The reactor is explained properly, which is more than most manage.

Six sittings. I came to this expecting to be irritated, because I have read a great many accounts of failures in my own field written by people who did not understand the thing that failed, and they are all the same account.

This one is not that. The positive void coefficient is explained clearly enough that a person without physics can follow it, and so is the graphite tip on the control rod, which is the detail everything turns on: the shutdown mechanism briefly makes the reaction worse before making it better. A safety device with a spike in it. Who signs that off?

The other thing he gets right is the shape of the failure. It wasn't one operator and it wasn't one flaw; it was a test postponed from the afternoon to the night shift, handed to a crew who had not prepared it, running a procedure altered in pen, on a machine with a characteristic nobody on the floor had been told about. Failures live in handovers. In 1983 I watched a temporary works arrangement get changed over a weekend by a man who did not know why it had been drawn that way, and we were lucky in a way those men were not.

Nine. The evacuation chapters are weaker than the technical ones, and I would still not skip them.