculations, corresponding M00-1 tests, and a FLECHT SET test.
Comparison of predictions To Data, Results of the data comparison to the pretest predictions perfonned with the RELAp4/ MOD-6 code indicated that the reflood behavior of Test S-07-6 was considerably different than expected. The reflood portion of the experiment was characterized by a prol6nged period in which refill and emptying of the downcomer occurred three times with corresponding relatively high sequentially decreasing oscillatory heater rod cladding temperatures.
The quench time at the peak power location was approximately 300 seconds longer than expected. The results from the Semiscale test are significant in the sense that the analytical models used did not predict downcomer voiding after downcomer refill and the oscillatory heater rod dryouts and rewets during the reflood period.
The chronology of the test is presented in Table 1.
The times at which the downcomer level decreases are shown in the accompanying ccmposite figure of measured key parameters. The important phenomena occurring during the test are summarized as follows:
(
1.
The fluid 1~evel initially decreased in the downcomer after
- ~
refill as the result of excessive metal heat transfer from the walls to the fluid. The fluid vaporized in the lower part of the downcomer forcing most of the mass inventory out the break. The temperature difference between the wall and fluid was sufficient to pmduce this local vaporization. The small diameter of the pipe accentuated the effects of steam generation and dcwncomer countercurrent flow phenomena.
2.
The lack of constant adequate fluid driving head in the downcomer prevented sustained reflood of the core which resulted in oscillatory clad temperatures.
' " * * " " * * " " " * " * " * " " * * * * * ~ *. " "
~ ? m m"* **"
NU
........m.............
......~....u......................*m."***"
~~ " = * = ~ ~
.. ~... " " ~ ~ ~ ~ ~ ~
~""~~~~ ~~"
ewenaAame >
........ ~. - ~ ~ ~.
~ ~ " " " " " " " " "
" " * * - ~ ~ ~ ~
eave >
t $
A
e h
I R. J. Mattson 3.
Successive emptying of the downcomer after refill by the low pressure injection system is the result of a combination of downcomer heat transfer which vaporizes the water from the hot wall and steam backflow from the core.
Figure 1-B shows that the fluid in the lower plenum reaches saturation temperature when the downcomer level decreases which indicates steam'ficw from the core. The flow instrumentation in the downcomer also indicate steam upflow at the time of mass depletion. Based on the results of a repeat Test (S-B7-6) of S-07-6 where the downcomer heat transfer was minimal after 200 seconds, steam backflow from the vessel and the resulting countercurrent flow prevented refill of the downcomer after mass depletion occurred.
Figure 1-C shows the steam generation rate from the vessel for
"~
3 Tests S-07-6 and S-B7-6. This figure indicates significant amounts of steam flow during the mass depletion periods which is unique to Semiscale and results from excessive core shroud ~
heat transfer.
Based on the analyses of the test data, subsequent tests and analytical evaluations, it has been concluded that the mass depletion phenomenon in the downcomer was primarily the result of excessive metal heat transfer from the downcomer walls to the injected ECC fluid for Test S-07-6.
The analyses of the differences between the experimental data and pretest predictions concerning the behavior of the Semiscale M00-3 system indicate that the wall heat transfer in the downcomer and vessel structure was not adequately modeled in the RELAP4 calculations. Changes in the modeling of the downcomer and countercurrent flow assumptions have resulted in significantly improved comparisons between the post-test calculations and data. The calculation was completed to 260 seconds after rupture and showed similar oscillatory downcomer hydraulic behavior and peak clad temperatures as the experimental data. The calculation was not extended past 260 seconds because of water packing problems and the assoc.iated excessive computational costs required to complete the calculation. Analyses performed with an advanced code (i.e. TRAC) are not expected to experience the water packing problem.
preliminary results from a TRAC calculation have been presented to your staff. The results appear promising since 2 decrease in the downcomer level was calculated. The results of the calculation Will be made avaliable to you in April.
- ===*
--a==>
A~
k'
/
ss R. J. Mattson Future Plans An improved downcomer insulator vill be installed in Semiscale in April 1979. Tests of the honeycomb insulation by the manufacturer indicated that the metal heat transfer will be significantly reduced. The design of the insulator should also permit an improved analytical description of the heat transfer in the downcomer. Two lower plenum injection tests will be completed to further confim that the hydraulic oscillations are the result of excessive heat transfer in the downcomer.
fio additional cold leg injection tests will be conducted in Semiscale until either the new downcomer insulator has been installed or an accurate analytical prediction of Test S-07-6 has been obtained.
Analysis efforts will continue until the differences between the calculations and data are resolved.
' original Signed bi T. E. Marley Thomas E. Murley, Director Division of Reactor Safety Research Office of fluclear Regulatory Research
Enclosures:
1.
Chronology of.S-07-6 Test 2.
Test Results t
DJ' RSR:$
RSR:W:SE N:W R6L RES RES m.c.,
WD[MAM1Tg/ca ilWSerkiz Johnson / Tong TE.iurley 3f)$'79
.3A7/79 3Lul/79 3/>>/79
- v.,
Chronology of Semiscale Test S-07-6 TIME (SEC)
EVENT _
0 Double-end break initiated.
3 High pressure safety injection system actuates.
19 Accumulator actuates.
30 Low pressure safety injection system actuates.
40 Flow reverses in broken loop cold leg. Condensation effects induce steam flow from supression tank as downcomer is filled (Figure 1-D).
47 End of ECC bypass.
51 Lower plenum refill begins.
60 Core reflood initiation.
65 Initiation of nitrogen flow from accumulator.
70 Flow in intact hot leg stops decreasing due to increase in core level (Figure 1-C). Minimum flow out broken cold leg as downcomer refills.
75 Initial mass depletion in downccrner. Significant downcomer heat transfer from walls to fluid.
(Figure 1-A).
96 Nitrogen flow from accumulator tenninates.
11 0 Fluid at top of lower plenum reaches saturation temperature for first time (Figure 1-B).
160, 270, 380 Minimum mass inventory in downcomer and core.
Clad tenperature peaks (Figure 1-A).
RELAP4' predicted hot sp(ot quench. intact hot leg) indic 180 Peak flow out of core 100, 190, 280, 390 steam generation in core to cause steam upflow in downcomer (Figure 1-C).
200, 290, 450 Maximum outward flow through broken loop cold leg as result of steam upflow in downcomer inhibiting refill by LPSI and flow from intact cold leg (Figure 1-D).
210, 320 Fluid in lower plenum reaches saturation temperature indicating steam flow' from vessel (Figure 1-B).
e Level in downcomer and vessel decreases (Figure 1-A).
275, 390 Condensation induced inward flow from the break (Figure 1-D) 450 End of data acquisition.
550 Hot spot quenched (recorded by strip charts).
3 1
I I
I i
i I
i 1
,1000 8:3 TD1ERATURE cy TN-C2-80 E
_m
=
55 FIGuRt i.4 g
r 500 t
C R
-'s, f
10WMC2 TEA LEVEL W
C
$ g S b i
8% 400
=
1e o W l1 M $
8 i
fa 200 i1 0
ORE LEYEL i
di o
d'
$50 TIMES OF DOWNCOMER l' ASS DEPLETION a
y 500 U
FIGURE 1-8 450 SATURAi!ON TE ERATUIE l 400 w
w FLUID TDPIRATURE AT TCP OF LOWER PLENm_
1 K
5-01-6 0
~
~
-f.
t '.
t
~
f
,4 o
\\
~
~
~
i g
5-81-6 y 20 FIGURE l-C E
..r.
0 FIGURE 1 h
TEST RESULTS
.,0 -
t we FIGURE i-D 3
L o5 30 a
g h
g o.0 l
V Ee.x k
i I
I I
1 I
I f
I
-50 0
50 100 150 200 250 300 358 400 450 500 TIMF ArftR RUPTutt ($rt)