Simple and cheap DIY exhaust heat to coolant heater idea - using air to transfer heat

[Daox]

I've been thinking about recovering exhaust heat again. As I've stated many times on the forum, I have a short commute of 7 miles. Engine warm up is critical for me. In winter, its not uncommon for the engine not even to get up to full operating temp, especially if I want heat. This really takes its toll on my fuel mileage. Thus, I love thinking about ideas on how to use wasted exhaust heat. So, here is my latest idea.

We've talked a LOT about exhaust heat recovery to heat the coolant here in my other thread on Mimicking Gen3 Prius Exhaust Heat Recovery. However, we really never came to a good conclusion on how to mimic it. IMO all the ideas were too complex or too expensive to be worth the time. So, whatever approach I took, it had to be simple, very simple.

This idea is pretty simple. We have a fan, an airbox around the exhaust manifold, and a radiator to transfer the hot air into the coolant. A lot of these parts can be salvaged from random junk, so a near zero cost system shouldn't be impossible to put together. The fan could be activated by a snap switch fixed to the block. No fancy electrical work, just simple. Yes, it will require a decent amount of fabrication to make the airbox and ducting. IMO that is the fun part.



There are no pumps, no coolant or exhaust diverter valves, no pressurized systems to deal with. You want heat, the fan turns on, you don't need heat anymore, it turns off. Air isn't going to break down if it gets too hot, or turn into steam and blow something up.

My main concern is pulling enough useful heat off the manifold. So, I'll be testing the warm up time on it when I get a chance.

Thoughts?

[oil pan 4]

I don't know if will move enough heat.
I actually tried something kind of like this once, where I tried to indirectly heat the coolant with warmed air. I took 2 space heaters, plugged them into my out door welding receptacles and ran extension cords to my car where I put both heaters on the passanger side floor.
The idea was to put heaters in the cab, direct them towards the vehicle air recirculater intake and use the heat to both thaw the ice and warm the engine a little faster with the engine running.
It thawed the ice but appeared to have almost no noticeable effect warming the engine.
This one of the reasons I have abandoned trying to air as a medium to heat anything beyond the inside of the cabin.

[Daox]

Yeah, thats my main concern too. Air isn't a great heat transfer fluid. What it does have going for it is its cheap, can handle high temps, and moving it is cheap and easy with a fan. Other than that, it can't carry much heat, and it needs large ducts to move small amounts of heat (vs liquid).

I'd like to run a test just to see how fast the manifold heats up on the car in normal driving. The only data I have says that in 5 minutes of idling, the manifold on my Paseo got up to 580F. I'd like to take an EGT gauge I have kicking around and throw it on the surface of the manifold and put some insulation over the top, then go for a quick spin to see what happens in the first few minutes of driving. If I can get 200F+ air in the first few minutes I think we'll be in business.

[whatmaycome14]

I assume:
The purpose of warming the coolant up sooner would be to save energy(by getting better gas milage).

So, the question would be this:
Wouldn't the energy used to power the fan offset any gains by warming the coolant?

[Daox]

Yes, it would offset it. The offset is just very marginal. In order to heat up the coolant faster, we need to move large amounts of heat. Heat is energy. The fan also uses energy, but its (hopefully) moving a LOT more energy than its using.

Its akin to a furnace blower. My home furnace blower fan uses ~300 watts of power, but it moves 80k BTU/hr which is the equivalent of about 23,500 watts.

Also, the fan is only going to be used during warm up. Once the engine coolant is warm, it'll turn off. So, its not a continuous draw.

[pgfpro]

It would cost a few bucks but a cheap Chinese turbo with manifold that is water cooled would work well. It would give you both hot air and hot coolant with the heat exchanger built in to the center cartridge?

This is my plan for winter on my car, no front mount inter cooler, hot charge pipe right to the T/B.

[Daox]

Yeah, that would work. But, getting heat doesn't seem to be the problem. One could simply wrap some copper around the exhaust pipe and be done. However, its now dumping a lot of heat into the coolant once up to temperature and the radiator how has to dissipate all that heat. More heat to the radiator means more airflow to it, which means worse aero. There are a lot of ways of disconnecting that heat, but this is the simplest I've been able to come up with.

[Xist]

Use the furnace blower?

[oil pan 4]

Trying to control heat flow between 2 fluids, now you are getting into industrial controls.
Which is what I do at work and there is no real easy way to do it.
(this is why I advocate the use of very direct heating via tube in shell heat exchangers and immersion heating elements)
That reminds me I have been meaning to do a spiel about how a vehicle is just like a production process and how all you have to do to make a vehicle more efficient is set it up to resemble a plant production process a little bit more.

[some_other_dave]

My question is: Why two separate heat exchange steps? You're gonna lose efficiency in each one, so better to keep those minimized IMHO.

For your copper tubing idea, you could have a bypass-type thermostat in the lines to and from the single exchanger. So when the coolant is cool enough, the thermostat would force it to the copper tubing, and when it was hotter it would close that section off and send the coolant into the lines coming back from the exchanger.

That's how an oil cooler thermostat works, generally, except backwards. I think you could make it work the way you want with some creative plumbing, though.

-soD