The use of lithium-ion battery cells in large energy storage applications is fairly new and the National Fire Protection Association (NFPA) wanted to know more about what would happen if one were to catch fire.

Tesla built its Powerpack with safety first in mind and was willing to put its battery system to the test. The company teamed up with NFPA last year and gave them two Powerpacks to set on fire. We got a hold of their test results.

Even though battery cells are unlikely to initiate a thermal runaway reaction and catch on fire, what can happen will happen, and with the number of cells Tesla is playing with, it needs to plan for it.

First off, a quick reminder about the Powerpack’s architecture. Tesla recently introduced the second generation Powerpack with new battery cells and some improvements to its pods and modules, but the overall architecture is similar to the first generation, which is the one that the company and NFPA tested.

A Powerpack is made of 16 individual “energy storage pods” – the same pods are used in the Powerwall, but with one pod per Powerwall instead of 16 per Powerpack. In the first generation Powerpack, each of those pods is made of approximately 900 battery cells separated into two modules, for a total of approximately 14,400 battery cells within the Powerpack – or 100 kWh of energy capacity.

Here’s a diagram of a pod and a picture of a Powerpack with 16 pods inside:

There’s a cooling system built into the door and an exhaust system linking each pod to a vent on top of the steel enclosure (that will come into play later).

For the first test, NFPA tried to simulate what would happen if one or several cells within a single pod go into thermal runaway. Will the pod explode? Will it propagate to other pods? Will the entire Powerpack catch on fire or explode? Let’s find out.

They installed a heater cartridge inside one of the modules of a pod placed in the middle of a Tesla Powerpack.

Here’s the setup at the beginning of the test:

They turned on the heater cartridge, which pushed the cooling system past its limit and while it could be detected that the cells/module was overheating, that wasn’t the goal of the test. They let it hit thermal runaway and the first “popping” sound of cells blowing up was heard after 12 minutes.

After just over half an hour, white smoke started to come out of the pack’s exhaust vent:

That’s gas created from the cells breaking down being evacuated.

The pops were still heard for about 15 more minutes, and once they had stopped the smoke began subsiding over 45 minutes.

At 1h30 after the beginning of the test, the smoke stopped coming out of the vent:

Nothing exploded. How boring.

They learned from the test that not only did the Powerpack not explode or catch on fire, but the fire created by the heater cartridge also didn’t propagate to other pods.

In fact, NFPA found that the 15 other pods were still functional:

“Following the test, it was determined that only one of the energy pods (the initiator pod) was damaged. The other 15 pods remained operational and had a full SOC. The energy pods were discharged and the Powerpack was recycled.”

That’s very good news. It means that a Powerpack cannot start a fire: even in the unlikely event that one or a few cells explode, it will be contained within the pod and will not unleash the entire 100 kWh of energy capacity of the Powerpack.

But what if the Powerpack doesn’t start the fire, but there’s a fire around a Powerpack? Will it explode and aggravate the situation? That’s the question they tried to answer with their second and more spectacular test.

They installed a propane burner to simulate a constant fire right on the side of the Powerpack.

Here’s the setup:

They turned on the burner.

After 20 minutes:

Nothing was happening for the first ~35 minutes of an intense fire burning on the side of the Powerpack, but then some white smoke started coming out of the vent.

Popping sounds started after 45 minutes.

After 1 hour:

Sustained flames started coming out of the Powerpack’s back panel.

After hearing pops for a good 15 minutes, they turned off the burner and let the thermal runaway of the Powerpack do the rest from here.

After 1h30:

The fire started intensifying at this point.

After 2 hours:

Steady pops are still heard as the 14,000+ cells inside the pack are still exploding.

After 2h30:

While the flames are very strong at this point, the pack itself is not exploding. The pods are encased inside a steel enclosure that prohibits any cell failure from projecting outside and in turn, the steel enclosure of the Powerpack is containing everything while the cells are individually exploding.

After 3 hours:

Fire starts subsiding and the last pop is heard.

After 3h44:

The last visible flame went out and the test was over.

Following the test, they unsurprisingly found out that “all of the energy pods were damaged and there was no stranded energy within the Powerpack.”

NFPA concluded that a prolonged fire outside the Powerpack could definitely induce the Powerpack into thermal runaway, but they found that the consequences were confined to the pack and didn’t propagate:

“However, no violent projectiles, explosions, or bursts (other than a overpressure release of the thermal door refrigerant) were observed during the test while the Powerpack was exposed to the burners, while it was in a free burn state, or after flames were no longer visible. Flames remained mostly confined to the Powerpack itself. Weaker flames emanated from the exhaust vent of the Powerpack, the front thermal door grill, and around the front thermal door seal at varying times throughout the test.”

Most importantly, they determined that the exterior temperatures at the Powerpack cabinet “would not pose a fire spread hazard” if Tesla’s installation standards are respected.

In conclusion, if a fire starts from inside a pod, it doesn’t propagate to the rest of the Powerpack. And if a fire starts outside the Powerpack, it won’t spread to other Powerpacks around it. Of course, there are also several safety features preventing those things from ever happening, but the NFPA’s tests were for worst case scenarios.

If you are interested in solar, we suggest you get quotes from more than one installer to make sure you get the best energy solution for your house or business. UnderstandSolar is a great free service to link you to top-rated solar installers in your region for personalized solar estimates for free.