Risk Thinking for the California Wildfires (Part 2)
Author: Ron Dembo
A year after the devastating Tubbs Fire in Northern California, in October 2018, I was driving through an area of burned forest the fire had raged. I had been invited to give a keynote speech at the conference of the California Independent System Operator (CAISO) to discuss risk thinking. As I drove, I saw work crews from PG&E installing new transformers on the burnt ground, in exactly the same part of the forest through which Tubbs Fire had spread and which was just as prone to drought and fire as it had been a year ago.
A Clear Risk
During my talk, I showed the below map of California. On it were the locations of transmission wire and transformers and the scientific analysis of where fires would be most likely to occur under the PC8.5 regime, as defined in the IPCC report for the evolution of carbon.
Place one diagram on top of the other, and you will see that fires are likely to occur precisely where critical electrical infrastructure resides.
At the outbreak of Tubbs fire, at least ten reports of downed power lines and exploding transformers had reached firefighters. The scenario was obvious: if there were strong winds coupled with drought and extreme heat in the future, which was highly likely given the Climate Change data available, PG&E should know that one of its transformers could ignite a fire.
A Chance to Hedge
PG&E’s normal hedge in this scenario had been to shut down electricity in affected areas in advance of dangerous conditions. This prevented transformers from sparking but plunged entire neighborhoods into darkness for days at a time, and it cost millions. Had the utility been able to measure the risk against the cost of the hedge, they would have found it totally inefficient from an economic perspective. Worse, the hedge was unable to mitigate the worst of the downside because there was no guarantee that PG&E could predict the timing of transformer sparks.
If PG&E had risk thought back in 2018, the utility would have seen the need to move transmission lines or, alternatively, bury those in the most dangerous areas. As it was, they did nothing — taking, probably unknowingly, a calculably astronomical bet.
The Predictable Consequences
A few weeks after that conference with CAISO, a spark from one of PG&E’s transformers caused Camp Fire, the deadliest in the history of California. 85 civilians lost their lives, 18,804 structures were destroyed, and the company lost 85% of its market value before going bankrupt just two months later. When it filed, it cited expected wildfire liabilities of $30 billion. On June 16, 2020, it pled guilty to 84 counts of involuntary manslaughter. Stunningly, as of August 2020, CAISO is still using rolling blackouts to cut power to a quarter of a million homes.
Have they learnt nothing?