DIY Wort Chiller

We’ve been talking about making a wort (“wert”) chiller for a while, and I finally got around to it this afternoon. It’s a pretty straightforward process and I found couple of how-to guides online. Some people are far more serious about making their wort chillers than others (e.g., blow torches and copper solder vs. tape and clamps). Given my nature, can you guess what path I took? We’ll see in a bit.

So a wort chiller is simply a device that rapidly cools down the wort (the liquid extract that contains the sugars that yeast will eventually turn into liquid gold alkeehawl) after the first boil. Wort cool down is needed because the yeast will die if dumped into the piping hot liquid and it needs to happen rapidly apparently because there are little nasty particles of wild yeast floating around in the air that could be toxic to your brew. So the shorter the time you can spend with your wort exposed to room temperature air the better. At Dave’s house we’ve had some epic battles trying to cool down hot pots of wort. One person would stir the liquid while the other gave it an ice cube rub down in the sink. So it was time for a wort chiller….

The principle of a typical wort chiller is this: submerge a hollow copper coil into the wort, post-boil. Copper is highly heat conductive, 400 W/m-K, compared to about 0.6 W/m-K for water and 1.5 W/m-K for porcelain. This means that any temperature sent through the coil will almost immediately cool the copper coil down to roughly that temperature. Hook the coil up to a cold water source (kitchen sink faucet or outdoor spigot). Water is used as a heat transfer fluid because it has a relatively high specific heat capacity (about 4.2 kJ/kg-K vs. about 1 kG/kg-K for air). This means that if you run cold water through the coil, you can carry heat away away about 4 times as quickly as you could with the same mass flow of air. The wort chiller designs I found online, and follow below, operate on an open loop, meaning that cold water goes in the submerged coil, gains some heat from the wort (in the form of a higher temperature), then the now-warmer water exits the chiller into the sink drain. This is admittedly quite wasteful, but will work for now. I think we’ll plan to recapture the outlet water at some point for reuse elsewhere.

Constructing the chiller was really easy. I spent about half an hour at the hardware store, but only 5 or 10 minutes building it. Here’s the shopping list:

20 feet of 3/8 inch outer-diameter copper coil – $25
1 10 foot washing machine hose (inlet) – $8
1 10 foot vinyl hose (outlet) – $6
a couple of hose clamps – $2
total – $41 + tax

You can buy one assembled for about $46 + tax + shipping. Was it worth it? You bet your ass it was!

Follow the pictures below to see how I made this thing. The copper coil came in a great shape for our pot straight out of the box. I also bought an adapter for the kitchen sink for it to accept washer/garden hose threads (3/4 inch). I whacked one of the ends off the washer hose to squeeze onto the inlet of the copper coil, then attached the vinyl tubing in a similar fashion on the outlet end. I bent both the inlet and outlet connections to orient them outside of the pot. That way, in case they leak, no water will go back into the wort. I clamped both ends and voila! The final pictureshows the final product. We’re about to use it over at Dave’s in brewing a Rye IPA (mini-mash), so we’ll see how much it cuts the cooling time down. However, because I couldn’t wait, I did some back of the envelope calculations to estimate how long it would take to cool down. See after the pictures.

You are now looking at the back of an envelope. Here it goes:

The wort should be about 90-100°C (194-212°F) just after the boil. We need to get it to 27°C (80°F) before adding the yeast. That’s a temperature difference of about 70°C (70 Kelvin), give or take. The amount of energy that represents in our 5 gallons of wort is about 5.6 kJ (Volume of wort = 5 gallons = 0.019 cubic meters. Multiply that times the density of water,  because the wort is essentially water (so x 1000 kg/m3) = 19 kg = mass of hot wort. We could have just weighed this. Oh well. Multiply 19 kg of wort by the specific heat capacity of water (~4.2 kJ/kg-K), and multiply that by the 70K temperature difference, and we need to remove about 5,600 kJ of heat from the wort (very rough estimate). That’s our target. Hang on to that number.

To remove this amount of heat (energy), we will send cold water through the copper pipe. James and I measured our faucet’s output with the wort chiller attached and came up with a flow rate of about 2 cups in 4 seconds, or about 100 mL per second. Not sure about Dave’s… probably close. Multiply that volumetric flow rate by the density of water (1000 kg/m3) and do a unit conversion, and we have a mass flow rate of water through the coil of about 0.1 kg per second. Now here’s the tricky part in the estimation: we can measure what the cold water inlet is, but we won’t really know what the just-past-inlet water temperature will be (this would require much more work, and math. People hate math. I don’t even like math.). But let’s take some rough numbers and give it a go. The cold water coming out of our faucet is probably about 60°F (let’s say 15°C). At it’s warmest, the temperature just of the wort just outside of the copper coil is about 90°C (delta T for this portion = 75°C = 75 K). Thus the “energy flow” of the water, at its best, through the wort chiller should be about 31 kW (31 kJ/s = 0.1 kg/s x 4.2 kJ/kg-K x 75 K). But, realistically, the 75°C temperature difference between inlet and outlet is probably much smaller because the water doesn’t stay at 15°C the whole time through the chiller – it warms up as it passes through. So let’s take some admittedly random number – the average between 15°C and 90°C – about 53°C; let’s use that as a delta T. So the new “energy flow” is something like 22 kJ/s (22 kW).

Divide the amount we need to remove from the wort (5,600 kJ) by the estimated rate of removal (22 kJ/s), and we should have this thing cooled within about 250 seconds (or 4 minutes). Although I’m probably doing something wrong, and the rate of heat removal by the chiller is probably much less as the inlet temperature will probably increase from 15°C to closer to 50°C or something within the first couple of centimeters. We’ll see how long it actually takes in the next few hours with Dave’s Rye IPA! I guess 10 or 15 minutes!

Update: See the comment below. Dave’s thermometer broke as we were testing the chiller, but we think we cooled the wort down within about 8 minutes or so. We’ll give it another test the next time around.

I also wanted to include a couple more calculations. The first is looking at the ability of the copper coil to transfer. The question is: is the coil a limiting factor in any way? Well, let’s see how much energy could transfer from the wort, across the coil, into the water. Multiply the thermal conductivity of the copper (U = 400 W/m-K) by the area of copper touching the wort (O/D = 3/8″ = 0.0095 m; L = ~15 ft = ~5 m; times pi -> A = 60 m2), and multiply that by the temperature difference between cold water and wort (75 K), and finally divide by the thickness of the copper coil (t = ~1/8″ = 0.003 m). The rate of conduction across the copper is therefore about 1.5×10^6 W or about 1500 kW (1500 kJ/s). Remember the entire amount of energy we need to remove from the wort was only about 5600 kJ (or 4 seconds worth of contact time). So copper doesn’t appear to be limiting.

Another calculation I thought might be helpful was to see how much contact time each pass of water gets with the submerged coil. This is quick: We have a water flow rate of about 100 mL/s or 0.1 L/s or 0.0001 m3/s. The cross-sectional area of the coil is pi times the inner diameter-squared (I/D = 1/4″ = 0.00635 m), divided by 4 (A = 3.16*10^-5 m2). The flow rate divided by area gives a velocity of about 3.2 m/s. Since there’s about 6 m of coil, the water spends a little less than 2 seconds in contact with the submerged coil.