Heating With Waste Heat

[sailorchic34]

Originally Posted by TeddyDiver

That's why I said big enough..

My original thought was a second heat exchanger inline with the engine tank. BUT... A centrifugal pump, which most engine cooling pumps (not raw water pumps) are will have a flow curve where as the pressure drop increases the flow rate drops. So changing the system head would reduce the GPM flow rate through the engine

After some thought, I realized that even 3 or 5 feet of head added to the engine by an oversized heat exchanger could be too much. This as the engine engineers size the pump for peak flow rate and pressure drop. They cut that pretty much to the bone. They do that to save money. So there is not much, if any excess head available in the main coolant loop.

With a typical water heater loop you have small 1/2" or 3/4" lines coming off the engine which divert oh 1.4 GPM ish to a water heater, etc. But overall pressure drop in the engine loop stays constant.

It's for that reason that adding a second heat exchanger inline with the main engine tank, may introduce too much head pressure for the engine cooling pump to safely meet the engine heat rejection rates, under all conditions.

[Dockhead]

Originally Posted by TeddyDiver

How I would..
Sorry crappy picture.

Yes, that would be the classical way to do it, and that was my original plan.

But it requires more hardware and a lot more hose, which will create an awful lot of clutter, which is why I'm thinking about doing it like in my drawing.

[Dockhead]

Originally Posted by sailorchic34

This will work up to a point. With 1/2" or 3/4" hose your limited by flow rate So you might get 2 GPM through 1/2, probably a bit less as rubber hose has really high pressure drops. Maybe 4 GPM through 3/4" hose.

The smaller hose will probably only give about 3 ish kW of heat. Why 1" hose/loop and a DC pump is what you need to get the full 10kW of heat

2 GPM would give you about 10,000 BTU"S (3 ish KW), 4 GPM about 20K BTU or 6KW. In order to get 10kw of heat to match the hydronic boiler, you need a flow rate of about 6 GPM through the heat exchanger with a 10 degree F delta T. In theory, 4 GPM might work with a higher delta T,say 15-18 degree's F. But that puts the loop temp down around 60 degrees C, which may effect heat transfer through the fan coils.

In other words, the air temp out of the fan coils would be cooler. The hydronic heating system is designed for 70-80 degree C and about a 5 degree C delta T (which is pretty standard).

OK, that's very clear and very useful. That's where real engineering knowledge steps in.

I would have built my system and then wondered why I got so little output. Thanks!

So I guess that's the death-knell of my drawing.

I guess I will need short separate loops from both main engine and genset with larger diameter hoses and pumps. And separate heat exchangers.

This is going to be much more cluttery and expensive than I had hoped

So I just cut the hoses before the thermostat and after the heat exchanger, right? And "t" in a 1" hose?

Did you write in a different post that this will interfere with engine cooling? How, if the electrical pumps are in parallel with the fresh water pumps?

[Dockhead]

Originally Posted by sailorchic34

My original thought was a second heat exchanger inline with the engine tank. BUT... A centrifugal pump, which most engine cooling pumps (not raw water pumps) are will have a flow curve where as the pressure drop increases the flow rate drops. So changing the system head would reduce the GPM flow rate through the engine

After some thought, I realized that even 3 or 5 feet of head added to the engine by an oversized heat exchanger could be too much. This as the engine engineers size the pump for peak flow rate and pressure drop. They cut that pretty much to the bone. They do that to save money. So there is not much, if any excess head available in the main coolant loop.

With a typical water heater loop you have small 1/2" or 3/4" lines coming off the engine which divert oh 1.4 GPM ish to a water heater, etc. But overall pressure drop in the engine loop stays constant.

It's for that reason that adding a second heat exchanger inline with the main engine tank, may introduce too much head pressure for the engine cooling pump to safely meet the engine heat rejection rates, under all conditions.

Oh, so you were thinking to put the heat exchanger in series with the fresh water pump and the sea water heat exchanger? Is that what you're saying? I wouldn't do that not only because it might interfere with engine cooling, but also it will only work with the thermostat open (unless you find some place upstream of the thermostat, but I don't know how you'd do that -- that's all inside the engine). That would ruin it -- since the thermostat is mostly closed anyway -- the coolant temp never gets more than 1C above the thermostat opening temp. That is, the engine never produces as much heat as the sea water heat exchanger is able to reject at the thermostat opening temp. So to tap into that heat, we must tap it upstream of the thermostat. And I can only see how to do that as a branch, not in the main flow, so this problem goes away, right?

[Stu Jackson]

Originally Posted by sailorchic34

When connecting the hose to the engine, Don't do it as you would for the colorifier. Doing so with a DC pump in series flow with the engine pump could adversely effect engine cooling performance. Take the supply and return off the same hose that goes to the engine heat exchanger. This decouples the DC pump from the engine cooling circuit and does not effect engine cooling flow rate.

That's a really good description of what sc said before: primary/secondary pumping system. Even these days, many engineers don't understand it, and Bell & Gossett, "THE" pump experts, have written an entire white paper on it. In 35 years of engineering, i still couldn't get a lot of senior engineers to understand it. Sheez... Remember, however, that it still uses the same water, so if there's a leak, both circuits will drain, they are separate only for functioning pressure and flow issues.

Good luck.

[Jd1]

Originally Posted by Dockhead

Oh, so you were thinking to put the heat exchanger in series with the fresh water pump and the sea water heat exchanger? Is that what you're saying? I wouldn't do that not only because it might interfere with engine cooling, but also it will only work with the thermostat open (unless you find some place upstream of the thermostat, but I don't know how you'd do that -- that's all inside the engine). That would ruin it -- since the thermostat is mostly closed anyway -- the coolant temp never gets more than 1C above the thermostat opening temp. That is, the engine never produces as much heat as the sea water heat exchanger is able to reject at the thermostat opening temp. So to tap into that heat, we must tap it upstream of the thermostat. And I can only see how to do that as a branch, not in the main flow, so this problem goes away, right?

Maybe I need to have another look at my system but I would think it would be essential to only tap in AFTER the thermostat. If you tapped in before the thermostat you could, in theory at least, extract too much heat from the engine. Running an engine cold is not a good idea .... everything being equal, a diesel engine would rather run hotter (195F thermostat).
You should only be able to extract extra heat that the system would have normally rejected through the raw water heat exchanger.

[Dockhead]

Originally Posted by Jd1

Maybe I need to have another look at my system but I would think it would be essential to only tap in AFTER the thermostat. If you tapped in before the thermostat you could, in theory at least, extract too much heat from the engine. Running an engine cold is not a good idea .... everything being equal, a diesel engine would rather run hotter (195F thermostat).
You should only be able to extract extra heat that the system would have normally rejected through the raw water heat exchanger.

I feel pretty sure about this --

you do NOT, I think, want to tap in after the thermostat.

The reason is that the fresh water cooling BYPASSES the sea water heat exchanger unless the thermostat is fully open. The thermostat modulates the diversion of cooling water to the seawater heat exchanger in order to keep the temperature right at 80C. Thus, the seawater heat exchanger does not get the full flow of circulating coolant, and so this is really not where you want to be tapping off coolant for heating. You want to get into the main flow which passes the seawater heatex by most of the time. I'm not an engineer, but it's very logical, and certainly that's the way car heaters work -- they get the coolant flow BEFORE the thermostat. The seawater heatexchanger -- or the radiator if we're talking about cars -- is actually in a BRANCH of the cooling system -- NOT in the main flow, which mostly bypasses the radiator or heat exchanger.

Yes, you could overdo it and cool the engine down too much. The only remedy to that is to watch the temperature and shut down your heating system if it's dragging down the engine temperature too much. And you don't want to switch on the heating system before the engine is well warmed up. If you get chronic overcooling, you will need to reduce the flow to the heating system.

[FlyingCloud1937]

Here is how I did what exactly you want to do, but I also have an oil fired diesel cook stove in the loop.

Lloyd

[sailorchic34]

Very nice system LLoyd.

OK, back to Dockhead's system.

So on the side stream heat exchanger BOTH supply and return to the side stream heat exchanger connect to the hose between the water pump and engine tank/heat exchanger (fresh water side). An option would be to connect both hoses to the hose from the engine block to the cooling pump.

Assuming we're going after the pump: from the pump you will have a hose with a tee to the DC pump suction, other end of the tee connects with a 6" length of hose and another tee which connects the return from the side stream heat exchanger. That tee then has a length of hose connecting to the fresh water inlet of the engine heat exchanger.

Both the supply and return connect to the same hose. It's magic...
Why to the same hose you ask. Great question. We will have about 8-10 GPM flowing in the fresh water loop at 80-85 degree C. We pull off 6 GPM of that and send it to the side stream heat exchanger and return to the same fresh water hose 6" down from the supply. The return mixes with the remainder of the engine flow and heads to the engine tank for additional cooling. The engine T-stat regulates flow to maintain engine temp.

By placing both supply and return hoses on the same engine hose, we decouple the two pumps (engine cooling and 12VDC) at that point. This is VERY important as otherwise the flows in the engine could get wonky, which would be unpleasant. This works where you have a side stream pump.

Without a side stream or secondary loop pump, you get a system like for a typical colorifier/water heater, supply is after the cooling pump and return is before the cooling pump. This works as there is only one pump in the system. You do not want two pumps in series with each other, which is what you would have with the DC pump piped across the engine pump. That as the ghostbusters would say "would be bad". The engine pump is variable flow based on RPM, the DC pump is constant flow. Not ideal, but by keeping the DC pump small, that is under the cooling pump flow rate life is good.

Because both hoses connect to the same engine hose, flow in the engine loop stays balanced and the engine is happy.


I pretty much memorized all of Bell and Gossett's design manuals. Even taught many a engineer how to properly size a pump and heat exchanger. I've had a lovely chat one day with the fellow that came up with the primary/secondary design for an hour or two one day, one on one.

We talked about variable flow primary and variable flow primary, distributed secondary. There is also Primary/secondary/tertiary. I've designed chiller systems to 10,000 tons/ 2500 GPM and learned on 20,000 ton systems. Cool stuff. Easy to get wrong too, and a bit of voodoo in there too. Many of the systems don't work as they were designed to do. But I know way that is too...

[sailorchic34]

The engine t-stat is generally connected to the discharge from the engine tank/exchanger. We do not connect there. You can connect to either the hose to the fresh cooling pump suction, that comes from the engine block, or after the engine fresh water cooling pump, as I described in the post above.

I'm been pondering and I think the two hoses will work slightly better if both are connected on the hose between engine block to the suction side of the pump. Though it should work ok too on the discharge side of the fresh water pump.

[Jd1]

Originally Posted by Dockhead

I feel pretty sure about this --
you do NOT, I think, want to tap in after the thermostat.
<snip>
And you don't want to switch on the heating system before the engine is well warmed up. If you get chronic overcooling, you will need to reduce the flow to the heating system.

Thinking some more about this, I can confirm that my boat system seems to take the hot water before the thermostat since with a hot engine and a cold hydronic system I can switch on the circulating pump and drag the engine temperature down.
I guess it depends on your priorities - if you want heat at all cost or if you want to be nice to the engine. In a car system, the customer wants heat at all cost and doesn't care what happens to the engine. Also, the heater coil that is being used is a fraction of the cooling load of the boat systems while the engine is usually multiple times bigger than the (sail)boat engine.
I would prefer to only use that part of the engine heat that is being rejected (after the thermostat) and will see if I can change my system configuration to achieve that. I realize that it will take longer to see the hydronic loop warm up but that is ok with me.
An alternative is to sense the engine temperature and only allow the hydronic circulation pump to run when the engine is at operating temperature.

[FlyingCloud1937]

Typically in a marine engine installation with take off heat for auxiliary use ie domestic HW, and or cabin heat.

Suction side of the diesel engine FW/closed system side, has a y, which is the return, ie FW suction pump, and the supply is taken above the t-stat.

By this method no HW is available to the auxiliary heat demand until the T-stat is open. This allows the engine to self regulate to it's own design temp. Any excess heat is then wasted on the D-HW, and or the cabin heat, then the sea water exchange.

This is a rule to live by...period

It goes the same for direct discharge or Heat-X system.

We only want the available waste heat that the engine does not need. In a proper designed system, in most cases this will be enough. If not, you must add additional heat source...not rob heat demand from the proper operation of the prime mover, whether it be drive engine(s), or generator.

Lloyd

Originally Posted by sailorchic34

The engine t-stat is generally connected to the discharge from the engine tank/exchanger. We do not connect there. You can connect to either the hose to the fresh cooling pump suction, that comes from the engine block, or after the engine fresh water cooling pump, as I described in the post above.

I'm been pondering and I think the two hoses will work slightly better if both are connected on the hose between engine block to the suction side of the pump. Though it should work ok too on the discharge side of the fresh water pump.

[sailorchic34]

Dockhead's load will be about 10% of the main engine heat and 20% of the genny heat load at peak demand. The engine T-stat will provide adequate regulation to keep the engine(s) happy. That after all is what it is there for.

One danger of placing the DC pump loop, pulling from the discharge of the coolant pump and returning to the suction side of the coolant pump is as follows. Say you have a engine coolant pump running at 10 GPM and a DC pump on the secondary loop running at 5 gpm. This has the effect of reducing flow from the engine block by ~ 5 gpm which will cause the engine block temp to rise.

By connecting the secondary side supply and return to one hose as I previously described, engine flow is not compromised and the pumps have been decoupled