The Cooler with Jason Barrett
In this episode, we are talking about process cooling. Later in the podcast, I bring in Jason Barrett out of Black Button Distilling in Rochester, NY and we talk about their distilling process and how to make a chemically stable cream based liquors and other unusual products.
Process cooling is what we use to remove heat from out distillate. The main places that this occurs are in our mash tun, fermenter and still. The process cooling equipment can also be used to provide space cooling if we have excess capacity or limited space for separate systems. Process cooling can be broken down into passive and active systems.
Passive cooling systems involve letting the warmed-up liquid radiate their heat into the hopefully cooler body of the room they are in. In the case of fermenters, this isn’t always possible since your distillery very well may be warmer than the maximum temperature the yeast can endure. In this simplest of passive systems, your yeast is putting heat into the liquid of your fermentation which is then passing the heat into your room. While this is considered uncontrolled fermentation most of the time there are things we can do to increase the rate of heat exchange including locating fermenters in a breeze way so they are continually getting new cooler air, decreasing the temperature of the room around the fermenters (although this may be considered active cooling and inefficient active cooling if it’s done by more than opening a door in December) and making sure the material of our fermenters is not insulating (metal over plastic or unjacketed over jacketed).
We can make a passive cooling system for our stills by running the cooling water from the chiller into a reservoir and then allowing it to pass heat into the room around it. Once this water is cooled then it is able to be pumped back to the chiller during the next distillation. If there is enough time between distillations this is about the least expensive way to cool your equipment but there are some caveats. Before we get too deep into them, let’s start off defining a BTU so that we’re all on the same page. A BTU is the amount of energy that it takes to increase the temperature of one pound of water one-degree Fahrenheit (see I told you load cells would be useful), one pound of water is about ⅛ of a gallon.
So, in this example we have 1,000,000 BTUs going into our still during distillation (about 3-5,000,000 during heat up or about a 5,000 gallon still) and we want to use our chiller to pull it all out so that our distillate is back to room temperature (in reality 85% works well for this number). To also make my life easy, we’re going to assume that there is no energy lost along the way, everything we put in we need to pull out. If we didn’t really want our reservoir to heat up we could circulate 1,000,000 pounds of water through our condenser every hour (again assuming perfect heat exchange in the condenser, don’t you love my perfect world?) it would heat up 1°F per hour or about 6F over the whole distillation. 1MM pounds of water is 120,048 gallons of water and we would need to be pumping at 2,001 gpm to move that through our still. If we had a 2,001-gallon tank it would heat up 60°F per hour or 360°F over the course of the distillation and we’d end up with 420°F which isn’t possible without tons of pressure not to mention from about hour two onwards the ‘cooling’ water would actually be hotter into our distillate, not the other way around.
Since that isn’t practical let’s try and bring this example a little bit more towards reality. If our tank starts off at 70°F and we don’t want to tank to get to over 140°F since at that point our distillate will vapor off some of its ethanol and any PVC that is contacted will start to soften and potentially buckle. With a 70°F allowed temperature increase we can start figuring out how large our passive tank needs to be, over the course of a six-hour distillation, we’ll dump 6 MMBTU into the tank so the tank has a minimum size of 85,715 pounds or 10,290 gallons.
As a note, I screwed up on the podcast here and gave the reservoir size as 15,000 gallons. That 15,000 was actually pounds of water needed per hour of the distillation. This is why you should read the show note and definitely not trust the numbers I’m saying during the show, just the concepts.
The large size required for the reservoir is the downside of this design buying a 10,000-gallon tank just to use as a reservoir, while cheaper than an 85-ton chiller or even a 30-ton chiller, is not cheap and it takes up a lot of room in smaller distilleries. This reservoir tank should be designed to maximize its heat loss: cooling fins, no insulation, or a large surface area in the direction of any breeze or in general. Even with these additions to the tank it will still take considerable time for it to reduce from 140°F back to 70°F and with a million BTU system that employs this technique you should really give the tank a week to cool off between uses (which considering this still will make 14 barrels in 6 hours probably isn’t terrible financially). For use cases where you want to distill every day it will probably be best to design a 30-40°F temperature increase or less, unless you’re able to put your reservoir outside on the north side of the building in Alaska in winter time then maybe we can distill twice a day into this system, but generally this is most practical for small distilleries.
When there isn’t time to wait we’ll need to do something on our end and this brings us to active cooling. There are several different levels of active cooling the cheapest will be designed to work partially as a passive system. If we take our 10,000-gallon tank above and simply want to ensure that it’s back to 70°F by the start of the next work day then we’ll need at most a 30-ton chiller assuming we’re not getting any passive help. If we want to eliminate the reservoir and just drop the coolant back to room temperature as soon as it comes off the still we’ll need an 83-ton chiller.
There are several types of active cooling devices with the most popular being glycol chillers and cooling towers. I’m not going to explain the refrigeration cycle that makes the glycol chillers work but Wikipedia has a nice write-up. Cooling towers are much simpler and they basically work like swamp coolers and are subject to a lot of the same limitations. Basically, the water in your cooling loop is either misted or pumped through a porous medium to minimize the droplet size and then air is blown over the water, some of the water evaporates removing large amounts of energy (and some water) from the system. The two main limitations are the humidity of the exterior environment of the distillery and that water is lost during this cycle (if water is a precious commodity there can be additional expenses for large energy loads). Very large cooling towers do exist but for distilleries, these are mainly practical for mid-size distilleries and glycol chillers are used where it’s humid and by the large operations. Most of the time I reach for a glycol chiller unless there is a special case for why the cooling tower would be better (mid-sized operation, limited cash, dry environment in the summers). With either type of active cooling, one nice thing is that these systems can be run in series so the 20-ton chiller you bought when you started out can just have another 20-ton unit added on as you grow rather than needing to buy a 40-ton unit. At least until you run out of space for them.
When sizing a chiller system, we first need to evaluate the load cases, in general, our cooling loads come from the mash tun, fermenters, and still but these loads do not occur simultaneously or at least don’t have to. With fermenters, the highest load case occurs during the first 72 hours so if we spread our ferments out so that #1 kicks off on day 1 and #2 kicks off on day 4 and #3 kicks off on day 7 then we will have a continuous load of a single fermenter (about 1 ton) or all three fermenters could be kicked off simultaneously on day #1 be finished by day #4 (about 3 tons) and maybe that last three days is when you’re running your still and mashing. These two systems will have different peak and base loads so the chiller needs to be sized appropriately. Crashing your mash will cause the single largest peak load on your chiller. If you finished your mash with a 140°F rest and need to crash it down to 70°F for pitching it is almost the exact same math as above. With a 500-gallon mash (we’ll pretend it’s all water for easy math) dropping in temperature 70 degrees we need (500*8.33*70) 291,550 BTUs of cooling but crashes typically happen in a half hour so we need twice the BTUs/hour or 583,100 BTUs (50-ton chiller). A 500 gallon still will typically require 600,000 BTU to heat up (if you’re stripping then during distillation too), and about 120-200 MBTU while distilling so that mash crash chiller requirement could take care of your still too, if they weren’t run simultaneously, otherwise you’d need a 60-75-ton chiller. In the end, the chiller requirement will vary between 80 tons on the high end and 50 tons on the low end for a continuously cooled 500-gallon distillery. Of course, if they’re only distilling and mash crashing 3 times a week it could be as small as 4-tons with an appropriate reservoir.
Slightly different from either active or passive systems are geothermal systems. They are passive in the sense that there is no active work required by the distillery but they are active in that you’ll still be running heat exchangers and actively moving heat out of your products. There are two main types of geothermal; one is circulating your cooling loop into the ground so that the naturally cool ground will pull out your heat (there is no practical difference between this and the active cooling systems above), or two where cool groundwater (or well water) is circulated through your distillery until it gets too hot to be useful (or it’s done all of the cooling you need it to). The second type is probably the most popular form of cooling across the country where streams have been diverted to cool worms for centuries. This method is incredibly cheap in most places where on the lowest end you don’t even need to pump the water and your artesian well can circulate it up through your condenser or heat exchanger and then it’ll gravity feed back to the stream. You can size these systems using your average ground water temperature (50-70°F), remember you’ll distill a little slower in the summer and a max of not more than 140 degrees and then back out your flow rate. If you don’t have access to high enough flow rates consider a tank to store water overnight and a pump. These systems are highly regional and they may not be legal and/or water & sewage fees may make it cost prohibitive in your area. If you can use them you probably already do and know more about it than I do and if you can’t one paragraph is probably all you can stand to read.
Instead of theoretical scenarios where we are removing all of the energy that we put into the still, we can be a bit more accurate and look at how much energy actually needs to be removed from our distillate. To do this we need to look at what we are actually producing and since this will still be an average calculation we just need to look at what we collect, for now, we’ll say we’re collecting 83 gallons of 120 proof whiskey at 68°F when we’re done with our 500-gallon single pass run (10% wort). What we’ll calculate is how much energy it takes to condense the vapored ethanol and water back to liquid and then to additionally cool that liquid to 68. So the first step is to say that it takes 971.6 BTU/lb of water and 363.7 BTU/lb of ethanol to condense them (this is another of the places load cells make life nice) that works out to 8,108 BTU/gallon for water and 2,395 BTU/gallon for ethanol (assuming the density at 60°F) at 120 proof we’ve captured 49.8 gallons of ethanol and 33.2 gallons of water so it will take 269,186 BTU and 119,271 BTU, respectfully, to condense them. Now the ethanol needs to be reduced from 173 degrees to 68 and the water needs to be reduced from 212 to 68 for that we need their specific heats which are 0.548 and 1 (BTU/lb °F). Once we multiply that out we discover we need 18,872.6 BTU for the ethanol and 39,824.1 BTU for the water. Add all of that up and we need 447,154 BTU for the cooling over six hours or 74,526 BTU/hr or 6.21 tons which is not a bad reduction from the 10 tons we were thinking of above. See sometimes it is worth while to do the math!
Once we know the BTU requirements for each piece then growing the system becomes like Lego pieces. One 500-gallon still is 7 tons, two 500-gallon stills are 14 tons, if we double the size of our mash we double the energy it takes to cool it. This really helps in expansion planning where if you have a 20-ton chiller that is just barely getting it done and you want to double your equipment then you can simply add a second equally sized chiller. On the other hand, if the chiller is only working at 75% load then maybe you only need a 10-ton chiller in your expansion. If you aren’t currently spreading you thermal load across the day and you have a continually reducing active system then you may be able to change your process during your expansion to consider thermal load and minimize the size of your new chiller. The biggest thing you can do is ensure that you don’t crash your mash while distilling and you can cut the size of your chiller in half.
Download the Podcast Here:
Rate us on iTunes!
Mentioned in this Episode: