ML19256D106
| ML19256D106 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 06/04/1969 |
| From: | Ross D US ATOMIC ENERGY COMMISSION (AEC) |
| To: | Long C US ATOMIC ENERGY COMMISSION (AEC) |
| References | |
| NUDOCS 7910170552 | |
| Download: ML19256D106 (6) | |
Text
.... _ _ _. _
~.
e$ 4,i UNITED STATES D
ATOMIC ENERGY COMMISSION SI* O b ')*
WASMNGToN. O.C.
20545
\\gs June 4, 1969 Charles G. Long, Chief, Reactor Projects Branch 3 Division of Reactor Licensing REVIEW 0F MET-ED FAN COOLER REPORT (THREE MILE ISLAND LGIT 1)
DOCKET NO. 50-289 Attached is a su==ary of a description of the fan cooler tests for Three Mile Island Unit 1.
My evaluation of each step is attached.
In summary the report is quite incomplete and shallow and will require considerable amplification.
The Met-Ed report has been circulated to RT.
Concerned RT personnel are requested to review the report and report their conclusions to RP.
't G
N D. F. Ross Reactor Projects Branch 3 Division of Reactor Licensing cc:
R. S. Boyd B. K. Grimes R. C. DeYoung R. R. Powell A. Schwencer A. W.'Droscrick M. Rosen V. A. Moore Orig: DFRoss (3)
PSB-3 Reading Docket /
khkb b0
- 7910170 9'7 2
/T
REVIEW OF REACTOR BUILDING FAN ASSEMBLIES FACTORY TESTS FOR THREE MILE ISLAND UNIT 1 A preliminary report covering tests of the reactor building fan coolers for IMI No. 1 was filed on May 27, 1969. It was noted in the accompanying letter that final reports on the emergency cooling coil and relief valves will be issued in July 1969. The fan and motor test report will not be available until January 1970, the approximate FSAR date.
The American Air Filter Company (AAF) is the prime contractor for the f an assemblies. The preliminary report, just filed, was prepared by AAF.
It contains outline information on the tests for the emergeticy cooling coils, the fan-assembly-housing relief valves, the GE motor, the Crane.aotor seals, the motor insulation resistance-to-radiation, and the fan assembly. The proposed tes ts are summarized below.
I.
COOLING COILS A.
Test Conditicns Steam-Air at a.
281 F, 68.3 psia b.
244 F, 44.7 psia c.
184 F, 24.7 psia B.
Me thod Full-scale testing will not be performed. The f ace length will be 24 inches,
instead of 45 inches. However the following test parameters will be identical to expected accident-values:
face height (16 tubes, or 24 inches) face air velocity (240 fpm) coil depth (8 rows) coil coolant velocity (5.5 fps) coolant flow rate (59.3 gpm) materials (5/8-inch OD copper tube, 0.049 wall, 0.007-inch copper fins)
A steam-air mixture will be circulated past the test coils in a closed-cycle arrangement. Wet and dry bulb vapor-mixture temperatures and flow rates will be measured before and af ter the cooling coils. Cooling water flow rate and temperature will be measured. Flow rate will be measured by a timed (s topwatch) flow into a collection tank, on scales.
)h4
. C.
Calculations Formulas are presented for calculating vapor mixture density as a function of partial pressures and mixture densities.
If heat balances compare
" reasonably," 1.e. air-side vs water-side, the test data will be marked and p rese rved. O the rwis e, the data will not be preserved.
D.
RP Evaluation Apparently each full-size f an cooler will have 128 tubes (16 high by 8 deep).
Then 128 tubes at 59.3 gpm per tube will result in 7600 gpm total flow per unit. The PSAR says 1780 gpm (page 6-13, Volume I). If the 7600 gpm value is correct, there will be an adverse impact on diesel loading due to a larger pump and motor.
There is an inconsistency in the tube velocity. A 0.625-inch tube, with a 0.049-inch wall has an ID of 0.527 inch$;and an internal area of (0.785)
(0.527)2 = 0.219 inch 2 = 1.52 x 10-3 ft the flow rate per tube is 59.3 gpm or f f ft 3
3/sec = 0.1316 ft /see the velocity is V = Q/A =
= 86.5
-3 1.52 x 10 fps; and 86.5 fps is 15.7 times the quoted value of 5.5 fps. Possibly the 59.3 gpm value applies to one bank of 16 tubes high.
The absence of definite criteria for test acceptance is disturbing. There should be statements about test replication. The coolant fios measurement system seems unnecessarily crude; stopwatches sometimes stop when they shouldn't. Also, the flow measurement is at best an integrated value.
There is a statement on page 5 that due to the likelihood that any one test run will not be successful it is impossible to prepublish a rigid test schedule. We need some forecast of events if CO or even the applicant or the A/E is to witness these tests.
II.
RELIEF VALVES A.
Description Relief valves are to be provided on the housing to open at 1 psid, and close
< 1 psid. Tests will be done at maximum design dip values, but at ambient air temperature. Design of the housing is to withstand 2 psid.
Static tests will be performed to determine relief valve flow characteristics. Two valve models will be tes ted: one full-size, and one smaller, for dynamic analysis.
ikh 1AAo/
. If a computer program (to be written), using data from the static tests, shows that the prototype valve meets system requirements, then only the smaller valve will be tested dynamically.
For the dynamic test, air at three atmospheres will be released to provide a test simulation. Capability for simulating the finite volume of the housing is provided.
B.
RP Evaluation It is not clear what basis exists to warrant full-scale vs part-scale dynamic tes ts. Three atmospheres of air (45 psia) does not simulate acci-dent conditions. Upstream and downstream pressures will be recorded on oscillosecpes, presumsuly operating in a single-trace mode with a synchro-nized Polaroid camera.
If the scale is say 20 ps{/cm, then there is no way to detect 1-2 psid by scaling the photograph. A & P transducer should be used to magnify the difference, as 2 psid is the design value on the f an housing.
III. GE MOTOR A.
Description Bill of Material on motor includes:
a.
Class H insulation b.
150/75 HP (2 speed) c.
Totally-enclosed, water cooled d.
Accident conditions are 108 rads I on design life 2800 F for 3-4 hours 100% R.H. plus water droplets slightly caustic 70 psia e.
Special grease suitable for radiation and temperature f.
Shaf t seal has two 0-rings Data were furnished by Crane Packing Company on the radiation resistance of the 2-1/4 inch Type 1 seal with Cranelast bellows. The seal was operated for 500 hours0.00579 days <br />0.139 hours <br />8.267196e-4 weeks <br />1.9025e-4 months <br /> with a total dose of 1.1 x 108 rad from Ov-60.
A reprint of KAPL-P-ll64 " Testing Electrical Insulation for Use in Gamma-Ray Fields," C. Mannal, June 1954 was presented.
General Electric specifications were presented. Again Class H insulation is mentioned, as well as a "special grease" for radiation and temperature. The seal design will be tested for the pressure, temperature, humidity, and radiation for the pos t-accident environ =ent.
F, 1 a
f 0
!J I
- A proposed qualification test is described. The motor will circulate air in a closed cycle and will heat air by f riction losses to 300 F.
Water will be injected to achieve 100% R.H.
The duct will be at 70-80 psia. The tes t will run 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. Then conditions will be reduced to 20 psia, 1500 F, 100%
R.H. and 3 more hours of f an-motor operation performed. Measurements will include motor temperature, air temperature, pressure, R.H., and voltage,
current, power to motor.
After the first test the motor will be disassembled, inspected, reassembled, and recycled through a second accident test.
The motor will be stopped and restarted during this test.
B.
RP Evaluation Accident environment is not well-defined.
Radiation is variously specified as 108 and 109 rad life time. Accident temperature is 281 F or 3000 F.
Water conditions are started as "slightly" caustic; there was no mention of sodium thiosulfate. The same seven pages o# GE specifications were printed in the report at two dif ferent locations. Type H insulation is specified for a totally enclosed motor, without regard for gas generation. It appears that radiation data on insulations should be more recent than 1953; hasn' t the industry developed new materials? What special grease will be used, and what data support its use? The motor test is extremely simplaminded. I don't see how air-friction will heat the sys tem to 3000 F.
Certainly steam rather than hot water will be needed, as well as duct insulation. How will the duct be pressurized? What is the 0-ring resistance to basic sodium thiosulfate? Why is only 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> accident testing considered sufficient?
IV.
ASSEMBLY DRAWING OF FILTER HOUSE WITH COOLING COILS AND FAN (AAF 1070-832295)
A.
Description The unit is approximately 17 feet wide,12 feet high, 26 feet long, and weighs about 37 tons in operating condition. Housing material is No. 11 gauge hot rolled steel, painted. Nine pressure relief panels are shown.
Inlet air filters Type HV2 are provided.
B.
RP Evaluation It is doubtful that the particulate filters can withstand a 1 psid, and it seems possible that such a pressure " wave" might reach them. The effect of filter failure should be considered.
It is not evident how external water connections are made, but if they come up through a concrete floor, then seismic forces caused by differential i446 152
,. motion may be a concern.
In that regard one wonders how seismic design has been included. The paint is not specified, but its compatibility with alkaline spray solutions should be assured. Water connections to the water-cooled motor are not shcwn. Hopefully it is not a closed cycle, where H2 evolution might occur.
e 1446 153