Immersion Chillers

[tampa911]

So I am sure that this topic has been beaten to death, and I apologize if I am just digging up an old topic for the millionth time, but I got to thinking about building my immersion chiller.

Thinking leads to drinking and drinking leads to the inevitable… Math

I have seen a good deal of debate regarding 1/2 inch vs 3/8 inch copper but really no discussion of 1/4 inch. From the rudimentary calculations I put together, 100 ft of 1/4 inch copper pipe gives you the same surface area as 50 feet of ½ inch with less displacement than 50 feet of 3/8.

50 ft of 3/8 copper has a total surface area of 4.91 Sq. Feet with a displacement of roughly .29 gallons
50 ft of ½ copper has a total surface area of 6.55 Sq. Feet with a displacement of roughly .51 gallons
100 ft of ¼ copper has the same surface area as 50 ft of ½” but with a displacement of only .255 gallons
(figures courtesy of http://www.aqua-calc.com/calculate/volume-cylinder-hollow)

Now logic would dictate that 1/4 inch pipe is going to have a higher head pressure (resistance) and therefore a lower flow rate than a wider pipe meaning more contact time of the water flowing through the pipe with the actual copper. Unless the water coming out of your chiller is raised to the temperature of the wort any additional contact time would lead to faster cooling. (how much faster is the real question)

I also did some rough calculations to see if you could fashion a coil of 100 feet of ¼ inch pipe to fit in an average brew pot. I use an 20 qt. pot with an inner diameter of roughly 12 inches. I believe that if you did 2 concentric coils you could fit the copper with no issues.

I brew roughly 3 gallon extracts currently which is 6.5 inches of liquid in my pot. 20 wraps of coil with an 11 inch diameter use roughly 60 feet of tubing (5 inches tall if stacked tightly or leaving 1/16 inch between the coils and not exceeding 6.5 inches). 20 10 inche coils inside the outer coils would use up an additional 52 feet meaning you would have room to tweak these dimensions significantly and still have everything fit. These measurements would leave a half inch of space all the way around the outer coil and pot and a half inch between the inner and outer coil with all of the copper submerged below the level of the 3 gallons of wort (before you even think about displacement).

Obviously you are going to spend more for 100 feet of 1/4 inch copper tubing, and will have more work creating your coil(s), but as I stated before I am a new member and new to home brewing, but I love theoretical discussion and experimentation, so I just figured I would throw this out for some of the vets to shoot down.

 

[Spintab]

If you're brewing extract, why not just make some "sterile" ice by boiling water and freezing it the night before then dumping the hot wort right on it? You'd eliminate the need for the chiller all together.

If not, 50ft of any size is kind of a lot in 3gal. I mean, it'd work for sure but usually people use that much in almost twice the wort. While it seems like slower flow rates through the chiller would work, you really want to just get the heat out of there as quick as possible. I'd have to guess most will advocate as fast of flow as possible. I personally use 2x20' concentric 3/8 coils for 5gal batches. I run about 5gal straight through into a bucket which I use for cleaning water later (nice and hot). Then I fill a 5gal bucket with ice and a little water and circulate through the chiller with a hearty pond pump. Usually takes about 15-20 minutes with some swirling of the chiller throughout. I've managed to get my water usage, for chilling, down to about 10gal including the melted 20lb bag of ice.

 

[tampa911]

I certianly won't be doing extract brews forever, so I would want whatver solution I devise now to hold up in the future. This was really more of a theoritical thought than practical application. I wanted to see if anyone had attempted 1/4 inch and if so where the pitfalls were in my theory.

In regards to your comments on flow rates, I have to disagree. If you are moving 10 gallons / minute of liquid through your chiller but the temp. difference between the wort and the exhaust water is 40 or 50 degrees or you are moving 1 gallon / minute with a difference of only a few degrees (given the same starting temp. for the wort and tap water) you are going to cool much faster with a longer contact time and slower flow rate. In the first minute you might see a greater percentage drop with a faster flow rate, but the overall time to cool a volume of liquid will be faster with a slower flow rate if the water moving through the chiller never gets as warm as the liquid it is cooling. (which it might with 1/4 inch I don't know....)

Think about it this way, if you had a red hot iron and dipped it in a bucket of ice water for a second, or took that same red hot iron and held it in a bucket of room temp water for a few seconds which would be cooler? It is kind of a reverse look at what I am stating, but the more time your fresh water stays in contact with the wort (through the coper tubing) the more heat it is going to draw from the liquid.

I thought of a better analogy after I wrote this... If you took a 20 lb bag of ice and dropped it in a 5 gallon bucket of water for a minute, or a 5 lb bag of ice in a bucket for 4 minutes which would be cooler? Volume vs contact time with all other variables the same. If the 5 pound bag melts completely in 2 minutes it is a moot point.

 

[Spintab]

Water will only give up it's heat so fast. A greater difference in temperature will cause this to happen faster, at a logarithmic rate. The cooler the liquid in the chiller is versus the wort, the quicker the water will give up its heat. Follow a single molecule from the in of the chiller to the out. If it is moving slow, it will be hotter by the time it exits the chiller but it also would barely be taking on any more heat. If the water is moving faster, it won't be as hot coming out but it will be maximizing the heat transfer from the wort throughout the length of the chiller. Granted you would end up using way more water unless you rechilled and recirculated it, but it would happen faster. The idea is to get the hot molecule out and away from the heat as fast as possible so others can take it's place. In your iron example, the thermal transfer in that initial second would outweight the few seconds following. Imagine taking three hot irons and dunking each in the water for a second versus only one for the entire three seconds. The total drop in temperature of the three irons would be significantly higher than the single iron.

 

[Spintab]

Check this out. I'm not trying to beat any horses, I just found this fascinating the first time I saw it happening in real time. I have probes in both the wort (blue) and fridge air (red). The purple shows when the fridge was on and off. You can see when the fridge shuts off, after it cycles, it quickly starts raising back but tapers off at a curve. Eventually, it will take a significant amount of time to raise even a hundredth of a degree. This same thing would happen on the way down and hit the max cooling of the fridge. Point being, the more often you are cooling on the "fast" side of the curve, the faster cooling will happen. This all at the expense of added energy (energy to previously chill water or freeze ice).

 

[TheZer]

but the overall time to cool a volume of liquid will be faster with a slower flow rate if the water moving through the chiller never gets as warm as the liquid it is cooling. (which it might with 1/4 inch I don't know....)

I think of it this way (of course numbers aren't actual):

At t = 1
With 100' of 1/4" copper, the water inside the tubing will be the same as the wort in 10 feet. Thus, the other 90 feet is doing nothing to reduce temp. This is wasted surface area.
With 100' of 3/8" copper, the water inside the tubing will be the same as the wort in 20 feet.
With 100' 1/2" copper, the water inside the tubing will be the same as the wort in 40 feet.

It is all about flow rate. Faster is better. The greater the T differential, the faster the heat can be pulled from the wort. And your faucet can only deliver so much through a 1/4" tube.

I would equate 100' of 1/4" to your 5 lb bag of ice melting in 2 minutes. Now, five 20' 1/4" ICs could do some serious cooling.

 

[PosterGuy]

"50 ft of 3/8 copper has a total surface area of 4.91 Sq. Feet with a displacement of roughly .29 gallons
50 ft of ½ copper has a total surface area of 6.55 Sq. Feet with a displacement of roughly .51 gallons
100 ft of ¼ copper has the same surface area as 50 ft of ½” but with a displacement of only .255 gallons
(figures courtesy of http://www.aqua-calc.com/calculate/volume-cylinder-hollow)"


How about using 200 feet of 1/4" copper as four 50 ft or even eight 25ft sections connected to a manifold at both ends so they can connect to 1/2" or 3/4" inlet and outlet tubes?

Assuming your displacement numbers are correct you'll get twice the surface area as 50ft of 1/2" copper with about the same displacement.

This might lead to higher efficiency.
Though I don't think I want to assemble the thing to find out.

 

[cupido76]

Check this out. I'm not trying to beat any horses, I just found this fascinating the first time I saw it happening in real time. I have probes in both the wort (blue) and fridge air (red). The purple shows when the fridge was on and off. You can see when the fridge shuts off, after it cycles, it quickly starts raising back but tapers off at a curve. Eventually, it will take a significant amount of time to raise even a hundredth of a degree. This same thing would happen on the way down and hit the max cooling of the fridge. Point being, the more often you are cooling on the "fast" side of the curve, the faster cooling will happen. This all at the expense of added energy (energy to previously chill water or freeze ice).

The graph on this post is great, and I'm going to refer another thread where the OP is having temperature control issues in his fermenter.

Which probe is your control connected to... liquid or air?

 

[cupido76]

Back on topic: I agree with most of the posters here that say you want that hot liquid out of the chiller as fast as possible.

When you put the chiller in and you've extracted almost all the heat you can in the first few feet of the chiller, the sooner you can replace that hot liquid with cold liquid, the better off you'll be. I.e. the shorter chiller with equal surface area should win.

Posterguy's idea is also interesting. Taken to the extreme, it might be best to have as many short sections as possible (i.e. 100 feet in 4 x 25 foot sections). But it would probably be a lot of work to make something so complicated.

 

[Conan]

While the math is interesting, I feel like it'd be more practical to just build a CFC- especially considering the cost and time involved with 100+' of copper.

What is the actual gain in cooling time you're anticipating with this much copper? My 3 gallon batches cool in 15 minutes with a 25'x.5" chiller, 25% of which is above wort level. I haven't measured outflow temps but it's too hot to touch, while inflow varies from 40-60, depending on the season.

I don't think there's any reason not to use your proposed dimensions, it just seems to be a lot more effort for the same result as a 'standard' chiller. Are you concerned about displacement in your current BK setup? Kyle

 

[tampa911]

Displacement isn't a huge concern, I have a 5 gallon stainless kettle and did a 3 gallon boil for my last extract leaving at a minimum 2 gallons of free space for displacemnt, I just wanted to ensure that you could logistically fit that much copper in a brew kettle before I posed the question.

My next boil will probably be 2.5 gallons as I was furiously fighting boil over with every addition on my first brew.

 

[Ryfi]

I use 1/4" copper. I cut a 50' length in to four equal lengths so i have four separate coils running in parallel. I can cool 5 gallons with approx 13 gallons of 50 degree water. Idk if that helps...

 

[Spintab]

The graph on this post is great, and I'm going to refer another thread where the OP is having temperature control issues in his fermenter.

Which probe is your control connected to... liquid or air?

Thanks. In this particular snippet, the control is the wort. I've gone back and forth between using that and the air. Using the air I can get the wort temp to stay more flat but I have to babysit it over time. Ultimately I've moved to using the wort, sacrificing half a degree in variance in order to set it and forget it. If you want to leave a link to that thread, I can jump in there rather than getting too far off topic here.