DIY Idea/Theory - Counter Flow Double IMMERSION Chiller

Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum

Help Support Homebrew Talk - Beer, Wine, Mead, & Cider Brewing Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

jezter6

Supporting Member
HBT Supporter
Joined
Nov 1, 2006
Messages
4,287
Reaction score
16
Location
DARLINGTON
So, my buddy and I start a conversation about a possible DIY immersion chiller, and little as we know about thermal dynamics, it turned into a lot of "I wonder if..." questions.

Our 14.5g brewpot is custom made (by someone I bought it from used off of ebay a year ago). It's kind of large, circumference wise, and the little itty bitty 3/8" immersion chillers I see in the store don't look like they could handle our pot very well (small tubing, and smallish concentric rings), and they definately don't look like they could handle a full on 10g batch. We are thinking about doing a DIY, but had some questions for some people a lot smarter than us.

1. Our plan was to use 1/2 copper pipe. Does that really work better than the standard 3/8" stuff?

2. Our plan was to have 2 sets of rings, one larger ring to go closer to the sides of the brewpot, and another ring (of possibly smaller copper, maybe not) that resides INSIDE the larger outer ring. Does that sounds like it will cool wort faster than a single 50' of 3/8" pipe?

3. If we ran that water clockwise through the outer ring, and counter clockwise through the inner ring, would it make a difference?

4. If we ran water in the other direction (top to bottom, bottom to top) between the rings, would it make a difference?

Discuss... :)
 
jezter6 said:
So, my buddy and I start a conversation about a possible DIY immersion chiller, and little as we know about thermal dynamics, it turned into a lot of "I wonder if..." questions.

Our 14.5g brewpot is custom made (by someone I bought it from used off of ebay a year ago). It's kind of large, circumference wise, and the little itty bitty 3/8" immersion chillers I see in the store don't look like they could handle our pot very well (small tubing, and smallish concentric rings), and they definately don't look like they could handle a full on 10g batch. We are thinking about doing a DIY, but had some questions for some people a lot smarter than us.

1. Our plan was to use 1/2 copper pipe. Does that really work better than the standard 3/8" stuff?

Not really. Smaller tubing is better because you have more surface area in contact with the wort than you would if you had larger tubing. The more surface area you have in contact with the wort, the more efficient the heat transfer will be and the faster yoou will cool the wort.

jezter6 said:
2. Our plan was to have 2 sets of rings, one larger ring to go closer to the sides of the brewpot, and another ring (of possibly smaller copper, maybe not) that resides INSIDE the larger outer ring. Does that sounds like it will cool wort faster than a single 50' of 3/8" pipe?

That sounds like a good idea to me. I've been thinking of doing something similar myself.

jezter6 said:
3. If we ran that water clockwise through the outer ring, and counter clockwise through the inner ring, would it make a difference?

After rereading the post it occurred to me that I had completely misread what you had written. I don't think it would matter if the inner ring went one direction while the outer ring went another.

jezter6 said:
4. If we ran water in the other direction (top to bottom, bottom to top) between the rings, would it make a difference?

This doesn't sound like a very good system to me. I think it would actually take longer to cool the wort than a regular IC. You would end up cooling the wort on one pass then reheating it on another.
 
eddie said:
Not really. Smaller tubing is better because you have more surface area in contact with the wort than you would if you had larger tubing. The more surface area you have in contact with the wort, the more efficient the heat transfer will be and the faster yoou will cool the wort.
I'll buy that! We thought pushing more water through, with a big surface area of a higher diameter hose would work.

eddie said:
That sounds like a good idea to me. I've been thinking of doing something similar myself.
I figured it would be more efficient (but costly in material) to have 2 chillers at the same time.

eddie said:
After rereading the post it occurred to me that I had completely misread what you had written. I don't think it would matter if the inner ring went one direction while the outer ring went another.
My roommate theorized something about heating/cooling water causes some sort of current in the pot. Not sure why, but he figured the cold wort would somehow move away from the coil and draw the warmer wort in for some reason. Again, I dunno, it was a crazy (drunk) theory.

eddie said:
This doesn't sound like a very good system to me. I think it would actually take longer to cool the wort than a regular IC. You would end up cooling the wort on one pass then reheating it on another.

I guess I didn't explain my idea well enough. The water would be split at the water IN side, pushing half the water into the inner, and half the water to the outer. The pipes would then meet up near the water out side and have a single water in/out hose controlling what is now 2 separate chillers split from 1 water supply.

My theory is if you start cold water at the bottom and push UP the coils on the outer, the top portion of the kettle is not as cooled. If you, in turn, push the water top DOWN in the other coil, then the top would be cooled by 1 chiller, and the bottom cooled by the other.

Does it make sense now?
 
I've made a chiller like you're describing and it works fine but it does obstruct your ability to whirlpool the wort for faster cooling. I figured that the two diameter coils would more widely pull heat, but you still need the wort moving to kill the stratafication.

Cooling is all about the tap water temp. If you have warm tap say above 70F, a longer IC doesn't really help because at some point the cooling water will hit 80F and now you're subject to diminishing returns for every additional foot of copper. I'd recommend 25' of 1/2" and a pond/utility pump to pump icewater if your tap is any higher than 70F.
 
Tapwater here isn't too bad, as long as you run the hose for 30 seconds to get the topmost warmer water out of the way.

It's not about not being able to cool it. It's about cooling it as fast as we can that this idea sprang from. Unfortunately, I'm not rich, so buying a buttload of copper to 'test' my theory is kinda out of the question.

My kettle is ~15" is diameter, so I was thinking something around a 14" coil around the outer, and maybe an 8 inch coil for the inner, giving 6" of clearance for whirlpooling.
 
I might be missing something here, but I think you would have more surface area with a larger diameter tube. The surface area of a cylinder increases as the diameter increases. I have been thinking about making my own IC using 1/2" copper, but it is cheaper right now for me to buy one from Midwest with the way copper prices are here locally.
 
eddie said:
Not really. Smaller tubing is better because you have more surface area in contact with the wort than you would if you had larger tubing. The more surface area you have in contact with the wort, the more efficient the heat transfer will be and the faster you will cool the wort.

TWilson said:
I might be missing something here, but I think you would have more surface area with a larger diameter tube. The surface area of a cylinder increases as the diameter increases.

You're both sorta right. Given the same length tubing and same fluid velocity, as tubing size increases, the relative volume of cooling water that actually contacts the interior surface of the tubing decreases. However, the outer surface area exposed to the wort increases. So, cooling ability will tend to increase greatly with small increases in tubing diameter, but a large increase in tubing diameter may actually have the opposite effect.

In practical terms, the difference between 1/4" and 3/8" tubing is quite significant and usually results in better cooling rates. The difference between 3/8" and 1/2" tubing may not be so dramatic and is probably not worth the cost difference.
 
I made a dual-coil immersion chiller, as I found the thin fridge tubing dirt-cheap at the Depot (when they couldn't tell me how much it was, they sold it for less than half price). It works great, I'm chilled to the mid-70s in not much more than 10 minutes, which as far as I'm concerned is just about as good as it gets.
 
the_bird said:
I made a dual-coil immersion chiller, as I found the thin fridge tubing dirt-cheap at the Depot (when they couldn't tell me how much it was, they sold it for less than half price). It works great, I'm chilled to the mid-70s in not much more than 10 minutes, which as far as I'm concerned is just about as good as it gets.

Do you have pics?
 
I like Bird's approach. Cheap and quick -- always a winner.

That said, my experience has been that the most important factor to quick chilling w/ an immersion chiller is to keep the wort moving. If you don't do that, you'll probably be dissappointed (I was) and if you do, unless you do something really stupid w/ you chiller design, you'll probably be content.

I know this was mentioned before, just want to re-iterate the point.

I've seen where some folks recirculate the wort with a pump. I just give it a good stir w/ a spoon every couple of minutes.

But yeah, a double chiller w/ smaller diameter tubing will likely be more efficient.

-Tom
 
I am pretty sure smaller diameter piping would be less efficient. It's not just about surface area, it's about flow. By using a smaller diameter copper pipe, you are restricting the flow of water through your immersion chiller thus reducing efficiency.
 
It's an ugly thermodynamics problem that will vary by wort temp, pipe diameter, pipe length, pipe wall thickness, flow rate, and tap water temp. Since of these 5 variables, two are unknown, and one varies during the process, it is impossible to accurately speculate about what would be more or less efficient.

Since we can't model results with any reasonable accuracy, we must derive our conclusion from experimental data. And since no one has done a controlled test (afaik), that means your best bet is to use real world experience from HBT members. And IMHO, there is a definite bias that 1/2" is a waste of money.

That said, my unresearched bet would be on more coolant mass per unit time (1/2") pipe. However, since I made a 3/8" first and it works fine (5g) I doubt that I will make another.
 
eddie said:
Not really. Smaller tubing is better because you have more surface area in contact with the wort than you would if you had larger tubing. The more surface area you have in contact with the wort, the more efficient the heat transfer will be and the faster yoou will cool the wort.

Sorry, but that doesn't make sense. A larger tube (1/2" vs. 3/8") has MORE surface area that would come into contact with the hot wort, not less. Thusly, given equal volumes of water, a larger diameter pipe, would have greater cooling properties. See Yuri's post above.
 
srm775 said:
Sorry, but that doesn't make sense. A larger tube (1/2" vs. 3/8") has MORE surface area that would come into contact with the hot wort, not less. Thusly, given equal volumes of water, a larger diameter pipe, would have greater cooling properties. See Yuri's post above.

Generally, though, you're be using a different length of tubing (more) if you went with smaller diameter. 25' of 1/2" of course has more surface area than 25' of 3/8" - but most people would be making a decision to either go for 25' of 1/2" or something like 40' of 3/8". Same total amount of copper, we all usually have a budget in mind when doing a project like this.
 
FWIW, 45ft of 3/8 is almost the same surface area as 25ft of 1/2"

Code:
Diameter	25	30	35	40	45	50
3/8		33.13	39.76	46.39	53.01	[COLOR="BLUE"]59.64[/COLOR]	66.27
1/2		[COLOR="BLUE"]58.90[/COLOR]	70.69	82.47	94.25	106.03	117.81
 
Hey, I was just guessing!

Has the thickness of the material been discussed? Lower thickness ought to result in greater heat transfer. Gotta believe the tubing with the smaller diameter will also be thinner. If my assumptions are right, you'd be better-off with the 45' of thinner material.
 
Total surface area is only one variable. You can use 300 feet of 1/2 but the truth is, the water inside will likely hit 5-10F below the ambient wort temp within the first 20 feet (just a wild guess). As I mentioned before, you're down to the law of diminishing returns. The lower your tap water temp (high differential between the water and wort), the more use you'll get out of a longer run.
 
I've got a ball value hooked up to the start of my chiller so I can control the water flow, so as to not be wasteful. Run it fast at the beginning to get as much heat away as quickly as possible, slow it down as the wort chills and the output is coming out cold. The fastest cooling would be to run it full-bore all the time, but I hate to see that much water running off (more than I need for cleaning, etc).
 
pldoolittle said:
FWIW, 45ft of 3/8 is almost the same surface area as 25ft of 1/2"

Code:
Diameter	25	30	35	40	45	50
3/8		33.13	39.76	46.39	53.01	[COLOR="BLUE"]59.64[/COLOR]	66.27
1/2		[COLOR="BLUE"]58.90[/COLOR]	70.69	82.47	94.25	106.03	117.81


I hope you don't mind me stealing this for a bit.

What this means is that the smaller diameter you use, the more tubing you can get into the pot and the more efficient your chiller should be. With this in mind, you must be careful not to go too small or too large with tubing you choose to use. If your tubing is too small, the coolant flowing through it will become saturated early and will fail to transfer heat effectively which will increase your time spent cooling unless you greatly increase your flow rate. If you go too large with your tubing, the coolant will be underutilized and won't transfer as much heat energy as it's potential allows. To compensate for this, you would have to slow your flow rate down so that it can become saturated before being expelled from the chiller which will also increase the length of time it takes to chill your wort. Using 3/8 tubing is a compromise between the two. It allows you to maintain an easily manageable flow rate while allowing the coolant to be fully utilized in the chiller.
 
eddie said:
I hope you don't mind me stealing this for a bit.

What this means is that the smaller diameter you use, the more tubing you can get into the pot and the more efficient your chiller should be. With this in mind, you must be careful not to go too small or too large with tubing you choose to use. If your tubing is too small, the coolant flowing through it will become saturated early and will fail to transfer heat effectively which will increase your time spent cooling unless you greatly increase your flow rate. If you go too large with your tubing, the coolant will be underutilized and won't transfer as much heat energy as it's potential allows. To compensate for this, you would have to slow your flow rate down so that it can become saturated before being expelled from the chiller which will also increase the length of time it takes to chill your wort. Using 3/8 tubing is a compromise between the two. It allows you to maintain an easily manageable flow rate while allowing the coolant to be fully utilized in the chiller.

I don't understand the concept of lowering the flow of the wort chiller so that the water going through the chiller can be saturated (with heat?). I would think an "underutilized system" (very large tubing) would still be more efficient as far as cooling goes compared to a smaller size tubing, but it would just be extremely inefficient in the amount of water required (and thus be a waste of water and unrealistic).
 
It just depends on which efficiency you're most concerned with... time or water. For the latter, you don't want water coming out of your chiller until it's damn near wort temp. However, if it is, you can be sure that the second half of the coil's length isn't really pulling all that much heat anymore because the delta is too low. Want fast cooling? full flow. Hug trees? slow flow.
 
iamjonsharp said:
I don't understand the concept of lowering the flow of the wort chiller so that the water going through the chiller can be saturated (with heat?). I would think an "underutilized system" (very large tubing) would still be more efficient as far as cooling goes compared to a smaller size tubing, but it would just be extremely inefficient in the amount of water required (and thus be a waste of water and unrealistic).

Someone has already mentioned water usage. Outside that, underutilized coolant is only an issue in a closed/split system like an AC unit. Underutilized coolant is an open system is just not an issue at our flow rates. That said, the highest heat draw occurs when the delta between the water temp in the tube and the pot is greatest.

That said, maximize surface area and flow as much as your system will allow.
 
Table with surface area above looks incorrect - it is for the volume, not surface area. d*pi*l is surface area, so at a given length, it should be a difference of .5/.375 , or 1.333 between the two lines - 25' of 1/2" = 33.3' of 3/8".

If you are comparing the same amount of copper (1/2" 25' vs. 3/8" 33.3'), my guess is that the 3/8" would be better. The 1/2" would hold 1.333 more volume (more fluid to transfer heat to) but the flow rate would be faster in the 3/8" tube - I don't have my fluid mechanics book nearby, but I think it is a wash. So, the smaller diameter would cause a more turbulent flow which will help cooling faster, assuming you dont get to equilibrium before the end of the chiller, and therefore would be the better choice.
 
Bearcat Brewmeister said:
Table with surface area above looks incorrect - it is for the volume, not surface area. d*pi*l is surface area, so at a given length, it should be a difference of .5/.375 , or 1.333 between the two lines - 25' of 1/2" = 33.3' of 3/8".

If you are comparing the same amount of copper (1/2" 25' vs. 3/8" 33.3'), my guess is that the 3/8" would be better. The 1/2" would hold 1.333 more volume (more fluid to transfer heat to) but the flow rate would be faster in the 3/8" tube - I don't have my fluid mechanics book nearby, but I think it is a wash. So, the smaller diameter would cause a more turbulent flow which will help cooling faster, assuming you dont get to equilibrium before the end of the chiller, and therefore would be the better choice.

The flow rate will decrease as pipe diameter decreases--your water pressure coming from your spigot is a constant. Think about it in term of water piping out of a water tower. A 2 inch pipe would provide a larger passage for water to escape than a 1 inch pipe. Or think of it in terms of a culvert under a roadway. If you have 20 acres of land draining to a pipe, it will need to bigger than a pipe with 5 acres of land draining it. A smaller pipe will restrict the flow.
 
iamjonsharp said:
The flow rate will decrease as pipe diameter decreases...

It's been a while since I took physics, but something in this doesn't make sense to me.

If the backpressure is constant (I agree with you there) then a decrease in pipe diameter (cross-sectional surface area) should result in an increase in fluid velocity and maintain the same flow rate.

There are other things to take into consideration, such as friction applied by the reduction to the copper tubing and friction of traveling it's entire length, the compressible qualities of water, etc. ... but I'm not sure they would account for an appreciable difference in this application.

Bernoulli's Principle:
http://en.wikipedia.org/wiki/Bernoulli%27s_principle


But it is 1am, so the chances of me being correct aren't as favorable as they could be ;)
 
I got a simple 25' immersion wort chiller from my LHBS for my system. The sweet thing about my system is I can recirculate the cooling wort from my brew kettle back to my brew kettle splashing hot wort directly on my submerged immersion wort chiller. It does still take me a good 15-20 min to chill my wort to 70*F but I get the same time stirring... this way there is ZERO STIRRING!!! and when I'm ready to pump it to my primary it's ready to do so! ...ugg, wish there was a high temp self priming pump... =( YAY!!! :ban: I have a spare immersion wort chiller I need to get around to putting in a bucket with lots of ice and salt water in to pre chill my water. THEN it'll drop the time more, and in the winter when the water drops 10 or so degrees it'll help more too.
 
D2T said:
There are other things to take into consideration, such as friction applied by the reduction to the copper tubing and friction of traveling it's entire length, the compressible qualities of water, etc. ... but I'm not sure they would account for an appreciable difference in this application.

You cannot assume continuity...headloss with pipes this small is HUGE. Using the Darcy-Weisbach formula and using the Blasius equation to get the frictional factor for smooth pipes with turbulent flow:

5958-headloss2.jpg


Granted this method may be beyond its limit to get very accurate headlosses, but you can get an overall sense of the trend: the smaller pipe diameter you use will increase the headloss dramatically which will restrict the flow of water through the pipe. Using 1/4" instead of 3/8" copper piping will make a huge difference in flow. There should be a noticeable effect in using 1/2" instead of 3/8" as well: the inner cross sectional area of the 1/2" OD pipe is about double the size compared to the 3/8" OD pipe.
 
if you really want to drop your cooling times, make sure you keep the wort moving over whatever chiller you decide upon, I now recirculate the wort in the kettle over my immerision chiller (which is a 25ftx2 chiller like you described) in a keggle and it really dropped my cooling times
 
jezter6 said:
My roommate theorized something about heating/cooling water causes some sort of current in the pot. Not sure why, but he figured the cold wort would somehow move away from the coil and draw the warmer wort in for some reason. Again, I dunno, it was a crazy (drunk) theory.


You'd be amazed how fast water convects heat. Yes, in theory the wort closer to the coil will be cooler but only very, very slightly. A cool excersize to show your roomate how fast water convects heat... Take a small paper cup (like the 2" tall desposible ones for the bathroom) put about 1/2" of water in it. Now, take a lighter and hold it lit right under the cup. It will convect the heat from the bottom and disperse it through the water so fast the cup will boil before the paper sets on fire. (just avoid the lip edge on the bottom with the lighter, it's all paper and will burn fast)
 
jezter6 said:
My roommate theorized something about heating/cooling water causes some sort of current in the pot. Not sure why, but he figured the cold wort would somehow move away from the coil and draw the warmer wort in for some reason. Again, I dunno, it was a crazy (drunk) theory.

Convective currents would be established, causing the cooler fluid to tend downward. But, that effect is very small compared to the reduction in thermal movement caused by the boundary layer of cooler wort near the pipe.

What that means is that (as said before), you maximize efficiency by keeping the wort in constant circulation around the coils.

Now, if you want to get super geeky, starting the cool water into the coil at the top, and removing the hotter fluid at the bottom, will yield a tiny bit of improvement. But only if you aren't stirring the wort....
 
davefleck said:
You'd be amazed how fast water convects heat. Yes, in theory the wort closer to the coil will be cooler but only very, very slightly.

Actually it's in theory and in practice. Test it for yourself like this; Leave an IC in an undisturbed pot of wort and run it for 1 minute. Test the temp of the IC output water or even just feel it yourself. Now, have someone stir the wort while you're measuring the output water. You get a HUGE jump in temp. Temperature stratification is pretty common. Ever swim in a lake?
 
I've played around with the wort chiller, stirring and not, leaving the flow constant. I've got a probe thermometer that stays in the wort, so I'm always getting a current reading. When gently stirring, I can see a stready, pretty rapid decline (it chills to pitching temp in about ten, maybe twelve minutes). When I stop stirring, it stops cooling - the rate of temperature decline is less than half what it was previously, probably significantly less than half.
 
davefleck said:
You'd be amazed how fast water convects heat. Yes, in theory the wort closer to the coil will be cooler but only very, very slightly.

Actually, the wort closer to the coil will be significantly cooler. As noted in the posts above, the cooling is so significant that further cooling will stop unless something occurs that disturbs that gradient.
 
the_bird said:
I made a dual-coil immersion chiller, as I found the thin fridge tubing dirt-cheap at the Depot (when they couldn't tell me how much it was, they sold it for less than half price).

i went to lowes to get my own, but found 3/8" x 25' roll at $60. how come you can buy a completely made chiller for less than that?
bill keiser
 
sharpstick said:
i went to lowes to get my own, but found 3/8" x 25' roll at $60. how come you can buy a completely made chiller for less than that?
bill keiser
Ouch -- I got mine for $20 CDN. I also got 100' of 5/8 copper tubing for $90. Maybe Canada needs to export some copper! ;)
 
hook it up, send it south

if you ship it to the states, put on the customs form "bong materials" that should get some laughs, hehe
 
sharpstick said:
i went to lowes to get my own, but found 3/8" x 25' roll at $60. how come you can buy a completely made chiller for less than that?
bill keiser

Home Depot - at least in my city - overcharges something horrific. I went to a more-local franchise (Menards if anyone has one) and their copper prices are HALF of HD's

I need to drive south to check the Lowe's they just opened.
 
Back
Top