DIY Cascade Refrigeration System (Part I)

foreign [Music] to expand my vapor compression refrigerator into a two-stage system using the ethylene I made in my last video as a refrigerant for a second stage with the goal of reaching -100c to recap from my last two videos a multi-stage refrigerator is basically two Refrigeration Cycles where the cold end of the first stage cools the warm end of the second stage the term warm here is relative and in the case of my second stage refers to a temperature of around -40c using ethylene as a second stage refrigerant condensation should occur at around 15 bar and if the flow resistance is tuned correctly the evaporator should allow the ethylene to expand to around one atmosphere which should bring it down to a temperature of minus 104 C now it is technically possible to get lower temperatures by using the first stage refrigerant in both stages but it requires the second evaporator to run far below atmospheric pressure and become sort of Impractical broadly speaking for most compressors to operate with a reasonable degree of efficiency if your evaporator is running at one atmosphere you want to be able to condense your refrigerator internet no more than say 15 or 20 atmospheres for a room temperature condenser propane is ideal and for a condenser running at -40c ethylene fits the bill other options could be ethane or carbon dioxide but ethane is more complicated to produce and carbon dioxide will freeze in the evaporator and clog it the second stage of the refrigerator is built pretty much the same as the first but with the addition of one key component the oil separator at the lower temperatures involve the compressor oil will freeze which will cause the system to clog and the compressor to eventually pump all its oil out overheat and die from lack of lubrication the oil separator causes the ultra fine droplets of oil in the loop to coalesce into large drops that drip down to a collection sump in the bottom which is then forced back into the inlet side of the compressor from the high side pressure this is usually done through an extremely thin capillary tube or solenoid valve that periodically opens because the rate of oil circulation is very tiny another component of the second stage is the after cooler this isn't strictly required tired but makes the process more efficient because the second stage refrigerant will still be heated from compression even if it doesn't condense into a liquid at room temperature and removing that heat takes a bit of load off the intercooler okay let's get to the build a lot of people in the comments on my last video pointed out that a real refrigeration system is braised rather than soldered so it got a more powerful torch and tried my hand at a few brazing joints they don't look pretty but let's see if they hold pressure easy way to test this is just to hook it up to a fridge compressor with a flare fitting and dunk the joint in water this compressor tops out around 500 PSI if there's any leak we'll definitely see it bubbling at that pressure I didn't see any bubbles though so I guess my technique is working even if it's not that pretty that was a phosphor copper alloy for brazing Copper to Copper or copper to Brass but the oil separator I got has stainless steel tubing coming off it for some reason so to brace that to Copper I'll need ssf 56 which is a silver solder with some sort of a fluorine based flux just FYI if you're going to use this the the flux fumes will make you horrendously sick so make sure you do this outside and wear a respirator and eye protection because that stuff is nasty the end result is a pretty nice clean joint between the copper and stainless and I didn't see any bubbles when I pressurized the separator and submerge it in my bathroom sink after that I brace the separator to the outlet side of the after cooler coil which had previously been the evaporator coil for the window AC unit I took apart in my vapor compression video to make the connections for pressure gauges I abrase a one millimeter inner diameter copper tube to a one-quarter inch MPT cap with a hole drilled in it copper to Brass doesn't require the silver solder but it does require some flux to stick right I put clear silicon caulk on threads before tightening down any fittings but going forward I should really use some of that gooey thread sealant paste stuff for this particular build the second stage will need yet another component which is a buffer tank the buffer tank helps to ensure my high side pressure can be as high as possible for a given pressure ratio by adding additional volume to the low pressure side of the refrigerant Loop if I didn't have the buffer tank what might happen is the compressor could pull the low pressure side down to well below atmospheric pressure which would drastically reduce the maximum pressure possible on the high side in this case it's much more important that I have a large high side pressure to ensure condensation occurs for the second stage compressor I use the fridge compressor that I had initially used in my vapor compression video the outlet tube was steel so I used the silver solder again to join it to the copper tubing from the after cooler for oil return I Simply Connected a needle valve on a capillary tube to the oil drain line on the separator and crush the larger tubing tight around the capillary this would probably only have to be opened for a few seconds once per run the crush tubing was then sealed up with solder Downstream of the oil separator is the filter dryer which will then connect to the intercooler turns out the setup is incorrect which I'll get to later the intercooler is a plate heat exchanger which is basically just a series of closely stacked plates that have very good thermal contact but the two sides are sealed against each other to connected to the system I'll use these three quarter inch caps which will have a one-quarter inch flare fittings braced onto them holes are drilled and flare fittings are inserted then I start bracing there's a pretty interesting range of colors going on here I made four of these fittings then slapped them onto the intercooler with some thread sealant I had to use my big wrench to get a good grip on them and here's the result the brace joints are a little ugly but ended up working pretty good doing a submerged leak test with my air compressor doesn't reveal any bubbles on the other side of the filter dryer I've got a flared quarter inch line with a flare nut on it to attach to the system the nice thing about the flares is that if something goes wrong or needs to be changed it makes it relatively easy to remove and reinstall the intercooler next I built the capillary tube assembly this connects to the outlet of the intercooler with a pressure gauge 25 feet of one millimeter capillary tubing and needle valve to bypass the capillary for fine-tuning the flow rate in retrospect I really should have used a smaller diameter tube so that I wouldn't need so much of it doing a quick test with the compressor and light open to the air I get about 70 PSI on the high pressure side with the adjustment valve open with the valve closed I get close to 180. not bad but I need to go a little higher so I added on an additional 25 feet of capillary tubing making the total length 50 feet definitely need a smaller diameter it also makes the whole assembly a bit messy with the additional capillary length the pressure tops out at around 220 PSI when open to the air with one and a half to two atmospheres of Inlet pressure it should produce more than enough pressure to condense the ethylene when it hits the intercooler next I move the buffer tank into position and hook it up with more flare fittings here's the evaporator coil it connects to the capillary Outlet on one end and has a t to connect to the buffer and the compressor Inlet on the other end I had a minor hiccup with leaks though and after starting out with a static charge around 55 PSI a few hours later I was down to less than 40. putting my gauge underwater for a pressure test revealed that there was a slow leak through the fittings that I needed to correct I fixed the leak but then realized I had to rearrange some things first off the filter dryer is supposed to be on the liquid side of the condenser which in the case of the second stage meant putting it on the intercooler Outlet also the intercooler is supposed to be vertical this way liquid refrigerant from the first stage can enter through the bottom and evaporate upward while gaseous refrigerant from the second stage can enter through the top and the condensed droplets can fall downward gravity is your friend in this case the intercooler also needs to be insulated very well this component isn't rejecting or absorbing heat from the environment it's transferring heat from the second stage to the first stage refrigerant any heat transfer to or from the environment would result in loss says here's a look at the whole system after rigging it up for the first stage things are a little messy so I've highlighted the components to clarify what's going on we have the first stage compressor first stage condenser first stage capillary tube the intercooler which is the first stage evaporator at the bottom you can see that the inlet line is iced up and the first stage return line moving to the second stage we have the compressor the after cooler the oil separator which is the Crimson cylinder hiding behind the high pressure gauge again the intercooler which is the condenser for the second stage the inlet is on top and the outlet is on the bottom the filter dryer the capillary tube with the adjustment valve with low pressure side gauge the buffer tank evaporator which is inside this foam box that the thermometer is sitting on and the return line here's a closer look at the intercooler in action the first stage Inlet and second stage Outlet are iced up while the first stage outlet and second stage Inlet pipes are basically room temperature I'm running this first test with propane in both stages just to verify that everything works once everything checks out I'll switch the second stage to ethylene as you can see the condenser pressure is running at a mere 90 PSI compared to the first stage is 200 and the evaporator is sub-atmospheric in this case with propane I bottomed out at -56c a good bit colder than the single stage unit but not as cold as it could be with ethylene as the refrigerant once I had confirmation that everything was working I pulled out the ethylene storage tank and hooked it up to my second stage fill port the initial charge used about half my collected ethylene meaning it consumed about 42 liters STP that seemed a little surprising to me but I guess the intercooler added a lot of volume to the system the initial charge got me 320 PSI on the high side running 25 psi on the low side no real cooling was occurring yet though maybe it needed more ethylene so I drained my tank down to 30 PSI from the original 120. the high side was now 350 PSI with a low side around 12. still nothing was really happening I drained my tank down to about 20 psi which brought my high side beyond the scale of the pressure gauge at least 420 or so PSI with the low side as high as 45. at this point I was seeing some cooling but the temperature never went below the evaporator temperature of the first stage meaning the temperature I was seeing on the evaporator of the second stage was probably just cold gas that had been chilled by the intercooler but not condensed this really left me scratching my head the second stage gas was being cooled to about -32c at this temperature the ethylene never should have gone beyond about 18 bar which is 250 psi but I was well past 400. maybe the problem was that I wasn't getting good heat transfer that seemed unlikely since these plate heat exchangers are supposedly extremely efficient but I needed to get to the bottom of this so I replaced the plate heat exchanger with my original coaxial coil that I built in the first Vapor compression video I ran the same test with propane in the second stage this time reaching as low as 75 PSI in the condenser instead of the earlier 90 which suggested that my coaxial coil was getting the second stage refrigerant colder than the plate stack was anyway it seemed to be working so I loaded my ethylene back in and tried again once again I went well over 400 PSI on the high side and things were definitely getting cold but not any colder than the first stage evaporator which again implied that all I was doing was cycling cold gas without any phase change so the problem probably isn't the heat exchanger there's actually two issues here for one the average temperature of the refrigerant on the high pressure side is much higher than the temperature where it comes out of the heat exchanger for example in this case the intercooler Outlet temperature was -32c but the propane pressure was 75 PSI or 6.1 bar but 6.1 bar corresponds to a temperature of 8.5 c not minus 32c similarly in the first stage the condenser Outlet temperature is basically ambient at about 20 4C but the pressure is around 200 PSI or 14.6 bar which corresponds to a temperature of about 43c not 24c I think the cause of this is that the pressure needed for condensation is a function of the average temperature throughout the high side the inlet temperature of the first stage condenser could be as high as 60 or 70c causing the average temperature to be significantly higher than ambient even if the outlet is ambient the second problem is that the ethylene I've produced is probably contaminated with hydrogen since I ran my Catalyst on the hot side the condensation and evaporation is based on the partial pressure of the refrigerant so for example if only 50 percent of the mixture was ethylene the partial pressure would be divided by two in this case if you were at a temperature where the ethylene would condense at 20 bar the total pressure with a 50 mixture would need to be 40 bar to condense which obviously isn't practical at all before proceeding any further I need to purify my ethylene and remove any hydrogen gas that's diluting it to do this I need to liquefy the the ethylene and then open it up to the atmosphere this would cause the hydrogen gas to float away leaving pure ethylene behind of course some ethylene would be lost to evaporation but what's left over would be very pure to ensure I could actually get my ethylene to liquefy I need to make sure that it was all at the same low temperature this would require having a system very similar to a Cascade refrigerator except the second stage would be static and open loop in other words a refrigerated tank that's pressurized with the second stage refrigerant the tank would be chilled to about -40c which should allow me to condense all my stored ethylene at under 15 bar since it's probably mixed with hydrogen though the actual pressure is going to be higher I didn't want to take apart the Cascade system I just built to do this so instead I went searching for something pre-built after venturing deep into the jungle I was able to catch a wild nocturnal ice maker while it was sleeping oh I stripped it down to its guts leaving just the compressor condenser and evaporator then I drilled a tiny hole in the evaporator to release the butane refrigerant inside and lit it up I bet you didn't know your ice maker is loaded with flammable gas then I wound my own evaporator coil with 3 16 tubing which would fit inside a one and a half inch pipe section A capillary tube is wrapped around the evaporator to get some sub cooling effect and drive it down to a lower minimum temperature the evaporator Inlet and Outlet will pass through this pipe cap in addition to a fill port and a well for a thermocouple the pipe cap with a single hole is for the drain line after some very messy braising here's what the evaporator assembly looks like and here's a look at that same evaporator crammed into the one and a half inch pipe section the edges of the fittings are soldered up to seal them I use solder instead of brazing because it would make the fittings a heck of a lot easier to remove in case I had to fix something no bubbles from the submerged pressure test at 300 PSI so far so good additional capillary tubing is wrapped around the evaporator return line to maximize the sub cooling effect and boost efficiency a needle valve is then added to the drain line with a soldered on brass fitting and another valve with a pressure gauge and flare fitting is added to the fill port for the pipe there's approximately 200 cc of internal volume inside the pipe and here's the assembly rigged back up to the ice maker system after spending a good 15 minutes vacuuming out the system I backfilled it with propane and let it run after a few minutes the pipe started icing up the system works but when I try to add more refrigerant the motor stalls out after a second this is kind of expected since the compressor was designed for butane or r600 which condenses at a very low pressure to remedy this I installed a one gallon buffer tank on the low pressure side of the system this added a lot of bulk to what had previously been a very compact assembly but allowed me to add a significant amount of refrigerant to the system while keeping the static charge pressure low enough that the motor would start up without any problems I also added a valve to this buffer tank so that I could adjust the amount of refrigerant I admitted to the system for a given run which allowed me more fine-tuning to further increase efficiency I enclose the cold bottle in a foam box which I stuffed with fiberglass wool and closed off I also replaced the propane refrigerant with map gas which is primarily propylene the boiling point of propylene is -48c compared to propane's -42 not a huge difference but certainly a noticeable boost this did bring the condenser pressure from around 150 with the propane to about 190 with the propylene but the compressor didn't seem to have a prop problem handling the change okay so here's the whole setup ready to go I chilled down the collection pipe and then start pumping in the ethylene from my beach ball at -38c the pressure is about 320 PSI or almost 25 bar this is another red flag that suggests the ethylene is impure because at this temperature condensation should occur at about 16 bar or 220 PSI based on these numbers I'd estimate that it's only about 70 percent pure with the remainder being either air or hydrogen another thing I realized is that while these little fridge compressors can put out up to three four or even 500 PSI the flow rate becomes absolutely abysmal past around 200. the fact that it took me close to four hours to pump down about 85 liters of the gas suggests that an average flow rate of about 0.35 liters per minute or 0.012 CFM is what we're working with the thermal load on the evaporator was so minuscule that I never even saw it move when I began filling the pipe this means even if I was successfully condensing ethylene at around 400 PSI in my Cascade system I'd be moving such a tiny amount of it that there would be practically zero cooling power in the evaporator going forward the fix to this might be to replace the fridge compressor with a rotator recompressor which has a much higher mass flow rate especially when the inlet pressure is raised anyway once the beach ball was pumped down I got my collection bottle and attempted to drain the condensed ethylene and it did kind of work the thermocouple showed minus 83c and you can clearly see how cold this was from the sinking condensation but unfortunately my flow rate was way too high and rather than dripping cold liquid into my collector I mostly just blasted it back out the inlet and it evaporated in the air unlike the dry ice I made in a previous video there's no Frozen chunks that get stuck to the walls in the container so everything just escapes pretty easily so yeah all that hard work basically just got pissed away in a couple of seconds by some bad planning kind of a good analogy for life actually I think I bit off a little more than I could chew so this project is definitely going to need a part two video where I correct some of my mistakes it's not a total loss though first of all I did technically achieve Cascade Refrigeration even if it was open loop and only for a few seconds you can clearly see from the frost and sinking condensation that we had an extremely cold liquid evaporating and not just cold gas being blown out the pipe I think I also got enough information to verify that my ethylene was indeed impure causing it to require a higher condensation pressure meaning the next batch will need to be made with a lower Catalyst temperature my little condenser unit also proved to be very useful and demonstrated that ice makers are a nice cheap way to create a small Vapor compression system going forward I can use this thing for making dry ice since I should only need around 130 PSI to condense CO2 inside my chilled pipe in the next video I'll attempt to make ethylene of a higher Purity and optimize my second stage a little bit better to hopefully get a cold side temperature of -100c once that can be achieved repeatably it won't be much more work before I can start making liquid methane and then liquid nitrogen thanks for watching and consider subscribing if you want to follow along with this project

Channel: Hyperspace Pirate

Views: 139,417

Keywords:

Id: T76UDLPv04Y

Length: 19min 15sec (1155 seconds)

Published: Sun May 14 2023