Inside a massive dome in San Ramon, just east of San Francisco, PG&E engineers are recreating a utility nightmare: power lines starting wildfires.
It’s a threat PG&E knows intimately. The utility has paid millions in penalties for transmission line-caused fires, including the 2018 Camp Fire — California’s largest fire, which killed 85 people and flattened the town of Paradise — as well as the 2019 Kincade Fire and the 2021 Dixie Fire. The financial fallout led PG&E to file for Chapter 11 bankruptcy protection in 2019 to manage billions in wildfire-related claims.
In light of the devastating fires in Los Angeles, those earlier fires are back in the headlines — though this time it’s a different utility caught in the nightmare PG&E knows well. In a terrible revisiting of history, Southern California Edison is facing allegations that its power lines sparked one of the fires currently ravaging the area.
PG&E’s territory doesn’t intersect with the worst of the fires, at least not yet. But nonetheless, 300 miles north at PG&E’s innovation campus in San Ramon, the work of avoiding more catastrophe continues.
At the Advanced Technology Services Campus — at the center of which sits the high-voltage dome once used to recreate foggy coastal conditions but now a testing ground for wildfire mitigation technology — PG&E is attempting to thread a difficult needle: accelerating the pace of tech innovations that the power sector increasingly needs to meet coming load growth and natural disasters, while both ensuring reliability and, ideally, not raising prices.
Inside a high-voltage dome, engineers are creating artificial lightning between power equipment to test their safety limits. Outside, they’re simulating wildfire conditions by dropping branches onto power lines to test the efficacy of Enhanced Powerline Safety Settings (one of dozens of technologies being tested on the campus, which will essentially replace proactive power shutoffs as a wildfire risk reduction tactic).
The ATS approach employs a structure that’s more familiar to PG&E’s Silicon Valley neighbors than to the utility sector. It includes rapid prototyping, close vendor partnerships, and rigorous stress testing. But in deploying this approach, the company is facing a rather unexpected challenge: the utility legacy of painfully slow movement has meant that most startups are prioritizing other markets, and ironically, aren’t ready to move as fast as the lab itself would like.
A unique structure
While most utilities outsource testing and validation of new technologies to entities including academic centers or external companies, PG&E leverages ATS and its nearly 130 experts — including mechanical and electrical engineers, welding and materials engineers, and chemists — to do that testing internally.
The lab is a “bill out” organization, explained Wen Tu, who leads integrated grid planning for the utility. In other words, every hour of work conducted in the lab’s various testing environments is billed directly to the department within PG&E that requested the work.
At current rates, the system costs the utility $138 per hour of work, Tu said. That’s a significantly lower rate than what the utility would pay for external consultants, he clarified, which can reach up to $500 an hour.
Because ATS isn’t a separate profit center, Tu said it remains very “nimble” to the changing needs of the utility.
At present, of course, the lab is very focused on the grid side of the business “because there’s some unique challenges from electrification and wildfires,” he explained. But that hasn’t always been the case. In 2010, for example, after a natural gas pipeline exploded in a residential neighborhood in San Bruno, ATS “tripled the work that we did for gas, based on the needs of the day,” Tu said.
Innovation catch-22
That nimbleness is rare in the utility world. And even within PG&E’s unique setup, despite best efforts to streamline, the rigorous testing and iteration process to get new tech ready for the grid remains an arduous one.
Part of the issue is that the startups ATS works with aren’t always structured with utilities as their primary market. As Tu explained, venture capital is hesitant to back startups primarily targeting utilities, in part due to the sector’s famously slow pace of adoption.
At the same time, utilities like PG&E are massive customers, especially for smaller companies. That means that in addition to offering a promising technology, vendors must also be able to manufacture millions of units
For example, inside ATS’ mechanical lab, engineers are testing out new technologies to streamline asset inspection and repair. When Latitude Media visited the campus late last year, engineers were testing a device for assessing the integrity of wooden utility poles — a critical task for a company with more than 2 million poles to maintain. The instrument uses smart sensors and a microbit to measure the resistance of the wood, to detect and measure cracks, holes, and decay. A software element then calculates whether poles need replacement, showing fieldworkers a graph of the pole’s resistance profile.
The benefit of implementing the tech would be multifold, allowing fieldworkers to more quickly and accurately measure poles, and also to avoid the much larger holes that traditional resistance testing creates. And in getting involved early with companies developing products for the field, PG&E can have a more productive feedback loop, weighing in on their specific needs for a tech solution. (One such partner is smart panel maker Span, which is making an intentional pivot toward utilities. Span has been on site at ATS testing out their hardware capabilities, Tu said, which has accelerated the feedback process.)
But in general, the “cycle of feedback” needs to happen a lot faster, he said.
“What we’re seeing is that some of our partnerships that we have here, I mean they’re great, but they’re also dragging their feet in terms of getting the hardware to us for testing,” he explained. That frustration is mostly on the hardware side of partnerships, he added, where the time between generations of a product tends to be longer.
What we’re seeing is that some of our partnerships that we have here…[are] dragging their feet in terms of getting the hardware to us for testing.
For instance, the smart drills for testing poles aren’t fully deployed yet, because extensive testing revealed several challenges, like the fact that it struggles in the harder wood of transmission poles, and also misreads smaller poles with internal rot. PG&E is still waiting on the vendor to make those modifications.
And that slowdown is not unique to asset inspection, Tu said.
“Even if the technology is great, there’s got to be a pathway to optimization at the utility scale that [most startups] don’t think about,” he said. That filters out a lot of potential partnerships at the start, he added. “A lot of these smaller companies we just can’t partner with…and so we’re waiting for them to be bought out by a bigger organization.”
The wildfire case study
Inside the dome at the center of the ATS campus, engineers are still working to perfect wildfire mitigation tech. They drag in sod at various levels of dryness, and various types of vegetation, mostly tree branches, from around the PG&E territory.
They’re essentially testing to find the sweet spot of enhanced safety settings. If they’re tripped too fast, small load increases could trip the lines, including when it’s not necessary — resulting in unnecessary power outages and customer discontent. If it takes too long for power to be shut off, though, they could start fires.
When PG&E first started testing the Enhanced Powerline Safety Settings, or EPSS, they were tripping “all the time,” Tu said. That’s a huge problem for the customers impacted, because those lines can’t just be turned on right away — the utility has to send out a patrol to view the damage, which can take hours.
“The testing we did here really reduced those false positives and made sure that the settings…are dialed in,” he explained. Today, both EPSS and downed conductor detection are deployed relatively widely in PG&E’s territory, in all high wildfire risk areas, Tu said. Moving forward, the ATS team is focused on secondary service lines, the ones that come into homes. Those lines, as of yet, don’t have safety mechanisms to prevent fires if they’re damaged. ATS is testing out new types of cables, with better protection.
When it comes to wildfire tech in particular, a lot of the products ATS tests have come to PG&E through international collaborations, including from Europe and Australia. But domestically, Tu said there’s been a lot of interest from other regulated utilities, both in the tech that ATS has validated, and in the lab’s overall approach to innovation.
“We’ve had lots of other utilities come visit us through industry organizations or just direct benchmarking,” he explained. Some of those conversations have focused specifically on the ATS business model, and how it could be replicated to fit the needs of other utilities.
“Sometimes innovation doesn’t require new technology,” he added. “Sometimes innovation is just thinking about doing things a different way. Sometimes it’s about looking at our old processes and asking…can we do something differently to get a better outcome?”
