ML20034H686
| ML20034H686 | |
| Person / Time | |
|---|---|
| Site: | 05200003 |
| Issue date: | 03/15/1993 |
| From: | Hasselberg F Office of Nuclear Reactor Regulation |
| To: | Liparulo N WESTINGHOUSE ELECTRIC COMPANY, DIV OF CBS CORP. |
| References | |
| NUDOCS 9303190230 | |
| Download: ML20034H686 (6) | |
Text
- - - - - -...
i Docket'.No.52-003' March 15, 1993 Mr. Nicholas J. Liparulo, Manager emp3"<'~'
""""~'~*3 Nuclear Safety and Regulatory Activities IWi I E 'i"i Westinghouse Electric Corporation
' Box 355 Pittsburgh, Pennsylvania 15230
Dear Mr. Liparulo:
SUBJECT:
EVALUATION OF SCALING ANALYSIS FOR AP600 PASSIVE CONTAINMENT COOLING SYSTEM (PCCS) LARGE SCALE TEST PROGRAM The Nuclear Regulatory Commission (NRC) staff and contractcr personnel are continuing their review of Westinghouse's scaling analysis and instrumentation for the AP600 PCCS large scale test program. Although our evaluation is not complete at this time, we have identified some preliminary issues and concerns regarding your large scale containment test program that we hope to discuss with Westinghouse during our March 23 and 24, 1993, meeting on containment testing.
+ contains a listing of these preliminary concerns.
contains two detailed contractor reports and a set of presentation slides that explain these issues in more detail. We are treating Enclosure 2 as Westinghouse-proprietary material.
Please note that both Enclosures I and 2 contain contractor conclusions and recommendations which have not been endorsed by the NRC and do not constitute the NRC staff's position at this time.
We request that Westinghouse review this material prior to our discussions on March 23 and 24, 1993.
Sincerely, (Original signed by)
Frederick W. Hasselberg, Project Manager Standardization Project Directorate Associate Directorate for Advanced Reactors and License Renewal Office of Nuclear Reactor Regulation e
Enclosures:
As stated f
cc: See next page
~
~~
DISTRIBUTION w/ enclosure:
Docket File PDST R/F FHasselberg TKenyon DISTRIBUTION w/o enclosure:
PDR DCrutchfield WTravers PShea RBorchardt
,1 Moore, 15B18 GGrant, 17G 1 ACRS (11)
/
OFC:
LA:PDST: AR PM:P R
SC:
T:ADAR NAME: PShea FHass Iberg:tz RB ardt 3
03/6/93 03 93
[
DATE: 03q OFFICIAL RECORD COPY: CONTTEST.RH G
y s
Mi 930319o23o 93o3 S W 4 d: s T,. -Ni "'*" W "2;kdi 4
[DR ADOCK 0520 3
g;h d%.
/
IIS
/Mr LNicholasLJ. Liparulo Westinghouse. Electric Corporation
. Docket'No.52-003 '
AP600 l
~
Mr.'BiA.McIntyre (w/ enclosures)
.cc:'
Advanced Plant Safety & Licensing
- Westinghcuse Electric Corporation Energy Systems _ Business Unit Box 355 Pittsburgh, Pennsylvania 15230 Mr. John C.- Butler (w/ enclosures)
' Advanced Plant Safety & Licensing
'4 Westinghouse Electric Corporation Energy Systems Business Unit
-Box 355
'Pittsburgh, Pennsylvania Mr..M. D. Beaumont-(w/ enclosures)
Nuclear and Advanced Technology Division Westinghouse Electric Corporation One Montrose Metro 11921-Rockville Pike' Suite 450-
' Rockville, - Maryland 20852 Mr. Sterling Franks-(w/o enclosures)
U. S._ Department.of Energy NE-42
' Washington,~D.C.' 20585-
'Mr. S. M. Modro'
'(w/o enclosures)
EG&G Idaho Inc.
Post Office Box 1625 Idaho falls, Idaho 33415 Mr; Steve Goldberg (w/o enclosures).
Budget Examiner 725-17th' Street,'N.W.
' Room 8002 Washington, D.C.
20503 Mr. Frank'Ros's (w/o' enclosures)
U.S. Department of Energy.
NE-42 Washington, D.C.
20585 f '
P L
AP600 Testing Program and Scaling Analysis 1.
Heat Transfer to Passive containment Cooling System (PCCS) in Transient t
Tests t
flow rates in transient tests are not representative of prototypic transients. The residence time constants in the test facility should be decreased (for example, by increasing the flow rate) if the blowdown phase is to be simulated.
For the post-blowdown phase, the residence time constants should be increased (See Table 3 of ERI report ERI/NRC 92-1222).
2.
Heat Transfer to Internal Structures and Mixing in the Containment Thermal time constants for the internal steel structures and the external steel shell are of the same order of magnitude for the test facility and-the AP600 prototype (see Table 5 of ERI report ERI/NRC 92-1222). The lack of internal concrete structures within the 1/8-scale facility is a limitation.
Simulation of concrete structures are important for two l
reasons (a) Natural circulation and mixing processes in the containment (b)
Partitioning of enerCy between the PCCS and the internal structures Hence, it is recommended that:
(i)
The simulation of concrete structures be included in the test-facility.
(ii) The geometry of flow paths in the containment be scaled.
3.
Turbulent Buoyant Jet Discharge of Hydrogen / Steam The buoyant jet discharge and mixing with atmosphere can be scaled by two parameters, i.e., the jet Froude number and the jet to containment density ratio (defined in ERI Report ERI/NRC 92-1222, Section 2.8).
It is possible to obtain similarity between the 1/8-scale _ tests and the AP600 prototype during the late phase of the accident. Westinghouse should utilize the scaling ratios mentioned above to properly scale the steam / helium mass flow rates in the tests.
4.
Natural Circulation of Air in the PCCS Concerns exist over the difference in the annulus design between the AP600 prototype and the 1/8-scale test facility, i.e., the riser and 1
f f
downcomer sections of AP600 prototype are separated by an air baffle.
Only the riser section is simulated in the 1/8-scale test facility with an ait baffle surrounding the containment.
The ability of the AP600 to sustain air cooling under natural circulation has not been established.
In addition, (a) the air circulation in the rounter-current, two-annuli AP600 prototype could be affected by heat conduction through the baffle o
wall, and (b) there is a potential for flow reversal and instability which could lead to degradation of the PCCS performance, especially under dry operating conditions (see sample simulation results in Figures A.2 -
A.5 in ERI Report ERI/NRC 93-202).
l The above concerns should be addressed by Westinghouse since natural circulation of air is the only heat removal mechanism for the dry containaent case. Degradation of the PCCS performance is also of concern for the case where the containment is striped or partially wet.
5.
Mechanistic Correlations in MGOTHIC l
Westinghouse has not validated some of the important. correlations used in the MGOTHIC code using the results of the heated flat plate tests.
The Colburn correlation for film heat and mass transfer is not validated by the heated plate tests (see WCAP-12665, pages 38 and 44).
Applicability of the forced convection correlrtions in the annular gap has not been demonstrated by Westinghouse through either experiments or analysis.
q 6.
MG0THIC Validation Using (Integral and Large Scale) Test Data In one of the small tests (Figure 22, page 109' of WCAP-13246), velocity fluctuations near the vessel wall indicate possibility of numerical instabilities.
In the MG0THIC computer code, the effects of subcooling are not modelled.
To account for film subcooling, wall ~ temperature data from the experiments are used in the code, and the Uchida correlation is used to model condensation on the inside of the vessel. This is not a sound modelling approach for the calculation of transients in the prototype.
The results from the MG0THIC code predictions for the small scale tests indicate that considerable stratification of steam / air mixture in the containment has occurred (Figure 27, page 114 of WCAP-13246). However, the concentration of steam in the niixture has not been measured to l
compare to the predictions. The level of accuracy of these predictions
J c
cannot be assessed. The effect of stratification upon condensation of i
steam in the containment can strongly influence the heat removal capacity of the system.
7.
Westinghouse Scaling Approach l
Westinghouse should perform a detailed scaling analysis to identify the major physical processes and dimensional scales. For exantple, the lengthscale for natural convection in the annulus is the gap width, and not the height of the containment. The effects of the jet mixing'can be scaled by the jet Froude number and density ratio (ambient to jet).
8.
1/8-Scale Test Matrix (a) Only two film flow / free convection tests have been performed; none are planned in the future.
(b) No dry, free convection tests will be performed.
1 (c) No striped / partially wet, free convection tests will be performed.
l (d) The two conditions described above could lead to high pressures (above design pressures) in the AP600 prototype.
(e) Only one test planned (test 10) to simulate severe accident i'
conditions.
(f) The purpose of tests 11 and 12 is stated to be the simulation of blowdown. However, no steam inlet mass flow rates are given. The objective and the method of performance of these tests must be-clearly stated.
9.
1/8-Scale Facility Instrumentation (a) The current off-site sampling method for the measurement of the air-and steam concentrations will not be adequate during transient tests, since the conditions in the test vessel will change substantially during the course of sampling.
(b) More anemometers in the annular region are recommended to adequately -
resolve the flow behavior, since the air flow pattern in the annular gap under natural circulation conditions is expected-to become l
complicated.
(c) For the planned severe accident tests, water film dryout could occur at high-vessel pressures. Hence, Westinghouse'should consider additional instrumentation for the measurements of the water film thickness in both severe accident tests and tests involving striped film flows.
l i
i
?
(d) Westinghouse should prove convincingly that the measured temperature profiles in the steel shell wall are linear at all locations (where temperatures are measured). Nonlinear temperature profiles in the shell wall can occur under conditions of large axial heat flux variations (such as those expected in severe accident. tests). This can be achieved by simply adding another thermocouple at a radial location midway between each pair of thermocouples used to measure heat fluxes.
- 10. MGOTHIC Numerics
{
(a) There is no discussion in WCAP-13246 of any nodalization studies to demonstrate the robustness of the MGOTHIC code and its sensitivity to the choice of grid size and the number of computational nodes.
There is also no discussion of time step size. How was it chosen and its relationship to the Courant condition? These results will improve our confidence of the code capability when it is used for full-plant transient calculations.
(b) The reason and justification for the 1/4 symmetry assumption in the-HG0THIC computer code is not clearly stateo.
b
[
9 4
r i
f 9
I
,