Back boiler system design

[RuK]

Hi I know there are already many threads on this topic, but my particular situation is that I've got a boiler stove with the intention of using it for hot water. My boat is heated just fine by a stove alone.

 

Only then I find I need a dump radiator working on thermosyphon. I have limited space to fit in a radiator or difficult pipe runs , so my question is does the radiator have be sized to equal the full output of the back boiler? I have about 4m run of 28mm copper to dissipate some heat. I assume my calorifier can't be treated as a heat dump.

[BEngo]

Your calorifier can't safely be treated as a heat dump because the tap water can only get so hot and after that the heat from the fire will go on boiling the water if it can't go elsewhere.

 

Ideally the radiator should be able to absorb the full output of the boiler, so that the water can not boil. If you can't get full capacity in one radiator, then you need to look at:

 

What is the heat loss from the 28mm copper? There are figures out there to estimate this.

 

Have I gone for the best radiator out there?- double, double finned high efficiency, or even fan assisted? Can you divert the output to a skin tank using a thermostatic valve?

 

What are the heat delivery conditions for the radiator I have. Typically these will be at given inlet temperature and a specified temperature drop. If the inlet temperature rises to 212F (boiling) then the radiator will deliver more heat to the boat.

 

Can I reduce the boiler heat output with some extra firebricks?

 

All these may get you enough heat loss, but if not....

 

Finally, how big is the header tank and how long will the water last if the boiler is boiling it away. If the content of the header tank (NOT the whole system) is more than the boiler can boil away while run flat out on one filling then you will be OK. Steamed maybe, but OK. If using this approach you need to ensure that there is a good clear run from the header to the boiler and that there is a 15mm or 22mm vent pipe off the top pipe run back into the top of the header tank. The tank itself must also be vented with a big pipe 15/22mm ideally to the outside. You do not want any way in which the steam pressure can build up in the system.

 

N

[RuK]

Thanks for that N looking at it I think I can prob get just enough heat loss in, and there is space for an extra heat loser in the engine room which seems pretty silly but if it's on a thermostatic valve then would make much more sense.

[BEngo]

You will need to use a thermostatic valve that opens when the water gets too hot rather than the normal radiator type. Not sure how easy these are to find. It might be easier to fit a pump which comes on when the water is too hot and pushes the water into the engine room- if you install it so the normal circulation through the extra rad is impeded, ( ie with a fall in the flow and a rise in the return) but is overcome by the pump this will work well.

 

N

[station tug]

we run a refecks derv stove running a small radiator and the clarofier the system is biast the hot water....22mm pipe...1 inch to the meter rise for a decent thermospen..if you get banging in the pipes add half a cup of washing up liquid this stops it..

[smileypete]

Going by this chart, 28mm copper probably loses up to 100-110W/m when the water inside is near boiling point:

 

http://www.engineeringtoolbox.com/copper-pipe-heat-loss-d_19.html

 

Normally a good way to go is have adequately sized gravity heat dump rads nearby or next to the stove, then more distant rads and calorifer on a pumped circuit. If the backboiler is large compared to water heating needs then partly covering it with a firebrick will reduce its output some.

 

Some details of stove kw output, backboiler output, cabin/boat size and calorifier size might get some suggestions.

 

cheers, Pete.

~smpt~

Edited January 8, 2014 by smileypete

[MtB]

Going by this chart, 28mm copper probably loses up to 100-110W/m when the water inside is near boiling point:

 

http://www.engineeringtoolbox.com/copper-pipe-heat-loss-d_19.html

 

Normally a good way to go is have adequately sized gravity heat dump rads nearby or next to the stove, then more distant rads and calorifer on a pumped circuit. If the backboiler is large compared to water heating needs then partly covering it with a firebrick will reduce its output some.

 

 

 

That way, about half the cooled return water from the pumped circuit goes backwards through the heat dump radiator, and the other half through the back boiler.

 

Very inelegant design even though I'd imagine it would work perfectly well.

 

 

MtB

[The Pipe]

we run a refecks derv stove running a small radiator and the clarofier the system is biast the hot water....22mm pipe...1 inch to the meter rise for a decent thermospen..if you get banging in the pipes add half a cup of washing up liquid this stops it..

Going back many years 'Fairy liquid' had a masive salt content so maybe not a good idea to chuck it in the system.

[RuK]

Thanks for your replies, and while we're on the subject - do I need a continuously shallow rising flow pipe to my rad, or a sharp rise then a horizontal run as some suggest?

 

Also, and I realise I may be opening a can of worms here, surely if you vent your Calorifier then the water pump will push water out of the vent every time it's on? A pressure relief valve would make more sense surely?

 

I may also add another heat source for hot water at a later date, maybe solar thermal but I may also never get round to it. I'd prefer to get a single coil cylinder and if/when I add another heat source just use 3 port valves to manually switch between systems. Is there any reason not to do this?

[BEngo]

A continuous rise to the rad is best- remember to allow for any change of trim as the water tank empties. A vertical rise then a horizontal pipe might work- sometimes it does, sometimes it doesn't, depending on the exact configuration.

 

You need to vent the heating circuit so it lets the air out, and doesn't pressurise if/when it boils, and have a PRV on the hot tap water side so that as the tap water expands when it gets hot the calorifier doesn't burst. The two parts are (should be) completely separate. St the PRV just above the water pump cut out pressure.

 

Single coil is fine, but may be slower than a twin coil plumbed in series. If using valves to swap heat sources you must ensure that there is no combination of valves which leaves a heat source unvented/without a pressure release.

 

N

[RuK]

Also I'm potentially looking at a pipe run rise of 35mm per metre on the flow, accounting for the trim of the boat. Any thoughts on whether this is an adequate rise?

[MtB]

Also I'm potentially looking at a pipe run rise of 35mm per metre on the flow, accounting for the trim of the boat. Any thoughts on whether this is an adequate rise?

 

That sort of rise will be fine. A little less will work satisfactorily. On reflection it will also work if the pipe falls downwards slightly from the stove towards the back of the boat PROVIDED there is a method of fully releasing all the air from the highest point.

 

The higher the flow pipe and lower the return pipe the better, and the bigger the diameter of each, the better, in any convection circuit.

 

 

Thanks for your replies, and while we're on the subject - do I need a continuously shallow rising flow pipe to my rad, or a sharp rise then a horizontal run as some suggest?

 

Avoid horizontal runs at all costs on convection systems. The rise/fall is needed to ensure the pipe is truly full of water with no air locks. You have no pump to expel trapped air, it has to find it's own way out by rising to the top of the system, where you will have installed a method of releasing it!

 

MtB

[smileypete]

 

That way, about half the cooled return water from the pumped circuit goes backwards through the heat dump radiator, and the other half through the back boiler.

 

Very inelegant design even though I'd imagine it would work perfectly well.

 

Only because the implementation you're thinking of is inelegant

 

The better way to do it, is to tee the pumped circuit right next to the back boiler. That way the back boiler helps to serve as a 'neutral point' and almost all of the pumped flow will go through it:

Back to the OPs question, I suppose they could have gravity rads that cover say 70% of the back boilers and then have at least 30% rads covered by a pumped circuit.

 

But if the pump malfunctioned they couldn't run the stove at more than 70% til it's sorted, whether that's good enough depends on stove output and boat heat requirement I s'pose.

 

A car heater matrix with some fans would output a lot of heat for a small space, I wonder if a few slow running 80mm fans across the bottom of a double rad blowing upwards would boost it's output some.

 

cheers, Pete.

~smpt~

Edited January 12, 2014 by smileypete

[RuK]

I am reassured, but a bit surprised by what you say about pipe rises /falls Mike, seems a bit counter to most of the stuff I've read on gravity systems. The flow at least needs to rise to encourage hot water to flow through it?

 

I will have 28mm pipework, rising continuously from boiler to vent for the flow. But I wonder whether I can have the return go lower than the boiler and then rise to connect to the boiler inlet? The Calorifier will be directly behind the boiler with 22mm coil, then the heat dump rad will be approx 4m down the boat with about 35mm rise per metr. .

I'm a bit worried the thermosyphon won't be strong enough to reach along to the rad and back. Also not sure as to best layout for pipes on connection to the rad - flow could continue straight up to the vent, with a T off to a falling pipe to the rad. Or I could have the flow go straight into the rad and take the pipe up to the vent from the other end of the rad. Sorry I've not worked out a way to upload sketches on here yet hope this is clear.

[BEngo]

Every inch the return pipe goes below the boiler is effectively the equivalent of an inch less rise on the flow pipe.

 

If you can get a level or better return then providing you balance the rad and the calorifier flow ( get the resistance to flow to be the same for both loops) then the 35 mm/m rise you have will drive the rad OK. If you have something like a thermostatic cylinder valve on the calorifier which closes when the water is warm then you don't need to balance the flow, otherwise install a ball valve on the calorifier return and the radiator return and close the one that is getting hot until both are doing OK ( returns the same temperature) . if you don't balance the flow either the radiator will probably not get hot until after the calorifier is pretty well heated up or only the rad will get hot. (I guess the latter is more likely because the cal coil is a big resistance to flow.)

 

For best effectiveness, both in getting heat out and in getting the best temperature drop to promote circulation, (pretty much the same thing), the rad should be plumbed for TBOE ( top and bottom opposite entry) where the flow is diagonally through the radiator. Since there is only one rad that means the top stove end can be connected to the flow, the vent comes out of the other end at the top and the return is taken from the bottom furthest away from the stove.

 

You will need to either arrange the pipework for the cal so that any air goes into the flow pipe and away to the rad and then the vent or fit a vent at the top of the calorifier circuit.

 

N

[MtB]

 

Only because the implementation you're thinking of is inelegant

 

The better way to do it, is to tee the pumped circuit right next to the back boiler. That way the back boiler helps to serve as a 'neutral point' and almost all of the pumped flow will go through it:

 

 

I disagree.

 

With your layout once the pump starts, the natural convection motive force will be opposed and overwhelmed.

 

The only way your layout will work as you expect is if you use an "injector tee" where the two return pipes join before entering the boiler. This way the venturi effect will assist the natural convection circuit. Installed as drawn, the pump will oppose the natural convection motive force.

 

MtB

[smileypete]

 

I disagree.

 

With your layout once the pump starts, the natural convection motive force will be opposed and overwhelmed.

 

The only way your layout will work as you expect is if you use an "injector tee" where the two return pipes join before entering the boiler. This way the venturi effect will assist the natural convection circuit. Installed as drawn, the pump will oppose the natural convection motive force.

 

Whoa! Lets start from the beginning, one thing at a time

 

Having seen the diagram do you still assert that:

 

That way, about half the cooled return water from the pumped circuit goes backwards through the heat dump radiator, and the other half through the back boiler.

 

cheers, Pete.

~smpt~

[Jon1]

Smileypete, re your diagram, is it possible to get gravity fed pump over into the expansion tank with this system?

[WotEver]

Pete's not been active on this forum for nearly 2 months and as this thread is getting on for 3 years old my guess is that he won't be replying any time soon.

[Jon1]

Thanks WatEver. Shame though as I've been enjoying reading some of his posts.