Index Home About Blog
From: (Bill Higgins)
Newsgroups: sci.astro,,sci.environment,talk.environment
Subject: J. Cuzzi on Cassini plutonium hazard
Date: 18 Sep 97 14:30:55 -0600
Summary: Simple approximate calculation of worst Cassini accident effects
Keywords: Cassini plutonium 238 Saturn NASA

I've just seen a message from Jeff Cuzzi, a prominent planetary
scientist working on the Cassini mission, and I thought I would pass
it along with his permission.  Opinions are his, not mine and not my
employer's. He offers this additional disclaimer: "Remember it is
not an official NASA position but a calculation ANYONE could do!"

Bill Higgins                            Internet: HIGGINS@FNAL.FNAL.GOV
Fermi National Accelerator Laboratory

Date: Tue, 16 Sep 1997 22:14:46 -0700
From: (Jeff Cuzzi)
Subject: Plutonium primer

Cassini Plutonium for the technically minded
by Jeff Cuzzi

I'm sure we will all have friends and relatives asking us what's up
with the Cassini Plutonium issue as launch approaches in early
October. Allegations of risk have arisen due to Cassini's onboard
RTG's (Radioisotope Thermal Generators) which derive electricity from
decay of 72 lb (33kg) of Plutonium dioxide fuel.

In anticipation, I wanted to provide some "derived from basic
principles" satisfaction that the Cassini health threat is negligibly
small even in the extremely small chance that anything does go wrong
with the mission (either at launch or at flyby). The Cassini project
has devoted more than a million dollars to a thorough analysis of the
problem, but the back-of-an-envelope analysis below is a little easier
to grasp and serves as a calibration and sanity check.

I am a Cassini scientist, and neither a health expert, nor a nuclear
physicist.  I do care about the health of the people of the world.  I
had several discussions with a physicist at the Nuclear Regulatory
Commission (NRC) concerning decay rates and comparative relationships
to health effects. I also had this reviewed by the President of the
Health Physics Society, a 6500 member national organization (who has
publicly stated that NASA has done a very good job and has, if
anything, OVERestimated the health risks).

For my initial health effect data I relied on Web sites maintained by
the EPA and the Agency for Toxic Substances and Disease Registry
(ATSDR; part of the Center for Disease Control - see references
below); my NRC and Davis contacts confirmed these values and
identified their primary source (FGR-11, 1988).  I suspect anyone can
reproduce the calculations below who can read a simple physics
textbook and the World Wide Web.

238-Pu decays by alpha-particle emission (like the longer-lived
weapons grade isotope 239-Pu, but 250x faster).  The decay rate can be
calculated from the half life (88 yrs) and the number of nuclei per
gram, and is about 6E11 decays/sec/gm, defined as 17 Curie/gm. A Curie
(Ci) of 238-Pu and a Ci of 239-Pu have the same radiation damage
potential (they emit the same alpha particles).  Because 238 decays
faster, it has a higher Ci/g rating by the ratio of half lives (about
250). The convenient unit is pico-Curies (10^-12 Ci = pCi).

Health standards are set by the International Commission on
Radiological Protection (ICRP), and found in FGR-11 and the ATSDR web
page.  The conversion factors between radioactivity (Ci) and potential
tissue damage in rem (Roentgen Equivalent Measure, or more often
millirem (mrem = 10^-3 rem) are from the FGR-11 (note 1). They can be
derived from values on the ATSDR page as well. The ATSDR quoted Annual
Limit on Intake (ALI) is 20000pCi/yr for "workers", and the
corresponding dose limit is 5 rem/yr, giving a conversion factor of
0.25 mrem/pCi (note 1), in good agreement with the standard value of
0.29 mrem/pCi tabulated in FGR-11.

Several expressions can be found for EPA-allowable levels of
radioactivity. The ATSDR web page gives a mixture of recommended
limits for the public and for "occuptional exposure" in rem, Annual
Limits on Intake (ALI) in pCi/yr, and in Derived Air Concentration
(DAC; pCi/m3) levels. These are generally consistent with a 10 times
lower limit for the general public than for workers, but my NRC
contact says the DAC's for the general public are maybe another 10
times smaller than can be inferred from this web page (probably
factors for time off-job as fraction of 24 hr, etc).

Also, it appears that the 500 mrem annual limit for the public cited
by ATSDR probably includes the unavoidable background level of 360
mrem/yr from Radon gas, cosmic rays, the dentist, etc. My NRC contact
thinks this would be consistent with his knowledge of an ICRP
recommendation for the public of no more than 100 mrem annually above
the annual background.

Presume a worst case scenario involving vaporization of ALL the Pu-238
that is in the RTG's. This 'astrophysical accuracy' calculation makes
no allowance for removal of Pu into the ocean, by rainout, deposition
onto uninhabited terrain, etc.  The 72lb of Cassini fuel is actually
nearly 30% oxygen and less active Pu isotopes, so is only 50 lb Pu-238
= 23 kg = 400,000 Ci (about 17 Ci/g). The volume of air in the
Northern troposphere and stratosphere (which receive 99% of the Pu) =
2 pi X (10 + 40) X 6000^2 km3  = 10^19 m3.

Dispersion of all this vaporized Pu in the northern atmosphere gives a
radiation density of about 0.04 pCi/m3, comparable to the allowable
DAC. The ATSDR numbers imply that you breathe air at about 0.1
liter/sec (plausible) so get 3000 m3/yr, or about 120pCi/yr. the
conversion factor above (0.25) gives a 50 year dose of 30 mrem from
each year of breathing this Plutonium - less than 10% of the annual
background.  You'd need to breathe it for 10 years just to get the
equivalent of one year of natural radiation. Meanwhile, of course, it
is being lost from the system so the real numbers are far smaller. And
this is using ALL the Plutonium.

Looked at another way, all the Pu settles out eventually, providing
2000 pCi/m2, probably over a few years. If a person has a cross
section of 1 m2 and inhales ALL the fallout in this area, he gets a
500 mrem 50 year dose. This is still considerably smaller than the
18000 mrem we naturally receive over the same 50 year period.

For comparison, 500 mrem total dose is about the same as one
mammogram. Of course, most of this settling Pu misses people's noses
and mouths, and if this amount of Plutonium were mixed into the top 1
mm of soil, it could be shipped as non-radioactive material. And this
is using ALL the Plutonium.

No credible indication has ever been found of increased health risk
even to the many people who worked milling Pu in the Hot and Cold War
days. The only documented health effects I have been able to find are
on the ATSDR web site (see references). Dogs (apparently beagles)
inhaled Plutonium at a rate of 1400 - 100,000 pCi per kg body mass in
a day, and suffered lung damage, even cancer, depending on dose, after
several months to years.

Allowing for 20 kg body mass, these dog martyrs consumed, in one DAY,
amounts which would be 14 to 1000 times the average person's share of
the entire Cassini Pu load as overestimated above.  The president of
the Health Physics Society has himself done extensive research on mice
that confirms these dog results.

Vaporization of all the Plutonium is, of course, a gross overestimate.
Forget (for a moment) the one-in-a-million probability that ANY kind
of flyby mishap will even occur which leads to reentry and
vaporization.  Even if a mishap does occur, only a tiny fraction of
the Pu is able to end up in people (this is the analogue of the fact
that there are enough germs in one sneeze to give a billion people a
cold - it's the distribution problem that stops this from happening).

The Cassini project and its consultants have done exhaustive analyses
of this problem. Atmospheric incineration and ground impact have both
been considered. The RTG housing itself probably does come apart under
entry heating, but the triple-protected modules (2 layers of carbon
composite, and an iridium cladding on each Plutonium golf ball) are
extremely durable, and designed to withstand atmospheric deceleration
and heating.  They hit the ground at terminal velocity - only 100-300
feet/second, or one-tenth the speed of a rifle bullet.  Rifle bullets
don't vaporize on impact. Neither do meteorites; they dig a little
hole.  So the units might dent the hood of your car pretty badly, or
make a hole in your yard, but won't spray pulverized plutonium all
over your house.  All this has been tested.

Factoring in these issues, the projects finds that the average
expected dose (per person) is only 1 mrem over the entire 50 year
lifespan of the at-risk population. Comparing this to the above upper
limit of about 500 mrem/50 yr, one gets a distribution efficiency
factor of about .002. If a sneeze had the same efficiency then each
sneeze would give 2 million people a cold.  So the project's
distribution efficiency factor, which includes the difficulty of
burning through the carbon-composite and Iridium cladding of the fuel,
is hardly unreasonable and actually seems quite conservative.

Given the low distribution efficiency, the "average" person receives
practically no Pu at all. So what's all the fuss about?  There is a
very narrow range of "hot" particle sizes (about 6-10 micron radius)
that is both large enough to have a significant radiation damage
potential (in the range that damaged dogs' lungs) AND small enough to
have any conceivable chance of being inhaled (but only a very, very
small chance - see note 2).

Because of the high density of the Pu (11 g/cm3), the aerodynamic
radius is 11 times the actual radius. That is, cigarette smoke
particles as large as 6-10 microns are inhalable with small
probability (a percent or less), but Pu particles of the same size
behave like 60-100 micron carbon grit. If ALL the Cassini Pu were in
this 6-10 micron size range, there would be 5 E11 particles to
distribute - "100 for each person" is what the critics might say.  But
in reality there are enormous reduction factors that must be

For instance, the fraction of Pu fuel that is actually vaporized is
probably less than 10%.  The fraction of all released particles that
lie in the narrow hazardous size range is perhaps 1%. The fraction of
Pu that ends up landing where people live (say, the 20 largest cities)
is roughly their area fraction or say 0.0001. The fraction of these
grit particles that are actually inhaled, because of their large
aerodynamic size of about 100 microns, is also small  - surely less
than 0.01 (note 2).  There is slop in these estimates, but they are
plausible "delivery inefficiences" and lead to 500 inhaled "hazardous"
particles worldwide, consistent with the Cassini project's far more
careful estimate of 100 additional fatalities over a 50 year period.

Recall that the probability of this happening in the first place is
one in a million; another type of celestial mishap with the same
probability, impact of a mile-wide asteroid, would kill over a billion
people. Also recall that a billion people will die from cancer
unrelated to Cassini during this same 50 years.

The health hazard numbers are even smaller for a launch-related
accident (even while it is perhaps 1000 times more "probable" at
1/1500 chance of Pu-release related to launch accident), because a far
smaller amount of Pu is vaporized and fewer people are exposed. The
RTG's have been exhaustively tested under conditions comparable to
such accidents; their Carbon-Iridium protection scheme is incredibly

Overall, I think the above simple arguments make the more exhaustive
analysis done by the Cassini project very easy to understand and
accept. The health hazard due to Cassini Plutonium really is
negligible.  Statistics in the World Almanac verify that a person's
risk of dying from Cassini is a million times smaller than his or her
risk of a fatal auto accident while driving one mile.


1) For the cognoscenti, all doses given here are effective (whole
body), equivalent (radiation type independent), committed (50-year)
doses (unless specified as annual). This is necessary to compare
different sources of radioactivity. There are factor-of-2 or 3
differences depending on how soluble the Plutonium is; the values on
the web page are appropriate for "insoluble" Plutonium such as the
Cassini ceramic form.  The basic constants are thus the 50-year
integrated effective (whole body) damage-causing dose in mrem from a
certain quantity of radioactivity in pCi.

2) The human nose is 100% effective at filtering particles that are 10
microns or greater and 95% effective at filtering particles over 5
microns. These particles can then be excreted easily. The critical
size for deposition in lung cells is 1-2 microns. Once inhaled, the
material is subject to removal processes involving incoproration in
mucous suspension and being swept out by the action of the cillia
which line the portions of the lung which are exposed to air
(Glasstone and Dolan 1977).


FGR-11 (1988), or Federal Guidance Report-11: "Limiting values of
radionuclide intake and air concentration and dose conversion factors
for inhalation, submersion, and ingestion"; K. F. Eckerman et al, EPA
Report EPA-520/1-88-020. This is based on standards developed by the
International Commission on Radiological Protection, and is endorsed
by the President of the United States.

Glasstone and Dolan (1977), Department of Defense Publication, "The
Effect of Nuclear Weapons"

ASTDR Web site:

JPL Cassini Home Page:

From: Henry Spencer <>
Subject: Re: RTG splash down?
Date: Mon, 8 Jul 1996 16:18:43 GMT

In article <4rh8ae$> (Jens Lerch) writes:
>>Does any body know of any RTG (radio thermal generator) accidents.
>Apollo 13's Lunar Module RTG survived reentry in athmosphere and lies
>at the bottom of the Pacific.

That was in fact the third US RTG accident.  In the first, the RTG burned
up in the atmosphere after a failure of some kind.  After that, the design
rules were changed to emphasize intact reentry rather than safe dispersal.
The second accident, a launch failure at Vandenberg, dropped a Nimbus
weather satellite into the water.  The RTGs were retrieved, refurbished,
and used on a later Nimbus.

The ALSEP lunar-surface-experiments package did include an RTG, although
the plutonium heat source for it was kept in a separate cask and moved to
the RTG as part of the ALSEP setup on the lunar surface.  Apollo 13's cask
survived reentry (as best anyone can tell) and went into a deep part of
the Pacific; in fact, putting it into deep water was a significant
constraint on the recovery location. 
If we feared danger, mankind would never           |       Henry Spencer
go to space.                  --Ellison S. Onizuka |

From: (George Herbert)
Newsgroups: sci.environment,
Subject: Re: RTGs versus solar
Date: 1 Aug 1996 12:57:50 -0700

In article <>,
Joe McIntire  <> wrote:
>>(1) There are no other demonstrated space power sources for this type of
>>mission.  It's too far from the sun to use photovoltaics.  Saying it has
>>to use something besides Pu-238 is the same as cancelling it.
>Solar electricity CAN be used as a power source on deep space probes.  We 
>are waiting for substantiating docs from NASA-funded Jet Propulsion 
>Laboratory documents that we have obtained  via the Freedom of 
>Information Act after the l989 Galileo launch in which it is acknowledged 
>that Galileo could have been reconfigured to use solar panels.  
>On Ulysses, NASA conceded in its Environmental Impact Statement that 
>solar would do the job. 
>In terms of an alternative to plutonium on Cassini we have the l994 
>"breakthrough" of the European Space Agency on solar photovoltaic for 
>deep space probes, and the statement from the ESA physicist (in USA 
>Today) that given the funds, ESA would be able to have solar available 
>for Cassini.  

Whatever these breakthroughs are, they have not been communicated
to the space engineering community as a whole, which makes me highly
skeptical of the accuracy of your report.  The economic impact of
being able to use such advanced solar cell technologies on earth
orbit spacecraft (weather sattelites, communications sattelites,
refitting them to Space Station Alpha, etc.) is so huge that nobody
would be able to keep such an invention quiet, nor would they
want to... it would make them tons of money and make many things
now impossible or too expensive or difficult in space much more
easy to do.

The basic problem here is something mathematically called the
inverse-square law.  Basically, if you get two times as far from
the sun, you get 1/4 of the sunlight on your solar panels (1/r^2).
Jupiter is roughly 5.2 AU from the Sun; that means the sunlight
there is 1/(5.2)^2 as intense, about 3.5% of what it is at earth.
Saturn is roughly 9.5 AU from the Sun, which gives it about 1%
of the sunlight intensity at Earth.

Earth gets about 1350 watts per square meter of sunlight on the
average, not counting weather (there is none in space 8-).
At Jupiter, the insolation (the technical term for amount of
sunlight energy) drops to about 50 watts per square meter.
At Saturn, it's about 15 watts per square meter.

Let us take a sample solar panel for use in Earth orbit and
look at its performance out at Jupiter and Saturn.  Let us
say it's a ten square meter panel with 20% end of life efficiency
of converting sunlight to electricity, about normal for good
gallium arsenide panels, the best we can do today.  The panel
will probably weigh about 100 kilograms or so (220 lbs).
At Earth, it generates about 2700 watts.  That's enough for
a moderately large workload.  At Jupiter, it generates about
100 watts.  At Saturn, it generates about 30 watts.

Large space probes require fairly large amounts of power.
Cassini is operating on about 750 watts electric power.
If it were aimed at Earth, you could meet its needs with
a fraction of that theoretical solar panel, perhaps 2.5
square meters of solar panel.  If it were a Jupiter probe,
it would need 7.5 of our standard panels, which would weigh
750 kilograms (1650 lbs).  For Saturn, it needs 25 of our
standard panels, or 250 square meters of solar panel
(over 2700 square feet... bigger than most homes!).
Those 25 standard solar panels will weigh 2500 kilograms,
or 5500 lbs, more than a loaded passenger van.

Right now, the whole Cassini spacecraft weighs 5630 kilograms
(over 12,000 lb) of which 3130 kilograms is fuel so it can
stop at Saturn and not keep going past it.  That means the
spacecraft now is massing 2500 kilograms by itself, without
the fuel.  What you are proposing is that it is feasible
to replace about 500 kilos of RTGs with 5 times that mass
of solar panels, in other words nearly doubling the total
dry (without fuel) weight of the spacecraft.  That is not true.
The resulting spacecraft with all of Cassini's instruments
would weigh a total of around 4500 kilograms (10,000 lb)
without any fuel, would require over 5600 kilograms of
fuel to stop at Saturn, and would thus mass over 10,000 kilograms
fully loaded.  There is no rocket launcher on earth which could
launch a mass of 10,000 kilograms on a transfer orbit
towards Saturn.  A requirement to substitute solar panels
on Cassini is not feasible, it is basically destroying the

It is just, barely, feasible to send solar powered exploration
missions to Jupiter.  The percentage of the vehicle which is
power supply is large to the point of dominating the design,
but it can be done.  Some people are looking at designing
such Jupiter missions with solar power, as it appears that
no matter how convincing the evidence that RTGs are not a 
health hazard some few people will insist they are too dangerous.
But you cannot realistically fly Saturn missions with them.
In particular, there is no way that existing engineering
technologies will allow us to convert Cassini to be
solar powered.  There is no solar magic wand we can wave
and change things.  If Cassini is to go, it must use the RTGs.

-george william herbert
Retro Aerospace

From: (George Herbert)
Subject: Re: Stop the cassini Disaster
Date: 20 May 1997 13:52:17 -0700

Mike Hall  <> wrote:
>>An RTG is designed to
>>survive a failure. Most other componants on a spacecraft are not,
>>however. Challenger was not designed to survive its ET exploding. It
>>could not have been designed to survive that... not and leave the
>>ground, anyway. It's a tradeoff between performance and the worst case
>>scenario. If a launch vehicle fails, you write off the launch vehicle.
>>An RTG is specifically designed for a massive failure. It's a small
>>enough item that it can get away with being armored like a tank.
>So you say.

Which part of this description do you not believe?

Describing RTGs as "armored like a tank" is litarary allusion
not technical description.  The M-1 Abrams has 18" or so thick
multi-composite armor and is certainly better protected than
an RTG is, though it's unlikely anyone will be shooting RTGs
with 125mm APFSDS anti-tank cannon rounds.

However, RTGs are built with quite good protection systems.
First of all, the Plutonium is PuO2, a hard ceramic oxide.
The pellets are pretty strong to start with.  They're covered
in a thin layer of Iridium cladding which is strong, heat resistant,
and chemically resistant.  Around that is a strong carbon graphite
impact shield that can withstand a terminal velocity impact with
solid rock.  Around that is a carbon-bonded carbon fiber sleeve,
and around that is a square aeroshell which will actually absorb
most of the heat loads should there be a re-entry.

These protective elements have been tested and performed to spec.
There have been 3 prior flight accidents with US RTGs,
one early one that was designed to burn up and did
(for some reason it was thought to be better to do that
than let it land in one piece); Nimbus-3, a Titan-launched
weather satellite which blew up on launch and whose RTG
was fished out of the Pacific, refurbished a bit, and flown
on the next Nimbus mission; and last Apollo 13's ALSEP RTG
which appears to have survived re-entry and landed intact in the
Marianas trench (no contamination has been observed, though it
might not have been noticed).

Could there be a design flaw?  Sure.  The elements of the design
have been tested, though, so it's not terribly likely.  Could there
be a manufacturing flaw?  Sure.  Also could be one in the 767
you get in to fly to your family's next reunion.  The RTGs are
inspected at least that well.  Nobody is perfect, and any engineer
claiming to be so is lying or incompetent.  But it is fairly
straightforwards to test and inspect and validate designs and end up
with a high confidence that they're reliable and will function right.
We do that every day with aircraft and spacecraft, cars, trains,
ships, bridges, highrise buildings, etc.

-george william herbert
Retro Aerospace

From: Paul Dietz <>
Subject: Re: RTG Plutonium
Date: Mon, 06 Oct 1997 21:56:02 -0500

Chris Stratton wrote:

> 1) How was this plutonium 238 made?   Did DOE go and load up a reactor
> specifically configured to produce it for RTG manufacture?  Or is it
> essentially a by-product produced refined from the products of breeder
> reactors configured to produce weapons-grade Pu-239?  Was it made
> recently, or dragged out of some warehouse where it's been sitting in
> storage since the "atomic battery in your transistor radio 50's?"

Pu-238 is made from Np-237.  Np-237 is isolated as a byproduct
from reprocessed military waste (it's the dominant Np isotope;
the others either have short halflives or very large thermal
neutron fission cross sections).  The Np is chemically separated,
then irradiated with neutrons in a special reactor.  IIRC,
the irradition is done in a particular manner: the Np is slowly
circulated into a target in the core, irradiated for a time, then
taken out to allow the Np238 to decay to Pu238.  The Pu238 is
then separated (otherwise it would tend to turn into Pu239 by
additional neutron capture).  The Np-237 is then sent back
in, continuing the cycle.

> 2) How pure is the PuO2?  Are there any nasty fission products, say
> some isotopes of iodine or strontium mixed in with it?  Were people
> exposed to these during its seperation?  What happened to the waste
> products?  How do you convert metalic Pu to PuO2 (do you "burn" it like
> you would magnesium?)

The separation is undoubtedly done in shielded cells, so there should be
little exposure.  The Pu must have very little in the way of fission
products, to avoid exposing NASA workers (and the spacecraft) to too
much gamma radiation.  I think you get the Pu oxide by heating Pu

> 3) For that matter, assuming that Pu238 gets produced along with
> Pu239, how do you seperate it?

See above.  You don't.  The production of Pu239 is kept down
by the online processing of the irradiated Np solution.


From: buckley@refuge.Colorado.EDU (Charles Buckley)
Subject: Re: RTG Plutonium
Date: 7 Oct 1997 17:03:30 GMT

In article <61c85g$>,
Frank Crary <fcrary@rintintin.Colorado.EDU> wrote:
>In article <>, Chris Stratton  <> wrote:
>>1) How was this plutonium 238 made?   Did DOE go and load up a reactor
>>specifically configured to produce it for RTG manufacture?  Or is it
>>essentially a by-product produced refined from the products of breeder
>>reactors configured to produce weapons-grade Pu-239?
>That's a hard question to answer. As far as I know, DOE hasn't publicly
>released anything about exactly when or how a particular lot of plutonium
>238 was produced.

Hmm. Not quite true. DOE does keep a document (forget the title) which
gives the specific characteristics of the Pu prodiced at each of it's
facilities. This makes tracking the matertials much simpler. Also, any
contraband materials can be readily identifiable as to nationality of

  Publically released is a harder question to ask. The users of the product
have to know what they are getting. How far the info is released beyond
that is fairly limitted.

Index Home About Blog