The Value of Safety

The Value of Safety

Today I had a conversation with some very well-meaning battery industry professionals.? We got onto how to justify the value of safety.? The conversation immediately turned to how we could use one safety innovation to potentially eliminate the use of a more expensive safety innovation, thus reducing the cost of the entire system.? Or perhaps we could eliminate two, or more safety systems, and get big cost savings.?

It’s as if, now that we have an airbag, we don’t need anti-lock brakes.? Let’s get rid of them, because when you hit something, the airbag will catch your face, and everything will be okay.? Right?

Not a Toggle Switch

This kind of discussion comes from thinking of safety as a toggle switch.? We make our products safe. Safety Toggle Switch on.? Can we keep it on and use less expensive safety measures?? Okay, then let’s save some money, because that’s what the boss wants in her spreadsheet, right?? As long as we keep the Safety Toggle Switch on, we can make whatever changes we want.

But hazards don’t work that way.

A Personal Story

I grew up in Hendersonville, North Carolina, a hamlet just south of Asheville.? I joke that, when I grew up there, it had a population of 5,000, and if you have a few minutes, I will name them for you.? Everyone knew my name, the owner of the bookstore knew what I liked to read, the owner of the hobby shop knew which things I had bought and what I might be interested in, and I never locked my bike because if someone rode off on it, everyone would tell him to give Brian his bike back.?

I lived there until I was 16, then wandered a bit for college and graduate school, and now live in Greenville, about 50 miles south of Hendersonville.? In all this time, since 1974, there has never been a hurricane to reach Hendersonville and do anything more than a little wind and rain.? Until a few weeks ago, when we made the acquaintance of Helene.

What is the hazard?

In an earlier issue of this newsletter, I wrote about how the magnitude of a hazard could be calculated.? It is proportional to the probability of the hazard times the magnitude of the damage it can cause.? While the probability of a Category 5 hurricane hitting my sleepy mountain town in North Carolina is very low, due to the steepness of the valleys and the channeling of the water into those narrow valleys, when it does hit the hazard is very, very high.

Batteries are different. The probability and the size of the potential damage are related.? Here’s how:

  • Probability: As the battery gets bigger, the likelihood of having a manufacturing defect, or a weak cell, or a part of a cell that ages differently from the rest, is higher, and that probability roughly scales with the size of the battery.? Thus, as the battery gets bigger, the probability of a fire also gets bigger, all other things being held constant.
  • Size of Damage: More straightforwardly, the size of the fire will also go roughly like the size of the battery, due to the fuel provided by the cathode, anode, electrolyte, aluminum current collector, and packaging material.? (The only thing that doesn’t burn is the copper current collector and perhaps the nickel tabs.)?

So, the hazard, which goes like the probability of the event times the damage the event can inflict, is proportional to the size of the battery squared.? An EV, which has a battery about 1,000 times the size of the battery in your phone, will have a hazard associated with it that is 1,000,000 times as high.? And an energy storage facility, with 40 times more energy than that (about 4 MWh), is 1,600 times more hazardous than an EV, and 1.6 billion times more hazardous than your phone.?

To put that into perspective, the hazard of a single shipping container of batteries used for ESS is more hazardous than all the iPhones in use today.? Think about it, it is 400,000 times more likely to catch fire than a single phone.? But when a phone catches fire, it is not (virtually) guaranteed to also ignite 399,999 other phones, which is what happens when a battery catches fire in an ESS container.?

Plan for the Worst

For a large battery, the worst-case scenario must also be considered and mitigated as much as possible.?

  • Phone: They use a hard case, give advanced warning when it’s getting hot, have shut down circuitry, vents and other protection that can prevent the fire.
  • Power Tools: The batteries are encased in thick plastic cases, and they use batteries with relatively low energy densities that will give off less heat and flames.
  • Electric Vehicle: There are all sorts of protections, cooling systems, warnings, and other features that make the fire less likely.? But they still happen.? More on that in a moment.
  • E-bikes: They do almost nothing.? Shame on them.

For phones, power tools, and e-bikes, the sizes of the fires are relatively small, so even though they are fairly common—read about them every couple weeks in this blog—the damage for each fire is relatively modest—if loss of at most a few lives per event is “modest.”

The problem is, when the size of the battery is large, the worst-case scenario is much worse.?

  • An ESS facility in Otay Mesa CA burned for 14 days, requiring round-the-clock firefighting and evacuations.
  • An EV that burned in a South Korean apartment building parking deck damaged almost 900 other cars and caused many residents to be homeless while the apartment is being rebuilt.? South Korea is enacting laws to prevent this from happening again.
  • A storage facility in Chicago caused an evacuation of over a square mile, with unknown broad health effects to those who inhaled the smoke, hydrofluoric acid and other chemicals that come out when a battery of that size burns.
  • A transport ship carrying EVs sank after some EVs caught fire, destroying $400 million in cargo and an estimated $25 million in the ship’s value, not to mention the spilled fuel, which is also significant.

How Do You Plan for the Worst?

It comes down to two things:

  • Reduce the probability of an event: Do everything you reasonably can to reduce the probability of an event occurring.
  • Mitigate the damage if an event occurs: Again, do everything you reasonably can to mitigate the damage that occurs when an event happens.

I’ll dig into both in later posts, sharing some of the specific things that are done, and can be done, for both of those strategies.?

A Personal Ending

The evening after Helene passed, I drove up to Hendersonville to see if I could get to my parents’ home, to make sure they were okay.? Cellphone coverage was out, so I didn’t have a map or traffic to guide me, and I was working with a 40-year-old mental map of the town.? I kept running into downed trees and flooded roads, getting completely lost and disoriented.? But it seemed everyone who owned a chainsaw was out cutting trees to clear the roads, and at every flooded road there was someone to help guide me back to a main road and safety.

It was the same small town I grew up in, 40 years later, with the same people, work ethic, and love for each other.

My parents were fine.? They left the day before with their 30-foot camper and drove a few hundred miles east, out of harm’s way.?

?

Ryan Fogelman

Partner @ Fire Rover: Changing The Way The World Fights Its Fires; COfounder/Partner @ COhatch: Strengthening Communities & Improving Lives;

4 周

I??agree. Mitigation of risk needs to be cumulative. The swiss cheese ??analogy is best. How many slices of protection do you have that if a hole perfectly aligns, you get the worse case scenario. Seemingly the more layers you cover the better unless it can be proven that one makes the other obsolete.

回复
Matthew Paiss

Technical Advisor, Battery Materials & Systems at Pacific Northwest National Laboratory - PNNL - Comments and opinions are my own -

1 个月

Brian, I think you are mixing up risk with hazard. Increasing components will increase the risk of a failure, but the hazard is very device and failure mode dependent. Component quality, workmanship, and abuse are far more predictable metrics for reliability than simply # of cells. A 15Ah E-bike battery with less than 40 cells far more likely to fail than a 5MWh BESS with 30,000 cells. As Jim says below, BESS installed indoors has led to much greater equipment losses simply due to the presence of water sprinklers (meant to save the building). Indoor installations are not easy to protect.

Jim McDowall

Battery energy storage consultant

1 个月

Brian, I agree with your numbers on probability regarding the cell capacity (or capacity of parallel-connected cells), but I don't think the same scaling works when you factor in series connections. A cell phone has a single cell, while an EV or BESS has many series-connected cells with thermal barriers between them to prevent propagation. If (big if) the risk of propagation can be eliminated, then the scaling stops at cell level. The Otay Mesa incident is not representative of today's BESSs. Stuffing hundreds of megawatt-hours of NMC batteries into a building is just not done anymore. The latest generation of BESS is entirely modular, with the intent that a fire would be limited to a single unit. Those units have been getting larger, now around 5MWh in a 20-ft container, so your Size of Damage number has been increasing, rather than the probability. Many of the recent BESS safety events have resulted from external failures like stormwater intrusion or coolant leaks, causing multi-module arcing. Such failures result in complete loss of an enclosure. The probability in these cases is not related to the cell size, because the battery is not the cause of the event. Happy to hear your parents made it out OK!

Paula Furlong

Research Supervisor at Diversified Product Inspections

1 个月

So glad your parents made their way to safety!

回复

要查看或添加评论,请登录

社区洞察

其他会员也浏览了