"That worst case would be a terrorist group hijacking a truck and obtaining a multi-kiloton hydrogen bomb."
Umm... is it just me, or does this really not seem like the worst thing that could happen?
You have a truck with nuclear bombs in it, travelling across America... isn't the bomb already in the place that a terrorist would want it to be?
I would think that the worst thing that could happen, is for one of these trucks to blow up - or am I missing an understanding of how nuclear bombs go off?
Perhaps someone more familiar with the topic than me can answer, but I believe the warhead casing protects from most external explosions (unless it is direct high-kinetic collision).
If it does penetrate the casing, I believe it would not trigger a nuclear reaction but rather become a dirty bomb. This is obviously still bad, but NOT AS BAD as you might think.
Side note, I'm not sure why the article says kiloton? Does the US still use kiloton warheads in its arsenal?
Either way, the reason for the reduced detonation (or imputed dirty bomb) is because multi-megaton nuclear detonations require hexagonal detonators arranged around in sphere with near-perfect implosion timing so it creates the appropriate nuclear reaction (which must implode before it can explode). If a detonator is triggered arbitrarily, I think a dirty bomb would be the only effect (yes, still bad). If the timing is off, then I believe the detonation would be kiloton, not megaton (reduced effectiveness).
The US does indeed have a lot of kiloton weapons. The warheads on the Minuteman missiles, for example, are 300-500kt. Larger numbers of smaller warheads are more efficient at destroying stuff if you can deliver them accurately, so weapons have shrunk quite a lot from their Cold War peak.
I believe all modern American weapons are one-point safe, which means "The probability of achieving a nuclear yield greater than four pounds of TNT equivalent, in the event of a one-point initiation of the weapon’s high explosive, must not exceed one in a million." (http://www.acq.osd.mil/ncbdp/nm/NMHB/chapters/chapter_7.htm) So, if it works as designed, accidental detonation from some external force won't even be in the kiloton range.
I understand "one point safety" to refer to the "detonation of the high explosives by means other than the firing system" as the firing system is designed to fire at the multiple points at once. Can anybody honestly claim that it's completely impossible that the firing system accidentally does the task that it is its main purpose and... fires?
Speaking just from the point of the computer engineer, bit flips or even more bit flips actually occur in practice. One specific "if" branch wrongly taken... one single specific bus line true instead of false... etc...
Only for the first few bombs, the human had to actually add the last piece of uranium or plutonium for the bomb to have enough radioactive mass inside to make the nuclear explosion physically possible. In all the more modern designs everything needed in already inside of the bomb or the warhead. The claimed safety seems to be "everything's there, but the electronics (or the computer) controls everything." How calming.
Regarding one in the million, the people who made the PBS documentary say:
For the accident in Damascus, Arkansas, US: "the air force claimed there's one in the million chance that the fallen socket would hit the missile. During our shooting of the recreation we dropped 12 sockets and 6 actually hit the missile":
"It is recognized
that it is very difficult to provide assurance that the < 1E-03 numerical requirements for a safety
subsystem have been met, let alone the < 1E-04 or 1E-05 assignments given to elements or
components within a safety subsystem. In fact, it is not possible to amass quantitative data that
supports such assertions with a high degree of statistical confidence across all relevant environmental
conditions. In other words, it is not possible to conceive of all possible abnormal environments, nor can all environments
which can be conceived be tested exhaustively in a repeated fashion to generate overwhelming
statistical certainty of weapon response."
You make some interesting points. Certainly, it seems tough to be truly confident of that one-in-a-million claim.
It's theorized that modern PALs incorporate the explosive timings into the authorization code itself. All those different explosive initiators have to be set off at just the right times, and if this theory is true, then the information of when each one needs to be triggered isn't even present in the bomb. This eliminates (well, makes extremely improbable, anyway) the possibility that the electronics accidentally trigger a full detonation due to some malfunction.
Of course, whether this is really true or not is unknown....
Umm... is it just me, or does this really not seem like the worst thing that could happen?
You have a truck with nuclear bombs in it, travelling across America... isn't the bomb already in the place that a terrorist would want it to be?
I would think that the worst thing that could happen, is for one of these trucks to blow up - or am I missing an understanding of how nuclear bombs go off?