How does the launch risk posed by plutonium compare to the launch risk posed by propellants?

For outer solar system exploration, virtually the only feasible power subsystem are Radioisotope thermoelectric generators (RTGs). These include plutonium, which may carry considerable risks (see also this question). According to Space Mission Analysis and Design (Larson and Wertz, Third Edition, Eight Printing, 2006), page 335:

We must also consider safety issues, but RTG sources are probably safer than most propellants.

How true is this statement? In the case of catastrophic launch failure of a spacecraft carrying several kg of plutonium, how does the risk posed by the plutonium compare to the risk posed by propellants? By risk, I mean the danger posed to human health and to the local and global environment.

Edit: I'm looking for quantitative calculations. The question I linked above links to environmental impact statements that NASA did for the launch of Cassini, New Horizons, and Mars Science Laboratory. These contain risk calculations: for certain catastrophic scenarios, they estimate the consequences to human health (latent cancer risks) and cleaning costs. I'd like to see similar calculations for the consequences of large amounts of propellants, or their partial burning, being released locally or through a larger region.

• James Wertz's SMAD is still state of the art and standard literature. It basically reflects current thoughts in the business. So risks ... compared to what? A crash involving a massive amount of hydrazine and derived substances? Under the fair assumption that there can be devastating problems in case of a really ugly crash with both, RTGs and hydrazine, the discussion about 'numbers' is a bit weird. How should an answer to this look like? – s-m-e Jul 23 '13 at 11:21
• @ernestopheles The Environmental Impact Statements from NASA for the missions Cassini, New Horizons, and MGS, and possibly others, contain calculations on risks to human life (such as latent cancer forms) and cleanup costs should the launch area get contaminated in case of a crash. A possible answer would contain independent calculations for the environmental consequences of a catastrophe, local and globally, considering human and ecosystem health and cleanup costs. I've edited the question for clarification. – gerrit Jul 23 '13 at 14:28
• Sorry, that should be MSL, not MGS. – gerrit Jul 23 '13 at 14:38
• Afaik, Plutionium is predicted not to vaporize in a rocket explosion. It would come down as a solid block, severely limiting the affected area. Someone look up a source for this? – Rikki-Tikki-Tavi Sep 30 '14 at 17:40
• You would need to look at the Environmental Impact Statements for the launch systems referenced in the linked documents. E.g., in the MSL EIS the references USAF 1998 and USAF 2000 cover the EELV program in general. – Mark Adler Oct 30 '14 at 4:34

Plutonium is astonishingly poisonous. But conversely, the amount of Pu used by a mission is fairly small. And it naturally decays.

Comparing double digit kilograms of horribly poisonous stuff (say 10Kg), vs hundreds of thousands (if not millions) of kilos of propellant (even Tetrazine, MMDH) makes it seem true.

Additionally, knowing how bad the Pu is, it is surprisingly armored. People forget that Apollo 13 dropped the LM's RTG into the deepest part of the Pacific, and the belief is that the containment vessel survived one of the fastest reentries to Earth's atmosphere. (I like how Top Gear America's first driver on their track was 'the fastest man they could find' Buzz Aldrin, but of course the Apollo 13 had him beat by a bit).

• Comparing double digit kilograms of horribly poisonous stuff (say 10Kg), vs hundreds of thousands (if not millions) of kilos of propellant (Even Tetrazine, MMDH) makes it seem true. — It's not directly obvious to me that "small amount * extremely bad" is worse than "large amonut * pretty bad". I'd like to see a more quantitative answer, yours doesn't add very much to the quote in the question.. – gerrit Jul 16 '13 at 23:43
• The propellant will be used either way. Using RTG's doesn't reduce propellant carried, it increases science payload carried. – aramis Jul 23 '13 at 9:52
• @geoffc: In terms of Apollo, you are referring to the RTG inside the ALSEP? – s-m-e Jul 23 '13 at 10:55
• @geoffc This answer needs some improvement. RTGs produce electricity. Although you could theoretically use to drive an ion engine, this is not done. So RTGs equal electricity for sciences and e.g. communication. Fuel (LOX), however, can be used for fuel cells as well, thus for electricity too. But in terms of robotic science missions leaving Earth's orbit, they are not used. Solar power, on the contrary, has been used for both, science payloads AND driving ion engines. I must admit, it is a bit confusing. – s-m-e Jul 23 '13 at 11:07
• @ernestopheles The point is, while Pu is dangerous. So is Hydrazine. but with Pu there is at most 10 kilos. Which is really not alot, in the scheme of damaging a particular area. Vs hundreds of thousands, if not millions of pounds of hydrazine. It is about relative risk, and the quantity affects that thinking. – geoffc Jul 23 '13 at 11:58

Solar is feasible for at least as far as Jupiter; NASA is using it for the Juno Mission. Solar panels are heavy compared to equivalent-power RTGs.

Fission reactors of several different kinds are also practical, but due to political concerns (and treaties) are not used by any space agencies to date. They too are heavy by comparison to RTGs, but lighter than Solar, in the typical configurations. They also have shorter lifespans, but much higher power output. (And, in theory, once expended, could have an RTG thermocouple system module which then turns them into an RTG.) The Uranium is not "safe", but isn't itself toxic other than its radioactive properties.

Radio-Thermal Generators (RTGs) have the advantage of having no moving parts, and being extremely light. There is almost no chance of failure that doesn't involve mechanical damage. They last for decades (the Pioneer and Voyager probes are all still operating †). The drawback is that the plutonium used is itself highly toxic, as well as radioactive. While there have been failures, the normal launch protocols have resulted in those failures not having significant population-affecting effects.

Versus propellants, the question is almost irrelevant. The total propellant used will be used pretty much no matter the weight of the on-probe power. This is because of (1) the various stages being general-use designs, (2) the normative design mode being to design a mission for the maximum payload capacity of the launcher for the trajectory desired, and (3) the desire to pack as much science on board as one can. Thus, the power source choice usually affects only what instrumentation is aboard, not how much fuel gets used; the exception being when it results in a mission being either scrapped as unworkable or having to be turned into multiple missions.

† Pioneer 10 was still broadcasting as of 2003; the broadcast antenna is one of the more high-power instruments aboard, and there isn't enough electricity for it to continue to operate the radio. But it's still got SOME power... over 30 years in flight and still under power. Just not enough to hit the 57W needed to operate the 8W transceiver. http://science1.nasa.gov/science-news/science-at-nasa/2001/ast03may_1/ and http://www.unmannedspaceflight.com/index.php?showtopic=2362

• I don't agree that the question is almost irrelevant. The two are independent, yes, but a risk analysis of one versus the other is insightful because it puts things into perspective. If we decide for solar rather than RTG's even though the risk posed by propellants is much larger, then perhaps we're being irrational. I just find that SMAD goes more rapidly over the safety issues than the toxicity and radioactivity of plutonium justifies. – gerrit Jul 23 '13 at 14:36
• @gerrit it's irrelevant because the power source for the probe almost never impacts what the launcher fuel load is; the launcher sets the design parameters for probes, not the other way around. Every design log I've seen published shows the launcher limits are picked FIRST, then the probe built to fit. I've never seen it the other way. – aramis Jul 24 '13 at 7:54
• I don't deny that the two are more or less independent design parameters, if the risk posed by A is negligible compared to the risk posed by B, then after deciding B, it's pointless to change A for safety reasons. – gerrit Jul 24 '13 at 10:22