Wort Chiller - How many feet start to be counter-productive?

[brettwasbtd]

I am looking to build an immersion wort chiller in the near future. I can't decide if I should go around 25' or 50' but I got to thinking... at what point (feet wise) does the initial stream of water equal the wort temperature? If you have boiling wort, and then start adding cold water, when the water exits the tubing is it the same temp (as the wort)? I am curious to see if anyone has experience with this. Usually I would think more is better, but not if the chilled water equals the wort water around 30 feet, then you are just feeding 20 feet of the same temp water around...

 

[slowbie]

There have been some threads about this in the past, but it is very dependent on wort temperature and water temperature. A 50' long chiller is not going to get your wort from 210 to 190 faster than a 25' chiller. It's when the wort temperature is closer to the water temperature that the longer chiller will be better.

IMO (even though it's really more of a guess than an opinion because there is a right answer to this one), if you buy 50' of copper tubing, you're probably better off making a 10' prechiller and 40' immersion chiller than you are making a 50' immersion chiller.

 

[mightynintendo]

I think I have a 25', 0.5" diameter copper chiller. It gets 6 gallons from boiling down to 70 in about 10 minutes or so. I never timed it but it's not long.

 

[brettwasbtd]

Ya, I am pretty confident that a decent chiller of 3/8" or 1/2" copper will get it cooled in under 25 mins, but I mainly am curious to see what length you reach a point where the water is just going through the coil and not being helpful... but I do see what your saying as far as the length won't matter at the start, but will later, I guess longer is better as you get the temperatures closer together.

 

[sonetlumiere85]

I would imagine it has a lot to do with the wort covering the copper or not. If the copper is simply sitting above liquid level, not being exposed to anything, it probably won't be as efficient as submerged copper.

 

[chrispykid]

I would imagine it has a lot to do with the wort covering the copper or not. If the copper is simply sitting above liquid level, not being exposed to anything, it probably won't be as efficient as submerged copper.

+1 It's all about the surface area of the chiller touching the wort. So the optimum length will be dictated by both your batch size and the geometry of your pot/chiller.

 

[brettwasbtd]

Maybe I didn't paint the right picture here, let me try again

Say your wort is 200 degrees, and your water that runs through the immersion chiller is 70 degrees. So for this arguments sake, say after the 70 degree water travels 30 feet down your coil, it is now the same temp as the wort (200 degrees). So the remaining 20 feet is pointless. BUT, as I type this example and as slowbie pointed out, the length won't make a difference for the first 20-40 degrees or so. The length will come into play when the wort gets down to a temperature that allows our 70 degree water to exit the chiller at a lower temp than the wort... so the Chiller Exit Temp < Wort Temp. At that point, the longer length will speed up the cooling process.

So i guess I have rationalized may way to the conclusion that more length is better? Anyone able to play devils advocate on this, I am having trouble figure out that side.

 

[Walker]

well... it'll never be COUNTER productive (meaning, it'll never HEAT your wort), so I'd say buy as much as you can afford without buying so much that it won't fit in your pot, under the wort.


You might be able to afford 100', but if you can't form it into a coil that can be submerged then you have definitely gone overboard.

edit: +1 on the idea of using some length to make a pre-chiller and then the rest for the basic IC.

 

[slowbie]

Maybe I didn't paint the right picture here, let me try again

Say your wort is 200 degrees, and your water that runs through the immersion chiller is 70 degrees. So for this arguments sake, say after the 70 degree water travels 30 feet down your coil, it is now the same temp as the wort (200 degrees). So the remaining 20 feet is pointless. BUT, as I type this example and as slowbie pointed out, the length won't make a difference for the first 20-40 degrees or so. The length will come into play when the wort gets down to a temperature that allows our 70 degree water to exit the chiller at a lower temp than the wort... so the Chiller Exit Temp < Wort Temp. At that point, the longer length will speed up the cooling process.

So i guess I have rationalized may way to the conclusion that more length is better? Anyone able to play devils advocate on this, I am having trouble figure out that side.

Well, let's put it this way: more length is never worse if we're looking at this strictly from a chilling time perspective. However, I do partial boils, and with the (borrowed) 25' wort chiller that I use, it takes roughly 10 minutes to chill. Would getting a 50' chiller speed that up by a minute or two? Perhaps, but that amount of time isn't worth the cash for me. However, if (when) I go up to full boils I'll be making my own chiller that will be longer than 25'. EDIT: AKA what walker said.

If your water is near 70 degrees I would like to re-highlight my prechiller suggestion. Although it looks like you might want to use at least 15 feet for it rather than the 10 I said earlier. Colder water will make your chiller work faster even during the times you're not utilizing its full length. A quick forum search will dig up some useful information on prechillers.

 

[l1ranger]

unless your running the water through very slowly, the water is not going to reach wort temps before it goes all the way through.
even just pumpin 1 gal/min through a 1/2" chiller is gonna get the water through a 50' chiller in about 30 seconds, and i don't think thats enough time with the water moving through the chiller to bring the cool water up to an even temp with the wort...not even close

edit - at no length will it become counter productive. but considering diminishing returns, there's a point where the price you pay is only going to save you a minimal amount of time in chilling and MAY not be worth it to you. and was noted before, if it doesn't fit in the wort, its not chilling and is too much line.

 

[slowbie]

unless your running the water through very slowly, the water is not going to reach wort temps before it goes all the way through.
even just pumpin 1 gal/min through a 1/2" chiller is gonna get the water through a 50' chiller in about 30 seconds, and i don't think thats enough time with the water moving through the chiller to bring the cool water up to an even temp with the wort...not even close

I have burned my hand on water coming out of my 25' chiller. There's another thread where people more ambitious than myself have done calculations and found that at 130 degree wort temp the water reached that in about 8 feet in at 3/8" diameter chiller. I'm not sure what starting water temperature he used. Copper is an excellent conductor of heat. However, if you have different calculations to share please do. As I said earlier, At some point later in the chilling you will be using the whole length of your chiller.

 

[Malticulous]

I used 20 ft of 3/8 ID tubing for some time. It worked alright for 10 gallons in the winter, but it the summer my watter is 68-70F so it took over an hour. I have to finish cooling in my fermentation freezer.

A while ago I made another chiller out of 50 ft of 3/8 OD tubbing. My 20 ft one fit inside of it. I was surprised to see my thermometer count down so fast. That was with 60F tap watter. I'm going to use the 20 ft ID chiller as a pre-chiller this summer.

 

[rico567]

My entry-level Midwest 25' copper chiller works just fine for me.....but we have 55F well water, which matters a great deal. I'm done in under 20 minutes, which is okey-doke with me. Most of the important points about chilling have been covered in this thread, one way or another. It does surprise me that as many homebrewers as there are out there with OCD (not me!), someone hasn't come up with a nomograph relating chiller material / surface area / flow rate / temperature of coolant to the time needed to go from boiling to 70F.

 

[Bmorebrew]

I have two 20' long 3/8" chillers - one is the prechiller, the other goes in the wort. I put the prechiller into my clean MLT, threw in all my ice packs and filled it with cold water. The other went into the wort obviously, and I just hit 75° in about 10 min tonight.

If we're looking for an exercise in math, I'm game - calculating deltas in the area of circles vs. circumference as diameter increases is always fun. But if you're looking for maximizing cooling with minimal chiller distance, the prechiller works awesomely.

 

[StAnthonyB]

There have been some threads about this in the past, but it is very dependent on wort temperature and water temperature. A 50' long chiller is not going to get your wort from 210 to 190 faster than a 25' chiller. It's when the wort temperature is closer to the water temperature that the longer chiller will be better.

IMO (even though it's really more of a guess than an opinion because there is a right answer to this one), if you buy 50' of copper tubing, you're probably better off making a 10' prechiller and 40' immersion chiller than you are making a 50' immersion chiller.

The more surface area the better. It's a quadratic relationship. The flow of the water on the inside of the pipe is logarithmic (I think?). That's laminar flow territory... I really trying hard to reminder my days as a mechanical engineering student.... So I'd make two coils from the 50' at 25' a piece as two wort chillers where each chiller is receiving the lower temperature water is better than 75' of copper receiving only a singular injection.

In a nutshell, two immersion-style wort chillers, a gardenhose Y-splitter, and wort pump to cycle to wort into a whirlpool.


See this is why I never graduated. But, that laminar flow thing tells us that sometimes it is more efficient to slow the water rate down so that the wort can become uniform temperature throughout the entire vessel.

 

[Malticulous]

Sounds cool. Parallel chillers seems like a good idea. I may try it.

 

[gxm]

Before I switched to no-chill, I used a dual copper chiller.
My 25' 3/8" ID copper would cool off a 5g pot in 10 minutes during the winter with 40F water. However, in the summer, the water was 65F and chilling could take 40 minutes.
I had 20' of 1/4" ID copper tubing lying around, so I used that in parallel with valves to control flow to each tube. The second coil made a significant difference, being more efficient when the water temp was close to the wort temp. If I recall correctly, most worts would get to pitching temp in about 25 min even with the warmer summer water.

 

[PorterIdontknowher]

well... it'll never be COUNTER productive (meaning, it'll never HEAT your wort), so I'd say buy as much as you can afford without buying so much that it won't fit in your pot, under the wort.
.

This doesn't apply to any practical applications of wort chillers, but as an engineer, when someone says "never" I start thinking of counter arguments...

While the coil may never heat the wort, theoretically you could reach a point where adding length to the chiller could theoretically make your cooling times longer. So far I've come up with 3 reasons while adding more tubing to an immersion chiller could lengthen the chilling time:

All of them assume that the tubing is long enough that the cooling water is already very close to the wort temperature when it is exiting the chiller. This takes either a really long chiller or a really slow flow rate of cooling water.

1. The chiller is placed in the boiling wort to sanitize it. Adding more tubing adds more copper... which increases the mass of material that you need to cool down. (you are removing heat from the copper as well) This doesn't happen if you sanitize the chiller and put it in at room temperature

2. As you add length to the chiller, you add pressure drop. Given that most water supplies have a fixed output pressure at the tap, if you make it longer, you slow down the flow rate of cooling water.

3. Here's the most complex one: If the tubing is too long it could put heat that it has already taken out of the wort back into the wort. If the cooling water is not well mixed (there is temperature variation along the length of the chiller) and the wort is well mixed so that it is all the same temperature. If the water entering the chiller pulls enough heat out of the wort to lower it below the temperature of the water near the exit of the chiller, heat transfer could start going the opposite direction...

All of these would likely require an insanely long chiller to take effect, but this is why you never say never...

 

[Walker]

I'm also an engineer, and there are times when I am perfectly comfortable saying "never".

Anyway, I would say that what you are describing is having a "sub-optimal" chiller, not a "counter-productive" chiller.

 

[viking73]

So basically what our engineer friends are trying to say is keep your input water significantly cooler than the final temp of your wort. So if you can prechill your input water to say 40 degrees, your chiller will still be effective over the length of the tubing.

Could just spring for a counter-flow and the whole thing is moot...

But then I'm low-tech and still put the pot of wort in an ice bath!

 

[Bobby_M]

The longer chiller will allow you to run the water faster and still maintain a decent cooling efficiency.

If you ran the water at a near trickle, it would reach equilibrium with the wort in a short tube distance. Any additional length is just a waste of copper. As you slowly increase the flow rate, that equilibrium will be reached further and further down the tubing until it gets to the very end. At that point, increasing the flow is just a waste of water.

What's the magic length and flow rate? I don't know but 50 feet of 5/8" is still short enough that full hose pressure doesn't have the wort coming out at wort temp.

For what it's worth, I think 50' of 1/2" is cheap enough and it's adequate for 10 gallon batches. If I went down to 3/8" tubing, I'd want to split it into two 25 foot coils.

 

[carp]

A significant parameter in determination of optimum length is whether or not the wort is being stirred or whirlpooled. I made a tuning-fork shaped contraption out of copper that I put in my drill and use as a stirrer, others here have used paint stirrers (I'm too cheap for the $40 SS one, and didn't want to use zinc) and stirring makes a huge huge difference in time. Heck even just moving the chiller around makes a big difference.
I would expect that one sees the benefits of a longer chiller if the wort is being aggressively stirred or otherwise agitated.

 

[Bmorebrew]

The most important aspect of all of this is the Delta T. Heat always moves from areas of greater concentration to areas of lower concentration. Having two coils immersed vs. one coil immersed, volume and temperature of incoming water being equal, two coils would be superior due to the increased surface area. Plate chillers work so well because of the increase in SA, and counterflow chillers work so well because of the large volumes passing by each other at every moment, that is the total amount of heat removed increases due to the large volume. Think of it this way - if you dropped a red hot bolt into a pitcher of water at 33°, it would warm the water a lot. But if you dropped the same bolt into a swimming pool at 65° there would be little change - water has a pretty high heat capacity.

For flows,

q = hA(Ts &#8722; Tb)

h is the heat transfer coefficient. This is larger for turbulent flows than for laminar flows, but for comparing differences in systems, it would be nearly constant since we're only talking copper tubing in direct contact with wort. It could effectively be removed from the equation.
A is the surface area of heat transfer (for copper tubing it would be 2 * pi * r * length)
Ts is the surface temp and Tb is the temperature of the liquid not near the exchange.


Obviously keeping the wort moving (i.e. homogeneous temperature) is ideal to avoid a heat gradient where it will be cool around the coils and warmer far away form the coils, thus reducing efficiency. Also, if I were to use dual immersion chillers, I would design one to flow from the bottom up and the other from the top down, again to maintain as high a Delta T between liquids as possible.

For my money, the prechiller is really worth it. I'm not an engineer, but I do play one on the internet.

 

[StAnthonyB]

The most important aspect of all of this is the Delta T. Heat always moves from areas of greater concentration to areas of lower concentration. Having two coils immersed vs. one coil immersed, volume and temperature of incoming water being equal, two coils would be superior due to the increased surface area. Plate chillers work so well because of the increase in SA, and counterflow chillers work so well because of the large volumes passing by each other at every moment, that is the total amount of heat removed increases due to the large volume. Think of it this way - if you dropped a red hot bolt into a pitcher of water at 33°, it would warm the water a lot. But if you dropped the same bolt into a swimming pool at 65° there would be little change - water has a pretty high heat capacity.

For flows,

q = hA(Ts &#8722; Tb)

h is the heat transfer coefficient. This is larger for turbulent flows than for laminar flows, but for comparing differences in systems, it would be nearly constant since we're only talking copper tubing in direct contact with wort. It could effectively be removed from the equation.
A is the surface area of heat transfer (for copper tubing it would be 2 * pi * r * length)
Ts is the surface temp and Tb is the temperature of the liquid not near the exchange.


Obviously keeping the wort moving (i.e. homogeneous temperature) is ideal to avoid a heat gradient where it will be cool around the coils and warmer far away form the coils, thus reducing efficiency. Also, if I were to use dual immersion chillers, I would design one to flow from the bottom up and the other from the top down, again to maintain as high a Delta T between liquids as possible.

For my money, the prechiller is really worth it. I'm not an engineer, but I do play one on the internet.

What an AWESOME post! :rockin:

 

[payton34]

go with 50 feet. that way if you go to ten gallon batches you will have enough to submerge

 

[Ryush806]

The most important aspect of all of this is the Delta T. Heat always moves from areas of greater concentration to areas of lower concentration. Having two coils immersed vs. one coil immersed, volume and temperature of incoming water being equal, two coils would be superior due to the increased surface area. Plate chillers work so well because of the increase in SA, and counterflow chillers work so well because of the large volumes passing by each other at every moment, that is the total amount of heat removed increases due to the large volume. Think of it this way - if you dropped a red hot bolt into a pitcher of water at 33°, it would warm the water a lot. But if you dropped the same bolt into a swimming pool at 65° there would be little change - water has a pretty high heat capacity.

For flows,

q = hA(Ts &#8722; Tb)

h is the heat transfer coefficient. This is larger for turbulent flows than for laminar flows, but for comparing differences in systems, it would be nearly constant since we're only talking copper tubing in direct contact with wort. It could effectively be removed from the equation.
A is the surface area of heat transfer (for copper tubing it would be 2 * pi * r * length)
Ts is the surface temp and Tb is the temperature of the liquid not near the exchange.


Obviously keeping the wort moving (i.e. homogeneous temperature) is ideal to avoid a heat gradient where it will be cool around the coils and warmer far away form the coils, thus reducing efficiency. Also, if I were to use dual immersion chillers, I would design one to flow from the bottom up and the other from the top down, again to maintain as high a Delta T between liquids as possible.

For my money, the prechiller is really worth it. I'm not an engineer, but I do play one on the internet.

I AM an engineer...you could have fooled me with that post. Did you stay at a Holiday Inn last night or something?

 

[Bmorebrew]

Did you stay at a Holiday Inn last night or something?

No, but if I keep talking to my wife only about beer, I might be living in one someday.

 

[ghack]

I've got 50' in two concentric coils. The two coils affect a much greater volume of the wort than a single would.

In my parts the tap water is currently coming out upwards of 90 degrees. I run that through for a few minutes and them directly pump ice water through. Works like a charm.

 

[plbrown44]

The science behind the immersion chiller is all about heat transfer (with a sprinkling of fluid dynamics). The more surface area of the cooling element you have, the more efficient the cooler will be. However, it has to be understood that a 100’ chiller won’t necessarily chill more efficiently than a 50’ chiller. This is because the cooling fluid (cold tap water in most cases) gradually warms up the further it travels through the spiral of the chiller so by the time the cooling liquid reaches the other end (or potentially half-way through) it is likely the same temperature of the warmer wort. Therefore an extremely long chiller may not be worth it if the cooling fluid warms up to wort temperature 25’ through the coil.

There are two kinds of heat transfer taking place. Conduction – direct contact of copper surface to wort – and convection – cold tap water moving running inside copper tubing (fluid dynamics). There is a big difference in cooling efficiency that depends on whether the water flow within the copper tubing is laminar or turbulent. Getting a little too complex here…

The heat transfer equation is q = m(&#916;T)Cp where q is the amount of heat transferred (Joules), m is mass (grams), &#916;T is the change in temperature (Kelvin), and Cp is the specific heat (Joules per gram*Kelvin). In our case we can calculate the amount of energy needed to chill the wort, q, since the other variables are known. Using water as our medium, m = 18,550 g (mass of about 5 gallons), &#916;T = 212 °F – 67 °F = 145 °F (80.6 K), and Cp = 4.1813 J/(g*K). It works out to be 6,251,587 Joules or 6,252 kJ. Now that you know the amount of energy required to chill the wort, a more complex analysis is required.

This concept is actually a really complex equation from an engineering standpoint that involves heat transfer, fluid dynamics, thermodynamics, and material science. Throw in a lot a calculus as well and basically you have invested more time in the science than is probably worth it.

Buy 50 ft. The object is to cool as fast as possible to minimize the opportunity for contamination. Subjectively (from what I've read), 25 ft isn't enough and 50 ft chills it plenty fast.

I'm trying to get our engineering college intern to bring this problem to one of his professors for a senior design project, we'll see how that goes. More to come...

 

[meatwad]

Ok, so I got a little lost there... but

Great first post PLB... keep it up, and let us know if you get the intern to work on this one. Would be cool to see the results from a semester of class work!

 

[Bobby_M]

The more surface area of the cooling element you have, the more efficient the cooler will be. However, it has to be understood that a 100&#8217; chiller won&#8217;t necessarily chill more efficiently than a 50&#8217; chiller.

Nice first post.

Here's the thing, people throw the word "efficient" around a lot when talking about IC lengths and diameters but I always have to ask exactly what the context is. That is, we have to define the priorities for which efficiency will be measured.

Efficiency measurement will be completely different if the goal is FASTEST chilling vs. Least water use vs. cheapest upfront materials cost.

Fastest? Assuming enough kettle capacity to deal with the displacement, 100' of 5/8" OD copper will absolutely cool faster than 50' of 3/8 but at the cost of materials and water usage.

When it comes to the question about coil length for a given diameter, longer is better up to the point where tap water pressure can no longer deliver the same flow rate. However, this ignores the question about the upfront cost vs. cooling gains.

I suppose you could look at efficiency at the maximum cooling speed with a sweet spot of water usage and upfront costs (the diminishing returns problem). If absolute efficiency is really that important, go with a plate chiller instead.

Specifically to the original post's question, the only way I could view a longer tube length as counterproductive is if you mean cost to value. For the most part, longer always means equal to faster cooling given a fixed coolant temp and flow rate. However, I read "counterproductive" as SLOWER cooling after a certain length. In the context of how we use these chillers, it's never counterproductive.

 

[audger]

people are getting bits and pieces of the equation, but in just skimming over this thread, i havnt seen all the pieces yet. there are four main pieces of the equation that influence the efficiency of any chiller, radiator, heat exchanger, etc...

-flow rate of the cold side
-temperature of the cold side
-flow rate of the hot side
-temperature of the hot side

given those four variables, we can conclude...

-when flow rate of the cold side increases, efficiency increases, but more water is wasted
-the greater the temperature difference (delta) between hot and cold, the more efficient the heat transfer
-if you either decrease the flow rate, or increase the time the cold water contacts the hot water (with a longer tube), the more efficient the design becomes (more heat transfered per unit of cold water), but the longer amount of time it will take to cool the hot side
-as you lower the temperature of the hot side, the temperature delta, and therefore thermal transfer efficiency, per unit of cold water, decreases. meaning your efficiency from start to finish is not linear but more logarithmic.

if you were actually doing the math, you would want to factor in the heat capacity and permisivity of the materials (kettle, copper tube) and other secondary factors, but for overall discussion, those four variables are the important ones.