Earlier this year the National Safety Transportation Board (NTSB) released a safety report documenting their findings of an investigation into four electric vehicle fires involving high-voltage, lithium-ion batteries.
Dr. Peter Lawrence, CDPT, Ed.D., sat down with Dr. Thomas Barth, Survival Factors Investigator and Biomechanics Engineer at the NTSB, to discuss the impacts of this report on pupil transportation.
Your investigation involved passenger cars, but clearly applies to all electric motor vehicles. Given the increasing numbers of all-electric school buses being added to fleets nationwide, please provide an overview of what your investigation found.
The research looked at passenger vehicle incidents, but the findings apply to any electric vehicle (EV) on the road.
The basis of the report used the existing emergency response guides that are available around the nation through the National Fire Protection Association. Those largely concern light passenger vehicles and various trucks. School buses are a little different. For school buses there are National Association of State Directors of Pupil Transportation Services (NASDPTS) emergency rescue procedures, for example, and then each state has its own emergency plan. That made school buses a bit beyond the scope of the report, but our findings regarding emergency responders apply to school bus EVs as well.
I would imagine now that electric school buses are being introduced, school districts will look at their emergency response plans and incorporate new EV considerations, and bus manufacturers themselves too will need to provide appropriate information.
The NTSB has a background with lithium-ion batteries. Back when high-voltage lithium-ion batteries became an emerging technology, we started to pay attention to them. A high-voltage lithium-ion battery, with its flammable electrolyte, poses fire risks that must be accounted for.
In August of 2017, a crash occurred where an EV passenger-car driver lost control of his vehicle. It went into a ditch, struck a concrete drainage culvert, which ruptured the battery, and then traveled across the street and crashed into a private home. The vehicle caught fire, creating challenges for the first responders.
The passengers in the vehicle escaped. When the firefighters arrived, they started fighting the fire, which had spread to the house. They were confused, however, because the flames would erupt again moments after being extinguished.
The vehicle was wedged into the garage and partly into the house, and they could not understand why the fire did not seem to go out. They realized that they had to tow the vehicle out of the house to understand what was happening. They did, and quickly realized that the lithium-ion batteries were reigniting.
It is called thermal runaway when the battery cells or modules heat up enough to cause the adjacent cell to also heat up and catch fire. It is basically a short circuit within the cells and causes a chain reaction of fires. The only way to stop it from happening is to cool the battery enough that the heat no longer transfers from one cell or module to the next.
It took the firefighters time to realize that they needed to apply water more than just to extinguish the flames, but also to cool down the entire battery and stop this reaction. It took them more time to realize that just pouring water on top of the vehicle was not working – they needed to lift the vehicle and spray underneath, directly on the
battery. By this time, nearly three hours had passed since the crash.
Then, after the vehicle was loaded onto a tow truck, the fire reignited again because movement of the damaged parts caused short circuits. After that fire was extinguished, the tow truck brought the EV to a storage lot with a fire truck escort. In total, almost six hours passed since the fire first broke out. That is a long summary of just one incident, but it prompted the NTSB to begin investigating these issues.
What agents are required to fight an EV fire?
The most important element is cooling down the battery, and water has been found to be pretty much the best thing. Sometimes responders will use a mixture of foam and water, since the foam will help to also put out the interior fire in the vehicle. But once the traditional fire is put out, water is most effective on the battery cells and modules.
Another event occurred in Mountain View, California, in March 2018. This crash is well known because it also involved an autonomous vehicle issue. The driver was on the highway using autonomous vehicle features and crashed the car into a median barrier. Unfortunately, he succumbed to his injures.
Firefighters got to the scene to put out the fire, and again encountered reignitions. There were issues with how to manage this damaged electric vehicle because responders were aware of thermal runaway and also an issue called stranded energy. That means you can put out the fire, but the vehicle battery still contains energy and high voltage. The crash closed a major highway in northern California for an extended time, with concerns of how to manage the risks of a fire.
They moved the vehicle to a tow yard, and then, three days later, the vehicle reignited in the middle of the night at the tow yard. We discovered that shifting components (from investigators working with the vehicle) caused a short circuit that was not immediately detectable but caused an ignition later that night.
There was another unique case that occurred in June 2018 in . This was not a crash but rather an internal failure of the lithium-ion battery. The driver was driving his vehicle down Santa Monica Boulevard, and drivers next to him noticed smoke emanating from under his vehicle. He parked the car on the side of the road, got out, and the vehicle slowly caught fire. Firefighters were confused about where to apply the water, because there was no obvious crash damage to identify as the source of the fire. That case illustrated that, even with no crash, a lithium-ion fire can still cause issues and risks to first responders.
Our report also looked at other fire incidents with different types of vehicles in other countries, reiterating the same concerns. In our report, we recognized that responders were having issues. We evaluated the emergency response guides from all the manufacturers linked to the National Fire Protection Association website. This included about 45 manufacturers of private passenger cars and some commercial trucks and buses. We identified two important safety issues.
First was that the emergency response guides are inadequate or minimize risks to first and secondary responders from lithium-ion battery fires. They do not provide sufficient information for some aspects. Second, we found gaps in the safety standards for high-severity crashes involving lithium-ion battery vehicles.
We made three basic sets of recommendations. First, we recommended that the National Highway Traffic Safety Administration (NHTSA) incorporate emergency response guides into the new car assessment program. This will provide an incentive for the vehicle manufacturers to improve their response guides. Also to NHTSA, we recommended that they continue their research on mitigating or de-energizing the stranded energy in electric vehicles. It is not an easy problem to solve, and so there must be more research.
Second, the broadest group of recommendations was to the electric vehicle manufacturers. This was to all the cars, trucks, and bus manufacturers in the United States (that make EVs). We recommended that they model their emergency response guides on ISO Standard 17840, which is a four-part rescue standard, but has one part focused on responding to and mitigating alternative fueled vehicle issues. We also recommended that they model their guides on SAE J2990, which is the SAE standard on hybrid and electric vehicles for first and secondary responders.
We also recommended that EV manufacturers provide vehicle-specific information for fighting fires, dealing with the stranded energy, and safe storage of
the vehicles.
The final group of recommendations we make in the report is to the responder associations. It basically asks them to inform their members of the risks and available guidance on this subject.
What can school districts and transportation directors start doing as they march on this path to electrification?
First, look at emergency response plans or emergency management plans – and ensure that there is a provision which provides guidance for electric buses. They will want to also see if their bus manufacturer has an appropriate emergency response guide for incidents.
In short: make sure your own district’s emergency guides are adequate for EVs, and make sure that the vehicles you are bringing into your fleet have the appropriate information provided by the manufacturer.
Some operators of EV vehicle fleets have done outreach to their local first responders, ensuring that everyone is aware and that any concerns are addressed. This would be a good idea for school districts operating EVs.
Is battery technology on a school bus different from the technology on a passenger EV?
There are different chemistries and configurations of lithium-ion batteries, but the batteries on buses are essentially the same as those used on other vehicles. Buses just have larger batteries due to bus power needs.
The differences come from battery configurations. Cars made by Tesla, for example, use small cylindrical cells. Thousands of very small cells are arranged in modules. This design is good at stopping thermal runaway and managing small defects, but also has many electrical connections making it complex – which can lead to more potential failures. It is a trade-off.
Other cars, such as those from General Motors and BMW, use a different design approach with fewer large cells. The less complex configuration with fewer connections makes quality control easier. But if these cells go into thermal runaway, it is going to spread faster and be a more aggressive fire event.
The industry is creating a variety of designs, and first responders have used different approaches. Some guidance suggests letting batteries burn themselves out. But in some circumstances, that is not possible. Our report tries to highlight those complexities and indicate that we need to implement very robust mitigation information and procedures.
The EV manufacturers have a big role, because a vehicle’s design affects the response. Responders also have a role in establishing the approach and tools needed. It needs to be a cooperative effort between vehicle manufacturers and first responders – with government playing a role as well.
What other ways can safety organizations help to mitigate risks?
That is an interesting idea. It would really depend on how the system is designed, and how the whole vehicle comes together as a concept. The battery configuration might already have robust measures to mitigate the propagation of heat during fire events. For example: in all but one of the examples we noted in the report, each occupant was able to evacuate the vehicle before a fire fully broke out. The vehicles performed well to protect occupants even after a severe crash.
A solution such as fire-resistant seats would first need a holistic understanding of the lithium-ion battery configuration. Then you would know if certain strategies will really provide a measurable benefit.
Containment is a major element of lithium-ion battery failures. The design dictates how quickly a fire could propagate through the structure of a bus. What is the packaging of the battery case? How is it designed to manage the venting or thermal runaway of the modules? Those are the critical questions to ask.
Should standard electric vehicles mandate any onboard fire suppression?
I think fire suppression strategies are potentially valuable, especially when you are talking about a bus. They might not work as well in passenger cars. Weight is an obstacle.
The problem with a lithium-ion battery fire is its resistance to foam-based suppressants. The flammable electrolyte has both the fuel and oxidizer co-located, so external oxygen is not required for fire propagation. The foam will be extremely helpful with extinguishing or suppressing the fire as it relates to materials on the bus, but you need to cool the battery down to put out the lithium-ion fire. Pure water is really important, and quite a bit of water is required.
In regards to smoke, is there any difference between a traditional fire and EV fire?
There have been many studies evaluating the toxicity and smoke from vehicle batteries. It is important to know the differences, especially if an EV caught fire in a tunnel or inside a parking garage. I do not have a simple or general answer. Most studies I have seen indicate there are some differences between the composition of an EV and a traditional vehicle fire’s smoke, but that the toxicity and quantity of toxic smoke is roughly similar between the two fire types.
Also, in general, the studies I have seen show that much of the toxic smoke is generated by the burning of synthetic materials in the vehicle rather than the battery. Although the battery certainly does create toxic and flammable components.
What about differences in emergency evacuations? We have procedures for evacuating diesel buses which catch fire. Does thermal runaway change how fast the fire spreads?
A lot depends on battery configuration, but EV fires are generally less volatile than fires from liquid fuels when it comes to evacuation time. When the liquid fuel burns, it all burns at once. Batteries do not ignite that way. That is a big advantage of EVs relating to fire risks. There is no real evidence to say that an EV is somehow harder to evacuate than a traditional vehicle.
We have had vehicles with internal combustion engines for a long time, and we have become adept at mitigating those fires. The message of this report is that we now must do our due diligence in responding to these much newer risks surrounding EVs.