Hot water tank construction

[Jim Cate]

It's like working with fiberglass.

More like fiberglass mixed with Mexican jumping beans! Don't turn your back on it!

Jim

[GILow]

Originally Posted by RaymondR

I think you have your red and blue arrows reversed.

Interesting concept.

Interesting...

My logic is that I am pulling hot water from the TOP of the tank (red pipe), cooling it in the radiators and returning it to the BOTTOM of the tank, "cold" in the blue pipe. What little temperature stratification I get from such a small tank dictates that I should pull hot water from the top of the tank.

Does that make sense?

[GILow]

Originally Posted by Uncle Bob

Hi Matt, somewhere in Tas there is a business that produces a heating system using a diesel fired heater & engine heating and is built to provide hot water and cabin heat. Probably very exy but may be a source of inspiration.

Thanks mate, I did look at those. They use (I think) the Planar heater, which, if I have it right, is the Russian version of the Webasto. Apparently a sturdy device and superior in some ways to the Webasto (simplicity) but not quite as reliable. The Tassie guys' system look neat, and they probably work well, but I never asked about the price...

Because...

The thing is, I REALLY liked ColdEH's approach to the problem. It's just very, very clever. I am ripping off his system totally, with only a minor tweak because unlike him, I don't live in Canada, but Australia, so I had to do all I could to minimise heat leakage into the cabin space. If I lived in Canada I probably would have taken his design unmodified as it has more thermal mass in the furnace circuit which is a big plus.

[GILow]

Originally Posted by TassieBloke

Don't forget a pressure relief valve in case of temp sensor failure. Route overflow down to bilge.

Good call, but the system is totally unpressurised. It will have an expansion bottle, but it will be the sort that does not create pressure in the system.

SailorChic made a good point about soft plastic, this system would have real problems if it ended up under pressure.

I am currently coding the temperature control system, which uses a picaxe chip and a DS18B20 sensor. The code has some built-in fail safes, which SHOULD shut things down if the temperature does not behave appropriately. So, say the temperature sensor dies in a way that causes it to give a constant reading, the code will detect the lack of temperature change as highly suspicious and shut down the furnace. The code is looking for a saw-tooth temperature profile, anything else is a sign of trouble.

Of course, I will bring the laptop with me because I am quite sure my code will be find a bug when it is minus 10 degrees C and gleefully shut down the furnace. If this happens, provided my fingers can still type, I can reprogram the picaxe appropriately. Nothing like debugging assembly level language while your fingers are turning blue.

[GILow]

Originally Posted by Lepke

There is closed cell pour/spray foam that doesn't absorb water. Fishing boats use it. You can fiberglass or paint it. UV light degrades the surface so coating is needed in direct sunlight. It comes in different expansions depending on need to carry weight. Amazon or ebay are good places to buy smaller amounts.
You don't have to pour all at once. Expansion is powerful and can bend/break things. Wear gloves and old clothes. It's like working with fiberglass.

Yes, I think that's the stuff Franziska was recommending. It "doesn't absorb water".... VERY MUCH. That is to say, it has a small percentage of open cells that will absorb water.

It sure is lively, as Jim suggests. I used the two-part foam recently to make my windvane and I, thankfully, was prepared. Thanks in part to watching a few YouTube videos where people were NOT prepared.

Messy, very, very messy.

[misfits]

Interesting idea using a buffer/thermal storage tank on a boat. I agree with SaliorChic, I wouldn't use a plastic tank. Other thing to consider, some two part foam products generate a tremendous amount of heat during the curing process.

Have to admit, sounds like a good science experiment.

[Typhoon]

Almost 5 years in with my hot water/heating system , no troubles what so ever . Only maintenance is changing the injector nozzle on the boiler once every couple of years .

Ps I have gone back to my Typhoon handle .

Regards John

[RaymondR]

Originally Posted by GILow

Interesting...

My logic is that I am pulling hot water from the TOP of the tank (red pipe), cooling it in the radiators and returning it to the BOTTOM of the tank, "cold" in the blue pipe. What little temperature stratification I get from such a small tank dictates that I should pull hot water from the top of the tank.

Does that make sense?

Ah, OK, but you need to put the out take down below the water surface in the tank on your schematic and maybe move the pump down to the input feed line of the radiators.

[GILow]

Originally Posted by Typhoon

Almost 5 years in with my hot water/heating system , no troubles what so ever . Only maintenance is changing the injector nozzle on the boiler once every couple of years .

Ps I have gone back to my Typhoon handle .

Regards John

Hello again John. I PM’d you on the ColdEh handle, but perhaps you could answer here instead.

I was wondering if you felt the Webasto circulation pump would be adequate in my adaptation of your design, given it is just pumping coolant to/from the buffer tank?

Also, are you still happy with 55 to 63 degrees C coolant temperature?

[GILow]

Originally Posted by RaymondR

Ah, OK, but you need to put the out take down below the water surface in the tank on your schematic and maybe move the pump down to the input feed line of the radiators.

Oops. My bad. I meant to colour in the area above in a sort of red colour to show it was full of hot water. I forgot. ☹️

[GILow]

Originally Posted by misfits

...
Have to admit, sounds like a good science experiment.

Hopefully more than just an experiment. Fingers crossed but John is happy with his setup so the physics are sound.

My plastic tank is possibly a weak link but early experiments with it suggest it will be fine. I can’t feel much softening when I filled it from the kettle. Waaay hotter than the planned temperature.

[Tricolor]

Stainless post mix tanks, for coca cola machines, they are 5 gallons and have a big lid on the top.

(also known as coca cola syrup tank)

[GILow]

Originally Posted by Tricolor

Stainless post mix tanks, for coca cola machines, they are 5 gallons and have a big lid on the top.

(also known as coca cola syrup tank)

Interesting idea. Smaller than I’d like though. John suggested a larger volume to stop the heater doing short runs. Can you think of something closer to ten US gallons?

[Tricolor]

Many moons agoo I used one with a 12mm 6m long soft copper spiral inside connected to a small Webasto diesel water heater as hot water boiler. I don't remember exactly but it took longer as expected to bring the water to 60ºC 20 minutes or so.

I don't know other usable tanks with double capacity.

[GILow]

Originally Posted by Tricolor

Many moons agoo I used one with a 12mm 6m long soft copper spiral inside connected to a small Webasto diesel water heater as hot water boiler. I don't remember exactly but it took longer as expected to bring the water to 60ºC 20 minutes or so.

I don't know other usable tanks with double capacity.

That was my original plan but Typhoon (aka ColdEh aka John) turned that design on its head.

From a thermal dynamics point of view, John’s version makes more sense. The rate of heat transfer from the large coolant tank to the copper coil does not need to be all that great as a shower should not flow that fast (at least not on a boat with limited water supply) and the kitchen sink is not all that large.

John’s design gets the heat from the furnace into the thermal mass as fast as the furnace design permits and keeping the overall system temperature quite low also assists by keeping the delta-T high in the furnace.

Trying to transfer heat from the furnace to the thermal mass using a copper coil would have the furnace running at MUCH lower delta-T and therefore much lower efficiency.