[daveswilson]

Placing blame is far less interesting, IMO, than considering this story in the context of our much-needed next-generation power grid.

One thing such a grid, and our management approach to it, has to do is not fail catastrophically, showering everything around it with sparks and cutting power to who knows how many customers. I assert that this is not a good thing regardless of whether a huge fire is started and people die as a result.

If you look at PG&E's outage map, it has a way of putting a smiley face on failure. You'll find it at https://m.pge.com/#outages, but of course you will have to look below the fold.

When you do that you will see the "CUSTOMERS AFFECTED" key. If less than 50 customers are affected by a power outage, PG&E calls that "green". There are always many green dots on their map in the Bay Area, whenever I've looked. That's another symptom of simply running things until they fail, and another indicator of the current state of our electrical infrastructure, and certainly an artifact of the culture at PG&E and the politics that allow it to exist.

As we reinvest in and modernize our grid, we need financial incentives for operators around reliability, and standards for both construction and monitoring that support their efforts to improve that. That could begin at the federal level, if we had a functioning federal government, in much the way the largely-successful things we did prior to 2016 have done (the NASA commercial crew program, what-used-to-be-the-CDC, NIST, etc.).

If we could put the right incentives in place and then take one more step - prioritize modernization of the grid in fire-prone areas - then we can look forward to a point in time when we're not shutting off people's power because the wind blows.

Sure, fires start for other reasons and we're living through them now. That's no excuse for keeping our profit motivations in the wrong place. Fewer fires is helpful and reliable power is helpful. Let's start paying businesses for reliability (and efficiency, and renewability while we're at it).

We're getting what we have because we're paying pretty much exclusively for the number of kilowatt hours delivered to where. A lot of things have changed since somebody came up with that. Seems we should blow that bit up, insert ways to measure additional things, and pay for performance.