ML20024A523
| ML20024A523 | |
| Person / Time | |
|---|---|
| Site: | 05000142 |
| Issue date: | 06/14/1983 |
| From: | Aftergood S, Dupont D, Hirsch D, Kaku M, Kohn R, Norton B, Plotkin S, Pulido M, Wayne L COMMITTEE TO BRIDGE THE GAP |
| To: | |
| References | |
| NUDOCS 8306170403 | |
| Download: ML20024A523 (42) | |
Text
{{#Wiki_filter:_ _ __ I l PA N E L- [ u hv/s3 l CNEMICAI, REACTIONS I l Introduction
- 1. In 1957,~a very serious fire occurred at a non-power reactor located at Vindscale, England. Although the reactor was a production reactor, it had a number of sinhities to the UCIA reacter-- fuel containing uranium metal clad in aluminum, with a graphite moderator / reflector, and normal operation at relatively low temperatures, which permitted build-up of stored "Vigner" energy in the graphite. Release of that r stored energy contributed to the cause of the fire, which resulted in extensive daanse and 20,000 curies of iodine-131 being released to the environment. Milk contaminated with I-131 had to be disposed of in an area of 200 square miles around the reactor because of the accident.
l 2 In 1960, the UCIA Argonaut-type reactor began operation. Its Hasards Analysis did not addrissa Vigner energy storage, and a brief paragraph
- dismissed the potential for fire largely based on the assertion that l
"none of the anterials of construction of the reactor are infh==mble." (p.62)
- 3. As the Windacale fire showed, and as shall be discussed in detail below, that assertion is dangerously untrue.* The graphite can burns i the uranium metal can burns the angnesium can burns even the aluminum under sono circumstances will burn. And ignoring Vigner energy can like-vise be dangerous.
4 It has further been asserted that the only chemical reaction of signif-icance to be considered for the UCIA reactor is a water reaction with aluminum, and that aluminum weald have to be in the form of metal filings l for such a reaction to occur. Sat, too, is not the case. 5 heh of-the destructive power excursions with aluminum-clad, plate
- typefuel(SPERT, BORAX,andSL-1)hasapparentlyresultedinsignificant metal-water reaction. Much of the core disassembly in those three cases can be traced to a combination of steam explosion and metal-water reaction.
He aluminum was in the form of fuel W M ==g most assuredly, not in the form of metal filings. A destructive power excursion, thus, could result not only in fuel molting, but in explosive disassembly of the core due to explosive steam and metal-water reactions.
- 6. Similarly, fire suppression response, particularly if ill-prepared as in the UCIA case, can vastly worsen the situation. Metal-water reactions between the aluminum, uranium, and magnesium can be explosive and liberate considerable energy if water were poured on those substances when burning.
Likewise with burning graphite. Furthermore, use of water in a fire-fighting situation could have unforeseen reactivity effects.
- Se original UCIA Hasards Analysis was apparently copied virtually verbatim from materials provided by A W and by the University of Florida, and makes a number of serious errors. As shall be shown further, so do i the neu analyses UCIA has copied. mis is one of the major dangers of copying analyses verbatim from others, rather than performing the analysis independently.
B306170403 830614 PDR ADOCK 05000142 T PDR
4
- 7. Se Potential for fire, Wigner energy release, and explosive or otherwise destructive chemical reactions has been avamined for the case of the UCIA Argonaut-type reactor. It is concluded that the UCIA reactor is not inherently protected against serious damage from these chemical reactions, and that significant radioactive releases to the environment could ensue.
Fire is a particularly dangerous hamard scenario for the UCIA reactor since it could release virtually an of the volatile fission products and could provide a powerful driving force for substantial release of particulates. 8 It is further concluded that the design characteristics of the UCIA Argonaut are most unfortunate from the point of view of inherent protection against such chemical reactions and their effects. Use of combustible materials such as angnesium, graphite, and uranium metal in an essentiany dry core, without the inerting or sealing employed in modern graphite 1 reactors, poses a substantial fire hasard. IAu-temperature normal operation makes Wigner energy storage a substantial problem (Wigner energy is not stored at Meh== temperatures), providing the potential for significant release of heat being triggered even in an otherwise minor accident. Bis is further complicated by the fact that the f ael and control bindes are ande of very low-melting materials. Se core constituent materials furthermore represent substantial potential for explosive or other destructive chemical reactions in situations such as power excursions or fire. Rose matters win be detailed below. FIRB
- 9. m original Nasards Analysis for the UCIA reactor dismissed the prob-ability of damage from fire resulting in the release of fission products as "very emall" in part because "none of the materials of construction of the remotor are inflammable." (1960 UCIA Reactor Masards Analysis, p. 62,
" Fire"). While other factors any affect the probability of fission product rolesse from fire, the. statement that none of the materials of construction of the reactor are inflammable is simply incorrect. A number of those materials- particularly the graphite, uranium, magnesium, and even the aluminum, among others- are, under the right conditions, most definitely combustible.
10 De first and most obvious of the combustible anterials used in the Argonaut reactor is the graphite-- used as moderator, reflector, and thermal column. Graphite will, under the right circumstances, most definitely burn, as the Hawley sped, correctly indicates. (Charcoalis,afterall, a graphitic substance, and it will, of course, readily burn.)
- 11. On page 82 of the Proceedings of the 1958 Atonio aner6y Commission and Contractor Safety and Fire Protection Conference (attached), held at AEC l headquarters in Germantown, Mary 1 sad, June 24-25, 1958, held in part to l analyse the implientions for reactor safety of the Windacale accident in l
which the graphite moderator and the uranium fuel both caught fire, Dr. C. Rogers McCullough of the USAEC is quoted as saying: By the way, this is an ==n=1ng point. Se belief had grown up on the part of many people in this country that graphite will not burn. His is nonsense. Graphite is carbon, and anyone knows that omrbon i will burn if you get it hot enough. But this glib remark, that graphite will not catch on fire, had become prevalent.
Graphite can, of course, burn in air, as the Vindacale fire unfortunately so clearly demonstrated. A belief to the contrary would be neither correct nor prudent. 4 12 As to the matter of the ignition temperature of graphite, it is dependent upon a number of factors such as the purity and density of the graphite, the amount of air present and the velocity of the air, the particle sine and surface-to-volume ratie of the graphite, and structural configuration influencing heat loss. Furthermore, there appear to be other uncertainties, as evidenced 17 Dr. McCullough's ocaments at the same page of the above-cited ; proceeding l Research work is going ons we are not satisfied that we know the ignition point of graphite.... At any rate, research is going on to learn more about the ignition temperature. It is a tough problem to solve, and we are exploring possibilities. i Mus, there are some uncertainties as to ignition temperature of graphite, and it might be wise from the point of view of a conservative safety analysis to place or establish the angnitude of error on whatever estimate of ignition temperature is used. Se Hawley . eye.;, uses a figure of 650'c as the point at which graphite will burn readily if sufficient oxygen is supplied. There are some uncertainties and some error limits might be ay,.eyante. Any temperature estimate is valid only for a fixed set of parnasters (density, purity,particlesise,airsupply,irradiationhistory,etc.).
- 13. Once ignited, self-sustained combustion of the graphite must be assumed if.the air supply is adequate. Although this depends upon con-figuration, and the like, Hawley assumes for purposes of his analysis that somewhere around 650*c is the critical temperature for induction of a self-sustained fire in the Argonaut reactor's graphite. Sie temperature is above a glowing red heat but below a white heat. Se reaction is exothermio, so if some of the graphite were ignited, it could rolesse enough heat to 1 ring other graphite to the ignition temperature.
14 In addition to graphite, the Argonaut remotor at UCLA employs meta nio uranium in a uranina-aluminum outeetic, clad with al==in==, Metallio uranium readily burns in air if ignited, and under somewhat more restrictive conditions, so can al==1===, Aluminum gives off more heat, pound for. Pound, than uranium metal when burned, but it is somewhat more resistant to burning. The fact that the uranium and the aluminum are in a eutectic , will not affect the ability of either to burn, alt"-h burning of the l esteetic win give off slightly lose heat than if the materials were not in a eutectie. However, the differesse is in=4gnificant. In addition, the fact that the outectic melts at a relatively low temperature (640'C-Ikuley, p.18), sin not affect the ability of the materials to burn. Se metals can burn as won in a liquid form as in a solid. In fact, molten metal can cause fresh aluminua, without the normal protective oxide layer, to be exposed to air, asking burning far more likely.
- 15. As to ignition temperature for uranium metal, again there are some uncertainties. Charles Russen (Reactor Safeauards, Pergamon Press, Oxford, 1962, p. H 5-n 6, eiting v.c. Reynolds, neport NACA TN D-182, " Investigation ture of of Ignition Temperatures of Metals") gives the ignitgen p20So soliduraniummetalinoxygenat1atmosphereas608F(3 C).
United Kingdom Atomic Energy Authority, in a report on the Windacale accident (in which both the uranium and the graphite were burning) states 1 that, "In still air uranium oxidises, i.e., the reaction is self-heating at ! 350'C the corresponding temperature with carbon dioxide is 6500-700'C." (Nucleonics, Vol. 15, No. 12, December, 1957. P. 91)
- 16. Turnings of remotor-grade uranium have ignited when being cut using a lathe, evidently from friction. Finely divided uranium ignites in air at room temperature. "**s the ignition temperature is a variaMe, depending on circumstances, but in general uranium metal mast be considered more combustible than graphite. The details of combustibility of the uranium-aluminua entectio certainly merit investigation, in both solid and liquid states.
- 17. Se control blades at h Ucu reactor are z.pdedly andmium-tipped and protected by angnesium shrouds. Magnesium can also burn, and when it does so it gives aff considerable energy. The ignition temperature of Mg '
metal is variable, depending on its particle sise, etc. If one spoeifies a desired ignition temperature, from 25' up, one can prepare a specimen which will ignite at that temperature. One should be ausre that slow oxidation occurs below ignition temperature. 18 M atum metal is a lou-nelting metal with a relatively high vapor pressure. The Handbook of Chemistry and Physics zipes its molting ten-perature as 3200C. If the control Mades are ande of the metal and not the oxide, it would seen prudent to analyse the reactivity and other possible consequences of an incident which resulted in the molting of the control blades. For example, an incident involving fire or other high temperature event might cause the low-melting control rods to melt out of the core region, increasing reactivity as well as making control difficult. Further-more, the volatility of endaina oculd potentially result in omdatum vapor being released in a fire or other incident involving elevated temperatures. If so, the cadmium vapor or its oxide would likely rapidly sendense in air as minute particles and could cause a potential hasard for fire-fighters or others due to the tozio nature of andmium. This, too, should probably be considered, it would seen, in designing fire-fighting plans and analysing potential accident seguenoes and consequences.
- 19. Ucu is requesting a license for 2 ouries of plutonium-239 in a plutonium-boryllium neutron source for the reactor facility. Vere this Pu-Bo source to boccas involved in fire, the consequences could verge on the catastrophic. Plutonium metal, of course, can burn, releasing minute particles into the aire dispersed by the energy of the fire. Fire-fighting would be extremely hasardous due to the presence of the plutonium oxide in the air, and the public health implications would be awful. (2euriesof Pu-239 is by no means an in-i==ificant amounts placed near the skin, it l will cause radiation burns in a few minutess hhmintion of even microgram amountsisexceedinglydangerous.) When Pu metal burns, it goes to PuC2 in limited air, to Pu3 0s in excess air, just lika uranina. Be is comparable to Al in its combustion, but is higher molting. Again, the chemical form of the material is important, i.e. uhether in metal or oxide. Be0 is volatile in steam at high temperatures.
t l Bmlesive Remotions in Fire i 20. The issue of hou to fight a graphite-uranium fire, leaving aside the possibility of ondmium and plutonium particles being released, has no ! easy answers and would require considerable prior analysis of the problems inherent and preparation in advance.in the form of emergency f === W . , There could be great danger, in particular, in employing either unter er, to a lesser degree, carbon dioxide to put out the fire. In either case, an explosion might occur, owing to the formation of combustible gases. 21 Dr. McCallough's report on the Vindacale incident, in the AEC document referred to above, describes how these fighting the fire tried various notheds over a couple of days to put the fire out, which involved both uranium and graphite, all to ne avail, and how they had to try, as a last resort, unters i Now they were faced with the decision either to use unter or to let the fire burn up. They decided there was nothing left for them to do but put unter in. There uns sono trepidation about this, as you can inngine, boomune they won knew that water en glowing uranium ankes . lva 8.s. Vater on glowing carbon makes hydrogen and Cop you have then a nice mixture of hydrogen, 00, and air, and you might have an explosion. But they had no other choice. They, in the end, followed techniques learned during World War II in extinguishing incenM e y bombs, and fortunately the smahle paid off. But they had no other choice, and rightly were extremely worried about the potential for an explosion. The fact that one did not occur at Vindacale does not get one around the fast that such an explosion is clearly possible, could be quite dangerous, and that unter should, if at au possible, not be used, or if used, used with the potential danger clearly thought out. As McCullough concluded: I think it took a great deal of courage on the part of these people to put i water on this reactor. They did it with fear and trepidation, and in talking with them they will not guarantee that they could do it a second time without an explosion. l It ~should be noted that the steam that ensued carried with it very significant quantities of fission products into the environment.
- 22. The potential for metal-unter or metal-stema reactions should be
*wamined in putting together fire-fighting plans. Aluminum, uranium, magnesium, and graphite au can react in a steam environment, producing large amounts of energy, liberating hydrogen which can cause explosion dangers. Russen indicates the Al-H 2O reaction liberates more than twice the energy of nitroglycerin, in calories per gram, and five times the energy of black powders the magnesina-water reaction just slightly less
! than aluminum and the U-H2O reaction just somewhat less than black ponder. 1 (Al + NHp O3 uns used as a cheap explosive in Vietnam, " Daisy Cutter.") 23 The use of CO2 on such a fire could also be dangerous. Graphite is oxidised by CO2 , yielding carbon monoxide, which is also explosive in the presence of air.
i
- 24. Simple carbon tetrachloride extinguishers that formerly were used for lab fires have a host of problems associated with their use, notably the toxic phosgene they give off when used on fires. And even some chemical fonas might have a favorable moderating effect that needs to be taken into account (this oma be gotten around, perhaps, by the addition of boron-containing compounds to such fosas).
25 A fire in this reactor raises other serious reactivity guestions as well(e.g.powerexcursionimplications). If unter, or some other moderating substance, were used to suppress the fire, a power excursion might result. If the control blades melted out of position, the equivalent of a large positive remotivity insertion might ensue. Furthermore, the positive ten-perature coefficient of the graphite means that as the temperature rose in the graphite, remotivity could ineresse as well. All of these factors could make a fire at the reactor even more serious. ,
- 26. Firefighters would have to be prepared to deal with potentially toxic substances such as cadmium fumes in the air, and work in an environment possibly contaminated with fission products and perhaps plutonium. They
, would need good information as to what materials had been released into the air and in roughly what oencontrations, good detectors for those materials, and ability to read and interpret that information. They would need ary.eyslate equipment to protect themselves from inhalation of the materials and from direct exposure. They would need to know the ayy.eyslate way to fight such a fire without ==IHar it worse.
- 27. It is not likely that a group of firefighters arriving on the soone of a fire at the UCIA reactor would have the competence to judge whether to use water, and if so, how, etc. This is acknowledged in the Fire Department's one-page fire response plan included in UCIA's proposed emergency plan, which essentially says that the Fire Department, upon arrival at the soone, will suppress the fire if the remotor is not involved, and will " confer" if it is. Without an emergency response having been carefully considered in advance, and without 'stookpiling of carefully chosen, non-moderating materials that could be used to smother the fire without reacting explosively with burning oore components, a fire could be ande innessurably worse. And instructions to not attempt to suppress the fire if the remotor is involved until conferring with reactor personnel and others (who might not be available, for azample, at night or on weekends), while sensible in the absence of a carefully-thought-out plan for safe response, could aesa substantial delay before suppression was attempted, permitting the fuel to be at greater risk.
- 28. Both the NRC Staff's Safety Evaluation Report (p. 9-2) and the Hawley report (p. 30-43) indicate that the UCIA reactor is not inherently protected against fire, and that protection against radiation release is dependent upon ys--yt,and sorrect fire department response. The Hawley report outlines a number of scenarios that have potential for la= Mag to such a fire, and indicates that, " depending on the length of time before discovery of the fire, the aluminum fuel boxes and fuel could be at risk for molting." (p. 43)
As noted above, discovery alone would be insufficient, beesume of the lack
. of preparedness and the difficulties involved in fighting a graphite-uranium-angnesium remeter fire without causing the reactor to explode.
~7-Fire Scenarios
- 29. Se Hawley report presents a number of potential fire scenarios.
Among them: welding torch accidentally igniting outer graphites power excursion sufficient to ignite a f1===mble solvent ( a common mode scenario for this event would be a power excursion caused by breakage of the sample container in which a large sample dissolved in solvent is being irradiateds removal of the neutronebsorbing material from the core could initiate the power excursion which, even though perhaps insufficient to melt the fuel j itself or ipite the grgphite, could ignite the solvent with its lower ' flash point): nuclear heating of inserted natarials "to a temperature high enough to ignite various flansable substances seems well sithin the realm of. Possibility": building fires and so on. One can suggest numerous others as well, but it is sufficient to indicate that the remotor is not inherently protected against fire.* Re h uley report indicates that a number of these scenarios oeuld put the fuel at risk if proper and prompt response were not made to suppress the fire. Se 2.p d also indicates that because graphite produces little smoke when it burns, the fire might go unnoticed for sub-l stantial periods of time. Sere is no procedure in the easrgency plan for actually fighting a remator fire. Given these factors, a remotor fire can occur and can put the fuel at risk. Fires are common events. 30 Se NRC Staff has asserted that a graphite fire in the UCIA reactor would occur only if an experiment failed and a general buildig fire occurred and the remotor's graphite blocks were exposed to a free flew of air. Se Staff cites pp. 41-43 of the kuley report. We samt not be readig the same zi p d. page 41 refers to a credible scenarie in which a hilding fire occurred while the shield blocks were removed: there is no mention of the necessity of a failed experiment as well. Credible common-mede causation is suggested by the. authors. page 42 of huley describes a credible accident scenario caused by a failed experiment alone. Se bottom of p. 42 con-tinuing onto p.43 describes another credible scenario, a simple kilding fire while the shield blocks were removed. Se Staff appears to have misread
.its consultants' report.
Sufficient Airflow for a Firs
- 31. Se current Safety Analysis Report of UCIA (1982) no longer makes the mistake of its ori=4==1 ksards Analysis in denying the combustibility
- In addition to the reactor itself estohing fire, significant basards could ocour through radioactive release due to other kinds of fire at the facility.
For example, there could be considerable danger if the plutonium souroe at the facility ass-involved in fire. Otherradioactivesubstances(forexmaple,
" hot" samples that had been irradiated in the core) could ignite, either
, in-core or outside. A fire in the "rabbLt room", where the samples return after being irradiated in the core, could be guite serious. (Seephotosof , the rabbit room, showing plastic bags containinf hundreds of plastic vials con +=i=4=a radicastive samples, being stored.) A fuel h adline socident could likewise involve fire, were, for example, the fuel placed in a vat of solvent to clean off surfaces for inspection and were the solvent to ignite. l
{ of the remotor's constituent materials. A new assertion is unde, howevsr, that there is insufficient air present for a fire, once started, to be sustained for extended periods. UCIA cites as basis for its assertion an asserted ! asasured airflow out of the core extract line. In so doing, they completely miss the point. 32 First of all, UCIA contradicts itself in several places as to the actual flew rate out of the core extract line. Secondly, it is simply incorrect to assert that the air flow rate in and out of the entire core is identical to the flow rate in the small diameter core extract pipe. If that were true, one would merely have to seal off the core extract line and there would be no Argon-41 emissions the radioactive asterial would decay away within the core and not need to be exhausted out the remotor stack. Se core is full of air, and that air passes in and out of the core through the many interstices in the graphite blocks and the eracks in the shield blocks and the numerous other passageways. A measured flourate in a small line is irrelevant to the flourate in and out of an unsealed core. ( 33. Iastly, and most importantly, the measured flowrate during non-fire situations is completely irrelevant to the flourate that would occur during a fire. Fires are self-feeding- they create convection currents that draw in and exhaust air. If this were not se, and the UCIA assumption were correct, no fire could ever ocour unless a moahanical ventilation system were feeding the fire with air. No house with closed windows could ever catch fire inside, if the UCIA assumption were correct, because no fans were present and measured flow rate inside was low. One does not need to provide a fire with airs it provides itself. 34 Sus, even were the University correct in its estimate of flow rate during normal conditions, in which an extract line fan produces forced circula-tion, such a measurement is irrelevant with regards the air flow possible in case of a fire. Se fire would produce convection currents, drawing air in and exhausting depleted air. After all, the airflow rate into a gas water heater is essentially sero when it is offs once the gas is ignited, however, the natural convection currents arented by the released heat provide the necessary airflow. And so it would be with the UCIA reactor.
- 35. ne airflow argument is spurious, seen if airflow were substantially restricted, that could well merely slow the rate of reaction rather than prevent it. Se airflow produoos two opposing effects-- it provides oxygen and removes heat. Restricted airflow will reduce heat loss, which can help to sustain the fire. Here are obviously lower limits to airflow capable of sus + mining the fire, but with the convection currents produced and the lack of:
a sealed structure, there is no evidence that those lower limits are approached for the remeter. (Furthermore,therearenumerousnoenarios involving the exposed graphite with a ready source of air-- insertion of experimental apparatus into the core, w= m ag near the thermal column, etc.) Not Inherently Proteeted Against Fire
- 36. Se primary anterials of the reactor (graphite, uranium metal *,
ungnesium, and so on) are combustible. Se reactor is not sealed it is
- In this respect.it appears unfortunate that the fuel is not an oxide, which would be far less susceptible to burning. l
essentially a pile of graphite and concrete H ooks with numerous penetrations for control blados, piping, and the like. Se core is diffused with airs otherwise there would be no Argon-41 problem from the activation of normal Argon in air. And the air within the core can readily be trans-Ported in and out of the cores again, if this were not the case, there would be no Argon-41 problem. Fires can occur in such graphite pile type reactore- witness the Windscale reactor fire, which occurred with the ventilation shut down. Modern graphite reactors are g====11y contained inside a leak-tight vessel in an inert atmosphere to prevent fire. Se UCIA reactor has no reactor vessel, isn't inerted, and has no containment. The UCIA reactor is certainly not inherently protected assinst fire. Vigner Energy 32 As iniianted earlier, the UCIA reactor's primary constituent material of construction is graphite, which serves both as moderator and reflootor, and urovides sono structural support. The reactor's fuel plates are cooled and additionally moderated hy light water. Se fuel is in the form of metalliouraniumalloyedwithaluminum,at134w/oU,formingthelow-molting outectie. Se fuel is clad with aluminum, which also asits at a relatively low temperatures in fact, both meat and clad melt at consider-ably lower temperatures than the constituents of most other reactor fuels. The control blades are also ando of a very low-melting substance, andmium, molting at 320'C.
- 33. Because there is no pressure vessel, contain===t structure, exclusion sone, or radiomativity removal system for use in an emergency to prevent fission products from reaching the puMio if released from W fuel, the nei==v barrier =6 min =t fission product release is the fuel aWan, 0.015 inch thick aluminum. Boomune of the los amiting temperature of the aluminum clad and the fuel meat, considerable attention has been given in analyses related to the UCIA remotor to the anximum temperature rise within the reactor that could sooompany various credible socident scenarios.
, 34. One of the potential sources of heat in such an sooident, either. singly or as one of multiple contributors to a temperature rise in moderator er fuel, is the emergy stored in the graphite due to its long-term bombardment by neutrons. Such bombardment causes damage in the graphite structure itself, kneeking carbon atoms out of their normal positions, and in the process storing significant amounts of energy. Sis is known as the "Wigner effect," after Engene Vigner who first predicted its occurence.
- 35. Sis stored emergy can be rapidly released if the graphite is heated over a certain threshheld temperature, beginning around 170'C. It thus poses a significant mooident potential, boomuss in the process of releasing the stored emeegy, more of the graphite is brought to the temperature where it can release its energy, and thereby asists a potentially dangerous positive feedback mechanism. Se more graphite that is heated, the more heat is released.
- 36. In addition to posing a simple thermal threat from Wigner release that could andanger the fuel's integrity, the graphite is, as indicated above, combustible. At be approxiantely650 $ )nCtemperatures
, it will ignite (estimated in the of in the presence Hawley air, inI an g hi, to exothermic reaction that releases large amounts of energy. Se Hawley Whi, (p. 34) indicates that the combustion of 1 g of graphite will raise 38 g to the
ignition temperature if no heat is lost, once again creating a dangerous positive feedback situation which, if started, couM re=M1v put the-l reactor fuel at risk of molting or of igniting. Uranina is likewise ! combustible, with an apparently consideraMy lower ignition temperature (~350*C). This is also true of the magnesium of the control Made shrouds. So, if sufficient Vigner energy were stored in W a relatively small initial temperature rise (to about 170 00) graphite, conM be sufficient to ignite or melt the core's centents.
- 37. The relatively low temperature required for =n===11ntr the radiation damage in graphite and relemming the stored energy points out one of the unfortunate aspects of the inherent design of the UCIA Argonaut. Whereas the low normal operating temperature of the reactor would be a favorable feature in most reactors, it has adverse effects in the Argonaut haa====
of the graphite design. Progressively larger amounts of self-annealing oceur at higher operating temperatures: eenversely, larger amounts of vigner energy I are stored at lower operating temps:stures, such as those found at UCIA. Rus, a low-temperature reactor such as UCIA's would be far more valneraMe to problems from Wigner energy than a high-temperature reactor, in which virtually all of h Vigner energy would be constantly annealed out of the graphite. 38 Furthermore, the small size of the UCIA reaeb does not necosearily mean that the amenat of Wigner energy absorbed per gram of graphite is likewise ==,11 In fact, were a large-sized remotor and UCIA's far smaller reactor to both produce 1 Mi-day of energy, all other things being equal, the amount of Wigner energy absorbed in each gram of adjacent graphite would be consideraMy greater in the UCIA reactor than in the larger reactor, for the simple reason that the larger reactor has far more graphite to absorb the same amount of energy, thus the energy absorption per gram of graphite is " diluted." All other things being equal, a large reactor with the same neutron flux as the UCIA remator, run for~ the same length of time, would prodnee the same amount of energy absorbed per gram of graphite as the UCIA remotor. And it is the energy absorbed per gram of graphite that is I the key to whether enough energy has been stored to bring any part of the graphite to ignition if enough air is presents given the proper config-uration, one unit of graphite ignited could release enough heat to bring naar additional units of graphite to the ignition point. Assessment of Visner Biern Storage is the UCIA Reactor
- 39. Se 1957 windseale accident- in which Vigner emergy release contrib-uted to ignition of both the ursaium and the graphite in the sore and resulted in substantial fission product release to the environment- pointed to the importance of recognising possiMe accident seguences involving stored energy in graphite.' It is thus neesenary to have an accurate idea of the
, amount of inach energy that might be stored in a reactor subject to irradiation l damage in graphite, particularly in remotors operating at low temperatures such as UCIA's.
n 4 -ga,a----.a J w A - '++ 4 A p - - 11-40 Se Hawley, Kathren, and Robkin review treats the Wigner matter in two brief paragraphs on page 37 of their report. They conclude that the amount of stored energy that may have accumulated in an Argonaut-type reactor like UCIA's .is approximately 5 cal /g, which they indicate is insufficient, if released, to heat the graphite by more than a trivial amount.
- 41. We Hawley, et al. estimate however is low by a factor of approximately 25-40 Se true level of Wigner energy that may be stored in the graphite of an Argonaut-type reactor such as that at UCLA is between 125 and 210 cal /g, given the calculational assumptions employed in the Hawley report and substituting numerical values that are more correct for the UCLA case than those used by Hawley. Such a level of stored energy is sufficient, if released, to raise the graphite temperature 600 to 1000 C above the temperature which had triggered the release, assuming adiabatic conditions. In sum,-
l an incident involving a relatively modest initial temperature rise in the graphite-- of roughly 120"C-- would be sufficient to trigger release' of _ sufficient Wigner energy to ignite the graphite or otherwise put the reactor fuel at risk of igniting and/or melting.
- 42 We Hawley report underestination is caused by a series of cumulative errors. First of all, the value chosen for the rate of energy storage at 30,C is low by a factor of gbetween 1.2 and 2. Next, the ratio of energy storage l at 50 C to that at 30 C is low by about 40%. In addition Hawley uses a thernal flux that is low by a factor of 3 3, tased on empirical measurements at UCLA. And he estimates a total operating history of 12 W-days, whereas the UCIA reactor has already run 19 W-days in its first 20 years and, if relicensed, can run an additional 37 W-days through the licensed period, given the operating restrictions at the facility. This is a further error of 4.7. The cumulative effect of these errors (1.2 x 1.4 x 3 3 x 4.7 = 26 to 2 x 1.4 x 3 3 x 4.7 = 43), a factor of 26 to 43, depending on which initial value is chosen for the rate of energy storage at 3000, is quite substantial.
%e errors are discussed in more detail below.
43 Se Hawley report takes the value of 0 5 cal /g per W-day /At as the best value for the rate of energy storage in graphite irradiated at 300 C, citing Nightingale's Nuclear Graphite, p. 328 However, on page 345 of the same text (attached), Nightingale states that "more accurate" values at low exposures range from 0.6 to 1.0 cal /g per W-day /At. 44 In order to correct these rates for the somewhat higher temperature found in the Argonaut's graphite, cited to be approximately 50 C, Hawley usesacorrectionfactorof3/5ths. Data given by Nightingale (p. 330) I for the change in the rate of energy storage with temperature, however, when graphed (see next page) produce an actual ratio of 5/6ths (ingerse 1.2).
- his yields storage rates of 0 5 to 0.83 cal /g per W-day /At at 50 C, as opposed to the 0 3 assumed in the Hawley report at this stage of the calculation.
1.e., assume an initigl temperature of 50 C and some incident which raises the0temperature, not 600 C to the molting point of the fuel, but rather a mere ^ 120 C to the temperature at which Wigner energy is released. Assuming no heat loss, the released 0 stored energy would be sufficient to raise the graphite to 770 to 1170 0, well above the ignition temperature of the graphite or theignition/meltingtemperatureofthefuel.
e _12-
- to_
' 9._ _- , 8.-
v---- 6 l benew eu -
- & = = 5. = -2_ - _- _y;- -
- _L_ --. -_ 5_ p- 4_ __ _ , _=_= _1 2 _ =& _= =- -. _=. M. . :-; . _ -; . q -;g _=g k_u z= m = a m. _ m-2. ==3 = - i =-3 = ---5 3 = - =
- - -. __ - _ _ _ . - - - - _ - - - - i=a__-
b-== = 1 bem-- i i z f I I I 1 I I I I ? I I . I e 1 1 I I I I I l I t f i I l l l . l . ' 6
! l I t I f +1 t * . I t . t . 6 . l . ; . . . , .i, , , . 1 . . , . . . 1 .
i e i g, . . 6 i . . i . . . i 4 . 6 . . . . . . ,# e i . . . . . 4 i e i i i . . 6 . . . i . , , ; . i . . . p 9 8-7 9-_ _ _=W5:__----_=.==-- .. = . . = = . .. - . .___ . - _ _ . -= .=:_=__ _=- . ;. . ._
= ___: _ _ = = - __
-_=- := ==_== r = ; ____ L - - - - _ _
--_:- - - :- :=- -_,5_=_r-=----
___ _= g=-gz _3 = ___3 3 _z_ -- _= = __ = --- _ 3 == .; --_3____ _ _ 33_ O-== n___ _ _ _ _ _ . _ _r : :_ _ _ _ __ __ _ 3.-- a.__ l \ I i i i i ' I f i i
! 1 I t ! .
t
- i( , ! i . e e e 6 9 . . p e i . ( , . . ! , . . . . i i , t , , i .
(, e i * * * . . 6 e .
- i 66 i e . e e e i . e i 6 . . . e 6 e
. i 6 ,
9
=
8 7_ - 6.- 9 (- h ==_7_ f
== -
$ M _= _-- .-N . . =
95 --* h =_=- _ l i-~=s. . -@C .-- l=O-iE
; _ . _ = = = _
T 2_53
- =.- .- _ = f=T=ATj Et----
= _ g E
3. h0- i ' 1 "t NO! . t 5 i L1 i I i N I i i l 6 N ! l 6 i l t .
- t t . e < l . e i * ! Ar i I i . . 9 . .l i . , , , , , i i i j
g_ i e e i e i e e i . . .ie e i i e e i e '(i . . . i i i i. . e e i . . . . . . . . . . . . 9_ S. _. S- ._ ..:- : .- - % 3 ._ m:" " Q RE - MMm
- i
*
- r e . i . . . i g i i i , , . . . , . . , , , . , , . , , , , .
0I roL1 @2 (k SLW MP
i
- 45. Using w equation given by Nightingale relating thermal flux and Mud /At (p. 328 of Nightingale), Hawley then obtained a rate of energy storage in the UCLA reactor. So Nigh+hrale approximation
- ist neraal avt (BEPO seuivalent) = 6.4 x 1017 M-day /At Inserting the correct values yields a rate of energy storage for graphite intheUCLAremeterof7.8to13x10-19 cal-on2/g-n,comparedtoHauley's value at this stage of 4.7 x 10-19.
- 46. Hawl en attempted to estimate integrated thermal neutron flux (nyt,in ) in order to convert, through the approximation provided ,
To estimate integrated flux, Hawley assumed a flux above, into cal rate of "about 10/gb m/cm2-sec."his order of magnit estimate was quite crude, as Hawley assumed the flux to be 1.0 x 1 , whereas actual measuronents ande at UCLA indicate neutros flux as high as 3 3 x 1012, **
- 47. Hawley then assumed that the reacter had logged 120 full power days, in order to estimate integrated flux (i.e., flux in n/cm2 per second as determined in M above, times number of seconds, to produce integrated dose.) However, UCLA reports (Amended Application, p. III/ )thatit had logged 19.4 W-days (or 194 full power days) in its first 20 years. In addition, Hawley failed to consider the next 20 years for which UCIA has requested the license. At a 5% operating limitation, as in the Technical Specifications, that would be approxiantely an additional 37 W-days, for a total of about 560 full power days to the end of the licensed period, in contrast to the 120 assumed in the Hawley report. ***
- 48. Inserting the more correct integrated thermal neutron flux into the relationship obtained fren Nightingale in 45 above one gets a potential stored energy ofa 560 full power days x 86,20 sec/ day x 3 3 x 1012,/,,2-sx7.8to13x10-19eal-on2 yieldingapotentialstoredenergyof125to208 cal /sofgraphite. nis is in sharp contrast to the 5 cal 7s estimated in W Hawley z + d.
- Hawley does not demonstrate that this approximation from Nightingale is univoraally applicable. It is used here only in following the Hawley methodology in order to demonstrate that given the methodological assumptions employed, but using more correct numerical values, a substantially different result is obtained.
** "Camma Flux Mapping of the UCLA Training Reacter" by George B. Bradshaw, h eters Resis, 1965, p. 53. The study measured'both gamma and neutron flux at a series of locations in the graphite. Se measurements were in limited locations and therefore even higher fluxes elsewhere in the core carmot be ruled out.
*** Furthermore, there appears to be some uncertainty as to the past irradiation history of the UCIA reactor's graphite-- whether, for example, it might have been previously used in another reactor prior to the con-struction of the UCIA reactor. Sus, the true anzimum exposure any be greater than the 56 W-days assumed here.
- 49. Thus, using the Hawley methodology and more appropriate numerical inputs, it is concluded that more than sufficient enerar can be stored in the UCIA remeter's araphite to produce if released. temperatures in excess of the ianition temperature of the graphite.__ magnesium, and uranium, and the molting temperature of the cadmium control blades and the aluminum-uranium fuel. ,
50 Mr. Ostrander, in his september 1, 1982, declaration asserts that it would take hundreds or thousands of years of operation of the UCIA reactor to produce enough Vigner energy storage to be of concern. He ! bases that assertion on the experience of the Fall == reactor, which was shut down because of swelling and cracking of the graphite moderator, and asserts that such deleterious effects were observed at Fall == after a far ! greater integrated fast flux than could be generated in the UCIA reactor.* l There are a number of flaws in Mr. Ostrander's assertion (among them, that it is not at all clear that the swellin as opposed to thermal or other effects)g ,was but due one to neutron need bombardment only aramine one of the errors- the ignoring of differences in operating temperature- to dispose of the matter. 51 Mr. Ostrander cites as tamis for his assertion above an answer by CBG to an interrogatory about the Fallan flux, but fails to mention the
- graphite operating temperature at Fallem cited by CBG in that answer. That
( normal temperature during operation is 600'C for the graphite, well above l the annealing temperature for the graphite. Above about 200 0 0, virtually l all of the radiation damage is constantly being annealed out of the graphite by the high operating temperatures. That is why high temperature grsphite reactors have essentially no Wigner problem. It is the low temperature graphite reactors, i.e. those reactors which operate at temperatures below which significant annealing of the graphite takes place, which must worry about stored energy. And UCIA's is a low temperature reactor. Fa l 1 = = uns not. The Critical Temperature for the UCIA Reactor 52 The Cort and Hawley analyses, as won as the Staff and UCIA reiterations thereof, are based on the promise that essentially no fission product release cgn occur should reactor temperatures remain in an accident below about 640 C, the molting temperature of the fuel anat. They therefore con-clude that if, in the case of Cort, airflow in the fuel boxes were cut off in a seismic event, the reactor would not be at risk because the maximum
- The graphite in the UCIA reactor has, by the way, apparently exhibited in the past some swelling or dimensional change, which could be physical evidence of Wigner energy storage.because, in addition to storage of Wigner I energy, radiation damage in graphite can cause dimensional changes in the
- graphite. Furthermore, the UCIA reactor is occasionally used to color diamonds.
If this effect is due to changes in the diamond's crystalline structure and not to impurities in the diamond, this would be further evidence of this reacter's capability of causing radiation damage in graphite, as graphite and diamond are the two crystalline forms of carbon and would react similmely to neutron bombardment. l l
c. 1 ( . temperatures attained would be below that critical temperatures likewise l in the Hawley xq wi, which indicates temperatures just below the molting i ! temperature in case of power excursion, and concludes that no fission pro- ! duct release would occur. l
- 53. However, all of these analyses ignore the crucial additional energy l
that conM be added to the incident.from release of stored Vigner energy in the graphite. Whereas Hawley indicates a power excursion couM produce fuel temperatures of 3900 0, just below that 0of the molting temperature, a graphite temperature rise of only about 120 0 is sufficient to release what appears to be enough Vigner energy to push the reactor far over the threehhold temperature for ignition and molting. Mus, were substantial Vigner energy stored in the graphite, an excursion not producing enough energy to melt the fuel alone any still have enough to trigger the Vigner Mlease, which couM sdd enough energy to bring the fuel to molting or ignition of either the graphite or fuel.
- 54. ne same is.true with the Cort analysis. Even accepting all of Cort's other assumptions *, peak temperatures of about 36000 are predicted.
While insufficient in and of itself to melt the fuel, such temperatures i would not necessarily be insufficient to push the graphite over the Vigner threshhoM releasing sufficient energy to melt the fuel or ignite the core. Similarly, heat sources deemed in the Hawley study insufficient to ignite the graphite by themselves any not be insufficient to cause release of the l Vigner energy, which couM then bring about such ignition.
- 55. %=___- a common-mode accident involving an incident insufficient in itself to bring about ianition or moltima could well trianer release of sufficient stored emerstr to bring about that result. And, in a sense, the concept of stored energy means this is an accident mode present through-out the lifetime of the reactor, just awaiting the triggering incident.
- 56. So, the critical temperature for the UmA reactor is about 170'C, the Vigner threshhold, not 6400 0, the amiting temperature of the fuel meat.
- 57. (Note that the A liestion (p. III/8-9) indicates that fission product release from alum aluminum-uranium alloys is eifnificant at temperatures of 400'C or higher. Furthermore, the Hawley study indientes the ignition temperature $of materials that any be placed in-core are substantially lower than the maximan temperatures Hawley assumes for a power excursion. And l none of the analyses ar=minas the effects of a'=ddian softening and volumetrio expansion that can occur at temperatures substantially below that of the outectie molting temperature. Even were there no Vigner potential, the critical temperature for this reactor would thus be considerably below the molting temperature of the fuel or the ignition temperature of the graphite.
Note mise, as indicated earlier, that uranium any ignite in air at temperatures well below that of the U-Al molting temperature, and th:.h nadmium metal centrol blades melt at around 320'C.) 9
- Note that Mr. Cort assumes no effect on thermal conductivity of either the fuel or the graphite due to irradiation effects. 21s erroneous assumption invalidates the final results, as they are dependent upon-the values used )
fcr thermal conductivity. I l
l 1 Conclusions as to Vianer Emerar
- 58. Aoospting the Hawley methodology and substituting numerical values more accurate for the USA case indicates substantial Vigner energy can be ;
stored in the graphite of the USA reactor during the license period. - This energy, if released, could raise temperatures well above ignition and l molting temperatures. Se energy release can be triggered by a relatively
==m11 initial temperature riser thereafter the reaction is self-heating.'
Thus, a number of seenarios of credible socidents which result in temperatures asserted to be below the amiting temperature of the fuel could actually result in putting the fuel at risk, due to release of the stored energy, through fire or molting, or both. l ' l corroaton, t%^^4== Damage
- 59. As indicated at the outeet, the primary barrier to fission product release at the UCLA reactor is 0.015 inch thick aluminum aladding. Severe corrosion of that eladding could produce substantial fission product release, including release of soluble, non-gaseous fission products.
l 60 Boomuse of the los utilisation of the reactor, fuel ori=4==11y installed in the core in 1960 oculd rammin there until the end of the proposed license period, the year 2000, due to the samil burnup rate. Forty years, amob l of which is spent in unter, could produce substantial corrosion of the I thin c1=AA4me, partigg]ggly if ygtgg ggglity 13 get adggggtely Emintained. l Failure to properly onlibrate or maintain the resistivity monitor for the primary coolant, and an inadequate secondary coolant monitor, could prevent
.tiscovery of substantial release until after it had occurred.
- 61. Se University now claims that the primary coolant leak that developed after the 1971 earthquake and required major maintenance and a long shutdown were not due, as cris4=11y stated, to the earthquake but rather to corrosion of the al==in== Primary coolant piping. If the far thicker aluminum piping unsase substantially cerroded in ten years of operation by exposure to the primary coolant, then the far thinmar aluminum fuel cladding could well be at substantial risk over the forty year period being considered.
(he University of kryland converted its remotor to low-enriched TRIGA fuel in part because of its inherent safety from power excursions and its non-proliferation advantages, but also becatae of conearn that its MTR-type fuel plates might be losing the in ty of their aluminua cladding after 11 years of use in a unter-cooled reactor.
- 62. Mmeh verk has been done on the attack of uranium ingots, elad in l aln=1===, through a pin hole. At elevated temperatures, air or unter l enters the pinhole, reacts, and the resulting oxide smells. His treaks l
acre Al skin, and the process continues fasters oxidation is retarded so much the ignition tem decomposition of UH )perature 3 can react is notliquid with reached. unterPowdereduranium(from and glow red, forming UO2 and 2H . Massive U metal anst be heated to react.
- 63. Numerous materials attack aluminua. Accidental insertion into the coolant of chemicals detrimental to aluminum, or experimental addition of some such material, or an attempt to remove anterial clogging parts of the coolant piping through addition of a flushing compound, or other acts could all result in severe and rapid degradation of th.e cladding and release
=_ _ _ . __ ._- . .. _ . . -
of fission products. Some materials react explosively with aluminum, which could be seen more devastating. Note also that hydrogen is produced when aluminua cerrodes, and that underunter aluminua tanks have exploded at reactors due to the explosive M8 gen-air mixture that evolved. (see Reactor Operating Erperience Report 70-3, attached.) Brolesive Reactions 64 We deseribed earlier the extreme danger of explosive reactions that could cocer were unter or CO2 employed on a fire at a reactor containing graphite, uranium metal, aluminum, and angnesium, such as at UCIA.
- 65. steam explosions and metal-nater explosive chemical reactions are possible if a power exeursion of the SPERT/ BORAX /SL-1 type were to occur at UCIA. The three remators all had their ceres explosively destroyed it such reactions, the onset of which and the initiating conditions necessary for which are not fully understood. It is not even certain, due to their l unpredictability, that such reactions couldn't occur even if the mart ===
l temperatures attained in the fuel were slightly below the molting temperature.
- 66. Other explosive reactions could result from the explosion experimental apparatus or irradiation of explosive anterials, ue d{pf to NEL improper experimental review, lack of adeguate pro sdures or supervisica, rule violations, failure to recognise the exp'esive nature of certain materials, or other mistake.
Conclusion
- 67. The ori=4=1 Hasards Analysis was in considerable error when it dismissed the risk of fire on the basis that none of the constituent materials of the UCIA Argonaut reactor was combustible. On the contrary, the graphite, uranium metal, ungnesium, and even the aluminum can burn.
In particular, the graphite, uranina and angnesium all have relatively lou ignition points (i.e., temperatures that could .guite credibly occur at some point in the reactor's operating lifetime, through accident, equipment unlfunction, building fire, etc.) 68 Subesguent analyses relied upon by UCIA are also seriously flawed in their assessment of the potential for fire and other destructive react $ons in the UCIA reactor. In particular, the estimates of Wigner energy that any be stored in the remotor's graphite are vastly undervalued: the potential for a graphite, uranium metal, angnesium fire improperly assessed: the predictions of peak reactor temperature that can be attained in an accident are far too loss and that consequently predictions of the magnitude of fission product release in case of accident are severely underestimated. l 69. The reactor is not inherently protected against fire. Substantial fission product release could ensue- over 90% of the gaseous material (those es volatile at the temperatures attained in the fire) and roughly of the particulate matter, dispersed by the driving force of the fire. I < _ . _ _ . ~ _ _ - _ _ _ _ . . . _ , _ _ . _ _ . . _ _ _ . - . . _ _ _ , _ _ _ . _ , - _ _ _ _ _ _ _ . - _ _ _ . . . . _ _ - - _ - _ . _ _ _ - - -
- 70. sufficient Vigner energy storage can occur in the UCM Argonaut to cause amiting and/or ignition, if released. Even mere a fire not to follen, fuel molting could ensue, releasing greater than 235 of the volatiles.
- 71. Se UCM remotor is vulnerable to numerous other reactions as nell.
including metal-water reactions during power excursions or fires and e6rrosion of fuel cladding. 72 Se threshbold temperature for the UCM reactor in accident is not S600 0, the molting temperature of the fuel, but 170 C, the trigger temperature for Vigner energy alesse. Sus, numerous accident scenarios which in themselves are not sufficient to put the fuel at risk any be sufficient to trigger the Vigner release, which could push core temperatures
- over the ignition and molting points.
- 73. He UCM Argonaut is not inherently protected against severe core damage from fire, Vigner release, and explosive and other destructive chemical reactions. In fact, numerous inherent design features anke the UCM Argonaut reactor uniquely vulnerable to serious accidents of the sorts described above: low temperature normal operation, lon-melting fuel and control blades, fuel ande of uranium metal rather than oxide, control blade shrouds ande of angnesium, moderator / reflector ande of graphite, all with ne sealing or inerting.
- 74. Iastly, we cannot emphasize enough that using water on a graphite-uranium-magnesium fire in the UCM reactor could.be disastrous.
l l l
CHEMICAL REACTIONS Exhibit List Exhibit Number Description C-II-1 "Se Windscale Incident", by C. Rogers McCullough C-II-2 Nuclear Graphite by Nightingale (excerpts) C-II-3 "Ca = Flux Mapping of the UCLA Trainirq Reactor" by G.B.Bradshaw (excerpts) C-II-4 L.A. Fire Dept. Energency Response Plan for fire at NEL C-II-5 " Aluminum Ank Explosion" (ROE 70-3) C-II-6 Photos taken within the Nuclear Energy *Imb
, .. u i
l i o i i - - . - , _ . . . , . . _ _ _ _ . . . _ _ _ _ _ . _ _ _ _ _ _ _ _ . _ , , _ _ , _ _ _ . _ _ . . .
kk EXHIBIT C-II-l 12 pages (cover sheat.
+ pas. 74-84) proceedings } Tio-7569 of the 1958 EM0lilEERillG & i!ArnTifA11r;R i .
Atomic ,'""N, sciEicEs tanAny Energy : -- ; c::a'ERITY OF Calif 00l!!A I Commission
* N' / LCS AliSELES 24 CAM-t
[L and (D@M18ITECE@E OMECE7 csranaf
. ;W 2.:
fu. .lre
. ~
i ..-l3:.g.e.fx.; 1
~@F@CebSTLOBR
.2 COMEREsCE held at
.i: / ..
Atomic Energy ~ t ,- UNIVERSITY OF cal.lFORNIA Commission S ANGE # Headquarters Building,. 1 . J 1959 . Germantown,'Macyland- _ sovT. eas Roou JUNE 24-25,1958,$ 0
/
c Af\ ' '. .
.-> ,: ,-l .;
f .. . f L UNITED STATE ATOMIC ENERGY COMMISSION .'. f 3 Technical Information Service . ' , . b$:f'h. .' k. _ . .$., s p: e
r j . . D The Windscale 'nciten: By C. ROGERS McCULLOUGH U. S. Atomic Energy Commission, Washington, D. C. To discuss the Windscale incident, I have broken my talk into two parts. First, I wish to talk about the accident and what actually happened. Then I would like to draw a few conclusions as to what this means and how we can profit by it. Let me state at the very beginning that this was an accident that was very serious insofar as the reactor was concerned. As far as I am aware, the reactor is still shut down, and the authorities do not know quite what to do with it. Whether it will ever be repaired and put back into service is a really serious question. However, the damage to people was zero or effectively zero. Exposures of plant personnel were minor. The highest exposure figure was 4.6 r, as I recall it. The damage to people off the plant site was even less. You are no doubt aware of the publicity given to the destruction of milk; this was a precau-tion that was taken in order to make sure that good public relations were preserved, that the confidence of the public in the Ur.ited Kingdom Atomic Energy Authority was preserved, and that there was no possibility of radiation damage. Actually the milk could have been held for a short time, and radioactive decay would have brought the activity down to tolerance (I calculate something on the order of two weeks), and there would have been no problem. Before reviewing the actual events of the accident, I want to quote a statement published in a British official paper: We are particularly conscious that the Windscale accident brought to the surface the latent public anxiety about the hazards of atomic energy work. Now that the na-tion is committed to a large nuclear power program, we consider of the utmost importance that the hazard of atomic energy should neither be exaggerated nor minimized in the public mind." I think this is the outstanding lesson of Windscale. Now, to go into the description of the accident: On Monday, October 7, they were carrying out what is called a Wigner release. The Windscale piles are graphite piles (a big cube of graphite, more or less a 50-ft cube). In this cube of graphite, there are channels, holes into which the cartridges or slugs of natural uranium Jacketed with aluminum are placed. These reactors when running at power are cooled entirely by blowing air through the re-actor. It is a British custom to push the air through the reactor; whereas in this country we pull the air through. (We have two air-cooled graphite reactors, a small one at Oak Ridge and a somewhat larger one at Brookhaven.) There are some consequences of this, as you will see. I do not mean to say that the British are wrong, because there are things to be said on both sides of the question. Before proceeding with the time table, let me add that you should realize that, when you have graphite subjected to neutron bombardment (which you must have to get the moderation), the atoms of the graphite are dislocated. The graphite swells and it swells nonuniformly. As a result, you will get a distortion of the holes into which the fuel slugs must be pushed, and after awhile the distortion is so bad that you will not be able to use some of the channels. 74
~
2% o
. r . i i
Now, Wigner predicted that this effect would take place in f,.a early days cf constructing i graphite piles in this country at Hanford. The Hanford reactors are water-cooled piles, by the t way. But nothing was known about how to take care of this situation. Quite by accident the British found out one time that annealing took place in one of their reactors. At the same time the graphite is distorted, nonuniformly or nonsymmetrically, energy is stored in the graphite and, if the graphite is heated up to something on the order of 400*F or so, the release of this energy can be triggered, and the temperature will go up still further. During this process you get annealing just like you do when you anneal' tempered steel; the en-ergy comes out, and the graphite returns to somewhat its original dimensions. ' So it has been customary in the case of the British and American reactors to periodically , anneal the graphite to avoid this distortion to channels. At the time the incident occurred they were carrying out one of their routine annealing op-erations. Now, as I v 111 point out later, you must remember that this is an operation for which the reactor pile was not designed originally. This is something that was added for good and sufficient reasons after the pile was built. Anyway, on Monday, October 7, at 7:25 p.m., they started up the reactor in the nuclear j sense. The castom was to keep the air flow down to a very minimum; actually, when they start, they turn the air off completely. (I will go through the time table and we will come back and j pick up the slides.) They ran this until 2:00 a.m., Tuesday, October 8, and then they shut the ' nuclear heating off accause the temperature had reached the point at which they knew, from ex-perience, that the triggering of the energy release would start.
, However, as they watched the thermocouples, which they had in the reactor, they began to fall; whereas in the normal annealing, the couples would rise slowly and level off, spreading the energy throughout the pile and uniformly annealing the graphite.
Now, remember that the objective here is to get all the energy out of the graphite. I may point out another thing, that, if you keep on storing this energy in graphite without annealing it, you reach the point at which therc is enough energy stored in the graphite, so that, if it were triggered, it wuuld go up to a temperature beyond control. In other words, the stored energy would exceed the specific heat of the graphite, and you could have a runaway ecndition; obvi-ously, you want to anneal in order not to have this occur. The objective therefore is to anneal the pile completely. You do not want any pockets of unannealed graphite building up this stored energy. When the Windscale people at least thought they had found thermocouples falling off, a de-cision was made that they had better give the reactor a second shot of heating. They applied l this second shot at 11:00 a.m. on Tuesday, the 8th. Then at 5:00 p.m. they shut it off again. During this second annealing, the operator slipped a little bit and ran the rate of reaction faster than the rules called for; however, this was not considered a very serious violation of the procedures. The rate of heating was not considered dangerous. It was merely a little bit faster than that normally carried out. At 2:15 a.m., Wednesday, the 9th (they had been watching these thermocouples all the time), they discovered they were slowly rising. Some of them had gotten to greater than 400*C, so they decided that they had better cool the reactor. This was part of the procedure they had been given. They opened the dampers. (There is a stack connected with this reactor which, when the dampers are open,lets air suck through them.) They opened them for 10 minutes or 15 min-utes first, and the temperatures came down a little and then came up again. l On Thursday, October 10, rignt after midnight, they opened the dampers again for 10 min-utes to try to cool it off. This cooled it very slightly. Again at 2:15 a.m. they opened the dampers for 13 minutes, at 5:10 a.m. for 30 minutes, again trying to let the draft effect from the stack pull air through to cool the reactor. This did not work, and at 5:40 a.m. they noticed an activity up in the stack-radioactivity. There are filters in the top of the stack; therefore there is a very high background, so the sensitivity of this method of detection is not very good. Nevertheless, they got a rise. This did not bother them too much because they had found in the past that, when you open the dampers, it pulls some radioactive dust through; this is more or less routine. 1 75 u
Than at noon on the 10th, th y found some activity on thi roof of the meteorological st>
' tion; this vas unusual, and they knew that something had gone wrong.
In previous experience, when they had slug ruptures in this pile, they had discovered that the activity that gets out and deposits on the ground around the station was the gross mfxture
- of the fission products. They immediately began monitcring and found this activity, but it was not very high, so they had no particular concern.
Then at 12:10 p.m., they opened the dampers again to try to cool the reactor off; now, they got a very marked increase in the stack activity, and they knew they had a slug rupture of some sort. At 1:40 p.m. they opened the dampers for 5 min, but still no real decrease in the temper. a tu re. At 1:45 p.m. they turned on the so-called " shutdown fans." They had decided that they had to do something to cool this reactor off; it was getting well above the 400* mark, which is dan-
; Jerous. It is apt to break the cladding of the uranium slugs which will be too hot and will catch tire. .
Now, another point here is that in the back of the pile they have a scanner that is supposed tr> d ?tect slug rupture. In carrying out the Wigner anneal, they cannot run the scanner, but, by turning on the shutdown fans, they can cool it down so that it will operate again. This turned out to be a vain hope. They also wanted to cool down the pile if they could. At 3:50 p.m. on the 10th, the temperatures-fuel temperatures, the thermocouples on the fuel temperatures-showed that these were hotter than the graphite temperatures. Then they took out some plugs. At 4:30 p.m. they took out some plugs in the shielding on the front face in order to look in. They found that some channels were glowing red hot, so they knew that they had a fire. Now, they decided that, first, they were going to push out the channels that were on fire, j and they got together all of the push rods etc., then the men put on the protective clothing nec-essary to work close to the reactor to try to push out the burning channels, but they would not push. The channels were plugged tight; they were red hot. The push rods came out with molten uranium dripping cff them. Remember, this is exposed uranium. The activity levels got so high in the charge hoist that they abandoned this operation. Then they decided they had to make a fire break. They pushed out all the channels surrounding those, roughly 150 channels, which they suspected be-ing on fire. This was on Friday, October 11. They realized now that turning the fans on full might either blow the fire out, or it might make it worse, and, if it made it worse, maybe the heat would get up the chimney and burn out the filters. If this should happen, then they would really have a mess. At this point they de-cided to install equipment to put on water if they decided they finally had to at the last ditch. About 1:38 a.m. on the lith they took a pyrometer reading on one of the channels and got a reading of 1300*C. They now had a real hot fire; they got some CO: from the Calder Hall re-actors, and they blasted it into one of the holes. It did not make a dent in the fire at all. Now they were faced with the decision either to use water or to let the fire burn up. They decided there was nothing left for them to do but put water in. There was some trepidation about this, as you can imagine, because they well knew that water on glowing uranitdn makes hydrogen. Water on glowing carbon makes hydrogen and CO; you have then a nice mixture of hydrogen, CO, and air, and you might have an explosion. But they had no other choice. They followed the technique learned back in the war days of putting out incendiary bombs. They put the water in at the top, using fire hoses. The water { getting in would trickle down and gradually cool as it advanced. It finally did cool in this way, i and they were able to put the fire out. l Then on Saturday, th;, i.:in, tka finally turned the water off. The fire was out.
' On the left-hand side of Fig.1 is the u.arol room and slug storage area. You can also see j the place called the charge hoist. This is the spot from which they had to work to charge the i pile and to examine it and try to find out what was going on.
In front of the charge hoist there is a double shielding wall. You see the charge space, f so-called. There is a wall between it and the charge face. That wall is thick enough to allow j 76 4
- le ,,E '
,.-;E,, - . ==
y -e . . J -- -- 3 , g _. j
I y "p a 6 . Or E W *^ ,, i d S '**W # 0 Y r 5_. _ _~ + {MDN$]C$. bfr'5'$/a.'.?_f.l.E_2.' . { __ ' ?NO.DYN__ _N_ ' m.dM.._ _ T . _"5b"4 M. '1.CY_.m@k.7 T.# i, p- --e % 7 m em. M i
. + . .a, . im. .a%
L
.. 3
~ $- p
'CCK ;_ L L ._;.~~ *.}. ;_'r_ f
, _. iI
,- m ., . , m .; H i I I T ~ i i i_T._I 1 _1{I T.--M:- _"_~'_.i"._ ._t r~.r~ r M.'r.: {' p* Jh - k. b;
~
.i.i i i i
'. i r -'rT..i
_~ .i"'_*
. _. T. _1 i
- 'l Iii i. I4_ i it a 6 i i i: l 1
. l __i_.l_l l.
- . l.i.!.I __._
l ' 24___..
~ _. - _ .
I i iIi l! Il Ill f 1 I lil
-- .lil. I
- i . .
Ill.li_,11 l f
.!!!- ! 1!!i n.,i.. .
ll
-+ t i D ii.i --
ITT;T
^
r'
;i 1 r i , ,
----.-.---..,..$. -~. } . -- , - . - .
j P
' i l l
-; ==cc. u s
I I o 9 i l i I
- cm. ,* . o. ,
5 l l $ " , - t , .U . l "g - -- s . I i y g 3 aa Rw l *3 d p.
'M- Eh .
l 5 y p ,., ._ 2 . , p! J.-.%. y. -. % .% l 3
- _ . . _ __ ... g_ = . ym _ . ;; , .. .
- we o j
[.____4,- d , c
, .s t
, S 13 v .# ,_ _ 3 0 m v_H o s t a__ _{_ iW, t
,, .pt. N l .- ,
5 , y, . - ,g ,, a T s =i f - 3
?! .-w = % .y -a
- - ~ J ' . ' . .n p. O g.-
a
/ ,
d{ni ,N', m -7 '-i -
- < [ l.t . , > r.b r 4L-'<
w ..
.,g I l
Qn $~ f4 " l
'.= p y,tj 1,QOp g.) , a,, q .-.
247
- a := = <
6 o M,0, ' g, .
.s u,,
l -
, //. . a
'~ "'
! d' '
T l l ld it ' 3 I / __ 4- l 33vd 20WVH3
.., [ _l2 E,[ TP": . ..! iFTTT- .i
- OmL ,, f
? _q gt; i
i e' / 5@ I ' e f s (+ Z r. r .=
'l \ ~ '
V
}' _ ._l'
/ N d--$ 7 - . un scooo i ~
3/~_ \ Cu- _ bp~ __
~
L* m 5$ - i - C. I' , J .-e man d
'~
b- og , h G8 Bo i e 4 - l E 3
- e. ,
77
3 the people to work th;r wh;n th? pila is shut dwn, but it is not thick cnough to shi;ld th;m with the pile running. There is another wall out beyond the passenger and goods elevator, f* i In order to work on the charge hoist, the shutdown fan must be run so that there is a pres. sure built up in the charge hoist that is greater than the air pressure in the graphite in order f to keep contamination from falling out into the charge holst. Now, the air goes through the pile, goes in the plenum chamber of the discharge face, goes ' to this chamber in the back where there are some exit air analyzers, which were not used, by the way, during this operation; then it goes through two chambers and up the so-called " vent shaf t," which is a chimney 40 ft in diameter and 400 ft high. At the too of the vent shaft are the filters. These are just ordinary glass liber filters, i something similar to those you put in your air-conditioning units at home. The next figure, Fig. 2, shows the slug storage, the goods hoist, and the charging platform where the men work. , The little black dots that can be seen on the wall are plugs that can be removed to charge f the reactor. Each plug services four graphite channels. Guide tubes to slide the fuel in at an ( angle are placed in these channels. Then it goes through the graphite block and into the plenum chamber in the back of the reactor, , There are monitors at the top of the stack, under the glass fiber filters, which show ac-tivity; but, because of the accumulation of activity on the filters, these are not very sensitive devices. Notice the control-rod system. (I might mention here that part of the trouble was that the Wigner release is a nonroutine operation, and procedures had not really been worked out, and some of the problems had not been foreseen.) The procedure was to put in the control rods at the top and actually run the reactor from the bottom half, but the monitoring chambers are at the top corners of this reactor, ', When you have this situation, the monitors, with the control rods all in on the top half, are not a reliable indication of power. In the second nuclear heating, they probably went up to a higher level and more rapidly than they really realized. The reason they adopted this procedure was that, if you generate the heat at the bottom due to convection current, a more uniform heating of the block of graphite is obtained. This, I think, contributed to their trouble, because they did not know exactly what they were doing in the nuclear heating. Figure 3 shows the fan pushing the air in the various chambers. The shutoff rods are at the top. Figure 4 show's a plan view; in one of the discharge ducts shown is the scanner, which en-ables them to take samples of the air coming out of every one of the fuel channels when the pile is in normal operation. This works very well in normal operation, but, when carrying out a Wigner release, you have to raise the temperature of this air in the back much higher than in normal operation; this jams the scanners. That is the reason the scanners do not work during the Wigner operation; they were not designed for this purpose. They had no particular worries about this because previous ex-perience had shown no reason why a slug should rupture during the Wigner release. The danger of slug ruptures is present when you are running the reactor normally,, or so they thought. There are sampling devices in the chamber which, again, they did not use be-cause they did not feel that there was any danger of a slug's rupturing, and they were not aware that they were in trouble until the situation had progressed quite a ways. Then they did try to run these samplers, but the indications were not very satisfactory; they did not quite know where they were. I think it took a great deal of courage on the part of these people to put water on this re-actor. They did it with fear and trepidation, and in talking with them they will not guarantee that they could do it a second time without an explosion. Let us now discuss the consequences. Remember, I made the point that all their previous experience had shown that, when you get activity out, you have the fission products in the pro-portions in which they exist in the slugs. Therefore, for some time they did not appreciate the fact that they had a probable escape of radioactivity because the general level was fairly low on the ground around the plant site. 78 s w..
- k YT
- m s h 1 4 m*An n . w n' m i n' i M w w m?s=M-
*
- E 3 d 1 r .
J N/*. g' l' {nh
' - a f i**?l -
l.. -
%i. t, 3 !".
f gI +6; * . g
,T, 3 i $ @DM f h .
S D o R l o a - T N O / c
, / b ,
j
/ /
/
/
Y E N [ ' M [ D - fH
- j I
/[/
/(d C -
L N l\ N xN le g% N'N ip e . Q , l 4 la d W f c s n i a fd o
,'q i
f, ' m ar
,i ,
g i a
- d
.'j
* ' ic r
% t e
%Me m
o Rl s
- I
,N
/ .
4
- 2 F
ig
' g w'%
.NTI
.' ,t >
Y S S' N S E S D R N = P Nl O E I E R B P D F M 5 A G N R F O T U H S H C N O I N IL O O C OE R T T T U O E L NP L L A W E E_R M O G GF R R T A A A H H L C CP a0 5
l [- I
~
.(...7___ ,_,_
I s. . 6 m. 4., s o ; I ; s a t > .. , , t--t- . I z ., . 1 op ,
. .l ,
e v . ',
* 'e
- f. - -.
in , t
- a
; ;' ;*U M M Hr ,;
; i T 1' ' f iW i .-
n 1 - 0
.f, h- . : / .r" .j-. Q :1 t4; M--- .
_- : 1 - e - .e i' .
,/ oaN' :
l ..
. W- : 4
_ _ _ __ i +L - .- w cc:
----. s!. ,a t .
u o e
' //,;,.
,, . , .j
, . o.,
J l l] . . . . .g. : _i: a w '. r-L . __ i: .:.......:..".e.,_... ,,
< z % ._
l
.. ... a _..2 ;:. y ... - _ __.
- g g.,.j. .
* ~~*~
6
' ?
* * * - O' *
.'r s na ;i_.gg, m; E, .i 1 - ,.i.'i h1
'...,,'l
.4 -
3 .,
-{ .
t
.0 ' ~.<* t "-.__:._.
aM .^
.. N ,
- f. g
*'e * = 4 4 -(= ' 2 \ '
'..j
..'., e l g H s'. 4- N. yl. NM 4
. Q=
1
- n. ,
m . . . l c
' .e u
i oy
*J t__. -G c
7 o ,l l a 1 l 5
< m. p.c [ . ._- = - o
'l L
,%_. n. .3- ,
o J ' bid' Y. O
- O l
]] .t\ oi byl.
i si . .. ' ,.n*. g E.! l l 80 C %.
..i M.5,$a$5&As.sk.40% L 5.1 agnf f*&.M&%;&i5k.nX GMMOOks -. w %.
+ ' .* * =L:muyl.Mwsm,sm,.
- ;.7.n;t m
. +3--: a;. ,, Q;g:i,,'%ig.w'*ir x *w _:onenth.c L,,fr ,m- > ,; ^
.. R % 'i-: -- i ?Wty
- d. Y.',:.,
. .e. ..
.:L '
'e*i'G,x
.[ %.m
- _ _- :=. .hvg5%
m- ~ _ - gg 7'.T;' y.w . p_A M. ~+ x.x.r.**y 7_
..,g '_:pg.fiWQt,1&g"m*7'i k.y:
_ ys _, .. ,.c.w:c
. ..x .'.y.~r. . , . .-:p,
,:. . .,. ,_ _. ..A_ ' . .,
me
'to 71 4ALLM
l
~~
7 r N i
/ s Y
vcur t swr g
/; ,
4 m y' , ,' l
"- HALF PLAN H ALF PL AN w ,
l 1' AT HIGH LEVEL AT LOW LEVEL [. 8- lJ itst ' wovst ac O \. ' l W
.. ' 'l. ., 1' ,-.
s l i ;:i- I- I I- - L' - -- '
, . r- I 2 m
. . , , i <
= ... ___ ___....___. . _ _ . . _ _ . . ,.__- .- - , i
~'. .s.. !
X " j . }4 a.......;..._,...._~.______.__ _wg g iggf-est ,y'H noos r._' T l ( i __ _ l ii
..n2 m
.' R co N? a o?x:-, fro a[_
~
n , j n
, .. , x -
. . . . . _ _ _ _-. 7 ._.. q ;
l r! . ,
.m ai o ser .
l [ j inn over ; l q......_.=.......
,lg_,_ _ ,
, ' IIllllllllillli - we rr i .
( p m l . . . . . _ _ . . _ _ _ . . _ _ _ _ _ _ _ _ . . . . . . .
'm" 'ist oc a -
Lr'
-r s -
- y. _ _ _ [ _L..;
g . I
! 4, ,
er _. _.__ l
-I- [ IE I 2 '--+'- [ -1'- L__] 1. J.
l [ N
) /co~ac<w rect \ !. , taur swives ;
accu , t noou h,, 8,...
. _ _ _ s j. e, c. ; {
.. ~
9 6 Fig. 4-Plan of Windseale pile. 81
o , . On Saturd2y afternoon, a day or two after they knew thzy had an accident on thIir hand:, they began to realize that they had an excess of tuline released. At this point they began to sample milk. They took, of course, the first milk samples near the plant site. The analyses of the milk took a little time. They began to find radioactive lodine and they kept extending the area of milk sampling. Volunteers came in to help them carry out the sampling program. He Army went out to help : them by gathering the samples; meanwhile, there were up to 200 workers in the laboratory an-alyzing milk samples. They finally ended up with 200 sq miles under a milk ban. ney analyzed samples from all the different dairy farms in the area. No tolerances had been set up for iodine in milk because this had not been foreseen as a problem, and they had to make a quick decision, which they made on Monday, I believe. They decided to set the limit of radioactive iodine in the milk at Y ie pc/ liter, which is an exceedingly low value. It turns out that a medical board has endorsed this figure, and I think the International Committee on 11adiation Protection will come along with a figure very similar to this, if not the same one. Now, in reference to the activity on the people, those who worked there worked at a toler-ance of 3 rads (or roentgens) to the worker in a 13-week period. Over the 13-week period up to October 24, they read the film badges. They did not have , any really good way of telling how much radiation the pecple received at the time of the acci-dent except the levels shown on those film badges covering the 13-week period. This was not an exactly accurate reading for the accident. Fourteen of the workers exceeded the permis-sible limit of 3 r. The highest figure was 4.66 r. Two workers were estimated to have re- . ceived 4.5 r and four others,2 r. . ; Thyroid surveys for iodine in the workers' thyroids were made, and nothing alarming was - found on that score. At the last report they were still surveying thyroids and were finding nothing very significant. The highest activity found in the thyroid was about Y pc. They offered to let the people of the vicinity come in on a voluntary basis and be examined for thyroid activity. Quite a number of people did this, and in no case have they found any-thing-the last I knew-very alarming. During the milk survey, they developed an instrument right on the spot that would enable them to take a milk can and check whether or not its activity was low enough to let the can pass without further examination. In this way, many samples were eliminated. Otherwise,it would be difficult to see how they would have gotten through. Again I want to point out that they did not need to dump this milk;in the first place, within a couple of weeks it would have been found to have decayed to tolerance. The highest sample was only 20 times the tolerance of '/ge pc/ liter. Moreover, if they had made the milk into cheese or some other dairy product, there would have been no real radioactivity problem. But, since children were consumers of milk, they wanted to be very sure, so they just dumped it. Now let us talk about the lessons learned. I made the point that these reactors were de- ^ signed in the early days and were not designed to can ry out the Wigner release. This # was an operation which had been added. l Therefore, the first point is that the Wigner release was an operation for which the pile was not originally designed. Point two: the Windscale piles had operated so well that con-fidence in continued operation without trouble had built up to a dangerous degree. The subtle-ties of nuclear reactors had been lost sight of to some extent (especially the possible diffi- ! < culties of the Wigner release had not been recognized). Three: there had been no systematic study of the accidents that could happen during the operation of the Windscale pile, including the Wigner release operation and the provision of adequate facilities to cope with burst slugs during the Wigner release. l I Four: the means of' detecting burst slugs during the Wigner release were not adequate. Five: means of measuring slug and graphite temperatures throughout the pile were not adequate. They had very few, relatively, thermocouples. l Stx: means of detecting the graphite fire were not provided. _ l 82
.__s,s
-. - ale- . - n'.s. A* = ,h *s t ,,
. m, 1
o . 3 < ' -NTW- 4 j~d g ' % 5- y,-9 ,G- Wy:'M 1* maga JYf s
. . 3" _
~;W. W. . , g_ $"'gd % es h y % ws %(dp
- h. 3.%%,$[s
- b_e p e @f [ .- - -
gi e Seven: there were no written procedures for carrying out the Wigner release with cri-teria determining the steps to be taken in case of an abnormal operation. Eight: there was insufficient technical manpower available to the operating crew to study the problems, if they became abnormal, and to advise of the actions to be taken. Nine: organization and procedures for dealing with the consequence of the accident when it occurred were inadequate. From the viewpoint of these lessons, we have examined our own reactor procedures (I told you earlier that we have two graphite reactors that are air cooled 1n this country). It did not take many minutes really from the time we heard about the Windscale incident until the operat-ing groups of those two reactors were right on the ball studying their own problems to be sure that they would not fallin the trap that had beset our British friends. Now, I want to make another point here. I think it is pretty obvious that the British had not foreseen what would happen. I do not believe that we can foresee everything that can happen, but I do think we can work very hard at it. As a result of having worked hard at it over quite a few years, I think the results of this care and foresight show up in our operating safety record. At the same time I want to warn you that we have been lucky, very lucky; I am sorry, but ljust do not believe that we are humanly perfect enough to avoid radioactive releases or acci-dents in an indus*ry as big as this. Thus we must be prepared to face some oversights that will occur. This does not mean, howeve,r, that we should not keep trying to keep our accident record as good as we possibly can. Thank you. DISCUSSION HAYES: Dr. McCullough, I am sure there may be some questions asked. I would like to ask you one, myself. I notice that as a last resort they turned to water; I did not realize that the British had to do that at Windscale. Do you have any recommendations or suggestions as to when we determine whether we can put water on the fire? We would like to know whether research is needed or whether, as a result of this accident, any studies are being carried on which will give us some clues to what should be done in the case of a fire. McCULLOUGH: Yes. We are making studies at Brookhaven on how you would go about putting out a fire if one occurred, but, first, I should say that we have assured ourselves that, barring some very extraordinary change of conditions, we will not experience a graphite fire. However, if a graphite fire should occur, we have taken steps to be able to smother it. So far, we feel this is a safer procedure than putting water on it, because, if you once get a fire in hot uranium and hot graphite, I do not quite know how you would get around the danger of a hydrogen or CO air explosion. Research work is going on; we are not satisfied that we know the ignition point of graphite. By the way, this is an amusing point. The belief had grown up on the part of many people in this country that graphite will not burn. This is nonsense. Graphite is carbon, and anyone knows that carbon will burn if you get it hot enough. But this glib remark, that graphite will not catch on fire, had become prevalent. At any rate, research is going on to learn more about the ignition temperature. It is a tough problem to solve, and we are exploring possibilities. HAYES: It seems too bad to have to depend on a bucket full cf sand as they did at the NRU reactor to put out a fire. Does anyone in the audience know of any research that is being done which will be helpful in the practical matter of putting out a uranium fire? We have heard a lot about the theories that have been developed. Dr. Quigley, have you any words of wisdom on this? Are there any other questions that you would like to ask Dr. McCullough? KNAPP: This was definitely a metal fire rather than graphite fire? McC ULLOUGH: Both. Metal was burning and graphite was definitely burning. KNAPP: Graphite did not produce any extinguishment problem, but the metal fire did? McCULLO UGit: Even a hot graphite fire and water will give you hydrogen and CO. KNAPP: Would your hydrogen necessarily have been dangerous? They apparently showed by putting the water on that they could have gctten rid of the hydrogen before it caused trouble. In metal it is so complicated you might get into trouble anyway. 83 i o
i
- l. . .
McCULLOUGil: That is right. They could not avoid a certain amount of air being present all the time. Incidentally, when they put water on, clouds of steam came out of the stack. ' Again the lodine, most of the lodine, wau caught by the filters, only a few thousand curies got . out past the filters. The base of the stack, by the way, was very hot. The stuff fell out, set- ! tied, in the base of the stack. KNAPP: Are there any figures on what the incident cost the British government? McCULLOUGil: Yes, there is a figure, but I cannot remember exactly. I can say this: it was less than $200,000. I remember two numbers. One of them is $80,000 and the other is greater than $140,000. I know it was under $200,000; that is neglecting the cost of the reactor, by the way. 1 q i i I
- r 1
a l
~
1 84
- + .:. n c< d; .k w w %.. / . ::4 d & 4Rterai-.. : .,=. ..#-
L I$r .2d.is bl. ?d&'kC +* hpa.&g&$t-l* llid$7. :_j&s:g&&f.Qg&;;
.. :: A ~~; c 5 =i.' h z : i 2;+ .K :.+ + wiy
- ;)g y ,- :- r ^ m ..;' . LQ&s; MD + :.9? .
.n _
'd ZZ'-:IEI~' C-II- ?
4 pacer (cover sheet
+ pts. 325,33q,345)
I
's i
NUCLEAR GRAPHITE e
^
Edit.d by R. E. NIG HTIN G ALE Hanterd Laborazones General Elecine Company ...
~ ~ ,,
....,.._..,,.....i...o.,.
oivi. ion o, T.cneie.i inform., ion unit.a se. .. Atomic En.rgy Comme.. ion 1 i u T a
-r AC A D E M I C PRESS ~1962 [
/43 New York and London t
l
?
l ; i
~
m ? . . ', , . . ,
,e'I * ,.c : - - ~_
.- 2 v.,.mam _ ;..
.- _.. _ 2. .3 1
l W r-L ,
+_ wom ene waew %..
* :.W;[r F%W;)pw a .
. ?S 5 5$# D 'g 3 P
4 J .
! i i '
P R. E. NIGHTINGA1.E 32S ;
~
h,l t : :
. I Two other methods for measurin total 3tored energy have b. < n -
h) ' but have met with little success. One involved an attempt to mea-v :
='
He free energy of graphite from electrode potentials established by gra. i
'W
_} l var:ous v!ution
- The -econd consi-tod in measurine the heat of n m l H - [.l 2 ll M with potasnum.' Although thtm att, metia wa- attr " 'n e We
.h- ad relatively low heat of react:on of graphite with potanium compar e lj li - -
I l j the heat of combustion ISI m 2 cal g at 66 to 95 C va 7500 cai ' .
?; 25'C), the excess heat of reaction of i. radiated graphite did not a -~ ..
nmsse } M the excess heat of combustion, nor was there a constant re!at on W'm. 3...- __q;ig .I' a
;p ._r Q; them for d!#erent sample < Furthewre. the axe " heat of reactio : .
7d:; ~ -- } ' decreate when an i=6ated <r.mpi- w a- anneni it u 4- m: - J
/ * l [ 0 the final state of the products of the potassmm-craphite reactica u- .in 1 - !' -
i m. . I1f -
's:- -
_ _ g -
,4 M ent for irradiated and unirradiated graphite both with respect to the m I position of the graphite-potassium compounds formed and with re-pw' gy --;;%,_
f' d o 1 h the structure of the residual :rsphite lattice. y nu >' In summary, we may conclude that it s possible to measure S , T
.i ,, _ if . i l
accuratelv bv diferences in heats of combustion if a ;reat deal of i c - i iy1 . taken. Often only one irradiated sampie :s available for a determinw. hteh a d. y' % - T r ygd . [i,;,y I In this case the pree:sion m a :s ,.imited in present techniques to =,> 'o ,-
,n , ,
cal /g. This is suficiently premis for most requirements, particuia:i> l
;. pl,f Q +
heavily damaged samples < >100 cal gi. Some improvement in preci- m
- ' l.
l, tj desirable for samples with small amounts of stored ener:;y.
.a i i. ;
- 0 i i ;
1 h 12-2.2 Bint.n-cr or TotAt. STonEn ENERGY TITH Exrost nE a , [ ;y 1.
- , c y The accumulation of stored enercy as a function of reactor irrmium
? -
l is sho'vn in Fig.12.2. The exposures for the measurements from the E T' L h l l.
. g; scale Laboratories were reported in units of Mwd /At. (Sec. 7-5 5-
! .1-recommended' con ersion 1 Mwd /At = 0 6 Mwd /At wasfail used. I{in+n'-
+i ;, s
- - . I y data" at 30 C, which are reported in thermal neutrons /cm' ,[T l' '
1 the 30'C curve of Fig.12.2 if 2,. , j < i& b
- M
', Thermal nyt (BEPO equivalent) = 6'4 X 10" w Mwd / At K.?
'g e
[, This is slightly higher than the value of 5 5 x 10" derived from othcr rm-p [,., G
,q j! erty changes (Sec. 7-5). The general character of the 30 C curve b -ma '
17 l
;j M h ay to that of most other radiation-induced property changes. The initN * [-
.} of accumulation is about 500 cabg per 1000 Mwd.'At. This decrear m ' E 3
tl 4 p Il h cal /g per 1000 Mwd /At in the 4000 to 5000 Mwd /At exposure range % A relatively large amount of energy 630 caLg, is <tored when gra p g
.I is exposed to 5000 Mwd 'At at 30'C. This is equal to 7.5 kcal aram ao l
jgh[;L - several times the estimated interlayer binding energy in unitradiate ! :' ite (Sec. 5-3) This amount of energy also corresponds to the mn er j9 K (Ch
, '!j, .
heat capacity of umrradiated ::raphite between 100 nnd 1550'C. rj j l ..
,;na a .; ]. 2)j '.'
E b
- 2+Q a r.
. Ifl
; j
~
*.ma wam .ege
- .e
_ -**TowsleA
*, I
- j
. i W w .
_, nerv
~ eQ._. %e.m.y ..,s. -:lGl.'1i
~- 't ~. , .;. ..
_ ;16
^ ' %*' '
~-
, - - + .
, w ;.6,
. . . ,,g, g; vg . , f ,;
*---S '.o. m . . -f~=n2% W " ~ c
% 4f 2. hyf$~h?.&&
V :; ... , . ! . ,
. .~-<.;. u ,
- .'G..s.5JL...A $ $ l &-U.---}.TrTs - &... D u.$%l:";d&* . ,nagg..u
.' : Y " ' ,.f:-Q*;%.$m.g: n ghg y ;
c
^~C.U'~./ENNj1,g9
.ns--. -
.y. t,m.,;_. e .6.eg{p,qr0"&%*Ep./g';gvpr "W s *g rhe ' ;
% .,.3 4 3 y # - - -w .r L 2
.::.:-v-emb;ss?;b if0M*CQ
- r W '#' "
~w:.:a:
. t'g
.'. .$'S.Q.yAtL7:2 .w,%*WX?e ..%,#
.n.w: m s.;;A w M.m-w 7@ k oc
.A.%. , 1 .r.
p$ y# fM 4 k . y wr
- ci-r-n ,
- .L;m. M.y5;g;agny7 w"Dhk p e. eT ! &a~I .e.~gm.lQ,.u,$%
~n
,n -
=, -
4i If ! NMN i
.bMS'I ll
~
. t g
3 i i 330 a. E. stairriscu.s , j- ) longer times are required to reach saturation at the higher temperature, , W - ' i gl ! The ratio kw/kr shown in the last column is equal to the dose required to
} -
j ? attain a given per cent of S. at T divided by the dose required to attain ti. fy ,} - same per cent of S. at 30 C. Thus at 200 C three times the 30'C dcm . QF n ; ('f
; necessary to get to any given fraction of S.. The 300:C curve seems to be an exception; however, at this temperature the amount of stored energy -
j,HV~ ;L ! 5 Of -{ n 3
,[!
l i 1 S. I small, and the values of S. and k are sensitive to changes in S of 5 in 10 cal /g. (*O' hfj] S ' Tabie 12.1 - PARAMETERS IVR THE EQUATION S= S., (1 - e ) I f lI % , I
,a
$ i$ ;
5 Irradiation S., , cal /g A, per 1000 Mwd /At Aw/A7 - Q .. .i temperature, .C m.- g3a iw4 ,
,.I , .!
m I 30 685 0.526 1
.K#1TQiBC '-
g l]!lh ' _ fl l ; 9 I 150 600 0.242 0.169 2.2 3.1 d.y 375 3 .d 200 T >N.. % Ie 250 200 0.151 3.5 ME ., (M.Mv:
,.ww , :
-l%
*-i m .1, )
!3 300 50 0.393 1.3 U ,
= :;+wn
;&p 1. ylI EgW.c - In the original derivation of Eq.12.1 by Newgard, h is the rate conc.nt d @ N ', .- -N. N:. -
for annealing displaced atoms. The numerical va'ue of k should inerm .- '
. wegg'J . 1 + g! p i i with irradiation temperature. However, except for the 300 C cu ve, the k i derived from Fig.12.2 decreases with temperature. There are several po--
%ggg-;phghr, -MM Q $ J yl j sible explanations for this apparent anomaly, all of which relate to the hke- j lihood that the simple model assumed in the derivation of the equatia -
-Q ny g
; '; p not adequate, particularly when the graphite becomes highly dame a 3
- g. p Therefore k deperds not only on T but also on E. The k in Table 121 j;jj
.-y,,.jh 4 :;
m ;lj should be regarded only as an empirical parameter and should not N ', i-gg . interpreted as having the physical significance that Newgard original y y G -s Y q( lj attached to it. j dd i . . - Il 1 i a The 150 and 200*C points on Fig.12.2 from the curves of David-~n' : a . -
; g g g 4"l p were obtained from samples irradiated in the Hanford controlled-tempen- , lgig y h' p ture facility (Sec. 8-3.2). These points fall below the curves at the cor-
[ u- -
- .m , -
Ebh N ?. f'. responding temperatures. The curves were obtained from irradiations car- i
?
n i kH . iU ried out inside hollow fuel elements in high-fiux reactors (DIDO, PIITA and DMTR) for which the intensity of the damaging flux was consideraUy l I t s.L _ _ - h .$ greater thao normal for a graphite-moderated reactor. It is possible that
'i .gdM p under such conditions the radiation erTects produced at a given total do-- .
9 are significantly greater (Sec. 7-5.6). In fact, an " equivalent Calder y - .
'. ~ ~ --9 [ l perature" (indicated in parenthesas in Fig.12.2) has been assigned *" to in
" ;p 4 y irradiations to convert to the irradiation temperature in a Calder ter'"-
{
- - 1 i H y -
'; at which the observed stored energy would be predicted. Althout;h 6 -
brings the points and curves of Fig.12.2 into better greement, fr-
-1:.G ,, ~, .. M WM=W I - studies will be necessary to establish with certainty the efects of S lh':.
Ydy' 4 . I M intensity on stored enerd and other radiation-induced chara;es in prop r" -
.YYM f[, !
5>. J$ ,4 m:;:m 3;,#."y
' ~ ' 21c '7p@" ~ T-
.. , yg
...e . w , - w . mvse. . m
~ .~ .; m m.
a.,:~M:2.
- ',. -$h-i N.3. % 14 @ @&E,5;QT. $$f
- w. W &). mw .m
I E " w. Q vp ^ - Wow Av.W h[f.*},*,,*9 '
**f T
- j %,M mp,g-g,h Sw e..
".",%**;@e N %= H5.4-i "* w. gg *- up -v _.e=m' ;,wg
.Ld5*.
m .. w@.-.u.. is e- ' 4x Q,,
,enn .3 p 4 %t .- - g-m-m
~ - " 2^"7 - ~ - - -
- 2 5 W 3 I"Q % c5 &,... % &.nvn:. &r?
[
~,
m.,m.
.f ue w A; -
v"'PJT frus- - -
- 1. .
f4 30 a OO 3
= , LC.w q' ,.
g'gl.. - _y _7 5
, . u
- 12. STORtD ENERGY 343 mates of the energy stored by interstitial-vacancy pairs give values of this
, 3,'3 g magnitude. The following values m kilocalories per gram atom have been 38'umed or calculated for H by various investigators: 220 ' Ref. 191, i
ul acane. pr. 230 ' Rei. lli, and 207 +1M 30
-t .:.ed t u t ..e nergy on.
Clustering will of cource. decrea e the average energy stored per
.in mtcrst:t:ai atom :s displaced atom. If it is assumed that interstitial C: croups have 2 h.mdmc -
j -
*cn.ed by the re!axation encr;y equal to that of C. vspor 1140 kcab molei." then the net increase in irbon atom is small The energy e ntent for the formation of' C. interstitials and vacane:es 's ratures su:ga-ts that the '140,2) = 170 kcal' gram atom.
., 240
. h e r+ .o rr-
. t. a jy rei:-
12 ~).2 Disi tacrutxT IsrT reon preatu-rxtsy .Titaarst trs rs atem and a vacancy (HJ For the calculation of displacement rates, the rate of stored-ener:y e energy of fortnanon on a build tp with exposure idSe dE) must be measured near :cro exposure at f H2 10. temperatures low enough to ensure that no clustering or annealmg to ur-s
;12 6) ;
faces has occurred. The alternative is to estimate the ener v stored for j each atom originally displaced at the temperature for which i.% d5 is naa mred exper: mentally. 3 measured.
~ he past as to the proper The lowest temperature for which dS,dE has been measured :s about
) kcalegram atom. 30'C. On the basis of property changes produced by irradiation at liquid-been measured and must helium and liquid-nitrogen temperaturas iSec.12-4.2). it is estimated that af a vacancy involvas the 30 per cent or more of the displaced atoms reinte; rate durm; irradiations The atom combines with st 30 C. The displacement rate is therefore tace. The annealing of a ge surface site, where it is Rsi3 H i
/!?.7)
I e 3hown in pirt b of Fig. At 30"C 1&dE from Fig.12.2 is 0.5 cal 31wd At. More-accurate values
- in graphite
- s 163 kcal/ derived from measurements at ve y low exposures range from 0.6 to 1.0 l - between a vacancy dif- cal, Mwd /At.t Assuming a value of I cal / Mwd /At, also Sec. G-2.6). If the 4 6 X 10:2 = 3.3 X 10" displacements stion enerw'. for acancy R 5 240 3 X 1000 g - Mwde At
.i probab,iy coes not ditier which, for nuclear graphite with a density of 1.7 g/cm 2 is z calculated by Dienes' l
.-wmation is tagen as 170 R = 17 X 10" du. placements i cmi - Mwd / At nc) annealing bezmning I ro. -ince. f rcm part b of ; Values of R calculated from displacement theory and from other prepern n of vacancies is apprcxi- chances agree within an order of magnitude with this number i5ec. 7.4).
l 12-6 Relation of Stored Energy to Other Rcdiction-induced
. 'icr and Hennig 2' state [
hert and Lee 2" a cume a i Property Changes 1Gntcr" plam the value Careful and time-consummg measurements are iequired for the a t er-
!; i>f 70 kM gram atom mination of 'tored energy. In addition, the radiation e:Tects are it 'c a -t partiaHy removed wl.cn stored-ener y release curve < are determmed and
- 1 atom Mcet other esti- 1.;ce Paf. 41 for 2 e ussicr4 c4 iS lE measurements l
l -m -- --.- __. c-m .
.y.m., ..,, ., ,,,,m.
. , - - . .m ,. 3 ...
e . .. . Arra ars.;.m v.
- , , a s.,. a.w lg
-Jd?/**.~,'.fh.g$f-MIIl*sQ NSfMN'M'
-i
, ,~<
?- . . . - .J~ . m
- ' ,& &,-* 4;,.-3K, .. . upyLig&.:
35 r, *. ,"/. 2~~yMq g .: % 'y g y q ,. 5.
. 2
. *. ~: ' ~~ =-cby;.
(
i -- EXHIEIT C-II-3 2 pages (cover sheet +
- p. 53) l
^
.L.. t.1
/ O *"n .'. T. m*1
* * ~LT~ ^UrsLAT
- * * *' L ""J'
. .' 'i s't .
Los Angelec I I Gamma Flux .apping of the ' U.C.L.A. Training Reactor
- A thesis cubnitted in partial catisfaction of the requirements for the degree :hster of Ccience in Engineering by George Brown Bradcha..
Commi *.ec in charge: l
- rofessor Thomac E. Hicks, Chairman Profescor F.olf Cchroeder Profecaor Paul C. Farrincton l
1C)S5 i l
gb , -a 3. , f..A s ,
;e Table 4 Total done Total neutron Dose due to % dose due measured, r. Plux, neu/cm2pec. Gamma neutrons to neutrons doce C.V.T. at 1 kw for 1 hr.
6" 19,000 1 600 10 12" 44,000
- 1200 1.12x1g0 8,000 *25% 42% 11,0001 20%
2.4x10" 17,300 " 18" 63,700 12500 2.9 20,900 " 39 27,000t205 24" 77,300 *3330 32 4 3,000 1 20%
*24" 64,000 e3000 3 35 " - 24,000 "
31 53,000 t 20% 64,000 *1600 1.85 "" 13,300 " 21 51,000 t 20% 30" 3.1 22,400 " 36" 40,000 11400 1.98 " 14,300 " 35 42,000*20% 42" 18,000 1 500 1.16 " 8,300 " 36 26,000h20% 48" 7,300 6100 46 9,700 t 20% 5.ox109 3,600 " 49 3,700120% 54" 2,630 t30 1,500 " 60" 1,020
- 20 ' 8.8x102.08 "8 57 1,100
- 20%
635 ," 62 390* 20% ~ West half of core, at 1 kw for 1 hr. ' 6" 73,000 *4000 3.oxlo" l0 22,000625%
**8" 57,2001900 2.6 " 18,700 "
30% 51,ooot 20% 14" 34,700 600 1.9 13,700 " 33 38,000t 20% !S 22" 13,800 11740 40 21,000h20% :d 8.ox109 5,700 " 41 8,ooot20% 30" 5,3801150 2.6 " 1,870 " U 35 3,500120% y
- East half of core, at 1 kw for 1 hr. , ,
< ?
- 6" 70,500 + 1300 2.9x1010- m 8" 53,800 1 2600 .2.5 " 20 39001 25%
18,000 "
~%
30 49,o004:20% 7 33 36,0006 20% l . u n 3, -! i, - -g \ , , :,.o: : s m
- Measurement made in Li1 thermal: neuthbri hhield m
**W.V.T. access port 16 cation.4 -
'.y i . O , , ' corrected fog.' epithermal neutrons only.
4
%9 O
~
'j i , ;II; -
M W; ,
- p. . .y .> y .
G to
+ = * * - me se p -
V . c:1n r o r.r
, - ( IY OF LOS ANGEL' 3 OCPARTMCNT OF FIRE C ALIFoRNI A
* 'Fi![10M MI55t ONEI3 0 00 N O H ' M AIN 57 4 U S rsO 3 2 47 LC% Af.GE Li & CALIF T0012 gp.c ,;w%,, s
. "'L','".',* 4,i,2-di 1 -
"T.:.'l .r. ..?.', " ' '?
l t..n-,......
- s r s, . , L,L r --l- y - ' ,,,,,,,,q,,,,,...,
A P, N f, E 15$ L A N E sosN c o AwsoN . o, o Exhibit C-II-4 DOMINCO NODftlGUEZ tom BRADLEY page 1 Of 2 MAYOR September 16, 1981 Donald W. Reichenbach U.C.L.A. Fire Marshal Community Safety Department Office of Research and Occupational Safety 112 Physical Plant Office 405 Hilgard Avenue Los Angeles, California 90024
Dear Mr. Reichenbach:
Emergency Response Plan, Nuclear Energy Laboratory at U.C.L.A. Attached is a copy of the plan which you requested. If you have questions or comments, please contact Captain Leslie E. Hawkes in the Planning Section, (213) 485-6034. Very truly yours, JOHN C. GERARD C)igf Engineer and, General Manager [ / 1
'\ l n u. ~ %~
ROSS L. WILLIAMS Battalion Chief Planning Section RLW:LEH:lmg i Attachment - cc: Battalion Commanders Brtttalion 9 AN EQU AL EMPLOYMENT OPPORTUNITY - AFFIRM ATIVE ACTION EMPLOYER
=
f September 1G, 1981 . page 2 of 2 E.5!ERGENCY RESPONSE PLAN, NUCLEAR ENERGY LABORATORY AT U.C.L.A. - The Los Angeles City Fire Department provides fire protection to the University of California,'Los Angeles. With regard to the Nuclear Energy Laboratory specifically, the following operational plans are in effect.
- 1. A first alarm assignment will be dispatched to s the Laboratory.
- 2. Upon arrival, the first Fire Officer on scene will contact campus or building security to determine the exact location and nature of the call,
- a. Fire Only - No Nuclear or Radiation Problem The Fire Department will handle as a routine structure fire.
- b. Radiation Problem The Fire Department Incident Commander will confer with the Laboratory personnel to determine the extent of the problem.
Notify the Fire Department's Operations Control Division who will notify the Los Angeles County Department of Health Services who, in turn, will dispatch a team to handle the control and decontamination problems. The Fire Department will also dispatch the on-A call Radiological Defense Officer to the incident to assist. The Los Angeles County Department of Health Services is under contract with the State of California to handle
. radiological incidents.
, The Fire Department will monitor the ,
incident and will make rescues if necessary. The Fire Department will g prevent entry into the area until the Los Angeles County Department of Health Services' team arrives on scene. The Los Angeles Fire Department will maintain command of the incident.
- e. . .- - - ==ex--n=== -.
hihiWt 0=II=$ p.s. 1 or 3 Nuclear Safety Informatur. Center i ASN0EMAL REACTOR OPERATING EIPERIENCES 1969-1971 1 U.S. Atomic Energy Coussission Division of Reactor Licensing , MOflCI - Tem rt ==s e et ==se W b m. .w sered II.E.d$!WM m h h c en. U ee.d Se.e
- m. ,
ee... , t it. U.ee.d 34
,.cen . e.,e As ees 3
,y w
= =i= ~ w . v.m . .ir en e. , .
I W , were ey .e emee.d..s om ag y hem innen, = re .appe enen, ser m. ==.rr , se ===. . c.m c , ww.me . m=e.=,.= d. = e m.: a m.
.samme ed am wri= me.e drun MAY 1972 OAK RIDGE MATIONAL LABORATORT Oak Ridge, Tennessee 37830 operated by UNION CARSIDE CORPORATION for the U.S. Aft 3 TIC INESCT (X3081SSION S
I
sas. 2 or 3 ROE 70 3 Alimide n Task Repleetes M A leek developed to en edesweter aleontem ballast task which initially contateed air at atmospheric pressure. h tank wee dreined and left over-eight to dry prior to weld repett of the leek. The mest day. W UI T M URLDER STEDE AN ABC. TM TAM EIPtASED. tegentisettes showed that as BIPl48??E ETBMSN-AIR HIEUTM MS EpcLVED FBDN COEROSION Oy TWE 4 LIS Cire estances i irradiated feel eteensta was equipped with as slaise 1 task to provide seen bueyesey for eene of sevement of the eart. The 3/4-toet: biais. rectangular task wee appromiantely 2 ft. m a ft. = 1 ft. Rech e ' I the two bung holes were opened to perest the leakag and the tank uns left la air to dry everst$t. The took wee open to the . etasophore the leak. for a estal of sheet 24 heere prior to starting weld repair of he the welder streek as are, the task espieded. N welder wee koecked ever backwarde. seat to the keepital for a abeeksp.Ala:heep he appeared to be meisjered, he Es significant lejertes were feued and the ash wee able te veters dont occurred. N emplosies sensed a severe to woth eering the ease shift la which the teet-belgie l task and a esoplete rupture aloeg see of the seems. g of the sides of the 1svestimaties chemistry eestice of the operettag ergeoisettee.As tevestigotte the sopleded teak wee found The interior eerface of to be emeted with a een-mifese layer of yellowish-dite crystallies esterial. i eesting wee tahme by sereptes the serfees with a opetela.A Aheet 1 gram sampleof of th ( esterial tedes. was removed from severe 1 areas totaling appeestentely 50 square emplesive ar.d insol dte to unter.The sample esterial was see-radiens graviestrie methods chosed that the enjer seestituent une Ala0s. taken.Samples of gases from bellast seeks used eder sietter esaditises were estely 22 by vetime of hydrogas, with the belases being air. l Free the appearenee of the fatorier eerfees of the emploded task. correeton by esterasemed the investigator wayer. that all 1storier eerfasse were gently espeeed to cording to the reesties The serveetes of aleteen psedeems hydrogen ae-l 5 Be4 + 2 A1
- Alset.23:0 + 3 se 0
--w- -
w.-._ -- ---- , . , - -_ ._ _ . .
pse 3 of 3 l 65 i RDE 70 3 The total interior eerface wee 3700 square taches. It was calculated that i et least 35 litete, or 153 by volume, of hydrogen wee produced. An addt- ' tiemal questity of hydrogen wee probably produced by redselytte decompost-ties of water, bet since the emment of water and its emposere to gemme rey energy wee met accurately kasus, as asenkgful calculattaa eeeld be made. This quantity wee conse:vettvely disregarded la the esapetaties. ! A calculattee of diffwise three$ the tuo venta showed that the stae-ophore la the tank after 24 heure contained apprestantely 4.SI by volme of hydrogen (the louer flamenh111ty limit of hydrogen la air la 4.12). The conclue16e of the tevestigstere see that the aset probable cause of the egleesoa of the fusi cart be11ast tank wee the ignities of a hydrogma-ear afstere in the tank by the solding are. I Cervective Actise 5l=ce this escurrence all almiam hatNt tanke used at this facility are betag replaced with statalase steel tanke uhteh will be filled with he11m. Alumin e tanks still ta use are new filled with hell e and persedic samples are tabas to assure matatemenee of am inert atmosphere. In addities, solding procedures ekteh require seepling for empleetve genes prior to sold-tag will he striatly entereed for all eleoed eyoteus and contatasse. 9
- , - - _ - - - - , - - - - ... , _ - _ _ _ _ _ - _ _ , __ , _ , , , . _ _ . _ _ ,}}