Cost-Benefit Analysis of the Spy Satellite Shootdown Attempt

February 20th, 2008

Posted by: Roger Pielke, Jr.

The Navy is apparently going to try to shoot down that wayward spy satellite sometime in the next 48 hours. The attempt to shoot it down is justified in terms of protecting human life from the risk of harm caused by the satellite’s uncontrolled reentry. This post discusses whether or not the shoot down attempt can be justified in cost-benefit terms. I don’t think it can, at least in terms of the formal justifications provided by the U.S. government. There must be other factors involved. The costs per expected life saved are about $2-$3 billion dollars! Read on for details.

CNN reports how the U.S. government has justified the shoot down attempt:

Pentagon officials argue the effort is worth the expense because of the slim — but real — chance that the satellite’s unused fuel, 1,000 pounds of toxic hydrazine, could land in a populated area.

Because the super-secret spy satellite malfunctioned immediately after launch in December 2006, its fuel tank is full, and it would probably survive re-entry and disperse harmful, even potentially deadly fumes over an area the size of two football fields.

Here are the details of a cost-benefit calculation.

In early 2001 I was asked by NASA to organize a workshop to examine the risks and benefits associated with the reentry of the TRMM (Tropical Rainfall Measuring Mission) satellite. NASA faced a choice between using the satellite’s remaining fuel to conduct a controlled reentry into the Pacific, or allow the orbiter to continue gathering data on tropical cyclones and other weather and eventually plunge to earth in an uncontrolled manner. What option had the most benefits as compared to the risks?

Our workshop report can be found here. What is relevant for this current discussion is that NASA prepared an analysis of the probability of casualty due to an uncontrolled reentry of TRMM, which is roughly of the same size as the spy satellite. The NASA presentation can be seen here in PDF.

The area affected by the re-enetering TRMM satellite was estimated by NASA to be about 110 square feet. Because the fuel tank was to be empty there was no concern about a toxic cloud of hydrazine, as is apparently the case with the spy satellite which is carrying a full tank of fuel. As the CNN article cited above notes, the concern is that the ruptured fuel tank might affect an area as large as two football fields, or about 115,000 square feet, or about 1050 times larger than the area that was expected to be affected by TRMM.

In the NASA presentation at our TRMM workshop a risk of human casualty associated with an uncontrolled re-entry was calculated to be 1 in 5,000 (please see the PDF for details of the calculation), based on the size of the debris field and the population density of the area under the orbital path. A risk of 1 in 5000 is pretty small, hence in the end NASA decided to let TRMM fall in uncontrolled fashion (which will happen in a few years).

But the much larger potential debris field associated with the spy satellite suggests a risk of casualty — using the NASA risk estimate for TRMM scaled to the larger debris field — of about a 1 in 4.8 risk of casualty, or a bit worse than the odds of surviving Russian Roulette.

Now at an estimated cost of $40-$60 million to shoot down the satellite results in an expected cost per casualty avoided of about $190-$290 million (i.e., $40M*4.8 – $60M*4.8). Given that the vast majority of the casualties might be temporary discomfort from inhalation of hydrazine gas this is a very high cost, this is a very high figure (I’d probably inhale a bit of hydrazine gas for $250 million;-). If we consider the cost per life saved, say, by avoiding a direct hit on a person by the satellite, the cost per life saved rises to $2-$3 billion! (That is, 5000*$40M – 5000*$60M.)

In any reasonable calculation of the costs and benefits to human life of the spy satellite reentry, it does not seem justifiable in terms of human safety according to standard measures of the economic value of lives saved (which tend to be measured on the order of $10 million). So there must be other perceived benefits as well — such as testing the U.S. military’s ability to hit a satellite, showing adversaries U.S. military prowess, and simply eliminating uncertainty associated with the satellite’s reentry. The odds may be tiny, but if it happens to land on the Russian Embassy in Ecuador, or some other highly inconvenient place, there would certainly be a huge diplomatic crisis.

Being rich sometimes means that you can afford to eliminate uncertainty via decision making, regardless of the cost. But don’t be fooled by the justification, it doesn’t stand up to scrutiny, which seems fairly typical of the current administration.

4 Responses to “Cost-Benefit Analysis of the Spy Satellite Shootdown Attempt”

  1. kmcolo Says:

    There is also the possibility that portions of this new, secrets packed, satellite might survive re-entry and land in, oh… China or Russia perhaps.

  2. 2
  3. kmcolo Says:

    There is also the possibility that portions of this new, secrets packed, satellite might survive re-entry and land in, oh… China or Russia perhaps.

  4. 3
  5. TokyoTom Says:

    Roger, this obviously irration decision from a public cost-benefit perspective rightly leads you consider other possible motivations. Sending a message to Russia and China would be top of the list, especially as China just did something like this itself.

    But you might also consider how private interests and public interests frequently diverge in the case of government decisions. No doubt the Navy was happy, some defense contractors are happy, and Republicans and the Bush are happy to make some public display of their potency and ability to defend Americans.

    A big government is easily manipulated in this kind of way.



  6. 4
  7. JoeWalters Says:

    The is an interesting article of some of the issues in the shootdown in the August 2008 IEEE Spectrum ( An interesting observation, to me, in the article is that the “experts” expected the hydrazine to be liquid at impact. It apparently was frozen solid due to the satellite failure and would have required so much heat to make the state change from solid to liquid that it would have likely been liquid at impact.

    The article has a higher probability of human injury (1 in 45 to 1 in 25) than used above (1 in 5000), or about a 100 to 200 fold change in cost per life saved. Would be interested in a discussion of that difference…why the difference? The article mentioned that it was much higher probability than had been used in the past due to the full fuel tank due to failure at launch.