Building the Ultimate chiller – AKA The GOAT (pics and data)

The forum for discussing all kinds of brewing paraphernalia.
setmash

Building the Ultimate chiller – AKA The GOAT (pics and data)

Post by setmash » Sat Jan 03, 2015 1:07 pm

Hi guys,
We all know the arguments for IC v Plate v Counter flow chiller. But some time ago FIl/Barneey /Cooky (amongst others) all built a new type of straight Counter flow chiller and named it a GOAT chiller (because it can be cleaned using a ‘GOAT’ brush). In my opinion these are the ultimate chillers. I considered for months my chill solution for my ‘Big Fat Shiny keg’ build and finally decided on a GOAT. So what advantages and disadvantages does a GOAT have?

Advantages

1. 100% inspect able
2. 100% cleanable
3. 100% demountable
4. Expandable
5. Very coolant economical – water meter and tree hugger guys will be happy
6. No threads on the cold side – can use a Blichmann inline temp gauge
7. Very easy to build – can be built without soldering (I will cover this and the cost in a later ‘how to’ post)
8. Very cheap to build
9. Can be used in single pass or recirculation systems.
10. Clean In Place (CIP)

Disadvantages

1. Can be quite large so best mounted on a rig or carrying jig.
2. Nobody really knows the ideal pipe combination and length for the chiller. Pipe: 22mm outer 15mm inner, 15mm outer 10 mm inner, 22mm outer 10 mm inner etc. and then there is turbulation of the coolant and wort. What effects does each of these have and how can they be achieved?
I’m trying to find the answer to the above. It’s costing me a fair amount in plumbing materials but hopefully I can come up with the ideal pipe combination and help people build their own Goats.

Let me first start by saying that it’s worth bearing in mind that a good chiller depends on what you are trying to achieve. Some guys are obsessed with super quick chill times and many manufacturers quote times to chill down from boiling to wort pitching in minutes for a set batch size. This is only half the story as they often fail to tell you the coolant temperature and more importantly (IMHO) the amount of coolant used (coolant flow rate). I’ve found that some manufacturers are using 22L/min coolant flow rates – mine at home is 4.9 Liters per min (particularly slow). For my application Id like my chiller to be 100% cleanable and inspectable – sorry plate guys no matter what you say you just don’t know whats in there. I’d also like to recycle my coolant by using it as my rig cleaning water with PBW (I actually use that brilliant stuff Malt Miller sells for half the price). I brew in 42 Liter batches. I don’t want to chill too quickly because my chugger pumps are so powerful they suck hops through my Copper filter and in to the FV if I set the flow too high. It takes about 10 mins for me to do a single pass transfer so my wort flows at about 4 Liters per minute. My rig vessels hold 55 Liters of water so my ideal chiller would chill 42 liters in 10 ish minutes and use no more than 55 Liters of water and the exit temp of the coolant would be as high as possible.
So what do we need to measure to compare a chiller? Obviously wort input temperature and output temperature but also coolant flow rate, coolant input temperature and crucially the amount of coolant used to achieve the desired pitching temp. One manufacture quotes a quick chill time of 6.6 mins for a 25 Liter batch but uses 145 Liters of coolant in the process OMFG !! My 12 foot immersion chiller uses 137 Liters of coolant to do the same job.
If anyone would like to add data to this thread I would be more than happy to add this to the data table I have put together. But please measure all the above parameters so we can do a like for like comparison.

In my tests so far I have measured the amount of energy (Joules) the coolant is absorbing and hence the number of Watts being sunk by each chiller (Joules per second). Its worth noting that the wort cools very quickly sat in the Copper after flame out – we must consider this as its relevant but bloody difficult to measure. If we don’t remember this then our chiller look over 100% efficient which they are not. Also to be fair to the big Duda plate chiller it monsters through the wort chilling 25 Liter batch in just 2.3 minutes so the wort sat in the kettle has less time to cool so in effect its having to work even harder.

So I have collated some data from other manufactures (at the top of my table) for reference. I have also included two plate chillers from DUDA – they have a useful table with all the info you need to help compare them thermally. The two Duda chillers I choose are based at both end of the spectrum – entry level plate chiller typically sold over here and then a behemoth 18 inch long 40 plate which would be at the very high end of a home brewers kit (if I was going for a plate Id go for this £200 well spent). I may add some more DUDA chiller data along the way.
So here is my GOAT test data base. As you can see I’ve tested 22mm outer pipe 10mm inner pipe at 4 Meters length, 8 Meters length and 8 Meters Coolant turbulated. I’ve also tested 22mm outer 15mm inner at 8 Meters length. Smooth means they have no turbulation (inserts to cause coolant swirling or indents to cause wort swirling). I’m still testing but the interesting thing is the GOATS use a minuscule amount of coolant matching the big DUDA plate. I’m going to do further tests by increasing my coolant flow rate to see how that effects stuff but also increasing further my GOAT lengths – as it seems they are scaling unbelievable well - 22mm/10mm smooth calculations. ....the 8M 22mm/10mm smooth goat is sinking 49 Kw per square meter which is 163% of the 4M version (30Kw) it shows up obviously in the amount of coolant used (47 liters v 101 Liters). I think this is really significant as the goat lengths being tested are not showing significant diminishing returns, i.e. by increasing the lengths further performance will increase, I think GOATS are the ultimate chiller – I will keep updating this thread with more results – please feel free to add comment and suggestions in our quest to find THE ULTIMATE GOAT

Image

Here is my 8M 22mm 10mm Goat under test (you can make them as long or as short as you like):

Image

User avatar
Hogarth
Under the Table
Posts: 1793
Joined: Fri Nov 30, 2007 1:30 am
Location: Brixton, London

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Hogarth » Sat Jan 03, 2015 1:47 pm

Looks great! Nice work.

I wanted to build one of these but simply don't have the space.

Also — sorry, but surely this has to be called the "Trombone"? :lol:

Fil
Telling imaginary friend stories
Posts: 5229
Joined: Sun Oct 16, 2011 1:49 pm
Location: Cowley, Oxford

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Fil » Sat Jan 03, 2015 8:23 pm

fwiw mines known locally as the flux capacitor ;)
ist update for months n months..
Fermnting: not a lot..
Conditioning: nowt
Maturing: Challenger smash, and a kit lager
Drinking: dry one minikeg left in the store
Coming Soon Lots planned for the near future nowt for the immediate :(

Frothy

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Frothy » Sat Jan 03, 2015 8:59 pm

whats a goat brush?

setmash

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by setmash » Sun Jan 04, 2015 12:46 pm

Lol re the trombone. You can make these much shorter and fairly compact.

A Goat brush is a long thin brush you can get down a 10mm or 15mm pipe to clean the chiller.

User avatar
Andy
Virtually comatose but still standing
Posts: 8716
Joined: Fri Nov 18, 2005 1:00 pm
Location: Ash, Surrey
Contact:

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Andy » Sun Jan 04, 2015 12:50 pm

I think I'll stick with my trusty CFC :)
Dan!

User avatar
Aleman
It's definitely Lock In Time
Posts: 6132
Joined: Sun Jun 03, 2007 11:56 am
Location: Mashing In Blackpool, Lancashire, UK

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Aleman » Sun Jan 04, 2015 1:09 pm

setmash wrote:A Goat brush is a long thin brush you can get down a 10mm or 15mm pipe
And the only place that has been found to have a 2m long brush is a supplier of stuff for keeping goats ;)

Of course The goat chiller is nothing more than a series of small CFC's ;) I still have my original 3m one (Single length) that I built something like 15 years ago (at least) . . . It didn't work as effectively as I wanted (15mm/22mm) only cooling to 35C with an incredibly slow wort flow rate, so I ditched it and was too embarrassed to post about it :oops:

I can't help thinking a series of 15/10s at 2.5m long screwed to the wall with parallel coolant feed and flow rates is a very good way to go. . . . but I'm not going there as I have a big FO plate chiller which I'm happy with

Fil
Telling imaginary friend stories
Posts: 5229
Joined: Sun Oct 16, 2011 1:49 pm
Location: Cowley, Oxford

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Fil » Sun Jan 04, 2015 4:35 pm

shorter and compact works out a lot more expensive in pipe fittings, i went low tech simple stupid with my length terminating fittings opting for bungs with holes bored thru em, but to accommodate 14 x 500mm lengths i still needed 28x everything to terminate each length..
viewtopic.php?f=6&t=61868&p=656050&hili ... ea#p654662
ist update for months n months..
Fermnting: not a lot..
Conditioning: nowt
Maturing: Challenger smash, and a kit lager
Drinking: dry one minikeg left in the store
Coming Soon Lots planned for the near future nowt for the immediate :(

setmash

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by setmash » Sun Jan 04, 2015 5:20 pm

Yes the longer the goat legs are the cheaper it will be to make for a set overall length as you need less fittings. My rig is approx. 2 Meters long so i work in 2 M lengths. Im going to fix the Goat to the back of the rig. If we discover that 15mm/10mm Goats are optimal then they are going to be very cheap to make (as long as you keep the legs a reasonable length - 1M). My initial results show that 22mm/10mm and 22mm/15mm at 8 Meters total length perform very similar. Im going to test 22 / 15mm at 12 Meters next. Then 15/10 at 4M, 8M, 12M lengths. I will then try 22mm/10mm at 12M - we will then have enough data to make a decision about which is optimal. I will then turbulate to measure the effects of that. Unfortunately I have to travel a lot at the moment so not getting much home time to do the tests. Hope to do some more this coming week.

A 10M goat 15mm/10mm would cost about £61 - with soldered fittings and rubber bungs. Pipe cost 15mm = £22 10mm = £16.54 Copper T's £12, Rubber bungs £6, Silicon tube £5 = £61 - thats for a 1M legged Goat total length 10M. Costs drop a bit if you make the legs longer as you wont need as many fittings. If you make it solderless then add another £16 for the push fit fittings. £61 for a chiller that can be 100% cleaned and uses very little coolant would be a great result.

Making your own kit, for me, is a big part of what our hobby is all about.

setmash

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by setmash » Mon Jan 05, 2015 6:49 pm

I decided to work backwards. I wanted to figure out how much thermal power is required to chill 25 Liters of wort and compare it to my calculations based on the coolant consumption. It turns out they are pretty close in value - for example I calculated that my 22mm/10mm 8M smooth goat sinks 12.26KW (at my Coolant flow rate of 4.9 LPM) - it took me approx 9.4 Minutes to cool the 25 Liters, working backwards it works out it should be 9.9 mins - very close and its the same for all the chillers. Formula is Temp difference (95-25) x Liters (25) x Specific heat capacity of water (4181) = 7316750 Joules divided by chiller power (12260 Watts) = 596 divided by 60 = 9.9 mins

I plotted some of the chillers on to the graph below. The graph below shows the challenges we encounter the faster we want to chill (exponential). It shows the thermal sinking power required to chill 25 Liters of wort from 95 degrees to 25 degrees in minutes (Im assuming wort has similar specific heat capacity as water for all my calcs). So for example to chill all of that 25 Liters to 25 degrees in 1 minute would require a chiller to sink over 120 Kw ! Dont forget chiller performance will vary with coolant flow rate and temperature. When I test my 8M 22/15 I will run it at 18 LPM coolant - expect to see a big difference

Image

User avatar
Jocky
Even further under the Table
Posts: 2527
Joined: Sat Nov 12, 2011 6:50 pm
Location: Epsom, Surrey, UK

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Jocky » Tue Jan 06, 2015 2:15 pm

Cool topic. Keep it coming.

When I do a wet run with my new plate chiller I'll take some measurements to try and add to your data.
Ingredients: Water, Barley, Hops, Yeast, Seaweed, Blood, Sweat, The swim bladder of a sturgeon, My enemies tears, Scenes of mild peril, An otter's handbag and Riboflavin.

setmash

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by setmash » Wed Jan 07, 2015 11:49 pm

Thanks Jocky that would be really great, the more data the better.
Ive had quite a night. Lots of data and need to analyse tomorrow....but i extended the goat to 12m using 15/10 for the final 4m....my GOD it worked !!!!!!

25 Liters 99 degrees cooled to 27 degrees:
Coolant in temp 10.2 degrees
Coolant exit temp 76.5 degrees :shock:
Coolant used 23 Liters =D> =D> =D> =D> =D>
Coolanmt flow 3.28 LPM - more on this in the analysis to follow
Time to cool 7.18 mins

What an amazing result - using less than 25 liters of coolant,,,,

and here she is in all her glory...what a cluster [censored] but shes a hell of a chiller :D ...and im knackered and wet lol

Image

User avatar
Aleman
It's definitely Lock In Time
Posts: 6132
Joined: Sun Jun 03, 2007 11:56 am
Location: Mashing In Blackpool, Lancashire, UK

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by Aleman » Thu Jan 08, 2015 10:13 am

The one thing that will rain on your parade is when the coolant temp hits 20C in the summer :(

I'm planning on getting a large HDPE container, that will live on a shelf in the 'brewery' that I will drop a product coil in and chill to 5C, using my python and use that as a COLD Liquor tank. Thats pretty much the only way to have a consistent cooling regime for your brewery.

setmash

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by setmash » Thu Jan 08, 2015 11:57 am

Last night I tested :
a. 22/15 8 Meter with higher coolant flow rate by pumping water with my chugger pump through the Goat.
b. I dont have enough parts to do a 22/15 12 Meter so I added 4M of 15/10 to give a 12 Meter long goat. The results were staggering.
All tests 25 Liter wort batch size 99 degree to 27 , coolant input 10.2 degrees.

22/15 8M showed some interesting characteristics. It shows that the goat restricts flow fairly badly. My source coolant from the tap is 5.45LPM the Goat ouput is 4.9LPM. HOWEVER if i used the Chugger which normally outputs the 17.6 LPM the Goat restricted it to 8.6LPM. Coolant flow rate is key - the more you pump through a heat exchanger the faster the cooling will be but you will use more coolant and the resultant cooling temperature will be lower. For example:

22/15 8M at coolant 4.9 LPM = 9.5mins to cool 25 Liters using 43 Liters. Increase to 8.6 LPM coolant and the chill time drops to 7 mins. Unfortunately i didnt measure the coolant out temp during this test as it was such a nightmare to perform the test i forgot to do it. So maybe if you want faster coolant flow rates then 22/10 would be better. However faster coolant flow results in lower efficiency.

The exciting bit. So I added a 4M 15/10 run on to my 8M of 22/15 resulting in a 12M long goat. What was interesting is that the 10mm soft copper tube was a sod to get down the 15mm as it wasnt exactly straight after all the test ive been doing. I wondered to myself how much this would restrict flow. No way of adding coolant turbulation to 15/10 - but i couldnt help think OMG this may aid heat transfer as its so tight in there.

Anyway the test blew me away. Coolant flow rate dropped to 3.26 LPM (I was using my crap tap pressure not chugger) but what came out was scorching hot, it burnt my hand at a blistering 76.5 degrees :shock: Chill time plummeted to 7 mins (I knew immediately it was working well as the wort flow was high) but what was utterly astounding is the amount of coolant used, a meagre 23 Liters =D> This is a very efficient cooler and meets my needs, i dont want a higher wort flow rate than that and i love the water usage and temp output. If you have good mains pressure the chill time would go sub 5 mins with this cooler easily, maybe even 4 mins !! SELL YOUR PLATE CHILLERS :mrgreen:

So the question ? Is the improvement down to length alone or did that 15/10 make a bigger difference. Hang tight Im hoping to build a 15/10 8M tonight !!

Results table:

Image

setmash

Re: Building the Ultimate chiller – AKA The GOAT (pics and d

Post by setmash » Thu Jan 08, 2015 11:31 pm

Man this is exhausting, last night i was happy...now im confused.

So the results of tonights experiments using an 15/10 8 Meter goat (im getting really quick at building these lol):

25 Liter batch from 99 degrees to 27 degrees
Coolant temp = 9.5 degrees. Coolant flow rate dropped to 3.0 LPM. Time to chill 8.5 mins, coolant used a meagre 22.5 Liters =D> coolant output temp 69.5 degrees =D>

These goats are incredible at low flow rates. The issue will be, as Aleman mentioned, in the summer. My worry is as water temps rise that 3LPM flow rate will kill me. Now im confused.

22/10 8M = coolant temp in - 9oC, Coolant temp out 44oC, coolant flow rate 4.9LPM, time to chill 9.4mins, coolant used 47 liters
15/10 8M = coolant temp in - 9.5oC, Coolant temp out 69.5oC, coolant flow rate 3.0LPM, time to chill 8.5mins, coolant used 22.5 Liters

WTF is going on.........I need to test 22/10 8M at lower flow rates...FFS

Post Reply