Hazards of Waste Storage at Hanford Works Part I

Page 2

peculations on the hazards involved i n the accidental dissemination of waete :terials a t Hanford Work are required for all foreseeable circumstances includn& enemy attack or sabotage. ff ect s of bombing on waste storage tanks, the regort hae been held up beyond be antic i pete d date. irst pzrt is concerned only with potential disaster to the nuclear reactors. ran the nature of the process, these are major waste etorage unite. Due t o delay i n obtaining data on the potential It is therefore being prepared i n two parts of which this sprat iona procees waete stored i n buried tanka or otherwiee will be c idered iter. Preliminary speculatiom on this topic have been given already PV, and t is not anticipated that much improvement in this phase will be effected. nall cases, it i s not expected that numerical conclusions vi11 be valid to a actor of two, but they should be reliable within a factor of 10. This is the st is ted eccuracy of the excellent British reports of the Plant Location Penel, iich have been extensively used in this campilation. lis report is a joint effort of the staff of the Health Instrument Divisions. Icp rti cul sr, F. G . Tabb is responaible for the calculations of fission yield, id aquatic biological data, reapectlvely. .- Storage of Redioactive Wastes i n a Henford Nuclear Reactor I n an assumed average operation of 200 days a t 275 MW,a Henford reactor will rste { 584 megacuries of fission products 220 metric tons of uranium 200 mejizcuries of neptunium 2.7 x 104 grams of plutonium By logarithmic classes o f half lives,the fission products are divided thus: -

Page 3

ram en atomic bamb is said to be 2lbe.,witha presumed activity of 10,000 of fission producte approximates 9lbs ; fia el on product release e decay of the total f is el on product6 i s specified by &?hi8 I s campatible with decay lave Activity = cat ant for the fire t5 hours he decay of atamic bomb fission products follows Ptera few hours, the release of all the activity fram the nuclear reactor would xeed that from an atamic bm bat the equivalent the. imea greet erin the present case, because of the higher relative content of longived fission products. The hazard leat all te initial activity of certain key fie a ion producte i s

Page 4

These have been diacuesed la the A.E.C. Reactor Safeguard Conrmittee report "Review of Certain Henford. Opsratiom"(2). The mac ti vlt y of the Hanford reactors is adequately controlled under all foreseeable condition8 except Planned e hut dm vithout addition of poison After the decay of t h e natural poison Xe135, the unit could becane reLOSE of water in o m tube * If accampanled by failure of the Bafety devices, the metal i n the dfect - ed tube would melt and be ejected with em accompanying minor steam axplowould be a serious local problam. access for addition of poieonmlght be denied, If the final dieposition of the act l vlt y prohibited the rapid dumping of shielding mter lal upon it. A n explosive run-&way 1s not vleuslizod in such a case. It a11 be assumed that 8UCh an opera ta error I s inconceivable. The rrvsnagement of the ejected 0.45 megacurlee of fie elm products If the reactor were then 8hut down, The loss of cooling water Increases the reactivity of the unit by about 850 in-hours, which can be held by the vertical safety rob (1700 lnhouro) or in emergency by the third safety device (1800 in-hours). uncontrolled condition requires the concurrent failure of both safety This i s visualized t o be poeelble under three circumetances: - There is no history of earthquakes in this region of sufficient Intens it y to disrupt the water supply t o the reactore and t o damage the eafety mechanisms. Nevertheless, there are records both in the western slopes of the Cascade Mountains and i n Montana, of earth movementa not e ip if i cant ly below that required t o produce auch damage. The relatively short recorded history of the Pacific Northwest casts legitimate doubt on the permanent stability of the unlts against could be affected simultaneously. developed in this report. This I s the one pleusible case In which all the reactors Thla extreme case hae not been The Divleion of Bioi-y and Medicine report6 information fram the National Mlitary Eatablisbment to the effect that although the reactors are considered relatively invulnerable t o bamblng attack, the slab cmt ruc tion of the main shield leaves ems poe el bill ty of dle - ruption of t h e shield. denolleh the vulnerable safety mechanlame, and damage the water aupply. Advance warning of a few seconds only would pennit safety t o be assured by insertion of all rods. with disruption of the unit ehauld not lead toa nulawny cad it ion. Evidently an enemy attack could, by chance, Subsequent bcnnbing, even -. --_ --' Tho PI2 ro pode adequztely diecuae tho CRBO In which tho ejected mots1 fran

Page 5

Bcmbing should hold a law priority In the foreeeeable hezard pattern. Sabotage by M eaeqy agent o r by an unbaknced pereon intrinsically loyal tothe United States appeara t o be the outatanding potential vould be negligible. There le no p o h t in elaborating here on the feasible ea bot qe methods. With two or more pereons in colhelon, the r lek of detection The general picture la that of rupture of t h e water mpply, together w i t h lateral motion of the reactor sufficient t o cloee the thlnbles t o the vertical or the piping tothe thimbles would be broken. safety ne cha aim is an alternate to the lateral motion. The fluid of the third safety device would fall t o enter the thlmblee, Dellberete tampering Kith t h e

Page 6

1. The Uncontrolled Nuclear Reactor The appraxlmately 6 cubic meters of water in the pile can be raised to the boiling point in about 7 seconQ a t normal operating power. cOntr8l tube8 would b o i l in about 3-4 eecand.8. reactor lncreaeee rapidly; the exceso reactivity above critical lee a lh to be Ke = 0.025 without water. On elementary reactor theory, the power i n - cresee sfb r lose af water l e The power loPel of the Pmer after t seca 20g,t: o The 250 tone of uranium inthe etructure l e boiled by about 1Clkale. in te gr atl an,thie vi11 occur in 13 eece. The actus1 couree of evente l e the boiling of the central uranium in a fraction of th ie time, vith poealbly the termination of the reaction. rapid that there can be no eu bst it ute control. The res1 rate may be affected by the unknown temperature coefficient, release of stored energy from the graphite, releaee of Itel35 ae t h e uranium melte. In any caee, the sequence of events is B O To determine the environmenfal hazard, we take two eimplifled plcturee. baaie: all the uranlum Just bolls : 1.16 IlG%al. an equal amount of heat goes i n t o the graphite, aluminum and or: 8- of the etared energy of graphlta (calculated ae 2.6 xl bl kal.at the preeent time) I s released, thus terminating the nuclear reaction at a laver integrated baeis: the Reactor Safeguard Cdt tee eta teeth at “it does not eeem possible for the mstal tempereture much t o exceed 10,0000 C at tho time when the course of the accident it e elf dierupts the chain reaction. Thie temperature w i l l be reached i n a power buret of about 7 . 3 x 107 KW eeconds” . Th16 is equivalent t o 1.8 x lGOc al., which le l neuf flcl ent t o boil the urnnium. the puwer burst required t o raiee the central very different phyeical conatante have been used for the metal. This report u ~ e s :- Preeumably the glvon fime refera t o Specific heat of uranium = 0.028 cal/gm a t any temp.

Page 7

X U the uranium boiler and approximately 1.5 x 10l1 cal. goes into the hot gas bubble. This l e equivalent energetically t o an explosion of 150 tms of high exploeive. The nuclear reactor w i l l be explosively shattered, although this le probably minimized by the heavy ahlelding. More probably the hot gaaee will be eJected violently fran each of the reactor tubes. I n either case we Bhallabendon the phyaicalmodel and proceed t o calculate the behavior. of t h e @e cloud a e If It a regular high explosive devrtonation, folloving O.C. Suttm ’ e method Behavior of the @e cloud The instantaneously genereted heat in the cloud l e about 1.5 x 10” cal. the volume V after the Initial rapid expanelan t o atmospheric preerure I s The Initial radius = 80 t o 100 metere Note t h a t the mean molecular weight of the exploeion game l a taken aa 25 for the normal productfi of erploaicm, and bears no rel atl on to much higher value I n the real model. The velocity of aecent from the original poeition after expansion I s given by velocity of aecont acceleration due to gravity = 3.8 metere/eec epocific heat of game produced = w 1/3 mean denelty of a i r over the range = *lo00 &uete?? index of turbulence 1.75 height i n meters above orl@nal expansion location We take a higher value because 3000’ C would not start a cloud of uranium

Page 8

me average Kind speed i n the Hanford region i s ject cloud riees t o Perhape an average value of C between Z = 0 and Z = 4000 i s applicable. Thie would send the cloud higher, and u e take the conservative low value. Activity in the cloud The cloud containe M = 584 megacuriee of fl sei on producte. It moves horizontally with velocity 9 meters/sec. bulence, the msxlmum concentration e t the ground i s Under average conhitions of t u r - 0.073M curie/meter3, where r meters is the distance frcm t h e cloud center t o ground. These concentratione occur ata tine, t seca. after the explosion (strictly after the ascent) given by The probable value l e about 5hrs.

Page 9

: The corresponb?ng distance from the reactor l a ,lo0 miles. Additionally, the activity after5hrs. has decayed t o approximstely one half the ln ltl al value. Asa man stands ata point 100 miles from the reactor, the cloud aubat anti ally contributes it a act lv it g to hh In 11 mlnutea." The time varia tim of concentration a t the ground l e shown i n Figure I. &k If the average energy of the fissian,products i a 1.7 MEV, the Integrated 310 (pcl liter) 0ec. x 1.7 rep = 0.075 rep 5- The variation of activity outwards frm the center of the moving cloud at this .- The germna-ray ccmrponent c a e s fran a range of ry 300 metera whlch Includes zones of higher ConcentratIan than a t the ground. "hie Isa -11 factor, and we can write The width of the cloud i n Sutton's arbitrary definition le terminated a t pointe where the activity is one-tenth of the central activity. ject cloud has a width of 7500 meters a t 100 miles frcm tho reector. It will be noted that sijiniflcant concentrations can occur a t points well beyond the convent i ml bounherlee of a radioactive cloud. CCBB, the effective wldth a t 100 miles wlll be greatly affected by wind shear. We hell elaborate the effect of meteorological varlablaa for8 much more dangerous cloud later. Assume 17 liters/min with 5% retention no e ign ifl cant hazerd If mterial is absorbed Plutonium - deposit 4 x 10-3 p - no hazard. formetion of Pu insignificant. * Neglected thraughout 18 the normelly loner uind speed near t h e ground; exposure time could be about twice the calculated value.

Page 10

tion ria ee away fran the reactor, e more significant hazard dat them intennediate distances. I n Hanf'ord terrain thi a for the present high cloud. For reference, the central fieelon proauct concentration l e given as a function of distance, together dth the width Oi the cloud, fran which the concentration a t any dietance and any elevation above the bee line can be estimated. Second order width = width t o polnts of 1/10 central activity. width t o points of 12central activity,etc. Example: A man on Rattlesnake Mountain is &lo00 meters abow t h e base-line of the reactor = 2000 meters below the d r i f t - line of the cloud, and 14 miles away. The central. concentratIan is 750 p c / liter, and the subdect point 1s beyond the fourth order eemi-width ( a 1250 meters below drift - Une) The concentratIan is below 0.075 ,pc/liter. The fall - out of particle e under gravity Kill be inconsequential vith an initla: microns will fall only 70 meters i n the poa tub ted 100 miles horizontal travel For example: a particle of specific gravity 5 and diameter 6 Inthe Hanford area, arain originating above 3000 meters l a so exceptional that one can dieregard the risk of rain-out of activity within aay 40 miles of the Plant. If rain-out did occur, it could be a potential major hazard. Sutton eatimatea fram Leiceeter deta that one-eight of the soot cantant of a cloud can be washed out by rain.We invoke two additional factore:- (1) Typical British rain Is obrioualy a more formidable scrubbing agent (2) The British rigure i s probably weighted by removal of the heavier on very amall particles, for which scrubber efficiency is low. than the Emford variety. Moat of the fi eel on product activity will be concentrated

Page 11

ebll aeeume a mar- rein-out of 5%. Initially, a deposition of om-rate of the order of 100 r/hr. ocatione up t o 100 miles or more from t h e reactor. A t 10 milea, the maxlmum deposition le ~ 3 0curies/eq.moter; Such a deposit @vee a radiation conclude that a rain-out could produce Intolerable conditione a t sporadic ais of the 2 x 10 calorie explosion Ccmputed for a flattened Hanford reactor, and thie time aeeuming all the heet goes into uranium heating, the distribution in arbitrary classee will be:- f-l% 55 tone a t the boi&lng point _ _5 55 tons just feil t o melt Behavior of the cloud Velocity of ascent - 13 meter/eec a t 200 metersCloud rises t o 0 8 0 0 meters. Activity inthe cloud The cloud contains uranium, 6 megacuries of Np, and 800 gms of plutonium. Under average conditione …