by Jeff St. John
October 18, 2019

Pacific Gas & Electric’s decision to black out the grid for millions of its customers last week has come under fire from all quarters of California, from Governor Gavin Newsom and top lawmakers and regulators, to millions of Northern California residents left in the dark — both literally and figuratively — by the bankrupt utility

But amid all this condemnation, it’s hard to argue with the fundamental premise of PG&E's decision. 

Yes, the decision led to disruption for millions of customers and billions of dollars in lost work and productivity. But during a week of hot, dry and windy conditions labeled high fire risk by the National Weather Service, in which blazes broke out near Yosemite, in the San Fernando Valley and elsewhere across the state, PG&E's power lines caused no fires.

There’s no way to prove that PG&E prevented a disaster. You can’t predict how a fire might spread from a spark that never occurs. Still, PG&E has identified more than 50 instances of weather-related damage to its grid, including downed wires and trees or brush touching power lines, that could have been fire risks. 

And, of course, it is an excess of caution that drove the California Public Utilities Commission to expand the scope of public safety power shutoff (PSPS) events for the state’s three investor-owned utilities late last year. The impetus behind that move was the November 2018 Camp Fire, which killed 85 people, nearly completely destroyed the town of Paradise, caused an estimated $16.5 billion in damages, and drove PG&E to file for bankruptcy protection in January. 

PG&E considered, but decided against, de-energizing in the days and hours leading up to the Camp Fire. But because its less aggressive PSPS plan last year didn’t include the 115-kilovolt transmission line that state fire investigators have determined caused the fire, it wouldn’t have prevented the fire even if PG&E had pulled the plug. 

This binary choice between blackouts or more fires has animated PG&E CEO Bill Johnson’s primary response to what he has conceded was a flawed execution of the state’s largest-ever PSPS event. "I do apologize for the hardship this has caused, and I think we made the right call on safety," Johnson said.

There’s little doubt that power outages will play a continued role in fire prevention for a state that’s coping with ever-increasing fire risks driven by climate change, poor forest management and increasing human development. But that’s not really the question facing California lawmakers, regulators and the general public. 

Instead, it is whether PG&E could have better prepared and executed a fire-prevention plan that, while mandated by state law, is largely up to the company’s own meteorologists, vegetation management teams, grid operators and customer outreach and support staff to institute. Or, more importantly, it's how PG&E can make sure it doesn't repeat the same mistakes next time. 

A long and uncertain path to better-managed fire-prevention outages 

On Monday, the California Public Utilities Commission directed PG&E to immediately address a number of failures in the execution of its PSPS, including its poor customer communications and coordination with state, county and local agencies forced to bear the brunt of dealing with the effects of its outage. 

These may be problems that PG&E can solve. While bankruptcy brings its financial stresses, the utility can be expected to ensure it has enough server space and call center staff for the next event, and to take the steps to assure integration with the state’s Office of Emergency Services, for instance. 

But the same can’t be said for the CPUC’s order to “take all possible measures” to limit the scope and duration of future fire-prevention outages. These include orders to prioritize high populations areas in its ongoing multibillion-dollar program to clear vegetation and inspect and fix power lines, and to strive to limit the length of outages to no more than 12 hours, rather than the 48 hours now set out in its PSPS protocol. 

Some of the CPUC’s goals on this front are backward-looking. For example, the CPUC has asked PG&E to summarize the grid-hardening work it’s done so far on infrastructure that was shut off during last week’s event, and whether they could have been adjusted to minimize its scope. 

But any demands that regulators set on how PG&E manages future events will be limited by the physical infrastructure of the power grid it has, the technologies it’s deploying to better manage it, and the internal depth of expertise to use these new technologies to best effect.   

PG&E lags well behind in these terms compared to San Diego Gas & Electric. The much smaller investor-owned utility has invested more than $1.5 billion in state-of-the-art fire prevention technology and programs in response to deadly wildfires in 2007, including tree-clearing, grid-hardening, and a full-time meteorology staff to analyze the data from 16 high-definition mountaintop cameras and 177 weather stations across its territory. SDG&E has also been able to sectionalize much of the grid network serving its highest fire-risk areas, allowing it to limit outages to tens of thousands of customers at a time. 

This purpose-built forecasting capability, the granular control over where outages happen, and a big investment into community outreach and support during outages “have been three pillars of success for SDG&E,” Elizaveta Malashenko, CPUC’s deputy executive director for safety policy, said in a review of last week’s event. “And they are currently sources of failure for PG&E.” 

At the same time, PG&E is much larger than SDG&E, covers far more mountainous forested terrain, and relies far more on long radial transmission lines to serve its more remote customers. These factors, along with PG&E’s “lack of granularity of its forecasting ability” — a reflection of its size and its relative dearth of sensors compared to SDG&E — mean that PG&E can’t “activate PSPS events as strategically,” the CPUC conceded. 

For example, PG&E reported last month that it has only completed clearing tree limbs, brush and other vegetation from about one-third of the 2,455 miles of power lines it had promised under its Enhanced Vegetation Management program. But PG&E and the CPUC will have to closely analyze this data to determine whether or not this lack of progress played a role or not in last week’s unfolding of events. 

Likewise, PG&E has largely completed the Wildfire Safety Inspection Program work it promised the CPUC, including inspecting and repairing its high-fire-risk power infrastructure, installing cameras and weather stations, and sectionalizing its circuits to limit the geographical reach and speed up the restoration after PSPS events, it reported last month. 

But whether or not PG&E could have used all those new cameras, sensors and switches to make last week’s outage shorter, smaller or more pinpointed to areas of specific risk, if only it had the years of experience SDG&E has in how to use them, is an open question. It’s possible that all that forecasting gear could have informed grid operators that it was safer to leave as many circuits off as possible, or that its new sectionalizing equipment happened to be in the wrong place to minimize the resulting outages. 

The misaligned incentives between power outage and fire risk 

It’s going to be critical for PG&E and the CPUC to find the data needed to provide fact-based answers to questions like these, if last week’s political backlash is any indication. Under a legal concept known as "inverse condemnation," California utilities are liable for damages from fires caused by their equipment even if they aren’t at fault. And PG&E can’t access the $21 billion wildfire fund created by the state this summer to backstop the state’s utilities against future wildfire liabilities until and unless it exits from bankruptcy by June 2020. 

But beyond complying with CPUC’s orders to invest in improving its PSPS response, PG&E doesn’t bear the costs of the economic disruptions of blackouts. For last week’s event, those include activating the state’s emergency operations center to manage the threat of losing power to railroad crossings, water delivery systems and other critical infrastructure, and the city and county responses to school closures, traffic accidents and aiding the sick and elderly people at real risk of death during a power outage. 

In simple terms, PG&E has no financial incentive in trying to make future outages as limited and strategic as possible, Andrew Campbell of the Energy Institute at the UC Berkeley Haas School of Business wrote in a blog post this week. “It’s time for the state to go beyond criticism of the utilities and take charge of utility shutoffs, while continuing to push forward with other efforts to protect the state from catastrophic wildfires.”

Just how this should happen is an open question, albeit one that’s likely to emerge as state lawmakers continue to pass laws aimed at tackling the wildfire crisis, and Gov. Newsom and the CPUC continue to press PG&E on improving its response. At the same time, state lawmakers have limits to their power. For example, while Gov. Newsom has asked PG&E to offer blacked-out customers rebates of $100 per household and $250 per small business, the state can’t compel the utility to do so, particularly not when its operations are under the supervision of a bankruptcy judge.

But any entity, organization or institutional framework seeking to better perform the responsibility of assessing the risks and making the real-time life-and-death decisions involved in fire-prevention blackouts will likely find that, like PG&E before them, they will face the responsibility and blame for the results.