A couple of thoughts, from someone who is a project manager for some reasonably critical transportation infrastructure:

1. You cannot supply enough alternate power using standby generators for more than a few percent of the facility's normal demand. You scatter around some plug-in circuits or focus on a very small set of rooms and then you make the BIG decision -- how many of these important spaces get air conditioning (or electric heat, if relevant)? The economics of generator size and operations quickly limit your possibilities.

2. In one of my projects, we sought two independent utility feeds to support some critical machinery. But when we discussed failure modes, there was always a situation where a switch was needed -- a common gadget in a single place. This kind of automatic transfer switch is a feature, not a bug, as it allows power to be uninterrupted inside the facility for a range of not-uncommon offsite situations (like storms). In our case, the utility provided the gear at an extra price on the customer's bill. Yes, there were ways to set up even more redundancy with a series of manual switches, but then there are the transformers downstream, etc. If you started separating the gear leading to these more thoroughly, you risked more frequent events where one concourse, etc. would go down. One of those events every ten years or one complete 12-hour shutdown every thirty years?

None of this tells us why there wouldn't have been more generator backup to properly light some critical areas, such as the pedestrian tunnel and support good communications. And a plan for the summer heat -- from what I saw, this event on a July afternoon could have been a real tragedy. A couple of large "cold rooms"? Another idea would be to support a concourse or two at a higher level of redundancy, so they could receive planes.

Of course, if we could imagine how every disaster unfolds, we wouldn't have disasters.







i