A few days ago, around Perugia, Italy, a compressed natural gas (CNG) bus caught fire. Fortunately, no major injuries seem to have occurred, so I won’t feel too bad about using this as an opportunity to learn about CNG bus safety devices, and, specifically, why a burning bus that is shooting massive columns of fire in multiple directions is actually what you want to see in such a situation. I know it’s hard to look at that thing and think, oh good, everything’s doing what it should be doing, but really, that’s what’s going on here.
I mean, technically, all that fire means that some things didn’t go like they should go at all, in a pretty big way, but those jets of flame are the result of safety systems performing as designed.
First, though, let’s marvel at the sheer raw, flaming, Death Metal-style violence of this:
According to sources
🔥 In #Perugia, #Italy, a methane bus caught fire on a provincial road.
Luckily, as it turned out, there was no one on the bus apart from the driver, which helped avoid a tragedy.✍️AJ#Europe pic.twitter.com/RGRxkJY121
— New_Vision_News (@N_Vision_N) April 20, 2022
Holy shit, right? That’s legitimately terrifying. But let’s dig a bit into just what is going on here. First, let’s look at how these types of buses are laid out:
The Italian bus shooting those flames is similar in design to this: rear-engined, with its pressurized tanks of natural gas mounted on the roof. In this diagram of an articulated bus, you can see the multiple tanks called out, along with an enlargement of something called a PRD, which stands for Pressure Relief Device:
(image: Researchgate)
You can see that PRD mentioned there, and it’s nice of them to call it out so big, because those are really the stars of that video up there. The whole reason there are multiple jets of flame shooting out of the roof of that bus and not, say, just one massive, shrapnel-and-bus-part-flinging explosion is because those PRDs are doing just what they were designed to do: relieve pressure.
The way these type of bus PRDs are triggered is actually via heat. Inside the valve is a plug of metal called the “fusible material” that is designed to melt at a specific temperature; once that temperature is reached, the metal of the fusible material melts, and opens the valve, so the gas inside the tank can then be vented safely. In these diagrams, the fusible material is shown in yellow: 
Here, you may as well watch the whole video where these diagrams came from, why not:
In a more ideal situation, the gas will just be vented to the atmosphere. But, since the valves were triggered by excess heat, it’s possible that there will be ample ways for that escaping gas to combust, which is how you end up with the flame-thrower situation seen above, or here, in a very similar event recorded by the Dutch Safety Board:
As you can imagine, massive jets of flame are not great to have around, particularly if you don’t want, you know, everything on fire. In tight urban streets or near heavy plant growth, this could be very bad; however, even these jets of flame are so much better than the tanks exploding.
A study with the evocative title CNG buses fire safety: learnings from recent accidents in France and Germany describes the situation very clearly:
The most unwanted event in case of CNG bus fire is the burst of one or more of the compressed storage tanks located on the roof of the vehicle. Tank burst is definitely not a tolerable option having in mind the tremendous amount of mechanical and chemical energy released in the course of this event.
The current safety strategy to prevent tank burst consists in fitting pressurized tanks with devices that release stored compressed natural gas as they fuse under the effect of temperature rise (fire). The melting temperature of these fuses is about 110°C. In practical terms, to prevent tank burst, internal tank pressure has to decay before the fire degrades the mechanical strength of the compressed storage. Experience shows that unprotected tank (inhibited pressure relief devices) can not survive a standard bonfire test for more than few minutes [4] & [5]. The main cause for a tank to burst is the decay of its mechanical strength and rise in internal pressure.
Therefore, pressure relief devices (PRD) should be capable of de-pressurizing a tank within a couple of minutes. According to experience, bus tanks can be exposed to fire for about 20 to 30 minutes which is an average time frame for a bus to be burnt out.
And, in case you’re not convinced that PRDs are the way to go here, the study gives some details on the force of exploding tanks:
Large quantity of mechanical and chemical energy are stored in compressed combustible gas storage. Sudden release of this energy in case of tank burst may cause some severe damage to the bus environment. When a tank bursts, observation shows [4] two consecutive pressure wave propagating in the surrounding environment.
The first one which is also the more severe is associated with the pneumatic rupture (gas expansion) whereas the second is caused by the combustion of the released combustible gas into the air (fire ball). It is therefore to be noticed that although the chemical energy stored is usually an order of magnitude larger than the mechanical energy, the sudden release of the mechanical energy induces greater overpressure effects.
Theoretically, the pneumatic burst of a 130 L tank at a pressure of 200 bar releases an energy equivalent to the detonation of about 1.85 kg of TNT (8.7 MJ). Windows can be broken within a 30 meters radius (50 mbar) and pressure wave induced lethality is to be foreseen within a radius of 12 meters (140 mbar). These calculations can worsen due to pressure wave reflection and pressure build up as well as to directional energy release (axial direction) due to the rupture mode of the cylindrical tank. Moreover, projectiles can also cause severe damages within a radius much larger than the one estimated above for overpressure effects. [5] shows that fragments of up to 14 kg (type IV tank filled with hydrogen at 350 bar, test conducted in open atmosphere / projectiles not hindered by bus equipment) have travelled a distance of 82 m from tank fire location. The mechanical energy released as the tank ruptured was equivalent to about 1.35 kg of TNT (6.3 MJ)
So, those tanks can explode with force equal to about 1.85 kg of TNT. For reference, that makes it more powerful than one kilogram of C4 explosives.
The point here is that while that video looks incredibly alarming, it’s a great example of safety systems at work. It’s also a reminder that in extreme cases, safety can mean the lesser of two really bad outcomes, in this case the choice of tree trunk-sized cylinders of flame as opposed to a massive explosion.
See? Everything is fine!
“In a more ideal situation, the gas will just be vented to the atmosphere.” No. It should be ignited and flared. A cloud of natgas vapor can ignite and that will be a kind of thermobaric explosion. Thermobaric bombs are of some of the most devastating weapons of war. In my engineering career the two most feaered refinery/chemical plant disasters were gas cloud explosions and BLEVE’s (boiling liquid expanding vapor explosion).
Good point!
I would like to purchase a retrofit kit for my SUV, for those times when other drivers are irritating me.
While the venting of the PRDs is good, and certainly releasing the energy slowly is good, venting horizontally is NOT good, and no longer the recommended approach. It used to be done to prevent water getting into the PRD outlet lines. Frankly, that didn’t work very well anyway. Now many jurisdictions require venting upwards, and many standards recommend that as well.
There are good ways to keep water out, and there are also other PRD technologies out there.
I’d assume a small plastic cap would do a decent enough job of keeping water out without any risk of accidentally containing the pressure.
Kind of a similar idea to AA batteries. Check the Slow-Mo Guys trying to blow up some batteries, and instead it all vented (relatively) safely out the top.
Are we sure this wasn’t Robert Maddox doing something crazy?
https://www.youtube.com/channel/UC8ui_J9ROP81qvp1my83KiA
I would rather have a fire/flamethrower than kaboom tank.
And in double speak “Tank burst is definitely not a tolerable option” = Explosion bad!
Granted from a yee-haw level, that would be awesome. From a blast radius point of view, no so much.
Having seen the aftermath of a gas explosion set off by a parolee inside a brick home, I can say wholeheartedly that the torch is vastly superior to the Hamster Huey option.
Based on the comments of this happening in a tight downtown area or something similar, I do wonder if it would be better to have it vent straight up? Where did that decision come from?
There are a couple interesting parallels to the worlds of rocketry and steam engines here.
In the latter case, most boilers also have a a plug of fusable material (traditionally lead). In normal operation, this plug is below the surface of water in the boiler and so kept cool. Running a boiler dry will lead to a catastrophic failure as, without water to distribute heat, hot spots will develop near the fire. The plug melts before the walls around it fail, letting the steam escape through a dedicated path, rather than by a metal-flinging explosion.
Rockets are equipped with a Flight Termination System (FTS). Most frequently this takes the form of a strip of explosives along the length of the outer casing. If, while it is still low in the atmosphere, the rocket deviates from its expected course to too great an extent, this explosive will be set off by either the Range Safety Officer or (more and more frequently) an automated system on the rocket itself. The idea is that, if something goes wrong, it is much better to have all of that fuel exploding high in the air at the edge of the safety zone and causing mostly lighter pieces of shrapnel to come down over a slightly broader area, rather than a single giant fuel air bomb to come down potentially closer to populated areas. It creates quite the fireworks show (see the launch of the Firefly Alpha last year), but is better than a similar spectacle on the ground.
(Unless you’re China. Then you just drop poison filled stages on villages and shoot anyone who complains.)
Username checks out. 😉
Depending on which way the wind is blowing you might not want to be running out of that bus too quickly if said safety feature/multi-directional flame thrower is activated…
Pneumatic testing of piping or tanks can be highly dangerous. If you want to dig into the math behind the equivalent energy calculations for a gas-pressurized vessel, check out ASME PCC-2 (Part 5 specifically) or play around here: https://www.piping-world.com/safe-distance-and-stored-energy-calculator-pneumatic-test. I’m somewhat surprised the linked study didn’t provide a reference for their stored energy calculations.
Fun fact, the 20-gallon/175psi Kobalt air compressor Lowe’s sells is worth ~0.03 kg / 0.066 pounds of TNT full pressurized.
Fortunately, those tanks are small and the relief valves appear to be pretty large in comparison. Fires surrounding larger tanks, such as those used to store propane, can cause detonations before all the material can combust or escape through the relief valve. Something called a Boiling Liquid Expanding Vapor Explosion (BLEVE) can then happen as the fire reduces the strength of the tanks walls, ending in a devastating explosion as all tank contents are released at once. A couple of good videos (one pretty old) up on YouTube explaining these.
Reminds me of a T80 cookoff
There was a number of people on twitter screaming that this incident is proof that EV’s are dangerous. I didn’t try to change their mind.
>> whereas the second is caused by the combustion of the released combustible gas into the air (fire ball).
Isn’t this a fuel-air bomb (AKA thermobaric)?
While the parts are roughly the same, not really. This is a sustained controlled release, in a specific direction. Thermobaric weapons are a momentary controlled release in all directions.
Thermobaric weapons create pressure waves as the fuel/air mix expands, sometimes a violent enough pressure wave to induce a brief vacuum in the area afterwards. The bus flame thrower will create a small pressure wave, but its in one direction, and is going to be dwarfed by the damage done by the flame.
Thermobaric bombs are carefully tuned to produce fuel-air *detonation* rater than combustion, which is a key distinction. It’s why the tank bursting in the case of the bus is more destructive than the burning gas itself, even though the gas contains more total energy–the speed of the expanding energy front in a detonation is supersonic, while open-air combustion is strictly limited by the speed of sound.
If you listen closely, you can almost hear the Burning Bus say “tell Pharaoh to let my people go”
I love that there exists a scenario in which “turn the bus into a giant flamethrower” is considered a safety feature.
I thought this was going to be a PDX bus once I read meth and fire but was pleasantly surprised.
So a guy named Torch writes about big jets of flame?
And lights! That didn’t sink in for me until I realized my bicycle headlight is made by a company named Torch.
Didn’t take long for some smoothbrain on Facebook to post this vid claiming it was a battery fire.
“Once that temperature is reached, the metal of the fusible material melts, and opens the valve, so the gas inside the tank can then be vented safely.”
“Safely.”
(Seriously though, this is cool and I learned something.)
Safely is a flexible term.
“Controlled Manner”.
Safety, sure it’s safe if you’re nowhere near it.
Meanwhile, it’s 4/20 and I can’t find my lighter…
Diesel fuel filler caps have lead inserts in them for the same reason, though the results are not nearly as exciting when they work.
I’m surprised the PRDs aren’t all vented straight up. Directly above a bus is more likely to be clear of fire-sensitive material than immediately beside it.
I was also wondering why they’re venting to the side. In an urban environment, you might have some overhanging tree limbs to worry about above the bus, but you’re almost guaranteed to have something to hit on either side.
I’d wonder if it has something to do with the way the contents still inside the tank behave as the tank vents.
I think it’s for visual excitement.
This is the same basic system used on farm LP tanks, and the visual of them burning is similarly fantastic.
Rural fire departments train specifically on how to handle the situation, but there are limits. Roughly mid ’90s I recall the news covering a wildfire that they were unable to keep sufficiently away from a nurse tank (the mega tank they fill the delivery trucks from) and when its vents went off sending flame jets of enormous size they pretty much set off the tornado sirens and noped the fuck out of there. It didn’t blow, for the record.
“…noped the fuck out of there.” is a keeper. Thank you.
As a former kid with matches I say
“WOW, way cool!”
I was bad for playing with matches as a kid. As an adult I ended up running a match factory for 10 years. True story.
I’ve seen a propane delivery tank truck do the same thing.
Back in the 80’s in San Jose. It was on the shoulder of the 101 northbound. There was a fire in and around the transmission of the truck. The relief valve was open and a column of flame about 8 stories tall was in progress. Sounded like a rocket motor.
Add a nozzle and it pretty much is a rocket motor
Well as a degreed engineer, I have this to say:
Cool.. .way cool!