home ||| current issue ||| past stories
about The Prism ||| volunteers ||| other sites

200,000 could be killed in Cassini accident

by Michio Kaku


In our March issue, the Prism reprinted an article "Racism meets Spacism," which included information about the impending launch of the Cassini space probe in October. This probe will carry 72 pounds of plutonium and will fly past Earth at 42,000 mph in 1999.

NASA bureaucrats have claimed that a maximum of 2,300 people could be killed (via cancer over a 50 year period) from an accident with the Cassini probe.

However, by carefully retracing their calculations, line by line, one can see that they have taken the lowest possible estimates at every step in their calculation. NASA is not considering the maximum possible accident at all.

The three steps where they underestimate the risks are:

  1. assume that only a few percent of the plutonium particles will actually reach the ground and be released into a populated area. (A more physically realistic estimate would conclude that perhaps 30 to 60% of the plutonium may be released in a populated area.)
  2. assume that the plutonium will fall within a small region of the earth. One NASA calculation has the plutonium confined to within about a square mile. But obviously, this does not factor in the wind, which is known to carry dust particles for 50 miles and beyond.
  3. assume a low population density, not a major urban center.

They are able to conclude that only 2,300 will die in a Cassini accident by consistently taking the lowest possible estimates at each point in this calculation, so I estimate that they can easily be off by 50 to 100 times, which would push the casualty figures into the hundreds of thousands, using their own calculation as the starting point. I conclude, therefore, that their calculation is self-serving, and borders on scientific dishonesty.

Their computer calculations are also suspect. They use the same discredited methodology (e.g. event tree analysis) which is used for nuclear power plants, which failed completely to predict Three Mile Island and Chernobyl. In particular, they don't factor in human failure, multiple failures, common-mode failures, and design flaws, which actually have caused most of the great disasters of the 20th century. In other words, their computer programs are largely a waste of tax-payers' money. NASA bureaucrats have thus committed a grave error: believing their own press releases, and placing ideology before the laws of physics.

Instead of using computer programs, which are misleading, one should look at the track record of actual failures. A chain is no stronger than its weakest link, which is the Titan IV booster, which has a failure rate of 1 in 20. The probability of a major accident is therefore not one in a thousand or one in a million, but 1 in 20.

  Dr. Michio Kaku is a professor of Theoretical Physics at the Graduate Center of the City University of New York and is author of Hyperspace and Introduction to Superstrings and several other books.  

home ||| current issue ||| past stories
about The Prism ||| volunteers ||| other sites

Send comments to prism@sunsite.unc.edu.