When a culture of risk aversion creates risk of harm

In a back meeting room, on a modest-looking floor filled with tall cubicles, PG&E opened my pitch with a long safety briefing on emergency procedures for an active shooter, incident threats, emergency contact numbers, and the evacuation plan. It was a cue that I was dealing with a unique organizational culture and this wasn’t going to be an ordinary pitch. The following is my December 2018 experience in presenting PG&E a technology concept for electrical arc prevention to mitigate the risk of wildfires.

Background

My call to action occurred when Camp Fire hit last November. The visuals of the mass devastation from the now truly unfortunate annual fall tradition of utility-triggered major wildfires made me want to be part of the solution. After consolidating some prior ideas, I wrote the senior leadership and technical teams at PG&E, California Public Utilities Commission (CPUC), Cal Fire, California Governor and Lieutenant Governor. PG&E, CPUC, and Cal Fire responded with multiple groups from PG&E reaching out.

First Impressions

During my pitch, I was thanked by one of the management representatives for reaching out as he indicated it was uncommon for companies to approach PG&E due to their risk aversion. It was an interesting comment that I now better understand. After multiple conversations with different groups, I found it very interesting that each time PG&E would send the proposal down to the technical teams for decision making, instead of to a central planning figure. Reflecting back, I see a great strength in gathering these different points of view, but a possible weakness in seeking consensus decision-making at the expense of vision.

The feedback I received after the pitch was that PG&E only collaborates with supply partners with a technology at TRL (Technology Readiness Level, NASA) 7 or 8, i.e. the technology needs to have been validated as part of a rigorous test at scale, which creates a chicken and egg dilemma.

The feedback I received after the pitch was that PG&E only collaborates with supply partners with a technology at TRL (Technology Readiness Level, NASA) 7 or 8, i.e. the technology needs to have been validated as part of a rigorous test at scale, which creates a chicken and egg dilemma as a development requirement would be to have access to a high voltage research lab.

PG&E explained the TRL requirement was in place to avoid investing in technologies that could create a false sense of security in the grid. Sound rationale, but also incomplete in that it omits the opportunity of playing off small strategic investments in various idea teams to create a faster, more redundant total solution, i.e. the concept of using hedging in uncertain environments. PG&E proposed engaging through an internal investment program known as EPIC that was off-cycle at the time and preferred single source arrangements with validated solutions, such as an existing relationship with Exacter in this same space.

PG&E explained the TRL requirement was in place to avoid investing in technologies that could create a false sense of security in the grid. Sound rationale, but also incomplete in that it omits the opportunity of playing off small strategic investments in various idea teams to create a faster, more redundant total solution.

The understanding I developed from pitching PG&E is that they are quite strong operationally, but a laggard in adapting technology that doesn’t arrive in plug-and-play form. The model of the Public Safety Power Shutoff (PSPS) in my opinion is a progressive concept, but it’s implementation success thus far has been dismal.

Early PG&E self-reported incidents suggest that despite the measure, the utility may have already triggered five fires this season, including the recent Kincade fire, while the measures themselves have placed the community’s vulnerable at risk and created mass resource uncertainty and inconvenience throughout the state.

I believe backlash would be acclaim if these shutoffs were administered by a semi-automated system based on real-time deterministic data on the health of the transmission line, instead of probabilistic data by meteorologists and ground observations.

One of the problems in this first conception is that PG&E is trying to administer it with manual operations, their strength, as opposed to technology, their weakness.

I believe backlash would be acclaim if these shutoffs were administered by a semi-automated system based on real-time deterministic data on the health of the transmission line, instead of probabilistic data by meteorologists and ground observations. Currently, it’s an unreliable method managing an unreliable system, so the results are likely to continue.

PG&E as a Function of Market Forces

In addition to the values of its leadership, companies are a function of their sector economics and operating environments. The public works, utility sector incentivizes monopoly and scale, which explains PG&E’s operational strength and lack of openness to new ideas.

Today, this makes PG&E and the electrical utility sector ripe for disruption in California, since a decades-stable operating environment flipped to be incredibly volatile to the threat of wildfires and that same scale advantage, is now a strategic barrier to delivering expedient, meaningful, local solutions to the wildfire problem.

It also makes sense why the state government appears hesitant to step in and drive safety outcomes as the liability for intervening is so high it’s likely more strategic to work with PG&E, so the tough talk continues for now. That said, I can understand why San Francisco has offered to buy PG&E’s local infrastructure: the segment is high-density, urban, accessible, and therefore low-risk. The question is who wants to be responsible for the remote, high-risk infrastructure with the current economics?

Today, this makes PG&E and the electrical utility sector ripe for disruption in California, since a decades-stable operating environment flipped to be incredibly volatile to the threat of wildfires and that same scale advantage, is now a strategic barrier to delivering expedient, meaningful, local solutions to the wildfire problem.

Market Viability

Despite the $10.5B single incident liability exposure of PG&E’s Camp Fire, the venture community informed me the concentrated number of players in the electrical utility space with the problem would create a difficult path to scale and would offer less favorable growth terms as other investments.

Raising capital is possible, but it required tangent applications to make it interesting to the community, which diluted the potency of the project.

For my solution the tangent space was 5G and continuous measurement, monitoring, and cleaning of electromagnetic interference (EMI) as higher speed, more EMI sensitive, frequency spectrum is assigned for consumer use.

To scale without friction and for the public good, it made sense to finance growth through a shared-benefit model of insurance premium reduction from underwriters such Cal Phoenix Re Ltd and Associated Electric & Gas Insurance Services Ltd who recently underwrote PG&E. Although today, I’m doubtful this model remains available from what would be an obvious reluctance to underwrite power transmission in California, there may be a time after system hardening where the operational risk would re-converge to acceptable.

For those interested in the idea:

First Principles

1. High voltage discharges (arcing) emit broad-band RF noise.
2. Four primary arc-based failure modes have similar frequency response (<10MHz) [1].
3. These electric fields can be detected by the predicate corona: the iterative process of the high voltage source and ground finding a connection.
4. The corona can be located by algorithm applied to neighboring sensors.
5. The voltage signal can be used as confirmation [1].

Product Features

1. Ultra-low-cost antenna arrays placed on transmission towers at defined spacing in a mesh-style network.
2. Powered by transmission line harvesting (notionally 12V, 1A) with onboard battery.
3. Operation in standard VHF frequency band (30MHz+).

Key Challenges

1. Measurement of edge cases and ensuring unrelated background events (e.g. lightning strikes) don’t trigger unnecessary blackout events, i.e. resilient to false positives.
2. Robust high speed communication for relay or substation interrupt.

A Year Later, Better Solutions

I see PG&E’s current operational strategy as a positive development, but at immense cost and anxiety to the local community. It is likely in the coming years when deterministic solutions are deployed it will be too late.

My sense is emergency investment may come shortly by way of a response from the Army Corp of Engineers or National Guard to harden the infrastructure, such as vegetation removal, burying cables, and deploying a similar established technology called synchrophasors. I also think the tertiary market is going to move faster than PG&E and government in creating a decentralized, distributed grid that will require a central authority to coordinate routing, which may be the next embodiment of the service electrical utilities provide in California.

Hard-Won Advice

My key takeaway from this experience is now the foundational value in my product team’s culture in that we lead every new conservation with product, not ideas. Simply, people acquire solutions to solve their problems, not ideas. No matter how big or small you are, solutions are the great leveler for new entrants to a market.

People acquire solutions to solve their problems, not ideas.

What do you Think?

How can we as a tech community do our part? I’d love to connect and talk more. [email protected]

Thanks to Chris Wu, Louis Virey, Stew Moffatt, & Miriam Christoph for their help with this post!

Technical References

[1] F. Zoko Ble, “Novel Methods for Arcing Fault Detection and Location in Power Distribution Systems,” Ph.D dissertation, Aalto University, Espoo, Finland, 2017.