Advice on antifreeze heater?

[flippineck]

Got a problem here with small 12V diaphragm pumps bursting in the cold.

They are for feeding water from an outside 1000L storage tank, into a static caravan's plumbing system.

Temperatures here in the UK drop below freezing regularly during winter. What's the lowest? I don't really know if I'm honest, a wild guess would be regular drops into the single figures below zero, sometimes into the teens below, and maybe exceptionally cold would be the rare drop to minus 20 or so? Could be wrong on those figures.

At first I just had the one pump lashed up with no protection against low temperature at all, a stopgap in autumn just to get the water 'on'. Coming into winter, I found that if I let the pump experience ice formation, it didn't fare so well, with leaks appearing upon thawing. Not entirely unexpected.

I discovered through the autumn that the one pump alone is okay if just one person is doing one thing like running one sink tap, but if several demands come on simultaneously (toilet cistern filling, someone washing the dishes, someone else filling the sheep's water trough etc.. the flow drops a lot. So, I rigged up five pumps in parallel on an 18mm thick plywood board. The water feeds are paralleled up through 15mm / 10mm copper pipework manifolds and the motors are electrically paralleled to the 12V lead acid leisure battery feed. As many pumps as necessary now come on in response to pressure drop, to keep up a good flow.

I made a hinged lid out of 18mm ply to fit over the 5 pumps on the board. The whole thing sort of shuts like a briefcase and is hung on the side of the 1000L water container.

So, I have these pumps and their associated pipework etc, shut up inside a wooden box 1195mm long, 345mm wide, and 70mm deep.

I'm planning to try and minimise ice formation in the pipes between the pumps and the caravan using foam pipe insulation / lagging. Hopefully the sheer size of the 1000L water container should keep solid ice away from the 'works' - the pipe outlet is right at the bottom where I can lag it well, and hopefully any significant ice will form at the top of the container.. as long as I keep it fairly well topped up hopefully liquid water should almost always be available at the bottom outlet.

What I was thinking might be a good idea, would be some form of small thermostatically controlled 12V heater inside the plywood box, along with the pumps - just to prevent temperature inside the box falling below zero degrees C?

Can anyone recommend the sort of parts I would be best using for this. Should I go for some kind of mini fan heater? Does anyone know of anything 'off the shelf' readymade for this purpose?

I thought of using filament lamps as heaters but then I thought, would they be inefficient? They'd be cheap enough but would the (unwanted in this application) light output be a significant waste of the limited battery power?

Peltier type devices crossed my mind briefly, another idea was a simple heating element made out of a length of resistance wire running around the inside of the box, maybe on some kind of frame.

I read about Polymer PTC heating elements, some of these seem to self-regulating, wondered if anything like this exists specifically to keep temperatures just above zero without the need for a thermostat?

Here's one heater I found which would basically do the job, but it's 220V.. http://www.ebay.co.uk/itm/252422314279

Also this trace heating cable seems interesting, it mentions it's self regulating, www.ebay.co.uk/itm/390986887824

How much heat energy would I need the heating arrangement to provide, in order to prevent ice formation under the coldest weather?

What would be the best method of sensing exceptional cold - so that say, if the temperature went below minus 20, power could be automatically removed from the entire set of pumps. What kind of thermostats are available for 12V operation, are they all electromechanical or is there some kind of simple semiconductor stat available? for instance I was looking at this ebay listing for a BEKO fridge / freezer stat www.ebay.co.uk/itm/201166213034 & wondered what the square blue component is - some kind of subminiature mechanical stat? Where can you get something like that?

Thanks for all and any thoughts.

 

[hevans1944]

I thought of using filament lamps as heaters but then I thought, would they be inefficient? They'd be cheap enough but would the (unwanted in this application) light output be a significant waste of the limited battery power?

You are overthinking this. ALL the electrical energy supplied to a filament lamp is converted to heat. The light output, too. In fact, most of the electrical energy is converted directly to heat and only a very small percentage appears in the form of light from the incandescent filament. You need to thermally insulate the inside of the box to avoid wasting the heat to the frigid outside world. Try using a few brake lights to heat the inside of your pump box, maybe wire them in series/parallel so the light output is diminished while obtaining the same power input to the lamps. Some trial and error is necessary to obtain sufficient power input to maintain a toasty temperature inside the box. Probably don't even need a fan (which wouldn't hurt to distribute the heat), but the power drain from the lights (heaters) will probably require a mechanical thermostat inside the box to avoid running the lights continuously. You might even try a bank of power resistors to provide heat.

Any electricity you provide to a resistive load is converted to heat with 100% efficiency. The question is (and you don't know the answer yet) how much power do you need? You should peruse a basic text on thermodynamics to help find the answer. The power you need depends on the difference in temperature inside the box versus the temperature outside the box and the "R-value" of the insulation between the inside of the box and outside the box. Bigger R-value leads to lower power requirement for a given temperature difference. Try to put some numbers on this.

 

[davenn]

ALL the electrical energy supplied to a filament lamp is converted to heat.

almost all ... around 97 %

 

[Alec_t]

..... but, since the light is bouncing around inside a closed box, virtually all of the light energy will also be converted to useful heat energy.

 

[hevans1944]

almost all ... around 97 %

Nope. Even the light is eventually converted to heat when it is absorbed by the surroundings.

 

[davenn]

Nope. Even the light is eventually converted to heat when it is absorbed by the surroundings.

Come on mate ... you know full well that's not what you referenced to in your post and I will quote again

ALL the electrical energy supplied to a filament lamp is converted to heat.

all absorption of the light is a later story

a more correct statement would have been ...
ALL electrical energy supplied to the filament lamp is converted to and is radiated as E/M radiation

 

[ldcarter]

Not sure about availability in the north of England but I used rigid foam with a foil covering on my work building. With the foil covering (probably aluminum, but not sure) facing "in" this would reflect both heat and light back inside your box. I faced mine towards the outside to keep outside heat and the sun's rays out. A university study I read said it didn't matter about the orientation but I liked the idea of having the foil side toward the heat source. If I remember correctly, it's recommended to install this with construction adhesive.

If there a breakers yard close by you could use hevans1944 idea and buy a used tail light with as many bulbs as you could find for the heat source. Save you some money perhaps and give you all the components needed to just wire it in.

 

[hevans1944]

... a more correct statement would have been ...
ALL electrical energy supplied to the filament lamp is converted to and is radiated as E/M radiation

Okay, that may be closer to what's going on, but it isn't the whole story. Some of those terribly inconvenienced electrons forced to journey through the lamp filament just transfer their energy to molecular vibrations represented by phonons. This is heat transferred to nearby objects or created in the object through which the current flows. The heat is transferred, always from a warmer body to a colder one, by conduction or convection or by radiation.

Of course anything warmer than absolute zero emits radiation. My point is just this: every electric current flowing in a resistive load requires energy, and this energy is converted to heat, of whatever wavelength, with 100% "efficiency" in that all of the electrical energy input eventually appears as heat energy. Sure, all of it will eventually escape to the environment, whether by conduction, convection, or radiation, but the sum total of "heat lost" is always exactly the same as the electrical energy input.