California’s Outages Are a Teachable Moment


Two keen observers of the power sector, Jigar Shah and Cheryl LaFleur, have noted that the responses to the rotating power outages in California on August 14 and 15 became a Rorschach test of individual preexisting biases. Before many facts were known, a favorite villain was chosen and a conclusion drawn: too much solar, not enough gas, underperforming gas, corrupt regulators, inept utilities, and failed market design are a few examples. Rather than jumping to such conclusions, however, we should recognize that extreme events and the responses that follow offer a high-profile teachable moment — and we should be learning and teaching all that we can.

Some facts surrounding the event have become clearer over the past several weeks, and more will be learned about the strengths and vulnerabilities of the California grid as historically hot weather and new wildfires continue to affect consumption and available generation. One thing is certain: The climate crisis is presenting increasingly severe challenges, and the need to effectively accelerate power sector transformation has never been more apparent.

We need to know how severe the controlled outages were and what prompted them. These outages are a signpost for potential larger-scale events, so assessing the need for action to protect system stability is important. Short-term actions to fend off additional supply-related outages have been implemented, and they are working. Long-term actions require the benefit of factual investigation and due consideration of fixes that accelerate market transformation while addressing any identified system vulnerabilities. Distribution utility practices, utility regulations, wholesale operating practices and market design improvements need to be considered to address identified vulnerabilities.

I am concerned that the investigation may overlook how this event affected people. Shah noted that the health impacts of outages tend to fall disproportionately on lower-income communities. The investigation should include a focus on how the outages affected people and whether they affected some populations disproportionately.

Above all, events like these need to generate lessons and dispel myths so that we can accelerate our progress toward power sector transformation. Key questions include:

  1. Who was affected?
  2. How severe were the simultaneous demand and supply contingencies?
  3. What does the success of voluntary load reduction on August 17 and beyond tell us?
  4. How did the California Independent System Operator (CAISO) market perform?

Who was affected? The CAISO staff provided a public briefing on August 17 to share information about the outages that occurred on August 14 and 15. The briefing indicated that on August 14, a total of 1,000 MW of load was shed for 80 minutes, with about 410,000 customers affected by rotating outages during that time. The August 15 event was much smaller — 470 MW shed for about 20 minutes. The customers affected by the rotating outages are determined by the distribution utilities (Pacific Gas and Electric, Southern California Edison and San Diego Gas & Electric). I have not seen a report on which customers were affected, nor have I seen assessment of whether any adverse public health impacts were experienced.  While the duration of the outages is modest from what we know so far, we should not miss the opportunity to learn whether they threatened public health or disproportionately affected certain low- and moderate-income communities. Setting a positive example assessing any disproportionate impacts on these communities is useful and will encourage similar assessments for other outages that may be more frequent and with greater duration.[1]

How severe were the simultaneous demand and supply contingencies? We know this was an extreme weather event that was prolonged, severe, and extended through the Southwest, California, and the Northwest, so we know that demand was higher than expected while availability of supply-side resources from the region was lower than expected. The CAISO staff public briefing summarized the specific contingencies that led to the need to shed load on August 14 (the unexpected loss of a 475 MW generation facility, combined with higher-than-expected loads and lower-than-expected imports) and August 15 (the unexpected decline of 410 MW of wind production, combined with an unexpected loss of a 400 MW generation unit). In addition to these discrete, incremental events, a joint letter to Gov. Gavin Newsom from CAISO, the California Public Utilities Commission (CPUC) and the California Energy Commission (CEC) identifies many areas for investigation that indicate additional contributing factors. Identifying these factors will provide insight into modeling, resource and program deficiencies that should be reflected more accurately going forward. Identifying these deficiencies will be helpful in the ongoing resource adequacy discussion in California and the West.

What does the success of voluntary load reduction on August 17 and beyond tell us? One commentator ominously predicted that the “wolf is at the door,” indicating a common expectation that outages would be worse on a Monday (August 17) than they had been on Friday and Saturday. Instead, there were no rotating outages needed at all. The CAISO staff briefing for August 18 shows that more than 4,000 MW of load was shaved from consumption during the highest-stress hours on August 17. That 4,000 MW substantially exceeds the 800 MW of emergency demand response accessed by CAISO on August 14; it exceeds the approximate 1,000 MW of diesel generation authorized to operate during the emergency and also swamps the magnitude of the deficiency (about 1,000 MW) that caused the need for outages on August 14. This tells me that demand response programs in California are broken, and the potential for an order-of-magnitude increase in demand response is reasonable. Severin Borenstein provides two straightforward measures to expand demand response in California: critical peak pricing and improved communication with customers.

How did the California ISO market perform? We know that the rotating outages were limited and did not persist into the following week, so there is evidence that policy actions and the market effectively limited more serious damage. However, we do not yet know whether prices effectively conveyed the urgency of the event leading up to and through the contingency periods. Prices were higher — but were they high enough to elicit a fast response to the contingency events? Further consideration of raising price caps to induce greater resource participation is worth consideration. The CAISO Department of Market Monitoring will fully vet any potential withholding, and that needs to be included in the assessment. At the same time, we also need to understand whether suppressed price signals originating from market design choices contributed to insufficiencies in demand- and supply-side resources.

A full examination of the facts surrounding the rotating outages is needed before long-term policy responses are implemented. The short-term actions have been effective, and the distribution utilities, the CPUC, CEC and CAISO should be commended for rising to the challenge. The joint agency memo to Newsom cites many of the questions that need to be investigated to identify lessons, expose myths and craft solutions. Most important, solutions must address the lead story: The climate crisis is at our door, and accelerating progress toward an equitable, decarbonized power sector remains the focus.

Note: Dr. Carl Linvill is a member of the Western Energy Imbalance Market (EIM) Governing Body. This blog post reflects his individual views, not those of the EIM Governing Body or CAISO.


[1] The question of how outages and blackouts affect communities is important to understand for all outage events, whether they are controlled, short-duration outages like these, longer-term outages arising from local distribution system deficiencies or more prolonged outages arising from more severe events. Outages on distribution systems are the most frequent, often last longer than an hour and tend to affect some neighborhoods more often than others. Wildfire season brings Public Safety Power Shutoffs (PSPS), which leave affected customers without power for an entire day or for several days as the utility protects us all from additional wildfires. And wildfire damage can deprive customers of power for days or even weeks.