Immersion chiller

[Tony]

Hi all
had a rather drunken conversation with a mate a couple of nights back - talking about immersion chillers (and apologies if this has been on here before, couldn't find it).

What is the best speed of cold water through the immersion chiller? I thought it would be as fast as possible, keeping it as cold as possible - and working on the same pronciple (only reversed) as heating things up - the hotter the better. My mate thought it should be at a more measured rate, allowing the heat to be transferred from the wort and carried away. He said that if the copper stays too cold, the heat can't transfer to it.

So .. what is it... fast flow of water through the copper pipe - or slow?

 

[Aleman]

Depends on what you are trying to achieve

If you have no care about the amount of water you use . . . Run it as fast as possible

If you want to minimise water use then run it at a rate at which the temperature of the water exiting the chiller is just below the temperature of the wort.

Both methods should cool the wort at the same rate however

 

[Tony]

Excellent!
Many thanks Aleman .. and because I am a fine upstanding member of the community (global), I'll not waste water and run it a bit slower.

Thanks!

 

[aneray]

We pay around £700 per year for our water in the South West, and not being on a water meter excludes me from the flow rate.

 

[BigYin]

I use a medium flow, and I use the initial 'waste', which is quite hot, to fill the FV, mixed with a sanitiser, to get it ready to accept the cooled wort :thumb:

 

[drut]

The transfer of heat is dependent on three factors:
1. The difference in temperature
2. The thermal conductivity
3. The surface area

The thermal conductivity can be largely ignored here is it refers to the copper of the immersion cooler (but see b below).
So for maximum heat transfer you maximise the surface area having the greatest temperature difference.
i.e. you make the whole of the cooler coil as cold as possible - run at maximum flow so that the outlet is as cold as possible.

I believe the most efficient cooling is when the outlet temperature is the same as the wort temperature, therefore in the ideal case you would reduce the flow rate as the wort cools. This is the best compromise if you are on a meter or conscientious. Proving this requires calculus, and calculus can hurt.

Things to consider:
a. You might not be able to get sufficient flow to cool more than the first few feet of the coil, the rest is then wasted.
b. You should stir the wort to stop stratification and cooler volumes forming around the coil (ref. Aleman in a different thread).
c. Having more coils in parallel, rather than one long one, will improve usually cooling.

If you want some references, start here:
http://en.wikipedia.org/wiki/Heat_transfer

 

[Tony]

:thumb:
Nice ... thank you!

 

[Darkbrewer]

Great post Drut. I had been considering a bucket of ice in between tap and chiller to bring my temp down faster. But from the way the equations work out its way overkill since my well water is already at 45 degrees.

 

[timbim]

The transfer of heat is dependent on three factors:
1. The difference in temperature
2. The thermal conductivity
3. The surface area

The thermal conductivity can be largely ignored here is it refers to the copper of the immersion cooler (but see b below).
So for maximum heat transfer you maximise the surface area having the greatest temperature difference.
i.e. you make the whole of the cooler coil as cold as possible - run at maximum flow so that the outlet is as cold as possible.

I believe the most efficient cooling is when the outlet temperature is the same as the wort temperature, therefore in the ideal case you would reduce the flow rate as the wort cools. This is the best compromise if you are on a meter or conscientious. Proving this requires calculus, and calculus can hurt.

Things to consider:
a. You might not be able to get sufficient flow to cool more than the first few feet of the coil, the rest is then wasted.
b. You should stir the wort to stop stratification and cooler volumes forming around the coil (ref. Aleman in a different thread).
c. Having more coils in parallel, rather than one long one, will improve usually cooling.

If you want some references, start here:
http://en.wikipedia.org/wiki/Heat_transfer

Reading the start of the thread, I was thinking down the same lines as you. I think the calculus required to prove any result properly will be horrendous as there are two independent variables, time and distance along the coil. I don't think I have the techniques to model such a system, I can do varying over distance along a pipe, but only in the steady state, which is pretty rubbish in this case.

The fastest flow will always result in the fastest cooling, so Aleman's assertion early on is I think incorrect. To minimise water use, you would turn down the flow rate until the point at which the water leaving the coil is at the same temperature as what you're cooling. You then want to be turning it down as the cooling progreses. More efficient than that would be to use some sort of recirculated system, pumping the water through a radiator from the coil and back into the coil again. This could only bring it down to room temperature, and in practice would take a long time to get it there. You could also cool the water in the coil in an ice bath, that would result in a very rapid cooling, and it would be an interesting calculation to see if the cost of making the ice is less than the cost of the water going through the coil. I suspect it isn't, but I might have a go at working it out at some point.

Tim

 

[hopin mad]

If you are worried about your water meter,
pump water from a butt that collects rain water and pipe it back to the butt,
that way you dont waste water.

 

[commsbiff]

I was thinking of putting 4 coils of 8mm copper in parallel using 25m tube in total for my 100L boiler - overkill? :hmm:

 

[ano]

I'm sure I've seen a triple coil immersion chiller here or elsewhere (I don't think I dreamed that one) which was found to be no better than a double coil in practice at cooling 25 litres.

Also the smaller bore will (as I suspect you realise) increase surface area of a volume of water, but that will only mean it will lose its cooling capacity sooner, i.e. the runs will need to be shorter. This may come around to be useful again as you reach pitching point?

Further efficiencies could perhaps be found to disturb the laminar flow of the coolant but I wonder if we're expecting too much from our chillers, I mean let's not forget how long those elements had to be on to heat the lot up to that temperature!

The specific heat of water is 4.19 kJ/kg. If you have 25 litres of water (lets not forget the specific heat of wort will be higher) then the energy you're trying to dump as it moves from 80 C (after the hop steep say) to 20 C will be 4.19 x 25 x (80 - 20) = 6285 kJ. That's quite a lot.

 

[Cyclops]

Our fridge freezer makes ice, so in a way it doesn't cost anything to make for me, I think I will make a load, put it in bags and then keep some in the freezer so I can have a bucket full. Leave part of hose in a coil in this bucket and then into the immersion chiller.

 

[timbim]

The specific heat of water is 4.19 kJ/kg. If you have 25 litres of water (lets not forget the specific heat of wort will be higher) then the energy you're trying to dump as it moves from 80 C (after the hop steep say) to 20 C will be 4.19 x 25 x (80 - 20) = 6285 kJ. That's quite a lot.

Specific heat capacity!

Our fridge freezer makes ice, so in a way it doesn't cost anything to make for me, I think I will make a load, put it in bags and then keep some in the freezer so I can have a bucket full. Leave part of hose in a coil in this bucket and then into the immersion chiller.

There's the cost of the electricity to make the ice. Not sure how much it would cost, I'll work it out later. On my way out now.

Tim

 

[ano]

The specific heat of water is 4.19 kJ/kg. If you have 25 litres of water (lets not forget the specific heat of wort will be higher) then the energy you're trying to dump as it moves from 80 C (after the hop steep say) to 20 C will be 4.19 x 25 x (80 - 20) = 6285 kJ. That's quite a lot.

Specific heat capacity!

Don't normally use wikipedia but http://en.wikipedia.org/wiki/Specific_heat_capacity

"the specific heat capacity, often called simply specific heat"

Guess it depends where you were taught your thermodynamics?

 

[timbim]

I know they are used interchangeably, but that doesn't make it right. To me at least, specific heat implies the current state of the material (in J/Kg) rather than an intrinsic property of an (ideal) material (in J/(Kg.K)).

I can't stand the use of the form- prefix to mean one carbon, for some reason. Acet- I'm not too fond of but don't really mind, despite the fact that it's often more ambiguous - acetone for example. I'm a fully IUPAC chemist :thumb: (at least mostly, don't try me on the IUPAC names of sugars :hmm: )

 

[BigYin]

This is very interesting in a geeky sort of way :lol:

I make twin coil immerssion chillers, from 10mm copper tubing. I figured that the extra set of coils almost doubled the total immerssed surface area/internal volume (but not quite because of the smaller radius of the second, inner, set of coils) of the chiller.

They (simplified) way I see it, is the larger the volume of 'chilling water', the more capacity to take away heat, and the greater the surface area of the heat exchange, the faster that heat will be exchanged.

Copper is a great material for this, since it has very high heat conductivity, is easy to work with, and readily available (but alas, not very cheap) to buy.

The ready availablity of 10mm tubing and fittings for it makes it a good compromise. A smaller tube would have an even better surface area to internal volume ratio (I think!) and therefore acheive a faster heat transfer - but if the internal volume is too small, it will 'max out' on what amount of heat it can take away.

So you can increase the available internal volume by increasing the flow rate.

The conclusion I drew from that was the finer the heat transfer coils, the faster the flow rate should be :thumb:

 

[timbim]

The conclusion I drew from that was the finer the heat transfer coils, the faster the flow rate should be :thumb:

Probably not that simple. The flow rate should remain the same to achieve the same cooling (a finer bore tube will result in a slightly better performance) but the major issue is that a larger pressure is required to realise the same flow rate in in narrower tube. Really need someone with a slightly better understanding of the thermodynamics and fluid dynamics to get to the bottom of this entire debate, I'm a biologist, not an engineer!

 

[BigYin]

I'm a computer programmer, not an engineer either :lol:

 

[timbim]

I'm a computer programmer, not an engineer either :lol:

Software engineer :hmm: