ML11355A248

New York State (NYS) Pre-Filed Evidentiary Hearing Exhibit NYS000283, the Hazard from Plutonium Dispersal by Nuclear-Warhead Accidents

ML11355A248
Person / Time
Site:	Indian Point
Issue date:	12/21/2011
From:	Fetter S, Vonhippel F Center for Energy & Environmental Research, Princeton Univ, Univ of Maryland - College Park
To:	Atomic Safety and Licensing Board Panel
SECY RAS
Shared Package
ML11355A240	List: ML11355A241 ML11355A242 ML11355A244 ML11355A245 ML11355A247 ML11355A248 ML11355A249 ML11355A251 ML11355A252
References
RAS 21598, 50-247-LR, 50-286-LR, ASLBP 07-858-03-LR-BD01
Download: ML11355A248 (21)

Text

NYS000283 Submitted: December 21, 2011 Science & Global Security, 1990, Volume 2, pp.2l-41 Photocopying permitted by license only Reprints available dizectly from the publisher

@ 1990 Goroon and Breach Science Publishers SA printed in the United Sta~s of America The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents Steve Fettef and Frank von Hippelb NUCLEARWEAPONS are carefully designed to have an extremely low probability of exploding accidently with an appreciable nuclear yield-even if they are involved in a high-speed crash, struck by a bullet or consumed in a fire! The principal concern when nuclear warheads are involved in such accidents is the possible dispersal of plutonium into the environment. In particular, an explosion could disperse a significant fraction of the plutonium in a warhead as particles of respirable size.2 There are two incidents of which we are aware in which the chemical high explosive (HE) in US nuclear warheads exploded and contaminated an area with plutonium:3

.In January 1966, over Palomares, Spain, a mid-air collision between a B-52 and its refueling aircraft resulted in the four bombs from the B-52 being released. The braking parachutes of two bombs failed completely, and they struck the ground at high speed. The HE exploded, and plutoni-um was widely dispersed. Cleanup and reparation cost $100 million.

.In January 1968, near Thule, Greenland, a fire broke out on a B-52. The bomber was abandoned and crashed into the ice at high speed and burned; the HE in the four bombs it carried exploded, spreading plutonium widely over the ice.4

a. School of Public Affairs. University of Maryland. College Park MD 20742

b. Center for Energy and Environmental Studies. Princeton University. Princeton NJ 08544 kc

22 Fetterand VonHippe!

Almost immediately after the Thule accident, the US Air Force stopped routinely flying its bombers with nuclear weapons. In addition, most US nuclear warheads designed during the past decade use "insensitive high explosive" (IHE), which is unlikely to be detonated by even high-speed impacts.

However, many of the older warheads still in the US nuclear arsenal and, for various reasons, a few of the newer ones-most notably the W88 warhead for the Trident II-still contain ordinary HE. Recently, the US weapon laboratories have chosen to raise the safety problems of these warheads as an issue.5 The purpose of this article is to offer some perspective on this concern.

Consider a hypothetical worst-case accident in which the HE in several nuclear warheads explodes. Based on experiments and calculations, it has been estimated that 10-100 percent of the plutonium contained in the warheads, with a best estimate of 20 percent, would be converted by such explosions into a PuO2 aerosol of respirable size (median aerodynamic diame-ter in the range of 5 microns or less).6 If we assume an accident involving all of the warheads on a missile carrying 10 warheads, and that each of the warheads contains approximately 3 kilograms of plutonium, then the resulting aerosol would contain on the order of10 kilograms of PuO2.7 HEALTHRISKSFROM PLUTONIUMAEROSOLS The principal risk from exposure to a plutonium aerosol is via inhalation:

most of the radiation emitted by plutonium is in the form of alpha particles, which have such short range (about 50 microns in tissue) that they cannot even penetrate the skin. External radiation from the passing cloud and from plutonium deposited on the ground can therefore be neglected.8 For an aerosol of 1-micron median aerodynamic diameter, about 15 percent of the inhaled PuO2 would be retained in the deep lung with a retention halflife of about 1.4 years.9 Health effects from radiation exposure can be divided into two categories:

illnesses and deaths due to high doses, occurring within a year or so after exposure, and cancers due to low doses, occurring during the remainder of the lives of the exposed population, starting a few years after exposure. As shown

The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 23 in the appendix, high-dose effects are unlikely even in a worst-case plutonium-dispersal accident-especially beyond the boundary of a military base. We therefore focus here on the cancer risk.

The principal hazard from exposure to lower concentrations of Pu02 aerosols is an increased probability of cancer of the lung and of other organs to which the plutonium is transported, particularly the bone. A recent review of the risks associated with low radiation doses from inhaled alpha emitters obtained a (very rough) risk estimate of one cancer death per 1,400 lung-rad or 900 to 12,000 bone-rad for inhaled Pu02!O The 30-year dose conversion factors in the literature for lung and bone-surface doses range respectively from 1,600-3,700 and 3,200-11,000 rads per inhaled milligram of 23%02 aerosol with a median aerodynamic diameter of 1 micron.ll We use the values 3,200 and 6,500 rads per milligram here!2 After correcting for the 6-percent plutonium-240 in weapon-grade plutonium (WgPu),13the lung and bone dose cOnversion factors are 3,800 and 7,600 rads per inhaled milligram respectively. The total (lung plus bone) cancer risk is therefore 3 to 11 cancer deaths per milligram of W gPu aerosol inhaled.

For comparison, the method for estimating cancer risk advocated by the International Commission on Radiological Protection (lCRP) is to use an "effective dose equivalent" (EDE), which is the weighted average of the dose to certain organs, with the weight assigned to each organ determined by the relative probability that a fatal cancer will occur in that organ after a uniform whole-body dose. The EDE for 23%02 aerosol is 15,000 rems per milligram inhaled, or 18,000 rems per milligram for WgPu. Dividing by the ICRP risk factor of 2,000 rems per cancer death14would give 9 cancers per milligram of W gPu inhaled.

In the appendix on high-dose effects, it is estimated from experimental data on beagle dogs that the risk of death from pulmonary neoplasia (a cancer) would be approximately 100 percent for the inhalation by a human adult of more than 0.08 milligrams ofWgPu, if early death did not occur first as a result of some other cause. If this risk were extrapolated linearly to lower exposures, it would correspond to 12 cancer deaths per milligram of WgPu inhaled Thus, these three different approaches to the problem of estimating cancer risk from the inhalation of Pu02 give risk factors in the range of 3 to 12

~

24 Fetterand VonHippe!

cancerdeaths per milligram of WgPu inhaled.

We make the usual approximation that the risk is linear with dose!5As a result, independently of whether a total of1 milligram ofPuO2were inhaled by 100 people(an average of 0.01 milligrams per person) or by 1,000 people (0.001 milligrams per person)there would be an expected3-12 cancerdeaths as a result.

Dispersal of the Aerosol For estimating the amount of plutonium aerosolinhaled by the population downwind of a release,one can use a Gaussianplume model(seethe appen-dix) and explore the dependence of its predictions for different assumed meteorologicalconditionsand populationdistributions. However,in situations like the present one in which the cancer risk is linearly proportional to exposure with no threshold, a much better "feel" for the estimates can be obtained by using an extremely simple atmospheric dispersion model, the "wedgemodel."16The simplicity of the results obtained with this model stem from the fact that, as noted above,for a cancerrisk that is linearly proportion-al to the exposure,the total number of cancersin the population downwind will dependonly on the total amount of the carcinogeninhaled by the popula-tion, not on the distribution of the doseswithin the population. The accuracy of the predictions of the wedge model in such applications is generally comparableto that of the Gaussianplume modelbecausemost of the cancers will ordinarily be due to very small dosesat great distancesfrom the release point where a Gaussian plume assumesa shapeapproximated by the wedge model.

In the wedgemodel, the concentrationof a contaminant is assumedto be constantin the crosswind direction overthe wedgeopeningangle e (typically ranging from 0.05-0.3 radians downwind17)and in the vertical direction throughout the height H of the mixing layer (typically 300-2,500 meters1B).

Under these conditions,the amount of plutonium inhaled I (in milligrams) by a persona distance r (in meters) downwind is

The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 25 I(r) = ~ (1) erHu where Q(r) is the total amount of plutonium (milligrams) that remains in the air at distance r downwind, b is the assumedbreathing rate (3.3'10-4m3S-1 for an adult male performing light activity),19and u is the wind speed(typi-cally between0.25 and 7 meters per second20).*

In the absenceof rain, the total quantity of airborne material declines with r becauseof depositionas Q(r) = Qoex{i) (2)

The averagedistance an aerosolparticle is carried before depositionis L =~ (3) v where v is the depositionvelocity. For most aerosols,the observeddeposition velocities range from 0.001 to 0.1 meters per second.21 The recommended values for plutonium aerosols from non-nuclear explosions of nuclear war-heads is 0.01 meters per second.22 We will assumea range of 0.003 to 0.03 meters per second.L in the absenceof rain can therefore range from tens to thousands of kilometers.

In the presenceof rain, there would be an additional exponential term associatedwith a characteristic washout time constant 't'ranging from 103 seconds(unstable atmosphericconditions)to 104seconds(stable conditions).23 The correspondingwet-depositionvelocity

• It should be noted that, in principle, it would be possible to reduce considerably the amount of plutoniUDl inhaled if the population stayed inside with windows and air intakes closed during the passage of the aerosol cloud and opened up and aired out the buildings immediately after it had passed.See,for example,Bernard Cohen,Health Physics, 32,1977, pp.359-379.

26 Fetterand VonHippe!

v w =!! (4)

't" is generally much larger than Vdwith a range of 0.05-1 meters per second.

Under rainy conditions, the deposition velocities would be added (v = Vd+ vw).

According to the wedge model, the amount of PuO2 inhaled by the total population downwind would be

1. = r-l.(r)Brp(r)dr (5) p Jo where p(r) is the population density at a distance r downwind averaged over the width of the wedge.

If we assume that the population density is constant, equal to Po,then we find that the total amount of plutonium inhaled is 1.p = ~~v (6)

The average population density of the 48 contiguous US states is about 30 Table 1: The number of cancer fatalities caused by inhalation during the passage of a plume initially containing 10 kilograms of PuO2aerosol for various deposition velocities and average population densities,for a riskfactor of 3-12 cancers deaths per milligram of Wgpu inhaled Deposition Populationdensity Po velocity km-2 meters per second 30 300 3,000 0,003 100-400 1,COO-4,000 -

0.01 30-120 300-1,200 -

0.03 10-40 100-400 l,OOO-4,{XX) 0.1 rain 3-12 30-120 300-1,200

1. rain 0-1 3-12 30-120

~

t ri\

i

7he Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 27 per square kilometer; in the most densely populated states of the northeast it ranges around 300 per square kilometer, and 3,000 per square kilometer is a mid-range density for urban areas. For a risk factor of 3-12 cancer deaths per milligram of WgPu inhaled, table 1 shows our resulting estimates of the number of deaths for various combinations of population density and deposi-tion velocity.

It will be noted that in table 1 the entries associated with very low deposition velocity are not filled in for the highest population density. The reasons are that, for v = 0.003 or 0.01 meters per second, L would most likely be hundreds of kilometers--much larger than any urban area.

We have checked the predictions of the wedge model against the corre-sponding predictions of the Gaussian plume model and obtained quite close agreement, independent of weather conditions.

Although the largest entry in table 1-four thousand cancer deaths--is high, the increase in the cancer risk for an individual in the exposed popula-tion would be small-typically on the order of one tenth of a percent. For example, for average conditions (H = 1,000 meters, e = 0.2 radians, U = 2 meters per second and v = 0.01 meters per second) the additional individual risk of cancer death would be 0.2-0.9 percent at a distance of 10 kilometers and 0.02-0.06 percent at a distance of 100 kilometers from the release. The small individual risk reflects the fact that the population risk would typically be spread among a very large population. For comparison, the individual cancer death risk in the US is at present about 20 percent. There may well already be many such large-scale cancer "events" occurring in the US due to the widespread release to the environment of carcinogenic chemicals that remain undetected against this large background..,24

• We note in this connectionthat approximately 3,000kilograms ofPu°2 was dispersed into the global atmosphere by atmospheric testing during the late 1950s and early 1960s, about 80 percent of it in the northern hemisphere. The resulting average inhalation by humans in the northern hemispherewas 0.13 nanograms.Using the above total cancerrisk coefficients, this translates into an incremental cancer risk of 0.4-1.6 cancers per million personsin the northern hemisphere. Assuming that an average population of 3 billion in the northern hemisphere was exposedto the plutonium fallout (corresponding to an average population density of about 10 per square kilometer) about 1,000-5,000 people have died or will die from cancerdue to plutonium inhalation, or roughly one person per kilogram of plutonium released.

28 Fetterand VanHippe!

Table 2: The radial population density in the direction of Seattle from Bangor Naval Base Distance from Bangor p{I)

Zone kilometers km -2 Kitsap County 0-18 130 Puget Sound 18-28 0 Seattle 28-38 2,300 Lake Washington 38-41 0 Bellevue 41-50 1,200 Eastsuburbs 50-70 1.200exp(-0.24(r-50>>)

Mountains 70-00 10 As an illustrative example, we have estimated the consequences if a hypothetical10-kilogram release of WgPu aerosol should occur at Bangor Naval Base in Washington state, one of the two bases for US Trident subma-rines, with the wind blowing towards Seattle. AS Bangor is located just 30 kilometers from downtown Seattle, this may represent a near worst case for such an accident. Table 2 gives the radial population density in the direction of Seattle and beyond as a function of distance from Bangor and table 3 gives the wedge-model estimates for the cancer deaths that would result from the release if the wind were blowing in this direction for different combinations of deposition velocity, wind speed and mixing layer height. The average wind speed in Seattle is 4 meters per second, so that the average value of Hu is about 4,000 m2 S-1, with a range from about 1,000 to 10,000 m2 S-1. The estimated number of cancer deaths under dry conditions ranges from 20 to 2,000.. These estimates agree well with the prediction obtained using a

• The wedge-modelapproximation will certainly break down when the plume reaches the mountains, but the contribution to the estimated number of deaths from the low-population-density region beyond 70 kilometers is relatively small.

~

The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 29 formula recommended by the Defense Nuclear Agency."

LAND CONTAMINATION After the plume passed, it would leave a swath of land contaminated with Pu°2. The main hazard associated with this contamination would be that the plutonium might be resuspended and inhaled. The concentration of the plutonium contamination 0"(mg m-2) at a particular point is simply related to the amount I that a person located at that point would have inhaled during the passage of the plume 0" = .!!i (7) b where, once again, v is the deposition velocity (meters per second)and b is the breathing rate (m3 S-1).

The ratio of the concentration of resuspended aerosol to 0"can be charac-Table 3: Cancer deaths predicted by the wedge model for a lD-kilogram release of WgPu at Bangor Naval Basewith the wind blowing towards Seattle Deposition Mixing height x wind speed Hu velocity v d S-1 meters per second 1,000 3,000 10,000 0.003 400-1700 180-700 80-300 0.01 300-1200 15Q--600 50-200 0.03 150-600 100-400 40-170 0.1 rain 20-90 45-180 30-120

1. rain 1-5 1-5 2-9

• Using a cookbook-stylemanual written for military commandersto assessthe effects of destroying nuclear weaponstockpiles during a war, we have calculated that, over the entire range of meteorological conditions, the number of expected cancer deaths in Seattle alone would be 30-1,000, which compares with the 10-900 cancerdeaths given for Seattle by the wedge model [Field Command, Defense Nuclear Agency, Estimation of the Hazard from Plutonium Dispersal, (Kirtland AFB, 1977)].

30 Fetterand VonHippel terized by a "resuspension coefficient" K. It is therefore easy to make a comparison between the amount of resuspended plutonium inhaled Ir and I if one knows the resuspension coefficient K (m-l) as a function of time:

~ = v (t K(t ')dt , (8)

I Jo The resuspension coefficient can be expected to decline with time as the plutonium aerosol sinks into the soil and becomes attached to larger particles.

Based on a review of the small amount of available data, a 1974 Atomic Energy Commission study suggested for populated areas an initial value Ko =

10-5 m-1 declining to a long-term value K- = 10-9, and interpolated according to the formula K(t) = Koexp(-5t) + K- (9) where t is measured in years.25Using this function, we find

~ = v (O.2Ko[1 -exp( -5t)] + K_t) (10)

Table 4: The ratio of the integrated inhaled dose from resuspensionto that from plume passage, for several values of the deposition velocity at 1 month and after 1 year Deposition Exposuretime 'f velocity Vd 1 month 1 year or greater meters per second 0.003 0.064 0.19 0.01 0.21 0.63 0.03 0.64 1.9

0. 1 rain 2.1 6.3 1.0 rain 21 63

7he Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 31 Table 4 shows values of Ir(t)/I calculated for various deposition velocities at 1 month and 1 year (after 1 year the resuspension dose rate will be negligible).

Note that resuspension can only be neglected for low to moderate deposition velocities ( 5 0.01 meters per second) and short exposure times (less than a month).

Evacuation and/or decontamination, as was done at Palomares, could reduce the hazard to that part of the population in the most heavily contami-nated area. In most cases, however, virtually all of the population dose would come from a very large area (on the order of 1,000 square kilometers for v =

0.01) of lightly contaminated land that might well be prohibitively expensive to either evacuate or decontaminate. The factors [1 + Ir(oo)/l] should therefore probably be used to multiply the cancer death estimates in table 1 except for the largest values of v ( ~ 0.1 meters per second) and in urban areas where the contaminated areas would probably be decontaminated. Table 5 gives the number of cancer deaths from inhalation during and after the plume passage under these assumptions.

CONCLUSION An accident involving the dispersal of kilogram-quantities of plutonium in aerosol form might, in a worst case (for example, an accident at Bangor Naval Table 5: The number of cancer fatalities caused by inhalation during and after the plume passage, assuming that urban areas (p = 3,(XX)km-2) and small, highly contaminated areas (v:?;0.1 meter per second) Ofe substantiallydecontaminated within 1 month Deposition Population density Po velocity km-2 meters per second 30 300 3,000 0.003 120-500 1,200-5,000 -

0.01 50-200 500-2,000 -

0.03 30-120 300-1,200 1,600-7,(xx) 0.1 fain 10-40 100-400 1,OOO-4,(XX)

1. fain 7-30 70-300 7OO-3,(XX)

~

32 Fetterand VanHippe' Base with the wind blowing toward Seattle with low wind speedand deposi-tion velocity), cause a few thousand cancer deaths during the subsequent decadeswith a probablyundetectableincreasein the resulting regional cancer rate. Even under worst-case conditions, no early deaths due to high doses would be expected-certainly not off the base. However, judging from the Three Mile Island and Palomaresexperiences,the psychologicaltrauma and the costs of reparations and decontaminatingthe most heavily contaminated areas might be enormous.

To get an additional perspective on this risk, let us assume that the probability of a near worst-case accident is 0.1 percent per year.26The expectednumber of deaths would then be on the order of one per year since, under average conditions, on the order of 1,000 cancerdeaths would result from a worst-caseaccident.

This risk could be reduced,but not completelyeliminated, by redesigning and rebuilding warheads with IRE-at a cost. A new warhead costs on the order of $1 million and lasts 20-30years. If old warheadscontaining sensitive high explosiveswere retired an average of 10 years early in order to replace them with warheadscontainingIHE, the extra costwould be at least $300,000 per warhead. If this were done for the approximately3,000warheadsthat are to be deployedon US submarines after the reductions mandated by START, the cost would be on the order of $1 billion, or $100 million per year of reduced risk. Given that the expectedvalue of the number of lives saved by such an expenditure is on the order of one life or less per year, the resulting costper life savedwould be 250-3,000times that for other investmentsin life-saving that the US is currently making.27 We therefore concludethat reducingthe hazard of plutonium dispersal by converting warheads to IHE need not be dealt with through a "crash" pro-gram. However,for warheadsthat a governmentexpectsto replaceupon their retirement, it would be desirable to have available replacement designs containing IHE. If such designsare not available already, their development should be given priority in any further testing before the achievementof a comprehensivetest ban.

Another hazard that has been hinted at in the recent press stories on warhead safety is that, under certain circumstances, detonation of the chemicalexplosivein a warhead might result in a nuclear yield greater than

The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 33 4 pounds of TNT equivalent. Specifically, there is apparently concern that this might occur in case of near simultaneous explosions of the HE in the W88 warheads of a Trident II missile, which are closely clustered around the third-stage rocket motor.28 It is difficult for us to provide any perspective on this concern in the absence of public estimates of the size of the nuclear yield that might result in such an event.

Appendix THERISKOF HIGH-DOSE EFFECTS Health Effects at High Doses Becauseit is relatively insoluble,a substantial fraction of inhaled PuO2will remain in the lungs for a long time. Early effects are therefore dominated by damage to lung tissue.

Experimentswith beagledogsindicate that, if a relatively large amountof aerosol were inhaled,the lung damagefrom the resulting alpha irradiation would causedeath from acute respiratory failure within a week. This would occurfor an initial alveolar deposition of about 60 micrograms of 239PUO2 per gram of bloodless lung,29which correspondsto a total inhalation of about 100 milligrams of weapon-gradeplutonium (WgPu) by an adult human.30 The samesetof experimentswith dogsindicatesthat, at lower doses,deathoccurs later because of respiratory insufficiency resulting from extensive fibrosis. The depositionof 2 microgramsof 23%02 per gram of bloodlesslung (correspondingto the inhalation of about 3 milligrams of WgPu by an adult human) will result in death within several months.31At still lower doses,fibrosis developsmore slowly. A least-(

squaresfit to the relationship betweenY, the initial alveolar depositionin micrograms ofplutonium-239 per gram of lung and t, the averagelength of time in daysbefore the death of dogsgiven that doseis32 Y = 560t-1.028 (11)

At the maximum lifetime of the beagle (15 years)this relationship gives an alveolar depositionof 0.09microgramsof23%02 per gram of bloodlesslung. (This also happens to be the lowestdoseat which a dog died of fibrosis in the experiment.)

Pulmonary neoplasia(a cancer)began to occurin dogs that survived 3 to 5 years after exposurein these experimentsand was the leading causeof deathin exposeddogs that survived more than 5 years. The least-squaresfit to the dose-longevitycurves of dogsdying of neoplasiais33

34 Fetterand VanHippe!

Y = 11,000t-1.416 (12)

This curve intersects the maximum beaglelifetime at Y = 0.04microgramsper gram, which correspondsto the inhalation of about 0.08 milligrams of WgPu by an adult human.

Dispersal of the Aerosol at Short Distances The inhalation high dosesthat are the subjectof this appendixwould occur,if at all, closeto the releasewhere the approxiniationsmade to obtain the wedgemodelwould not hold. For the purposesof estimating the high-dosehealth effectsfrom an accident we have useda Gaussianplume modelto estimatethe dispersalof plutonium at short distances.In this model, the time-integrated ground-levelconcentrationof plutonium (mg.s m-3)downwind from the releaseis given by x<x,y) = ~exp no:y o:zu

[~2~

-~

2~

J (13) y z where x is the downwind distanceandy is the crosswinddistance (m), Q(x)is the mass (milligrams) of plutonium aerosolremaining in the cloud when it arrives at x, o:yand O:zare the horizontal and vertical standard deviationsof the cloud concentrationat point x, u is the mean wind speed(metersper second),and h is the centerline height of the cloud. Formulas for o:yand O:z for a point sourceare given in the ReactorSafety Study for atmosphericconditionsranging from very unstable (class"A") to very stable (class "F");34we have modified these formulas so that they give the initial standard deviations of the explosion-formedcloud, ~ and ~, at x = 0, and so that they are appropriate for an instantaneous(rather than a continuous)release.Sincethe mixed layer of the atmospherenormally has a finite height H (typically 300 to 2,500meters),

we have alsomodified equation13 to preventplutonium from diffusing abovethe mixed layer and to accountfor reflectionsfrom the top of the mixed layer and the ground.

The amountof plutonium remAining in the cloud at distancex is given by

[

Q(x) = Qoexp-f!

~ ~

!.. rZ~exp u Jo o:z

(~ )]

2~

(14) z where Qois the initial amount of plutonium in the cloud (milligrams) and v is the depositionvelocityof the aerosol(meterspersecond).As mentionedabove,a worst-case accidentmight involve the detonationof the HE in severalballistic-missile warheads

1he Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 35 (and perhaps the missile propellantsas well), releasingas much as 10 kilograms (107 milligrams) ofWgPu as respirableparticles. As noted above,dependingon the sizeand compositionof an aerosol,v can range from 0.001 to 0.1 meters per second;a typical value for plutonium aerosolsis 0.01 meters per second.

The initial height (h) and size (~ and ~) of the cloud dependson the amount of explosiveenergy released.(Cold,ground-levelreleasesof significant amountsof PuO2 aerosolfrom smoldering chunks of plutonium are not consideredcredible.35) Sincere-entry vehicles typically weigh 100 to 200 kilograms,warheads probably contain 20 to 50 kilograms of HE. Sincenormal HE is nearly twice as energeticas TNT, an accident could result in an explosionequivalent to between40 and more than 400 kilograms of TNT, dependingto the numberof warheadsinvolved.The detonationof the propellants in the third stage of a missile would be equivalent to an additional 4 to 8 tons of TNT.36Estimates of the initial cloud heights and radii for low, medium, and high-energy releases are given in table 6.37All else being equal, smaller explosionsare more dangerousbecausethe plutonium-bearingcloud remains closerto the ground.

Experimentshave shownthat approximately5 percentof the radioactivity in the cloud is initially found betweenthe groundand T14,where T is the cloud-topheight; 30 percentbetweenTI4 and T12;40 percentbetweenTI2and 3T14;and 25 percentbetween 3TI4 and T. Wehave modeledthis situation by using four cloud sourcescontaining the abovefractions of plutonium with centerline heights of T18,3T18,5T18,and 7T18;~ of R14,R13,R12,and R12,where R is the initial cloud radius; and ~ equal to T18.38 The amount of plutonium that would be inhaled by an individual (milligrams) standing on openground is given by I(x,Y) = X(x,y)b (15) where b is the breathing rate (m3S-I). As noted above,for an adult male performing light activity, b = 20 liters per minute = 3.3.10-4m3S-I.

Table 7 gives the results of the Gaussianplume modelfor a 10-kilogram releaseof Table 6: Cloud-top height and mean radius in meters for low, medium, and high energy releases from an accident involving the detonation of the HE in one or more nuclear warheads and possiblythe third stage of a ballistic missileas well Yield Cloud-top Mean Estimate TNTequivalent height radius kilograms meters meters Low 40 230 19 Medium 400 410 44 High 4,000 740 110

36 Fetter and Von Hippe!

Table 7: The amount of plutonium inhaled by an individual (milligrams)during the plume passage at several points downwind on the plume centerline for unstable, neutral, and stable conditions, under worst-case assumptions about windspeed (u

= 1 meter per second), plume height (4Q-kilogramequivalent TNTexplosive energy release) and thickness of the mixed layer (H = 300meters).

Distance downwind Dose kilometers milligrams unstable neutral stable 0.1 0.08 0.07 0.08 0.2 0.1 0.06 0.06 0.5 0.07 0.04 0.05

1. 0.03 0.03 0.03

2. 0.02 0.02 0.02

5. 0.003 0.014 0.011

10. 0.002 0.011 0.006 WgPu under conditionsthat result in the highestdosesto individuals (u = 1 meter per second,H = 300 meters) for unstable, stable, and neutral conditions.39 A maximum doseapproaching0.08 milligram&-thelowest doselikely to causethreshold effects-occurs only at close ranges ( < 500 meters) and low wind speeds ( ~ 1 meter per second). Therefore doses exceeding 0.08 milligrams are highly unlikely to occur anywherenear civilian populations.

Summary If the area around the releaseis evacuatedafter the plume passesto avoid chronic exposureto depositedplutonium, there will almost certainly be no acutehealth effects from a worst-case accident, even close to the site and under worst-case weather conditions.This is especiallytrue of civilian populations,which are usually no closer than a few kilometers from locationswhere missiles or nuclear weaponsare loaded or stored. Even if the surrounding population was not evacuatedand the land was not decontaminated for long periods of time (conditions that are highly unlikely), the maximum off-site dosewould exceed0.08 milligrams only under a very limited set of weather conditionscombinedwith a very low cloud height. Therefore,for all practical purposes,threshold health effectsfrom such accidentscan be ignored.

The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 37 NOTESAND REFERENCES

1. US nuclear warheadsare designedso that "in the eventof a detonationinitiated at anyone point in the high explosivesystem[rather than multiple points, which would occ~ providing the authorization codeswere properly entered and the environmental sensorsregisteredthe design launch to target sequence],the probability of a nuclear yield greater than 4 pounds[1.8 kilograms] of TNT equivalent shall not exceedone in one million." US Arms Control and Disarmament Agency,Fiscal Year 1979 Arms Control Impact Statements,p.92.

2. A surveyof experimentsinvolving the burning of plutonium metal in hot fires found that the fraction of plutonium convertedinto a respirable PuO2oxide aerosolranges from less than 0.001 percent to a few percent. Ralph H. Condit, Plutonium Dispersal in Fires: Summary of What is Known (Livermore, California: LaWrenceLivermore National Laboratory,1986),p.iO.

3. Ibid, p.11.

4. For a report on the measurementsof plutonium contaminationand of the cleanup, see USAFNuclear SafetyStudy 65(Kirtland AFB, New Mexico:Directorateof Nuclear Safety,1970),part 2, "Project CrestedIce."

5. R. Jeffi-ey Smith, "Defective Nuclear Shells Raise Safety Concerns,"Washington Post, 23 May 1990,p.A-1. In a letter to SenatorEdward Kennedy,dated 9 July 1990, Brent Scrowcroft, President Bush's national security adviser, argued that "recent revelationsregarding safetyof certainwarheadsunderscorethe importanceof testing."

6. SupplementaryDocumentationfor an Environmental Impact StatementRegarding the PantexPlant (Los Alamos National Laboratory, report # LA-9445-PNT-D,1982);

Report on the Safety Criteria for Plutonium Bearing Nuclear Weapons,(Washington DC: US Atomic Energy Agency,report # RS/5640/l032,23 January 1973; declassified with deletions,9 January 1989),summary,p.10.

7. We infer from note 6 that at least somenuclear weaponscontain plutonium-238to power a radioisotopethermoelectric generator(RTG). The amount of plutonium-238 dependson the power requirements for nuclear weapons;1 gram of plutonium-238 generates about 0.57 watts of heat, which could be convertedto no more than 0.06 watts of electric power.Although 14.5gramsof plutonium-238 presentsaboutthe same health hazard as 4 kilograms of plutonium-239 (assuming the hazard scales with radioactivity), it could only be used to generate about 1 watt of electrical power. If powerrequirementsare greaterthan 1 watt, then plutonium-238shouldbe considered in a hazard analysis. We do not, however,know what the power requirements of nuclear weaponsare, and in any casethe plutonium-238 in the RTGshould be much better protected from dispersal during an accident than the plutonium core. We therefore ignore the contribution of plutonium-238 to health effects in this paper, although we flag the issue here.

38 Fetterand VanHippe!

8. The amount of the aerosol inhaled is proportional to its concentration Coin the air (measured in mg m-3) and the length of exposure (To). A 30-year dose due to inhala-tion will then be DibT cPo,where b is the breathing rate (about 3.310-4 m3 S-1 for an adult male involved in light activity) and Di is the dose per milligram of plutonium aerosol inhaled. Below, we will see that Di = 3,800 and 7,600 rads per milligram respectively for the lung and bone-lining cells--the organs in which the cancer risk would be greatest following the inhalation of Pu°2. The external whole-body dose from the cloud would be D.coT0 and the external dose from plutonium deposited on the ground would be DFoTovT. Dc = :1..210-8rad m3 mg-1 S-I, Dg = 6.610-10 rad m2 mg-l S-1 for plutonium-239, v is the deposition velocity (of order 10-2 meters per second in the absence of rain), and T is the duration of exposure to the contaminated ground in seconds. The ratio of cloud to inhalation dose is therefore D/(bDJ = 10-8 and the ratio of the ground to lung inhalation dose is (DgTu)/(Dib) = 410-12T. Since the lung would account for a significant fraction of all cancers caused by external whole-body gamma radiation, the cloud dose brings with it negligible additional risk, and it would take about 10,000 years for the integrated ground dose to equal the inhalation dose. The ratios for plutonium-240 are similarly low. See Steve Fetter, Internal Dose Conversion Factors of 19 Target Organs and 9 Irradiation 1Imes and External Dose-rate Conversion Factors for 21 Target Organs for 259 Radionuclides Produced in Potential Fusion Reactor Materials, (Idaho Falls: Idaho National Engineer-ing Laboratory, report # EGG-FSP-8036, 1988).

9. Reactor Safety Study (Washington DC: US Nuclear Regulatory Commission, report

1. NUREG-75/014, 1975), appendix VI, pp.D-2-D-7.

10. Health Risks of Radon and Other Internally Deposited Alpha-Emitters (BIER N)

(Washington DC: National Academy Press, 1988), p.332.

11. According to table 23 of Ionizing Radiation: Sources and Biological Effects (New York: United Nations, 1982), the average dose commitment to the human lung and the bone-lining cells from the inhalation ofPuO2 from atmospheric nuclear testing was 1.6 and 4.8 millirads per becquerel respectively. Since plutonium-239 has a specific activity of 0.06204 curies per gram, this translates to 3,700 and 11,000 rads per milligram.

Reactor Safety Study (Washington DC: US Nuclear Regulatory Commission, 1975),

table VI D-2, gives a 30-year dose to the lung and bone of 2.9108 and 5.2108 rems per curie. Since the study set Q = 10 rems per red for alpha particles, this translates to 1,800 and 3,200 rads per milligram.

D.E. Dunning Jr, G.G. Killough, S.R. Bernard, J.C. Pleasant, and P.J. Walsh, "Estimates of Internal Dose Equivalent to 22 Target Organs for Radionuclides Occurring in Routine Releases from Nuclear Fuel-Cycle Facilities," vol. III, ORNU NUREGtrM-190N3 (Oak Ridge, Tennessee: Oak Ridge National Laboratory, 1981),

pp.43-44, give 50-year dose conversion factors for lung and bone surfaces of 580 and 4,160 rems per microcUrie. After dividing by the assumed Q = 20 rems per rad for  :

alpha particles and multiplying by the ratio of the 30-year and 50-year doses given by ,t Fetter, this translates to 1,600 and 9,300 rads per milligram. t

~

I i

]he Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 39

12. Fetter, Internal DoseConversionFactors,addendum,gives30-yearlung and bone-surface dosesof 2.9.108and 5.2108 reInS per curie respectively.Since Q = 20 reInS per rad for alpha particles in this study,this translates to 3,200 and 6,500rads per milligram of plutonium-239.

13. The specificalpha activity of fresh weapon-gradeplutonium is 1.17times that of pure plutonium-239. After one halflife (14 years) ofplutonium-241, the activity would increaseby an additional 1 0 percentbecauseof the alpha decayof americium-241.

14. International Commissionon Radiological Protection, Recommendations for the Commission-1990,ICRP/90/6-01,draft, February 1990,p.3-13.

15. For a discussionof this approximation, see Health Risks ofRadon and Other Internally DepositedAlpha-Emitters,appendixII.

16. "Report to the American Physical Society by the Study Group on Light-Water ReactorSafety,"Reviews ofModem Physics,47 (1975),p.S45.

17. Reactor Safety Study, table VI A-I.

18. Ibid, figures VI A-4 and A-5.

19. Health Risks ofRadon and Other Internally DepositedAlpha-Emitters,p.147.The ratio of breathing rate to lung mass doesnot vary by more than a factor of two with age (ReactorSafety Study, tables IV D-4 and D-5).

20. ReactorSafety Study, table VI 5-2.

21. ReactorSafety Study, table VI B-1.

22. GA Schmel,"Particle and Dry Gas Deposition:A Review,"AtmosphericEnviron-ment,14 (1980).

23. Reactor Safety Study, appendixVI, p.E-13.

24. Four thousand cancer deathsin the lifetime of the US population of 250 million correspondsto an averageindividual risk of about10-5.This is in the range where the US Environmental ProtectionAgency(EPA)tends to set the limits of acceptablerisks for the carcinogensthat it regulates.[Seefor exampleEliot Marshall, "EPA'sHigh-Risk CarcinogenPolicy," Science,218 (1982),p.975.]

25. U.S.Atomic EnergyAgencyProposedFinal Environmental StatementLiquid Metal Fast BreederReactorProgram,(WASH-1535,1974),appendix II-G.

40 Fetterand VanHippe!

26. The two large plutonium releasesfrom US nuclear weaponsthat occurred in the first 50 years of the nuclear age were in areas of low population density, and the practicethat resultedin thesereleases-routine flights by nuclear-armedbombers-has been discontinued.

27. Bernard Cohen,"Reducing the Hazards of Nuclear Power: Insanity in Action,"

Physics and Society,16 (1987),p.2, quotes cost estimates in the range of $20,000-140,000 per life saved for various types of cancer screening,$400,000 for kidney dialysis, and $30,000-$300,000per life savedby various highway safetyimprovement programs undertaken by the US Departmentof Transportation in the early 1980s.

28. R. Jeffrey Smith, "DefectiveNuclear Shells Raise Safety Concerns,"Washington Post, 23 May 1990,p.A-l.

29. W.J. Bair, J.E. Ballou, J.F. Park, and C.L. Sanders,"Plutonium in Soft Tissues with Emphasis on the Respiratory Tract," in H.C. Hodge,J.N. Stannard, and J.B.

Hursh, eds.,The Handbookof ExperimentalPharmacology,Vol.36: Uranium-Plutoni-um-7ransplutonicElements(NewYork: Springer-Verlag,1973),p.548.

30. The specificalpha activity of fresh weapon-gradeplutonium (6 percentplutonium-240)is 72.5microcuriesper milligram-l.17 times that ofplutonium-239.The bloodless lung of man weighs about 500 grams. The fraction of inhaled material initially depositedin the alveoli ranges from 5 percentto 50 percent for aerosolswith a mass median aerodynamicdiameter (MMAD)of 10 to 0.2 microns; for a MMAD of 1 micron, the fraction is about 25 percent. Thus, 100milligrams ofWgPu inhaled is equivalent to an alveolar depositionof 0.06 milligrams of plutonium-239 per gram of bloodless lung. In a human being, about 60 percentof this depositionwould remain in the lung with a retention halflife of 1.4 years.

31. The dose-mortalitycurve in Bair et al., p.548,gives an initial alveolar deposition of 2 micrograms per gram for a survival time of 6 months, which correspondsto an adult human inhaling 3 milligrams ofWgPu.

32. Ibid, p.548.

33. Ibid.

34. ReactorSafety Study,table VI A-I.

35. Report on the SafetyCriteria for Plutonium-BearingNuclear Weapons,appendix, p.28.

36. Assume a third stage carrying 10 warheadsweighing 100 to 200kilograms each gives them a velocity increment of 2.5 kilometers per secondand has an exhaust velocity of 2.5 kilometers per second.Also assumethat the postboostvehicle (including propellants) weighs as much as the warheads,and that the total stage mass is 1.1 times the propellant mass.Then, using the rocket formula, the propellant mass in the

The Hazard from Plutonium Dispersal by Nuclear-warhead Accidents 41 third stage would be equal to (e -1).2.10/(1.1 -0.1e) = 40 times the mass of a single warhead, or 4 to 8 tonnes in all.

37. H.W. Church, Cloud Rise from High ExplosiveDetonations (Albuquerque,New Mexico: Sandia National Laboratory, report # Till 4000, UC/41, 1969), gives the following formulas for the cloud-top height T and the cloud radius R: T = 76wo.25, R = 3.5WO.375,where T and R are in meters and W is in poundsof TNT equivalent.

38. SupplementaryDocumentationfor EnvironmentalImpact StatementRegardingthe Pantex Plant: DispersionAnalysis for PostulatedAccidents, LA-9445-PNrx-D (Los Alamos, New Mexico:Los Alamos National Laboratory,1982).

39. The dose at a given point is approximately inversely proportional to u and, at distancesof less than 10 kilometers,is relatively insensitive to factor-of-tenincreases in v andH.

~