ML20198C344
| ML20198C344 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 11/08/1985 |
| From: | Standerfer F GENERAL PUBLIC UTILITIES CORP. |
| To: | Travers W Office of Nuclear Reactor Regulation |
| References | |
| 0355A, 355A, 4410-85-L-0227, 4410-85-L-227, NUDOCS 8511120059 | |
| Download: ML20198C344 (9) | |
Text
s
)
GPU Nuclear Corporation NggIg7 Post Office Box 480 Route 441 South Middletown, Pennsylvania 17057 0191 717 944 7621 TELEX 84-2386 Writer's Direct Dial Nurnber:
(717) 948-8461 4410-85-L-0227 Document ID 0355A m
??
Novembe'pl8, 1985 CD E!E ap n
90 TMI Program Office
[
y Attn: Dr. W. D. Travers O
E Acting Director to US Nuclear Regulatory Commission c/o Three Mile Island Nuclear Station Middletown, PA 17057
Dear Dr. Travers:
Three Mile Island Nuclear Station, Unit 2 (TMI-2)
Operating License tb. DPR-73 Docket No. 50-320 Quality Assurance Issues Relating to the Defueling Canisters Your letter, dated October 17, 1985, provided comments on quality assurance issues related to the defueling canisters. GPU Nuclear letter 4410-85-L-0210, dated October 28, 1985, provided our response to your comments with one exception; information concerning the concrete filler material in the fuel canisters was not available. Information relevant to that issue has been received and, accordingly, our response is attached.
Sincerely, 0
4
}, I t
F R. Standerfer Vice President / Director, TMI-2 FRS/RDW/eml Attachment 8511120 0 S')
09
\\\\
GPU Nuclear Corporation is a subsidiary of the General Public Utilities Corporation
~
ATTACHMENT 1-(4410-85-L-0227)
NRC COPMENT 8 Describe your testing program to assure that the cement filler material in the fuel canister will not degrade as a result of 1,mersion in water. In particular, address the effects of immersion on the structural strength of the material and the potential for leaching material from the cement that may act as a contaminant to the RCS, the fuel pool, or the recombiner catalyst.
GPU NUCLEAR RESPONSE Effect on Structural Strength
~ Attachment 2 provides the results of strength tests performed by Babcock and Wilcox (B&W), the defueling canister designer, on the concrete filler material used in fuel canisters. These tests were performed to address the effects of water immersicn on the structural strength of the material. These tests demonstrate that there was a reduction in compressive strength when compared with the original design phase testing.. This reduction may be attributed to a much shorter curing time for the concrete used in the immersion tests.
Furthermore, it should be noted that the design basis of the canister is met with the reduced compressive stength.
Effect on the Reactor Coolant System and the Fuel Pool B&W performed a series of leaching tests on a sample of cured LICON concrete similar.to that used in the fuel canisters, to determine the effects on the Reactor Coolant System (RCS) and the fuel pool. A standard test method (ASTM D3987-81) _ was employed to determine the concentrations of leachable species
- under standard test conditions for leaching periods up to six (6) days. The LICON concrete was first crushed to expose large surface area and approximately 300 grams of rubble were contacted with one (1) liter of demineralized water. An analysis of the leachate for cations, anions, and pH followed. One test was performed with pH adjustef boric acid at 5000 ppmB.
In general, the results confirmed low concentrations of leachable chlorides and sulfate ( 2 ppm each) in the leachate. This corresponds to 7 ppm of each anion per gram of LICON concrete. TMI-2 specifications for leachable chloride, fluorides and sulfate materials in the RCS is 200 ppm. No other anions were detected in any of the leachates using ion chromatography analysis techniques.
The leachates were also analyzed for cationic species by atomic absorption.
Relatively high concentrations of Na, Ca, and Al were measured in the leachate (Na and Al at approximately 200 ppm and Ca at approximately 40 ppm). This is' not surprising since LICON is calcium aluminate, empirically Ca0 x 2.5 Al 02 3 with approximately 0.5 weight % sodium impurities. The relatively high concentrations of cations caused the pH of the demineralized water leachate to be on the order of 11. The high pH probably resulted in more aggressive leaching conditions than would be encountered in the RCS due to the high buffering capacity of the coolant. For example, a 200 ppm increase of
r ATTACHMENT 1 (4410-85-L-0227) sodium concentration in the RCS will change the pH by only approximately 0.1.
Other cations were found in only trace amounts in the leachates. The leachable aluminum and calcium could provide a slight impairment to visibility due to the presence of insoluble hydroxides but should be removed by DWCS filtration. The sodium and calcium may compete for active sites during demineralization which will slightly reduce the throughput; however, it is questionable that it will be a measurable effect.
Assuming five open fuel canisters in the vessel, each containing 350 lbs. of LICON, the leaching tests would suggest that the chloride or sulfate content in the RCS would increase 0.1 ppm and the sodium and aluminum content would increase 5 ppm. Both calculations conservatively assume all the LICON is available for leaching. After the canisters are closed, the leachable ions cannot migrate into the fuel transfer canal or A Spent Fuel Pool.
Thus, as a result of these tests, the leachable material from the LICON should not contaminate the RCS or fuel pool to levels of concern.
Effect on Recombiner Catalyst Rockwell Hanford Operations performed testing for EG&G to determine the effects of the fuel canister cement filler material (LICON) on the TMI-2 mixed-bed catalysts. The sample of LICON concrete provided to Rockwell was pulverized, mixed with water to form a thin slurry, and allowed to stand for a few days. The slurry thickened to a non-pourable paste and was diluted with water to again form a slurry. Slurry was added to the two catalyst beds, in the underside of the test vessel lid, so that the slurry covered the catalyst. The catalyst absorbed water from the concrete slurry, causing it to thicken.
The screens were then placed over the catalyst beds and secured. As the test vessel lid was turned over, some of the thick slurry flowed down from the bed and through the screen; much of it was retained in the bed and screen.
Photographs were taken which determine the extent to which the concrete slurry filled the catalyst bed and screen. The test was conducted under 2 atmospheres of argon gas with 0.3 liter / hour of stiochiometric hydrogen and oxygen flowing into the test system. After 17 hours1.967593e-4 days <br />0.00472 hours <br />2.810847e-5 weeks <br />6.4685e-6 months <br /> of testing, the oxygen concentration was 3.4%, the hydrogen concentration was 6.7%, and the recombination rate was slightly higher than the design rate of 0.11 liter / hour. Since the gas mixture was below the lower flammable limit (5% for oxygen) when the design recombination rate was exceeded, the test showed an acceptable result even under these highly adverse, conservative conditions.
10-19-8513:18_T-BMFD LYNCHB' RG
- 485 PS2 m ATTACHMENT 2 J
(4410-85-L-0227) 6.Pages t.
Babeock & Wilcox e..,r
.,o w.i.n
~
n
- McD*rmott **weat jsgs,ofalgs,tnona Lynchburg.VA e45o6 o935 1
(a04) sa5 20oo BW/BTL-85-80.
.0ctober 10, 1985 Telecopy Mr. R. L. Rider GPUN Design Engineering c/o Bechtel North American Power Corp.
15740 Shady Grove Road Gaithersburg, MD 20877-1454 D
Subject:
LICON Application in Fuel Canisters
Dear Mr. Rider:
The purpose of this letter is to transmit re:;ults of tests run by B&W on LICON and to respond to questions on its application in the fuel canister. The attached letter report addresses questions about the behavior of LICON when saturated with water. The conclusions of the tests reported in the letter show that LICON is an excellent product for the desired application. One factor that is creating an illusion of difference when none may exist is the fact that the original LICON sample were tested after about 8 days of curing whereas the recent samples cured only one day before being submerged and three days before being tested. Although one day curing will set the shape, full strength in concrete increases for about 20-30 days after pouring. The required speed of the test program may be producing lower values than the actual canisters will have. Although the strength of the LICON apparently decreasedwhensaturatedwithwater,thekeydeterminationwasthgeven then reduced strength was sufficient to prevent significant Boral
, shroud deformation.
After the bulkhead is welded to the lower body, the LICON is contained between the shroud, which is seal welded to the bottom support plate, and the shell/ bulkhead. The shroud is not welded to the bulkhead because of possible thermal expansion differences between the shell and the shroud and to allow any gases generated by radiolysis within the LICON to have access to the recombiner catalyst within the canister.
The gap between the shroud and the bulkhead is a nominal.072 inch.
e
,-Y-
.?,_,-n
-m.-
10-12-9513:19 T-BW4FD LYNcHB'JR3
- 415 PB347
- a BW/BTL-85-80 Mr. R..L. Rider Page 2 October 10, 1985 If you have any questions on this matter, please contact the undersigned '~
on(804)385-3609.
Very truly yours.
THE BABCOCK & WILCOX COMPANY P. C. Childress Project Engineering Manager Canister Development Project THI-2 Recovery Operations PCC:pby
Attachment:
Letter Report cc: w/att.
F.R. Standerfer/T.F. Demmitt W.H. Linton P. Bradbury D.R. Buchanan D.M. Lake M.K. Pastor L.H. Lilien A. Stowe 8.J. Short t
T e
10-10-95 13:20 T-BunFD LYNCHBURG C41D PO4/97 etc saa Babcock & Wilcox a****rch =ad o. var.pm.at oi i.i.=
a McDermots company Lynchburg Mosearch Center Lynchburg, Virginia 24506 1165 73 P. C. CHILDRESS e.
J. M. SIORICtyR. A. NMNER can.
- m. w..
5251 sumi.
o..
LICON COMPAESSIVE STTLENGTH October 9,1985 T m. w w.e..
.ne mini wv.
LICON is the trade name of an ultralight concrete developed"by S&W in conjunction with the IMI-2 Canister Program. It is a lightwelght void filler for the ruel Canister, and as such provides deformation resistance for the boral shroud insert in case of a transportation accident.
Several questions regarding the use of LICON as a void filler in the Fuel Canisters have arisen. These questions are in regard to LICON's sensitivity of certpressive strength to water omtant, Img range stability in a water environment, adequacy of a 24-hour room temperature cure, chemical composition and abrasion resistance. To address these questions a series of tests were performed at R&W's Lynchburg Research Center (LRC).
Chemical Cornposition LICCN is conposed of a refractory grade CA-25C high purity, high i
alumina cement and hollow, unicellular glass microspheres; deionized water is used to prepare the concrete. The cement sets up to achieve high strength in under 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, curing at up(storage in plastic bags which are to 90 degrees F (hydraulic bending),
.. 30. percent RN. Under proper conditions resistant to moisture vapor transmission and sorption of carbon dioxide),
without loss of mixing,pheres mixture typically has a 12-month storage life the cement / glass micros l
placing or performance properties. Exposure of the cement mix to both water vapor (humidity) and carbon dioxide can cause loss of hydraulic bonding strength over prolonged periods of storage.
The cement used in the formation of'LICON is CA-25C,an ALCoA reduct.
It is composed of 80% A10,,184 Cao, ainor inpurities and 1.5% vofatile 3
material. riuxing impurItfes are restricted to ve low levels through the selection and unique processing of very high rit lime and alumina raw natorials. Typically the oxides of iron (re silica (Si
) are below 0.3 and 0.2 percent respectively. A t! le)al chemical anal sis of the t
comer.t is given in Table 1.
Composition condoras to the empiric 1 aclar fornula ca0*2.5A10 The predominant h% raulic bonding phase is monocalciumalumiMe. (Cao*Al 0 with a m secondary phases of 2 3 calcium aluminatis ()ref.1).
3 12Cao*7A1 0 1
i
- [~
10-19-85 13:21 T-BWMD LYNCHBJG
- 41S P05 4?
- y*.
F. C. Childress page 2 Oct. 9,1985 Table 1.
Cement chemical Analysis (4) l A1 0 79.7 23 Cao 18.4 Mgo 0.4 310 0.2 2
0.3 re,03 Na O 0.5 g
s Compressive Strength 'hsts R e compressive strength tests were done using an MTS system. S e sarple configuration tested was a constrained cylinder, 3.0 inches in i
diameter and 2.7 inches in height. 'No different concrete compositions were tested. The LICON cement mix sold to NES was used in all the tests, but the water content (delonized) of the mix was varied. In order to determine the effect of high water content on strength the cement was mixed with 6% excess water based on the water to dry cement weight ratio.
Two LICON concrete mixes, one each of standard and hips water ocxquiticas, were prepared and cast into samples. Five constrained samples of each composition were prepared, allowed to cure at room temperature for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> and s'.tenerged in deionized water for 48 hours5.555556e-4 days <br />0.0133 hours <br />7.936508e-5 weeks <br />1.8264e-5 months <br />. he average cocpressive
.... strengths for these samples are presented below in Table 2.
Table 2.
Compressive Strength Conosrison (psi) x+o CCNSTRAINED CYLINDERS Compressive 4 Change Strength (psi) From Original original LICON Data (not sunnerged) 3031 1 164 standard Congosition:
2171 1 403 28.4 i
RI-4mTER Cosposition:
1967 1 215 25.1 8
g e
d
.,n,,
,-.-...-----e n..
10-19-85 13:22 T-BU/NPD LYNCHBJts C415 P06/27 P.C.O[ildress Page 3 Oct. 9, 1985 CONCLUSIONS 1.
Compressive strengths n e compressive strengths of all the LICON samples that were submerged and tested are less than the previously reported values, but the values remain significantly above the 500 psi design goal.
he effect of increased water in the concrete mixture also appears to decrease the compressive strength of the LICON. In the constrained cylinders the nominal comprere 1 strength of the HIeTER camposition standard water content LICON concrete.
concrete was 9.4% lower that t
21s indicates that when tn..ater is being added to form the LICON concrete, the quantity of water added should be carefully controlled. Both normal and HI%ER LICON had an acceptable ccepressive strength when compared to the design goal. S e strengths obtained are high enough to justify previous drop test results using LICON (ref. 2).
W e test samples are cured at room temperature for 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. During this time the sagle surface is covered with plastic. Se plastic prevents the abnormal rapid evaporation of water from the surface of the sagle, which could cours a flaky crumbly surface. In reality, only a very small portion of the concrete surface is exposed to the air and that effect should be negligible.
2.
Adequacy of a 24-Nour Room Temperature Cure
~
h e formation of a hydraulic bond in CA25-c cement occurs at room J:enperature. We higher the terperature up to 90 degrees F the faster and more conglete the bonding. If the ambient terperature is below 60 degrees F, additional curing tima may be required. At curing temperatures near freezing the strength of the LIcoN concrete can be expected to be significantly conpromised.
.i 3.
Abrasion Resistance LICON's composition is such that it does not provide hlqh abrasion resistance. Its as-prepared composition is presented below :,n Table 3.
4 g
e g
e
.._,__.,,____,,..,_,__,,.,_.,-,_,_,,,.n,,
12-19-85 13:24 'T-BMFD. LYNCHBUR3 C413 P27/07
- /,*
- p. C. Childress Page 4 oct. 9, 1985 i
Table 3 Component Wat. 4 Vol %
CA-25C 60 23 Microsphere 11 43 D.I. Water 29 34 i
since the glass microspheres represent 434 by volume of the mix, the surface of the concrete can be expected to have a low abrasion resistance.
4.
Chemical and Physical stability of LICON under Water Based on experience and product information associated with CA25-C concretes, underwater applications of LICON should have no impact on the concrete's long term chemical and physical stabi3ity.
J. M. Storton R.R. U)%
R. A. Wagner s
References 1.
CA-25 Extra-Migh Alumina Refractory Cements Product Form F35-14200, Aluminum company of America, Chemicals Division.
2.
Sree Mile Island - Unit 2, Drop 'hst!ing of Defueling' canisters - Final Report, prepared by B&W's Nuclear Power Division for GFU Nuclear Corporatlon, February, 1985.
l l
I q
l
,--,--,-----,,wy.,,-,w,,,
,m,,,_.-
,-,___.-,,---,..,-.--n_.
,