ML20008F537
| ML20008F537 | |
| Person / Time | |
|---|---|
| Issue date: | 09/15/1980 |
| From: | Hendrie J NRC COMMISSION (OCM) |
| To: | Starr C ELECTRIC POWER RESEARCH INSTITUTE |
| References | |
| NUDOCS 8104210156 | |
| Download: ML20008F537 (1) | |
Text
_
4 g* @ M:tig#
e
- o UNITED STATES
[ * ) gd Ek FJUCLEAR REGULATORY COMMISSION g,
WASHIN GTON, D.C. 20555 o*
September 15, 1980 h
8If
- sse*
OFflCE OF THE N
f COMMISSIGN ER g
'lPR.I 51981s A1 3
. s v., %
=
4 N
,cq 4'
Mr. Chauncey Starr 4.24.
Vice Chairman
'N' 4#
m Electric Power Research Institute 3412 Hillview Avenue P. O. Box 10412 Palo Alto, California 94303
Dear Chauncey:
Thanks for the draft paper by Levenson and Rahn.
It is a good piece and I hope it gets wide circulation when published.
There is a similar thrus' from Stratton, Malinauskas, and Campbell on iodine assumptions in acd dert releases contained in a letter from them to the Commission.
I ent.iose a copy.
I plan to ask for a Commission, public briefing by Stratton, et al., in a month or so on their proposi-tions. You, Milt, and Rahn might want to join in and present your views as laid out in your draft paper.
It would be a way of focusing some attention on the subject of these accident release assumptions.
I will be in touch, or have the Secretariat contact you if I can get the s'
Cnmmission to agree to the briefing.
Also, Herb Kouts and his people at BNL have been looking at some of the assumptions about iodine and cesium behavior in accidents and are coming to the view you hold that we grossly overestimate these releases.
Sincerely, ph M. Hendrie Commissioner
Enclosure:
As stated 8 104210156
=-
t August 14, 1980 e
i l
1 Chairman John Ahearne l
U.S. Nuclear Regulatory Commission 1717 H Street j
Washington, D.C. 20555 l
Dear Chairman Ahearne:
We wish to. bring to your attention a matter that may be a very important devel-l opment in reactor safety analysis.
We believe that sufficient evidence has -
accumulated to show that the behavior of iodine during nuclear reactor accidents Iodine is not correctly desetibed by existing NRC models-and Regulatory Guides.
volatility is grossly overes,timated by -these modelq for ' accidents in which sub-stantial amounts of water are.present, and escape of iodine to the environment will be extremely small (as it was at Three Mile Island) as long as reasonable l
containment integrity is also maintained.
As a consequence, the risk to the j
general public presented by iodine is lower than estimated, perhaps by ' orders of
~
magnitude.
Our concern with this issue originated with our involvement' in the several Technical Staff Analyses for the President's Commission on the Accident at Three Mile Island.
The mechanism for the behsvior of iodine that we propose here was i
derived from those analyses, from further examination of experimental and theoretical studies involving the chemistry of iodine and cesium fission pre-ducts in light water reactor fuel and systems, and from the observed behavior of l
iv?fet ruhsequent to fuel failures during accidents and incidents at other reac-
~
tor sites.
We believe that the explanation presented here will change the pre-concepts of the haza,rds involved during and subsequent to reactor accidents sent and, therefore, will require a critical reexamination of how these hazards and risks are caldulated, and the' criteria to which engineered safeguards are designed and installed.
Although the Three Mile Island (TMI) reactor core inventories of xenon-133 and iodine-131 were comparable, between 2.4 and 13 million curies of xenon escaped to the environment during the accident, while only 13 to 18 curies of iodine similarly escaped!
This great disparity was identified as a matter of crucial importance early in the investigation by the President's Commission, and an effort was made to find the explanation.
It was clear that we could not claim 6
5 to 10 ) y,,
t understand the accident until this discrepancy (a factor of 10 explained satisfactorily.
Further, it was recognized that the physical and chemical conditions during the accident at TMI may not have been unique.
(We note that, generally, radiciodine is the controlling fission product species with respect to site safety analysis as well as the design and operation of certain engineered safeguards.)
{
' g
' p q) raps 86/dCD3D g
T,[Ja
Chairman J. Ahearne August 14, 1983 The explanation for the very low escape of iodine that developed during the l
investigation by the President's Commission was that, as the temperature of the core increased, iodine diffused out of the fuel rods thr6 ugh the failed cladding and vaporized.
The iodine escaping, if not already in the iodide form, then encountered a chemically reducing environment which converted it to iodide.
The L
iodide subsequently went into solution as iodide ion when it contacted water.
It was recognized that additional experimental work was needed to provide a quantitative description of the iodine behavior.
Nevertheless, this explanation accounted for the.nuch smaller escape of iodine that was observed at TMI com-pared to the amount predicted to escape if elemental iodide had been present,,as in assumed in the Regulatory Guides.
We believe that this description can be strengthened and-made more definitive..
Although the present data are not absolutely conclusive, we believe that iodine emerged from the fuel as cesium iodide, already reduced to Lodide.
The reactor system environment then sustained this chemical state.'
Furthermore, it would have converted other iodine species, should they have been present, to iodide.
Cesium iodide would be expected to condense or " plate-out" when it reached metal surfaces at temperatures at or below 400 to 500*C, and it would finally enter into solution as iodide ion as soon as water or condensing steam was encoun-l tered.
The reactions of iodine species in water, and the fact that iodide ion is the dominant species, ensure that iodine volatility will be very small (compared to'that implied by the Regulatory Guides, for example).
A reaction causing oxidation of iodide would be necessary to increase the volatility of iodine.
Additional experimental work is required to provide a quantitative description of iodine behavior, but this qualitative picture is consistent with the small escape of iodine observed in a number of incidents when water was pre-sent, such as at TMI.
This echanism is supported by the folicwing observations, as well as By measurements made at THI:
I l
1.
Iodine and cesium are released congruently from PWR leakers during oever transients (the iodine spiking phenomenon).
i 2.
Thermodynamic calculations performed at several sites indicate that CsI is the stable form of iodine in URR fuel.
Further, the fission yield of cesium is larger than that of iodine, and cesium is always present in great (about tenfold) excess over iodine.
i Irradiated fuel has been caused to fail in experiments performed under simu-3.
lated accident conditions, and the iodine released is recovered predo=i-nantly as Cs1 rather than as molecular I
- 2 l
l l
l o
n -
Chairman J. Ahearne August 14, 1920 4.
Ihe chemistry of iodine is such that, if water is accessible, iodine will interact with the water so that its concentration in the gas phase will be
]
much smaller than its concentration in the water.
5.
In other incidents that have led to the destruction of fuel in water systems (NRX, Spert-1, Snaptran-3, SL-1, MTR, ORR, and PRTR), we understand that a much smaller amount of iodine escaped from the systems than would be pro-jected by the existing models.
Data are hard to come by for many of these accidents and' experiments, and our investigation is continuing.
In sarked l
contrast, a large fraction (20,000 curies) of the iodine escaped to the environment during the Windscale accident, which occurred under oxidizing conditions and in the absence of water.
,r.
The significance of this mechanism for iodine escape and transport can hardly be overemphasized.
We assert th'at the unexpectedly low r~elease of radioiodine in the IdI-2 accident is now understood and can be generalized to other postulated accidents and to other designs of water reactors.
We believe that an accident i
involving hot fuel and a water or steam-eater environment will have the same i
controlling chemical conditions as did the TdI-2 core and primary system.
The iodine will emerge as CsI (and possibly some other iodides) and enter into the solution as soon as wet steam or water is encountered.
It will persist in solu-tion as non-volatile iodide ion as long as oxidizing conditions do not prevail.
Although we feel that the evidence is sufficiently strong to justify this letter, it is important to qualify our position.
Iodine chemistry is very complex, and definitive experimental and analytical studies of iodine behavior during and following loss-of-coolant accidents are lacking.
Nonetheless, it is clear that the behavior projected from the existing Regulatory Guides is wrong.
1 l
The current NRC assumption, that elemental iodine is the chemical form of the radioiodine released is regar.ded as a conservatism, but in this case the assumption of i wrong, chemical form nust be' regarded as an error which has com-pounding effects.
If, af ter due consideration, the NRC is satisfied that our description of iodine behavior is valid, we recommend,that an urgent study and assessment be made of all available inf ormation, and appropriate actions be undertaken.
With due
~
respect we point out four consequences should our position be correct:
1.
The frequently quoted fission product escape assumptions (from TID-14 844 in 1962 to the more recent Regulatory Guides 1.3 and 1.4, and the Reactor Safety Study, WASH-1400) should be reexamined.
The present assumptions grossly overstate iodine release from a reactor site. in many types of loss-of-coola nt accident, and safety criteria based on these assumptions should be reevaluated.
e
..a C3 airman J. Ahearne August 14, 195 2.
The dispersal of radiciodv in the biosphere may no longer do=inate and control consideration of accidents and the design of safety syste=s.
3.
Many, if not most, accident sequences must be reexamined in detail.
The iodine risk to the general public may, in fact, be lower than previously estimated, possibly by orders of magnitude.
The. impact of a reduction of iodine risk on the requirements for evacuation is particularly important at this time.
4.
The engineered ' safeguards designed for iodine control should be reexamined to assure effectiveness and optimization for the actual iodine behavior rather than the behavior currently assumed.
Finally, we realize that a major revision of NRC assu=ptions relative to acci-dent analyses, dose calculations, and design of safeguards should not take place without an adequate base of technology from both experiment and theory, and especially until the Co==ission itself is convinced that it is appropriate to from fuel accept a rc ised physical and chen'. cal description of iodine transport to the environment.
On the other hand, the impact of vrong assu=ptions is so serious that an intensive effort should be made to establish the facts.
We are ready, to of fer more detailed information or further assistance should the NRC request it.
We vill be pleased to brief the NRC staff or any review commit-tees you may appoint.
Sincerely, W te M i W. R. Stratton Los. Alamos Scientific Laboratory fl ?
w A.'P.
Malinauskas Oak Ridge National Laboratory e
sj D. O. Campbell Oak Ridge National Laboratory l
f cc:
G. W. Cunningham, DOE-WASH D. M. Kerr, lASL H. Postma, ORNL 1
a
ELECTRIC POWER RESEARCH I,N S T I T U T E September 2, 1980 2
a The Honorable Joseph M. Hendrie Commissioner Nuclear Regulatory Commission Washington, D. C.
20555 Dear Joe As you are undoubtedly aware, prior to the Three Mile Island accident, the probability of a public catastrophe resulting from such an event was considered to be negligible, although it always has been a continuing subject of professional study.
Since mI, there has been a fresh flood of wide-ranging reassessments of the public risk. The enclosed draft study is an attempt to step back and take a realistic look at the basic scientific processes which are the fundamental deteminers of what these public risks might really be.
The main thrust of this study is that the natural laws of physics and chemistry substantially limit the distribution of radioactive of fluents fran any nuclear accident, no matter how severe. This study makes the point that the empirical information that can be garnered from a variety of large-scale accidents that have already occurred, when combined with known physical and chemical laws, tends to confirm that the theoretical
" source term" traditionally used in nucle'ar risk evaluations is one to two orders of magnitude greater than the realistic magnitude which might actually result from the ultimate accidents.
Because of the relevance of this issue to the current flurry of federal and state emergency planning and evacuation criteria, I am forwarding this draf t to you prior to publication for both your consideration and your comment. The study will be presanted as a paper at the ANS International meeting in Washington, D.C., November 17-21.
Sincerely, Chaunc'ey Starr Vice Chairman CSimi Enclosure o
D %"~' E gojooBol>B g r95 a
.=uJH L b Headauarters: 3412 Hillview Avenue, Post Office Box 10412, Palo Alto, CA 94303 (415) 855 2000 Washington Ottoce 1800 Massachusets Avenue NW Suite 700. Washington. DC 20036 (202) 872-9222