Fault Tree Analysis

Back in the 1950s the DoD instituted a program that  became Fault Tree Analysis. Namely with large scale systems they analyzed all predictable failure modes and then determined the signals that would be associated with them and what actions would be taken. It was then adopted by the nuclear forces to be applied to nuclear weapons as well. Clearly it could not anticipate terrorism of stupid humans, but for the most part it became a doctrine.

By the early to mid 70s I was applying it to our satellite communications and then to the early Internet structure in the ARPA Net world. But the Internet dealt with faults via simplicity and redundancy, unlike the telephone networks which were hierarchical.

Now to power networks. These are single threaded networks with single points of failure. There is no Fault Tree like approaches and their network management is an after the fact approach. Namely when a colossal failure occurs then then try to find out what happened.

The reason for this is that power utility companies draw their people from the least technical in the entire pool of resources. Thus the tendency for massive power failures. Worse is yet to come if an adversary can attack this system. It seems all too simple by focusing on SCADA elements connected to the Internet. Why they use an open and compromiseable network is beyond belief but they do.

Perhaps there should be some Congressional hearing on this but frankly who would testify? All that's left are those who sit and wonder why the lights are out.