A Thermos filled with coolant sounds like it could be part of a really cheesy murder plot. Some guy goes out on a mountain, takes a swig from the Thermos, dies of poisoning before he can get back to the ski lodge. Soap opera stuff. But what if the entity taking a swig from the coolant Thermos was a 2004 to 2008 Toyota Prius? No really, bear with me on this.
It’s no secret that starting an engine from cold isn’t great for emissions. While there are a litany of reasons why cold starts are bad for the environment (high losses from thick lubricating fluids, high losses to metals due to high deltaT between combustion reaction and surrounding metals, etc), but let’s focus on how fuel behaves when cold. Gasoline from the pump is generally (unless you’re doing something really wrong) a liquid, but before it’s injected into a combustion chamber, it’s atomized to ensure nice mixing with air and therefore efficient combustion. Unfortunately, cold fuel is hard to get to vaporize; it tends to condense on cold engine components much like water vapor in the air in your room will condense against the outside of your cold Coke can. The result of the condensed liquid fuel dribbling down your cylinder walls is a much less effective burn than if the gasoline had been properly-atomized. Think lean conditions, unstable flame fronts, that sort of stuff. To combat fuel condensation, engines tend to just add more fuel to the fuel/air mixture when the coolant, and thus the engine, is cold.
Coolant temperature enrichment seems to work. Yes, it can stink like the aftermath of a Taco Bell binge sometimes and it often requires secondary air injection to get a complete burn prior to reaching the catalytic converter, but hey. It’s cheap to tune and easy to implement. However, coolant temperature enrichment is not necessarily the best solution. Ideally, coolant temperature would never get properly frigid, and warm coolant would cycle through the engine upon cranking — this way you don’t have to enrich your fuel mixture. But how do you keep coolant warm when cars are switched off overnight? Enter the vacuum flask, more commonly known as the Thermos.
If you like the soup in your packed lunch to be piping hot or have ever enjoyed hot cocoa in the snow, you’ve likely used a vacuum flask. It’s a pretty simple bit of kit, essentially two concentric flasks joined together at the neck with a near-vacuum in between them. Conduction and convection don’t work very well in a vacuum, so anything in a vacuum flask typically retains temperature very well. It’s perfect for storing cocoa, chicken noodle soup or coolant. Maybe not all in the same vacuum flask, but still.
Toyota saw fit to equip North American examples of the 2004 to 2008 Toyota Prius with a coolant vacuum flask. It’s a common trope that overseas manufacturers save their best tech for their home market, so a North American exclusive is a cool thing to see. So how does the vacuum flask work? Well, hot coolant gets stored in the vacuum flask upon engine shutdown, thanks to the closing of a valve. Upon cold start, there’s a brief delay between when a Prius powers on and when its engine fires up. In those few precious seconds, an electric water pump shoots warm coolant from the “coolant storage tank” into the cylinder head. The vacuum flask can’t store much coolant, but Toyota said in a slide deck that it’s effective enough to raise intake port temperature to 104 °F (40 °C) upon startup, a temperature that aids fuel atomization and cuts cold start emissions.
[Editor’s Note: As I was once a cooling system engineer, I’ve got to hop in here. Prepare for the longest Editor’s Note in the history of journalism.
Let’s take a closer look at the cooling loops for the Prius. Out front there’s a single radiator with two distinct circuits (I can’t for the life of me remember what we used to call such a shared heat exchanger at Chrysler). The bottom circuit part of the heat exchanger cools the two motors in the Prius’s transaxle as well as the inverter, and receives water via an electric pump. The top section of the radiator is the engine radiator, and is fed water via a mechanical water pump on the car’s accessory drive. Let’s have a look at some slides I found on some random website:
Here’s a quick look at the transaxle cooling system that feeds the bottom of that rad. MG1/MG2 refers to the motors in the hybrid transaxle:
Obviously, we’re more interested in the engine cooling circuit. Here’s a high-level overview via YouTuber EyeOnAiman:
You can see there’s a mechanical water pump circulating coolant through the engine, out of a conventional thermostat, and through a throttle body and heater core. In orange is our storage tank, an accessory water pump, and a water valve.
Let’s focus on that orange section:
Let’s look at how this works. When the engine is running, and the storage operation is off, no coolant needs to go to the storage tank. And since there’s a mechanical water pump running off the front-end accessory drive, the electric water pump can shut off, too. So this means hot coolant comes out of the top of the engine (where it says “Engine Coolant Temp Sensor” in the image above), and then heads straight into a water valve that looks like this:
From there, coolant goes through an electric water pump (which again, is off) into the heater core and back into the engine, where it picks up more heat to deliver to your cabin for maximum comfort. Here’s a diagram showing that “engine running (w/o storage operation)” condition. You’ll see that coolant is bypassing the thermos:
I’m not entirely sure I understand the conditions under which this next condition would occur, but it seems like coolant storage can happen when the engine is running, per the diagram below. In this instance, the electric pump would remain off, since the mechanical pump is still circulating coolant, and the three-way valve would just open both outlets, sending coolant to both the heater core and storage tank:
Here, the document talks about why you’d have the engine on and still send coolant to the thermos. I don’t really get it, though it seems the “max.: 4 times” is done to preserve the life of the coolant valve:
Speaking of the valve, here are the signals that go to the valve to determine where to send flow. As you can see, coolant can go from the engine directly to the thermos, it can go to the thermos and heater core, or it can go just to the heater core:
Oddly, I don’t have diagrams of engine-off operation, but those diagrams above would be largely unchanged, except the electric water pump would be on. Coolant would come out of the top of the motor, it’d be circulated via that little electric pump, and it’d be diverted to the heater core and/or the storage tank.
Other fun stuff from this presentation I found online includes this explanation for why this storage tank exists in the first place — again, it comes down to the vaporized fuel mixture condensing on the cold metal::
And there’s also a good explanation from the publication Toyotatech:
As part of its PZEV emissions rating, the system includes a thermos (called the “Coolant Heat Storage” or CHS tank) that can store coolant at 180 degrees F for up to 3 days.
When readied on, the driver may hear a faint whine while a third electric pump (at the thermos) runs for several seconds (prior to engine crank) to push hot coolant to the block, heating it to reduce fuel condensation, therefore lowering hydrocarbon emissions during startup.
Likewise the whine may be noticed when the car is turned off, pumping hot coolant from the engine into the thermos for storage.
To manage both exceptional cases — cabin heating and engine warm up — the Gen 2 has yet another exceptional component: the so-called “water valve” (a.k.a. “three way valve”). This valve determines when hot coolant is circulated to the heater core, the thermos, or both. The three way valve is the gatekeeper to make sure heat isn’t wasted, but its existence in the system adds yet another layer of complexity regarding diagnosis and bleeding.
Anyway, back to Thomas. -DT]
Are there any drawbacks to the vacuum flask system? Of course, nothing is perfect. The first issue is a matter of cost. The coolant vacuum flask on a 2004 to 2008 Prius retails for around $1,700, with a wholesale value closer to $1,200. Even if there could hypothetically be a 100 percent profit margin on the wholesale price of each coolant vacuum flask, that’s still one seriously expensive part. It’s also a fairly hefty part to package. We’re talking a cylinder roughly the size of a rotisserie chicken. As such, Toyota tucked it under the front bumper cover just in front of the left front wheel. If you’re thinking that this placement sounds a bit vulnerable in a crash, you’d be right.
In the end, Toyota abandoned the vacuum flask tech for the third-generation Prius. Still, the drawbacks of vacuum flask tech don’t hugely detract from its impressive potential. Imagine a cost-no-object scenario, like on a BMW 7-Series or Mercedes-Benz S-Class or any bewilderingly complex six-figure, 155 mph land barge for plutocrats that you’d never want to service when out of warranty. One of these tire-frying gin palaces could have two vacuum flasks, one for emissions and one for piping-hot heat as soon as the owner pushes the starter button before an electric resistive heater has any chance to get going. Doesn’t that sound marvelous?
Lead photo credit: Toyota
And that explains the weird noises emanating from my ’06 Prius that sounded like wheezing and wailing.
I wonder if they compared the gains from a warmer start vs the fuel used to drag that weight around. It seems that most engineering solutions for fractional increases in fuel economy add excess weight.
This is really cool, I wonder how it does in those places like Siberia and parts of Alaska where you cannot turn the engine off because you won’t be able to start the engine again.
If modern ICE cars are going to be complicated and less durable you might as well give them neat features like this one.
Well I guess there is one perk of living in the desert, 6 months out of the year the car is baking near 100F no Insulation needed. Sometimes I can start it and the coolant already registers 1-2 ticks.
Even the winter isn’t that bad as the unheated garage sits around 65F
I’ve been wondering what that sound is! We got a gen 2 prius for my partner and it always makes this weird whirring noise on engine start/stop!!
Two quick things:
1. There are four Priuses (Prii?) of that vintage at the local yard, wonder if I should go snag that part off of them?
2. I’ve used molten salt for heat baths in the medical device industry for heat treating nitinol for arterial stents. Operating temps are very high, and it is a stable medium. Cool to look into the furnace to see the lava-like movement
“It’s perfect for storing cocoa, chicken noodle soup or coolant. Maybe not all in the same vacuum flask, but still.”
Yeah, you definitely fit in here! 😉
I recall reading some years ago that one manufacturer (VW?) was looking at using molten salt as the medium for heat storage to improve the thermal efficiency of their engines. I remember it because it was one of the first times I’d heard of using molten salt but apparently there are some properties of it that are very attractive for storing thermal energy. Curious if the idea ever got past the concept stage.
More for heat extraction than for thermal storage, but sodium-filled exhaust valves have been used for decades.
It’s used in some high-temp applications, such as thermal solar storage (which allows thermal solar plants to produce consistent power through the night) and some nuclear reactor designs. One of the big problems is that molten salt freezes at over 200 degrees, meaning that it is a solid at most engine temperatures and all normal room temps. You can use it for storage in a car, but not reliably for a circulating heat transfer medium. This means that you need a coolant-salt heat exchanger that is made out of something non-corrosive enough to live with being constantly covered in hot salt.
My understanding is that there is a lot of this going on with the gen 2 Prius because the primary design goal was not great fuel economy (though that was a happy side effect), but the lowest possible emissions. Another example is that when the engine is cold and revving it would cause a lot of emissions, the car tries to use the battery for propulsion as much as possible to give the engine time to warm up before it gets pushed hard. If you hook up something like Torque to the OBDII port and keep an eye on the powertrain you can sometimes see when this happens. The power draw from the battery suddenly drops and the engine revs up instead.
In pracice I’ve found that if you’re just parking the car overnight this is about as effective as a block heater in terms of warming things up more quickly. The starting coolant temp tends to be very similar in my Prius compared to my truck with the block heater.
That said, there are drawbacks. That three-way coolant valve is a common failure point on these cars and when it fails you may lose cabin heat because it no longer sends coolant through the heater core. Not so much fun when it happens in the dead of winter. :-/
Still, all of this cool engineering is part of why I bought an old Prius so it’s a tradeoff I’m more than willing to make.
*practice
Also, edit button! 😉
Lol, DT put in as many words as Thomas.
Hmm may have to snag a couple of these tanks for my cabin heating system
I was just thinking about how I could use one of these to heat as an on demand heat exchanger for my vanagons shower unit.
Its the Insane amount of over-engineering that makes me love Toyota . So much effort went into Hybrid synergy drive(THS) that even though first gen Priuses(Priuii?) are extremely spartan, I am still seeing them around today. Looking at the Toyota newsletter that leaked last year were expecting the next gen Prius to get a bit of a power boost. If they can clear 150 hp and 130 ft-lbs of torque and not manage to look like a UFO its going to be an attractive offer, especially if its standard PHEV.
Also, whenever HSD is mentioned i feel obligated to drop this absolute gem of a video explaining the drivetrain in the Rav 4 hybrid, though he has videos on the Prius as well https://www.youtube.com/watch?v=O61WihMRdjM&t=0s
And the fact that not only did they introduce all this tech but made it incredibly reliable at the same time as opposed to various german brands that get credit for over-engineering half baked technologies.
Hey, this is a timely article!
I just bought one of the valves from a Prius to use with the Vintage Air system that I’m using with my GenV LT swap project right now.
The Vintage Air system comes with a servo controlled heater valve that completely blocks off the heater core circuit when the heat is off to keep the cabin cool. However LS and LT engines aren’t supposed to have a completely blocked heater loop. You can use a bypass, but I _also_ have a Holley water pump instead of a GM water pump, and it has no provision to supply coolant to the in-pan oil cooler (which is also an oil _heater_). So I’m mixing/matching the servo from the Vintage Air valve with the valve body from the Prius valve to have a variable diverter. Always 100% flow through the oil cooler, but variable temperature control to the heater core.
Now I know what the valve was actually for!
Hmm. I wish we could edit comments here, so I could go clean up ugly grammar and sentence structure after clicking POST and then re-reading to see how dumb I sound. 🙁
Nice. I always did think my mom’s old Prius was nice and quick to heat up; I always thought it was just because the engine was small (and therefore had less thermal inertia) but now I know that it was probably at least partly because there was a bottle of piping hot coolant sitting in there, ready to go. It’s an interesting and kind of elegant solution to a problem I’d never really thought much about before.
From what I can tell, the engineers left a little unpicked fruit. It seems as though the tank gets replenished with cold antifreeze as it is outputting hot, thereby mixing the two fluids and tempering the potential heat output. Similar effect to most folks’ water heater being actively cooled as it outputs hot – necessitating a larger tank. The solution to this could have been a pressure bladder type arrangement (surge tank) that outputs 100% heated fluid until it is all gone. The original configuration could have been a nice opportunity to throw in some phase change material to really store some heat and reduce vacuum tank size. Eh, who knows, probably would have cost more.
Probably this was just the simplest solution that allowed them to hit their design targets. Could they have taken it farther? Sure, but the job was already done and there were other things on deck. No point in fiddling around, making it more complicated than it absolutely had to be.
Blending cold and hot water in your house actually has the opposite effect – it makes the required tank size _smaller_.
That is because you can store more heat in less volume in the form of much higher temperature water. Then you use a thermostatic balance valve to bring the overall temperature down to something usable without scalding. This also handles the less deliberate blending that occurs when you replenish your hot water tank with cold water.
I actually have my hot water tank outfitted with thermistors at 4″ intervals from bottom to top for…. reasons. You would probably be surprised how little temperature blending occurs in the tank. It can be piping hot at the top and room temperature at the bottom.
LOL that last paragraph, ivan256. How to say you’re a mechanical engineer without saying you’re a mechanical engineer.
Oh shoot, hate to go down the theoretical road, but… If I have a 60 gallon surge tank filled with 160F water and I output 60 gallons (as in, to a couple shower heads in 30 minutes), I get 60 gallons of 160F water out of the tank and down the line. If you have a conventional 60 gallon water heater at 160F (and an entering water temp of 60F), you do not get 160F water, not even close.
As far as mixing goes, that’s what you do at the point of use valves, which is also where the anti-scald protection is located. In fact, it makes zero sense to rely on the tank and entering water mixing to prevent scalding (the initial hot water is still going to be scalding).
You are certainly right on the part where you can just up the tank temp — but then you lose efficiency to boot.
I, too, know the ways of the water heater, as my own water heater operates exactly as described above and is especially terrible in winter months.