compact peltier cooled carboy setup

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.
I suppose a compressor and fill valve could be scavenged from any old, defunct junker fridge? What’s the lowest coolant volume you can get away with? Presumably that’s a function of your coil size, which in this case is much smaller than a full fridge.
 
you can calculate the capillary length with an app from Dancap, the layout of the refrigeration components are readily available on the internet, it is preferable to get a professional to gas up as it needs the correct pressure and some oil but mainly needs vacuuming prior to gassing, this also makes sure that there are no leaks.
That's amazing! I never thought about that, but that is a good solution. I guess the only tricky part is the dimensioning of the capillary? How do you determine how much refrigerant to put in? Do you have to calculate the volume of your plumbing, or is it just filling to some pressure at a given temperature?
 
I suppose a compressor and fill valve could be scavenged from any old, defunct junker fridge? What’s the lowest coolant volume you can get away with? Presumably that’s a function of your coil size, which in this case is much smaller than a full fridge.
you can use an old compressor but make sure you use the correct refrigerant, I used a plastic cooler box for a six pack something like 6 lts coolant with 2 of glycol, personally I preferred to go the brew way with the components, nothing more annoying to get it working and a free months down the line have it fail.
my coil was 2 meters long 1/2 inch pipe, coiled around a wooden former, filled the pipe with salt and duck taped the ends before winding. othe 1/4 inch pipe was easy to wind up but I used too much, some 4 meters and it offers too much impatience to
IMG_20190815_163859.jpg
 
Wouldn't the compressor be pretty much the only thing that can break on a fridge/freezer? I'd pick one from a working fridge, or find out why exactly your donor fridge/freezer was dumped.
 
Here is the assembled and taped fermenter&keg enclosure:
IMG_20200808_172433 - Copy.jpg

Two inches of polyisocyanurate, which adds up to an R (F * square feet / (BTU/hour)) of 13. Long term this stuff degrades a bit, by about 10%, it seems (compare LTTR = long term thermal resistance and metric R column in the earlier picture). The box fits my 7.5 gallon carboy and my two 5 gallon kegs. Hopefully this will cut down on condensation (dew point indoors is currently about 14C), and allow the peltier system to run colder.

Edit: corrected; thanks, garzlok! Also, the polyiso (safer to abbreviate...) might only be R 11 at beer temperatures.
 
Last edited:
I’m sure you meant Polyisocyanurate Foam Boards....nobody wants anything to do with an isocyanate. 😖 (Nice build, nice use of space/materials)
 
Looks really good - I see that you have a lip on the lid - that ought to help with air infiltration and thus condensation. I had wondered about different ways to seal a lid. The magnetic rubber strips in fridge doors are pretty nice, but I wonder if some sort of a drape over the lid, secured by magnets, would be an idea? Some kind of plastic or rubber skirt?
I’d be tempted to fill all the empty space between kegs with foam packing peanuts.
 
I think best would be a dry purge. But for some reason they are running neither dry air nor nitrogen lines in our condo...
 
So this build has got me reconsidering.. I like the ability to reconfigure your fermenter for just the cost of some sheet insulation. Recently, I bought about 8x 1gal jugs from uline with the idea of doing splits. I’m wondering how hard it would be to build a multi chamber fermenter capable of independent temp control of, say, 4 jugs. The scheme would probably be to use a common cooling loop with 4 independent heaters. Do you think this could work? What would a good cooling loop geometry look like?
 
This is a fascinating idea. So, I am not an engineer, but I DIY a lot of stuff using plans from people much more technically-educated than I.

I use the Anvil Cooling system for my fermenter. The last brew I had to put about 10# of ice in the 20+Gallon cooler 1-2x/day, with ice cream salt and H2O as the coolant to keep my brew in the 65-67F range for a week. That was a lot of ice to buy and because of the salt to lower the temp more, I couldn't just reuse it to water the garden.

1: Looking at what has been posted on this thread, would a Peltier like this work to keep a 6-pack cooler with glycol cold enough to lower the beer temp 20-30F below ambient for fermentation (assuming that is enough coolant to cool the beer, this idea did not work with just ice water and I haven't tried glycol yet):

Semiconductor Chilling Plate: TEC1-12706
  • Size: 40 * 40 * 3.75mm
  • Internal Resistance: 2.1~2.4Ω
  • Max. Temperature Difference: 67°C Above
  • Rated Voltage: 12V(Vmax. 15V, Starting Current 5.8A)
  • Working Current: Imax.=4~4.6A(Rated 12V)
  • Refrigeration Power: Max. 50~60W
  • Working Environment: -55~83°C
2: Assuming it works to cool well enough, should the Peltier fluid exchanger be in series with the tubing and aquarium pump of the anvil cooling system, or as a separate system?

3: The refrigerator pump pictured a few posts above also looks like a good idea. Would that be more energy efficient in a 6-pack cooler with glycol than the Peltier?

I am also not in a position to set up a fermentation chamber using an A/C unit or mini fridge at this time, though I have looked at plans and designs to build them.

Keep posting cool stuff!

Reevesie
 
So, while I like and run my system, I must say it is marginal and not robust. I am using nine (9) of those TEC1-12706 peltiers (or whatever it is you actually get when you order them from the cheapest supplier on Amazon), several of them are already broken, and it seems with the fermenter (empty) and both kegs in the box, I can run the almost full Weissbier keg at 8-9C, while the Pilsner keg (still fermenting) is not hooked up to the cooling loop and just tagging along at 18C, finishing the fermentation from 1.03 further down, which is good enough for me, but I shouldn't have to use nine (seven or eight working) 12706 peltiers and 270W to do that.

This is with 270W electrical power when on, and just a little bit of cycling off, i.e. the 8C setpoint is pretty close to the lowest temperature this system can reach. (I'm running 4 of the 9 at 6V, hence the lower power.)

Maybe there are TEC peltiers that meet their specs, the ones I bought certainly don't. At 12V they draw about 3A.

Condensation (running the cold side below the dew point of your house/apartment) can take away an essentially unlimited amount of cooling power.

What I would do if I started over:
-keep the copper cooling loops wound around the kegs, attached with stretched aluminum tape
-do the same type of external cooling loop for the fermenter, or use an off-the-shelf stainless steel cooling insert, or switch to fermenting in a keg (which is what I'm doing with the Pils right now)
-keep the box: it is a bit of work, but quite a bit cheaper and lighter than a freezer
-keep the cheap pumps, temperature controllers and hot side radiator (the radiator only unless using a mechanical compressor, option c below); self priming pumps would be nice, but venting the loops every few disconnect cycles is not that bad
-maybe try push-to-connect fittings instead of the somewhat expensive quick disconnects?

For the actual cooling, I see three options:
a) Combine hot and cold loop, put the radiator directly in front of a window A/C unit outlet, and see how low that gets (no Peltiers).
b) Higher nominal wattage peltiers, possibly running at lower voltage: this would be another experiment, using 12710 or 12715. If using peltiers, I would definitely recommend the Meanwell power supply.
c) Converted window A/C unit or home made setup (see dtashmore's posts above) using a commercial compressor to provide the cold liquid (in a cooler, with the pumps sitting in the bath, which solves the self priming/loop venting issue).

I don't want to sound too pessimistic, I guess it all depends how much experimentation and fiddling you are prepared to deal with. After all, we are not running a life support system, it's just beer...

Argyll, regarding the one loop/several heaters idea: I'm afraid a peltier based system will never have cooling capacity to spare. The pumps are pretty cheap, so maybe run several independent loops against the same hot side heat exchanger? (separate peltiers and cold side heat exchangers).
 
Forgot to answer JAreeves' questions 2 and 3:
2: in series, I'd think, i.e. the cold side directly plumbed into the anvil loop. One pump only, obviously. Putting a reservoir and two heat exchangers in between (or just a reservoir, used by two separately pumped loops) just introduces more heat.
3: a factor ten (10) more efficient. (A mechanical system produces about a factor three more cooling power than the electricity it consumes: 3500 BTU/hr = 1kW, but you typically get at least 12 SEER = BTU/hr per W; a peltier system is maybe 30% efficient - very dependent on the temperature difference it is trying to produce).
 
How much heat load do you estimate you are dealing with from heat leak through your insulation, and from the actual fermentation? I’ve seen estimates that active fermentation can generate 10-50W.
I imagine that peltiers are the preferred solution below a certain cooling power level where there aren’t many options for compressor systems
 
I think it's all heat leak and condensation (on the lines outside the enclosure), except for 5-10W.

I'm more and more convinced that motor bike electrics and home brewing heat gain do not follow the normal laws of physics.
 
So, while I like and run my system, I must say it is marginal and not robust. I am using nine (9) of those TEC1-12706 peltiers (or whatever it is you actually get when you order them from the cheapest supplier on Amazon), several of them are already broken, and it seems with the fermenter (empty) and both kegs in the box, I can run the almost full Weissbier keg at 8-9C, while the Pilsner keg (still fermenting) is not hooked up to the cooling loop and just tagging along at 18C, finishing the fermentation from 1.03 further down, which is good enough for me, but I shouldn't have to use nine (seven or eight working) 12706 peltiers and 270W to do that.

This is with 270W electrical power when on, and just a little bit of cycling off, i.e. the 8C setpoint is pretty close to the lowest temperature this system can reach. (I'm running 4 of the 9 at 6V, hence the lower power.)

Maybe there are TEC peltiers that meet their specs, the ones I bought certainly don't. At 12V they draw about 3A.

Condensation (running the cold side below the dew point of your house/apartment) can take away an essentially unlimited amount of cooling power.

What I would do if I started over:
-keep the copper cooling loops wound around the kegs, attached with stretched aluminum tape
-do the same type of external cooling loop for the fermenter, or use an off-the-shelf stainless steel cooling insert, or switch to fermenting in a keg (which is what I'm doing with the Pils right now)
-keep the box: it is a bit of work, but quite a bit cheaper and lighter than a freezer
-keep the cheap pumps, temperature controllers and hot side radiator (the radiator only unless using a mechanical compressor, option c below); self priming pumps would be nice, but venting the loops every few disconnect cycles is not that bad
-maybe try push-to-connect fittings instead of the somewhat expensive quick disconnects?

For the actual cooling, I see three options:
a) Combine hot and cold loop, put the radiator directly in front of a window A/C unit outlet, and see how low that gets (no Peltiers).
b) Higher nominal wattage peltiers, possibly running at lower voltage: this would be another experiment, using 12710 or 12715. If using peltiers, I would definitely recommend the Meanwell power supply.
c) Converted window A/C unit or home made setup (see dtashmore's posts above) using a commercial compressor to provide the cold liquid (in a cooler, with the pumps sitting in the bath, which solves the self priming/loop venting issue).

I don't want to sound too pessimistic, I guess it all depends how much experimentation and fiddling you are prepared to deal with. After all, we are not running a life support system, it's just beer...

Argyll, regarding the one loop/several heaters idea: I'm afraid a peltier based system will never have cooling capacity to spare. The pumps are pretty cheap, so maybe run several independent loops against the same hot side heat exchanger? (separate peltiers and cold side heat exchangers).
hi just a small update from my side which may help, I am not too interested in making calculations, so my first effort into cooling was only a minor success, I made a keggerator from a broken fridge which had morning but the condenser left working, I used the cheapest compressor I could get, 1/10hp, it produces just 70w of cooling power and uses 100w of 230v ac power, this is cooling a
 
sorry accidental posting.
it is cooling a 30lt keg, it still manages to get it down to 8 degrees C with 35 degrees of indoor ambient. it is running continuous at this. the other chiller unit is 2.5 times this size and can easily get down to 5 degrees with about 50 percent utility.
 
dtashmore, stupid question: these horse power ratings, they are in 745.7 Watt= 1 hp horse powers, and they are what the compressor draws electrically, not its cooling output, right?
But you are probably right: too much calculation can spoil the fun.
Autocorrect garbled your post a bit; what was the 'morning' actually supposed to be?

I'm now thinking of simplifying my setup, removing the old individual 40mmx40mm heat exchangers completely, and replacing the TEC1-12706's with three TEC1-12715 running at 8V each, mounted between the two 120mmx40mm heat exchangers. This would simplify the plumbing a lot. Assuming those guys draw 8A at 8V, this should yield the same cooling power as eight 12706's drawing 3A each at 12V, but with only 2/3 the waste heat produced. Well, let's not jinx it with numbers.
 
Last edited:
dtashmore, stupid question: these horse power ratings, they are in 745.7 Watt= 1 hp horse powers, and they are what the compressor draws electrically, not its cooling output, right?
But you are probably right: too much calculation can spoil the fun.
Autocorrect garbled your post a bit; what was the 'morning' actually supposed to be?

I'm now thinking of simplifying my setup, removing the old individual 40mmx40mm heat exchangers completely, and replacing the TEC1-12706's with three TEC1-12715 running at 8V each, mounted between the two 120mmx40mm heat exchangers. This would simplify the plumbing a lot. Assuming those guys draw 8A at 8V, this should yield the same cooling power as eight 12706's drawing 3A each at 12V, but with only 2/3 the waste heat produced. Well, let's not jinx it with numbers.
yes that is the convertion factor. sorry about the error couldn't go back and correct the error after accidentally hitting the post reply button, the output power is around 75 to 80 percent of the input power so for a 1/10 hp compressor the input power will be 100 watts for 75 watts output. the word morning should have read "nothing"
 
Pretty exciting: the three TEC1-12715 in series connected to 24V (seeing 8V each) are indeed drawing 8A. With a short-circuited cooling loop and the hot side radiator drawing in 78F/25C room air, the system now reaches 32F/0C!
That's with about 220W electrical power - still need to measure the actual AC side draw. With the old 12706 setup, under the same conditions the system would reach 43F/6C drawing 270W.

With the radiator mounted in front of the 63F/17C A/C outlet, the new setup reaches 26F/-3C with the short-circuited cooling loop.
It's now again hooked up to my Weissbier keg inside the polyiso box, with the set point at 6C.

In anticipation that this might work I had also ordered some food grade propylene glycol; I'm running a 1:3 glycol:water mix that should be good to 10F/-10C.

I should post some pics later. The plumbing is a lot simpler now.
 
Last edited:
Some info on power usage:
After the change to three TEC1-12715 running in series on 24V, drawing 8A, the power supply delivers 192W (from current and voltage measurement) and draws 230W (from outlet power meter); that gives a power supply efficiency of 84%. The setup used 1 kWhr in 7 hours, so that should be about 3.5 kW hours or 50 cents per day.
I'm now running a single 5 gallon keg with a set point of 6 deg C and a 2 degree deadband (cooling off at 6 C, on again at 8 C); the temperature readout fluctuations felt a bit too large to run with 1 degree deadband.

It seemed that the problem of air getting sucked into the cold loop was getting worse with the lower running temperature. The hot side does not have this problem. I assume this has to do with the plastic tubing getting less pliable as it gets colder, and sealing at the connections getting worse. I think I also lose small amounts of coolant and introduce a bit of air every time I reconnect a keg or fermenter with the quick disconnects. Therefore, instead of trying to seal all connections on the suction side of the pump, and still having to vent and refill every few quick disconnect cycles, I added a small closed reservoir inside the peltier enclosure and mounted the cold side pump in it. This vents air bubbles out of the cooling circuit every time the pump starts up (and actually continuously while it is running). It should be a while before I have to add coolant.

Some pics:

cooling enclosure with two 5 gallon kegs and a 7.5 gallon fermenter (partially visible in the front):
IMG_20200807_143242 - Copy.jpg



side view of peltier enclosure; 24V power supply on top, cold side coolant reservoir just below, with cold side pump inside:
IMG_20200814_183133 - Copy.jpg


top view; the plumbing inside looks a lot cleaner than before - the outside wiring could still use some cleanup....:
IMG_20200814_183206 - Copy.jpg


less plumbing, less wiring; the two pumps split 24V in series, and so do the three peltiers (this would not have worked with a 12V power supply); I'm monitoring one of the three 8V peltier voltages. The cold side of the peltiers uses some of the left over keg enclosure insulation (on the right). Cold (left) and hot (right) loop connections at the bottom of the picture.
IMG_20200814_182700 - Copy.jpg


close up of the cold side coolant container; the pump output line is fed through the lid directly, the discharge line into the reservoir plugs into a barbed Y-fitting (because that is what I had left...):
IMG_20200814_182557 - Copy.jpg
 
Last edited:
Just to add how nice it is to see the second 5 gallon keg cooling down from 14C to 6C in just a few hours! I hooked it up to the cooling loop at midnight, expecting to wake up to a slightly warm Weissbier (since the loop will exchange heat between the kegs), instead the cooldown was already done.
This is the keg with more copper cooling tube windings and the better insulation (which shouldn't matter so much in the new enclosure). Also, I did not have to vent the cold loop after putting the second keg inline.

I find that the cheap controllers I'm using for temperature readout have offsets as large as 2.5 C between them. Luckily they seem constant.
 
I should probably update the parts list; the main changes to the original list are the polyisocyanurate enclosure, the switch to the 15A Peltiers that I run at 8V/8A, the Meanwell 24V 350W power supply, and the 120mm by 40mm heat exchangers.

In steady state the system uses about 3.1kWhr per day. Yesterday when it was cooling down the second keg, it used 4.6kWhr.

Updated parts list:

Aideepen 5pcs TEC1-12715 DC12V Heatsink Thermoelectric Cooler Cooling Peltier Plate Module 40x40MM $28.69 (used three of them)


ARCTIC MX-4 (8 Grams) (Current Edition) - Thermal Compound Paste, Carbon Based High Performance, Heatsink Paste, Thermal Compound CPU for All Coolers, Thermal Interface Material $11.76 (used very little of it; I moved away from the thermal silicone adhesive so that the cooling stack would be repairable)


Propylene Glycol Half Gallon (64 oz.) by Pure Organic Ingredients, 100% Pure, Food & Pharmaceutical Grade, Hypoallergenic Moisturizer and Skin Cleanser $10.95 (used about 100 cm^3 of it; this just seemed the best deal)


MEAN WELL LRS-350-24 DC Power Supply, 24V 14.6A 350W for 3D Printer, LED Strip Light, Industrial Control System NES/SE/S $29.95


Quickun 5/16" Hose Barb Fitting Equal Barb Cross 4 Way Splicer Joint Mender Adapter Union Fitting (Pack of 20) $15.99 (used the two-way unions for feedthroughs, and one Y-joint for the cold reservoir return)


2x BXQINLENX Aluminum Water Cooling Block for CPU Graphics Radiator Heatsink 40x 120(White) $12.99x2


25 Feet Vinyl Tubing- Food Grade- 5/16" ID 7/16" OD 10 Ft, 25 Ft, or 50 Ft FDA Approved
Sold by: Nolk Co $13.45


AGPtek® 12 Pipe Aluminum Heat Exchanger Radiator for PC CPU CO2 Laser Water Cool System Computer $17.99



GDSTIME 120mm Case Fan, Increase Air Pressure Fan, 120mm x 25mm 12V Dc Brushless Cooling Fan Dual Ball Bearings $15.99


MOUNTAIN_ARK 2 Pack 63 GPH (240L/H) Submersible Water Pump DC 12V 3.6W 9.8ft Lift for Fish Tank Pumping, Rockery Water, Bonsai Fountain $13.99


HDVD 10 Pack DC Power Male Pigtail Connectors 2.1 x 5.5mm Copper 10 inch Length $6.50 (not all used)


HDVD 10 pack DC Power Female Pigtail Connectors 2.1 x 5.5mm Solid Copper 10 inch length $6.30 (not all used)


R-Tech DC12V 1A UL-Listed Switching Power Supply Adapter for CCTV - 5 Pack - Black $13.99 (only used one of those)


Temperature Controller Module with Case, 2PCS XH W1209 Display Digital Thermostat Module with Waterproof NTC Probe $7.99

total ~$220
I'm not listing the plastic enclosures ($11 for a whole bunch of them), wiring, zip ties, quick disconnects, and probably some other stuff.

and for the keg and fermenter enclosure:
  • 1 sheet 8' by 4' 2" thick polyisocyanurate foam, AP-foil faced on both sides, for ~$30. I used two sheets of 1" thick material because that is what was in stock. It helps to have your cut list in hand, since not all cars will fit a 8'x4' sheet, and Menards has cutting tables available in the building materials area.
  • aluminum tape ~$6
  • construction adhesive ~$10 ; I used three cartridges of Liquid Nails Fuze-It, but using a single 2" sheet one cartridge should be sufficient.
total ~$46
 
Last edited:
I tried removing the hot side radiator fan and just having the A/C blow cold air (17 C) through the radiator, but the temperature started to creep up (cold feed is at 0 degrees C, Pils is at 5 degrees C), so I put the fan back. I moved it to the other side of the radiator so it now blows through the radiator, which makes it a bit quieter. My office A/C is super quiet, so the radiator fan is actually the dominant noise source: 62 dB at 1 foot, 56 dB where I sit; background noise floor is about 50 dB.)

The next things to improve would probably be a larger radiator and a larger and quieter radiator fan; but I think I am at the point of diminishing returns, and I am happy with the setup as it is now.
 
I installed a quieter radiator fan; it's also cheaper. Only downside: the blue light. (Actually, it looks pretty cool straight through the radiator, and is otherwise mostly shielded from view.) Apparently the kids these days demand LEDs in their fans.

Cooler Master SickleFlow 120 V2 Blue Led 120mm Square Frame Fan with Air Balance Curve Blade Design, Sealed Bearing, PWM Control for Computer Case & Liquid Radiator $9.99

It provides a bit less air flow, but the coolant temperature is the same within 1 degree C. I will have to check the power usage to see whether the system runs less efficiently.
The old radiator fan was nominally 92 cfm, but actually only provided 31 cfm. ( 230 W 192W divided by 11 K radiator outlet - inlet temperature difference divided by 1.2J/K/Liter heat capacity of air gives 39 31 cfm.)
The new fan leads to a radiator outlet temperature of 33 C, so 16 K Delta T, so the air flow is 31 * 11/16 = 21 cfm. Nominally this fan is supposed to provide 62 cfm. That is the same derating factor (34%) as for the GDStime fan.

Update: energy usage with the new Coolermaster fan is 20% higher than with the old GDStime fan: 4.4 kWhr per day vs. 3.7 kWhr per day.

Update 2: A website I find useful for Peltier cooling estimates: TEC / Peltier Element Design Guide
In particular this graph tells me that with running 8A through a 12A nominal Peltier I'm not too far from the optimum for a Delta T of about 35 K:
1599085061746.png

So I am probably getting 60-70W cooling power out of my system (Qh/Qc between 2.75 and 3).
 
Last edited:
I've got a "catalyst" fermenter, it's not polycarbonate but similar, and so I can cut holes in it. I use one of these stuck in the side: IceProbe Thermoelectric Aquarium Chiller - Nova Tec The stopper in the top of the fermenter is dual hole with a thermowell, and a temp probe connects to an inkbird to turn this chiller on. Kind of a small version of one of these: BrewJacket Immersion Pro - Fermentation Temperature Control Systems which I discovered afterwards.

I agree sometimes space is important, and I don't have anywhere for a keezer either. I don't brew lagers, just ales, and just need to keep them running 5 - 10F below ambient at times, so this has been good enough for me. It's been through at least a dozen brews and seems to be working well. There's a chance the surface in contact with the beer is changing a little, so hopefully I'm not getting dain bramage from it if it is.

I love what you're doing, just throwing out some ideas in case you feel like experimenting in a different direction.
 
That's a very compact setup, tracer bullet. Sticking the cold side of the peltier directly into the liquid is certainly the most efficient way.

I like that mine is modular; I can cool the fermenter and kegs (and anything I stick into the enclosure, like wine bottles, if necessary). I think for the next brew I will remove the interior cooling loop from the fermenter, and just run it parasitically (no directly attached cooling loop) inside the enclosure that holds the cooled kegs. I keep the kegs at 6-8C, so it might actually be a little bit too cold for my next Weissbier. We'll see.

Update: The sanitizer in the fermenter was at 15 C, which is actually pretty much ideal. I want to ferment the Weissbier at 16-18C, so with a little bit of heat from the fermentation that should just work out. I was able to extract the cooper loop in one piece, and taped it to the outside, in case I need a bit more cooling oompf.
 
Last edited:
Nice!
Confess I dont quite understand if there are more than one similar project here
could you chill a 10L keg running an outside copper around the keg without gaving a cold water bank? I mean, could it be possible to install the peltier on that water block, it will refrigerate the liquid that will run through the coil around the ouside of thekeg and then return to the peltier water block to be colded againa?
 
Christian, I'm not sure I understand 100% what you are saying, but you also need to cool the hot side of the Peltier(s). So just the cold side loop from the Peltier to the keg is only half of the work. Getting rid of the heat from the Peltier is as important as cooling your beer or wort.
(The third half is preventing condensation...)
 
Gruel, yeas I undertanded that you need to cold the hot side of the peltier, but what is the efficiency of the system just with the fan/dissipatort (a expensive good and a cheap bad)?

the dissipator/fan would chill the hot side of the peltier, think about 5L or 10L,
added a child paint image made by myself of how im thinking this system. A 4 peltier with 4 fans and disspipator and a blos of water chamber. And a bar that I think its made with peltier, that you put on the fermenting beer and it will cool
also there are the minibrew system, that cools like the way im saing

sorry for my english, its a little rusted
 

Attachments

  • example.png
    example.png
    19.5 KB · Views: 13
  • peltier cooling.png
    peltier cooling.png
    55.5 KB · Views: 13
  • techspecs_400x.jpg
    techspecs_400x.jpg
    12.9 KB · Views: 13
  • images.png
    images.png
    60.1 KB · Views: 13
Hi Christian, I think calculating the actual efficiency is a bit involved, but the curves from Meerstetter in my post above tell that for example for a 30 degree (Kelvin or Celsius, not Fahrenheit) temperature difference you can get 40% efficiency if you run the Peltiers at half their maximum current. This 30K is the temperature difference between the hot and the cold side of the Peltier, not between your beer and the ambient air. You are heating the ambient air to get rid of the heat from the Peltier, so at best the hot side of your Peltier is just a little bit above the temperature of the hot air coming out of your cooling system. In other words, how much of that 30 K temperature difference is then left for actually cooling your beer depends on the thermal resistance of your hot side cooling system. Directly cooling the hot side of the Peltier with air can be very efficient if you have enough cooling fin area and create enough air flow.
Here is a link on dimensioning heatsinks: https://www.electronics-cooling.com/1995/06/how-to-select-a-heat-sink/ .

The heat capacity of air is only about 1.2 Joule per liter and Kelvin at normal pressure, so you have to heat quite a bit of air to remove a few hundred Watt.

For example, to get 100W cooling power at an assumed Peltier efficiency of 40%, you need to remove 250W.
That means you would need to heat about 20 liter of air per second (or about 40 cfm) by 10 degree (K or C). In that case you would still have 20 degree 'left' to cool your beer below ambient temperature, since you 'spent' 10 degree on the air. (This is neglecting the thermal resistance from the Peltier to the heatsink.) How much air you actually need to flow past your heatsink depends on the efficiency of the heatsink; the link above has some numbers on that.

For my system I'm using an intermediate step transporting the heat through water, because that allows a very compact heat exchanger at the Peltier and a large surface area radiator at the ambient air side. But if you can fit a large enough heat sink directly on the Peltier, then that makes for a very efficient and less complicated system.

TL;DR : use the link above to dimension your heatsink.
 
I moved my hot side radiator outside the window, since A/C is off since a few days. In principle if it's cold enough outside I could just run the pumps and leave the Peltiers off, but that would require quite a bit of rewiring. It would also make the system less flexible to deal with varying outside temperatures.

Best would actually be to bypass both Peltier liquid heat exchangers, and run a liquid loop directly from the kegs to the outside radiator. I'll keep an eye on the electricity use of the system. For the last month the average was 4.4 kWhr/day or about 60 cents per day.
Beer cooler electricity use 2020-10-03 114326.jpg
 
Update: On average the system used about 140 kWhr per month since August, or $20/month at the rate I'm paying. It ran very reliably. I had to top off the hot side once since I redid the tubing connections. I could hear the pump getting a bit noisier with a few air bubbles circulating.

I now turned the system off, since I'm not drinking enough beer for it to be worth running. Maybe I'll start brewing again in the spring or summer.
 

Latest posts

Back
Top