ML19242A397
| ML19242A397 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 07/30/1979 |
| From: | Pearlman H ROCKWELL INTERNATIONAL CORP. |
| To: | Powers D Office of Nuclear Reactor Regulation |
| References | |
| 79ESG-7531, NUDOCS 7908010559 | |
| Download: ML19242A397 (4) | |
Text
'.
m Energy Systems Group 8900 De Soto Avenue Canoga Park, CA 91304 Telephone: (213) 341-1000 Rockwell TWX. 910-494-t237 Te e>c ist0i7 International July 30, 1979 In reply refer to 79ESG-7531 Dr. Dale A. Powers Office of Nuclear Reactor Regulation Nuclear Regulatory Commission Washington, D.C.
20555
Dear Dr. Powers:
The purpose of this letter is to document our telephone conversation of July 24 about information we may have here from the SNAP Program, that could possibly bear on hydriding of the zircaloy cladding in the Three Mile Island incident.
As I stated, the hydriding conditions in that program were sc stantially different from those that must have prevailed at TMI-2.
To begin with, we prepared SNAP fuel by hydriding a metal rod whosg surface was main-tained scrupulously clean while being heated to 900 C in a vacuum furnace.
The hydrogen gas used in the hydriding reaction was very carefully purified prior to admission to the furnace.
I have not located any documented experience with hydriding under conditions where impurities, especially oxygen and water vapor, were known to be present.
None of our own hydriding data, therefore, was obtained under conditions that matcg the probable situation at TMI-2: metal temperatures greater than 1500 C, the presence of H 0 and its dissociation products, and also 9
the presence of an ionizing radiation field. However, evidence from our over 20-year-long program in hydriding technology, and relevant observa-tions made at other sites, do show the following:
(a) When oxygen or oxygen-bearing compounds cre present in the environment surrounding zirconium or zirconium hydride, a surface layer of oxide forms.
(b) The oxide layer is a very effective barrier to the transfer of hydrogen between the solid and the gas phases in either direction.
\\
462 001
\\6 h o
\\
29 08020 5 p V
v
79ESG-7531 July 30, 1979 Page 2 Four instances of these phenomena are cited below.
(1) H Loss from Zirconium Hydride in " Vacuum" 2
Sampfesofhydrige(H/Zg)1.05wareheatedinvacuumat1150F (620 C) and 1350 F (730 C) for 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> Between 3 and 4%
3 hydgogen was lost from the sample at 620 C. but the sample at 730 C lost essentially no hydrogen.
In the latter case, the residual oxygen impurity in the vacuum rapidly formed an oxide coating on the surface that prevented hydrogen loss. A reference for this information is Report NAA-SR-8617, dated November 15, 1964, Page 1.11.
(2) Hydrogen Retention in Failed SNAP 8-ER Fuel Elements Post-irradiation examination of the fuel elements from the SNAP 8 Experimental Reactor revealed that a substantial number had cracked cladding. These cracked elements retained hydrogen surprisingly well.
The fuel composition was approximately 90 Zr-10 V (93% enriched, hydrided to a P/Zr ratio of about 1.7).
Becausg fuel element temperatures reached peaks in excess of 850 C, it was expected that substantial hydrogen wouid be lost from elements with cracked cladding. Metallo-graphic examination of such cracked elements revealed the formation of a " rind" on the surface.
While this was not analyzed chemically, the conditions of operation suggested that this was an oxide layer.
In order to contain hydrogen, the inside surface of the cladding and end caps had been coated with a layer of glass prior to assembly. The glass was a mixture of aluminum and barium silicates and represented a source of oxygen, which the liquid metal coolant that invaded the cracks could have carried to the surface of the fuel rod. A reference for this information is NAA-SR-Memo-12210 1ssued November 30, 1966, Pages 39-42.
(3) Hydrogen Locs from Zirconium Hydride as a Function of Environ-ment Purity Samples of zircogium hydride (H/Zr)1.7 were exposed for 100 hours0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> at 600 C in the following environments:
very high vacuum, very pure liquid sodium, a.id liquid sodium with a known oxygen impurity.
Hydrogen loss from the first two fm 082 n
79ESG-7531 July 30, 1979 Page 3 6
environments was a factor of about 10 greater than that from the impure sodium.
In that sodium, an oxide coating was formed on the hydride surface that effectively prevented transfer of hydrogen. A reference for this information is the same report as in Item (2) above, Page 42.
(4) Formatio.1 of 0xide Layer on Zr Metal by H O Vapor Impurity 2
In order to eliminate the induction period for the ini'iation of hydriding, Zr metal is heated in vacuum at elevated tempera-ture, prior to the introduction of hydrogen ggs (see baiow).
In one instance, the metal was heated to 1000 C, which caused desorption of Hg0 from the apparatus walls, and gettering tc the hot Zr. Oxidation of the surface was severe enough to prevent subsequent reaction with hydrogen.
The reference for this information is NAA-SR-1508, issued October 15. 1956, Page 410.
The fact that the presence of a surface oxida layer effectively prevents hydriding of Zr metal was well known from the literature, even at the beginning of our SNAP Program. We made an exhaustive study or hydriding technology at that time, on which basis we set high standards of purity for both the metal surface and the hydrogen gas.
We found that the oxide film formed even in ambient lat *atory air wag sufficient to impede subsequent hydriding, at temperatures of 500 C and above, in gaseous hydrogen at 1-5 atm pressure.
The remedy was to prgheat the metal in vacuum for several hours at temperatures up to 800 C, which diffused the surface oxygen into the bulk meta 1 Favorable hydriding 3
kinetics could then be obtained, usually at 300 C, or even at lower temperatures.
It is emphasized that the preheating treatment must be performed in a sufficiently high vacuum to insure a reglible chemical activity of oxygen-bearing gases (see Item 4 above).
Our report, NAA-SR-1508, is also '.he reference for the abc/e information.
This 834 -page document, which lists 848 references, contains a thoroud dis-cu'sion of the fundamental studies on the Zr-H system by Dr. Earl Guibransen and his colleagues.
It includes a detailed evaluation of the e'fects of surface oxide layers on the kinetics of hydriding.
In none cf these studies was the initial oxide surface replenished.
Contrariwise, conditions at TMI-2 seem to have been very favorable for replenishing an oxide surface layer.
462 083
79ESG-7531 July 30, 1979 Page 4 I believe that hydriding of the core cladding at TMI-2 could not occur to any significant extent.
If you have questions about this or any other aspect of this letter, please do call me at (213) 341-1000, Extension 2233.
Sincerely,
\\[
d t L'uta w H. Pearlman, Manager Materials and Physics Technology Energy Systems Group klp:727 cc:
Dr. W. V. Johnston, Chief Fuel Behavior Research Branch U.S. Nuclear Regulatory Commission Washington, D.C.
20555 462 084