DIY Vapor Compression Refrigeration System

foreign I'll explain how a vapor compression refrigeration cycle works and how to build one for yourself this is a continuation of my ongoing project to create a DIY cryo cooler for liquid nitrogen production after some ill-fated attempts to make hydrogen to use as a refrigerant for my GM type pull tube cooler I realized that if I'm gonna have to dabble in chemistry to make this project work there's some easier approaches these involve either cascading multiple Vapor compression Cycles together or having a mixture of hydrocarbons in a jewel Thompson cycle either of these systems can be made relatively easily from standard Refrigeration or air conditioning components with a little bit of modification let's talk about the vapor compression option first if you're watching this there's a decent chance you're already familiar with how a vapor compression cycle works but I'm still going to explain it because it's important for context when this project gets more complicated in future videos if you've ever traveled to a city at a high elevation you may have heard that water boils at a lower temperature there because of the lower ambient pressure so if you did the opposite and went down to some place that's way below sea level where the pressure was a little higher than usual would that increase the boiling temperature and the answer is yes let's look at an extreme case of the high elevation example what would the boiling point of water be if you went extremely high up in the atmosphere or even out in space where there's almost zero pressure turns out the boiling point would be so low that your water would cool down and then freeze to ice and then even after that it would keep boiling and cool off through sublimation in fact liquid water is non-existent in space for this reason it can only exist as a solid or gas this is all because phase change of a liquid is dependent on temperature and pressure here's a detailed diagram that shows the relationship between temperature pressure and phase of water here we can see the familiar numbers of zero C and 100c for the freezing and boiling points respectively but that's at one atmosphere as we travel down on the graph we can see the window between solid and Vapor getting narrower and narrower until about 0.6 percent of one atmosphere where a liquid can't exist anymore in the case of our chunk of ice in Space the pressure would be near zero so it would still be losing heat through sublimation and settle somewhere around -60c so how can we use this whole pressure temperature relationship to our advantage to produce Cooling in the case of water we can actually build a refrigerator by having a liquid and then pulling enough of a vacuum on it that it boils at a temperature below ambient and takes heat from the surroundings effectively creating the outer space situation but right here on Earth in fact this is exactly how a freeze dryer works but that's not very efficient in terms of refrigeration so we'd like some other substance preferably one that's a gas at room temperature but with a relatively modest pressure could have its boiling point raised enough that it condenses to a liquid near room temperature and this is where we get the common refrigerants you've probably heard of like Freon the pressure versus temperature for that substance is shown on the graph here there's also R22 which has a similar pressure temperature curve but it's being phased out for environmental reasons ammonia known as r717 has a very similar Behavior though also extremely similar is a much cheaper and more commonly available substance which is r290 although you probably know it as propane its phase curve is almost exactly the same as R22 or ammonia the only downside is that it's flammable now it may sound kind of nutty to use a flammable substance as a refrigerant but if you own one of these tiny mini fridges or ice makers you're using a flammable refrigerant known as r600 or butane here's butane's pressure temperature line on the graph as you can see the boiling point at one atmosphere isn't quite as low as the others bottoming out at about zero C but it takes much less pressure to condense it which means it's favorable for tiny low power devices so yeah using a flammable substance as a refrigerant really isn't that crazy you might not want to charge the lines in your home air conditioning system with several pounds of butane or propane but it's certainly not a big deal in a small system besides these systems are sealed off from the environment so unless you have a bunch of monkeys hitting your refrigerant lines with axes the risk is actually pretty minimal fortunately I've exterminated the wild monkeys on my property so that won't be a concern it's time for some propane and propane accessories let's look at one of those face diagrams again but just for propane if we've got gas phase propane and we want to condense it at room temperature of about 25c we'll need an absolute pressure of 9.5 bar or 125 psi gauge however for heat to flow out of the system we can't be at exactly the same temperature as the surroundings we need to have a temperature differential so a more realistic number is maybe about say 45C which calls for a pressure of 15.3 bar or 210 PSI gauge I want my evaporator to get as cold as possible meaning I want the minimum pressure possible however some compressors don't handle sub atmospheric pressure very well so we'll Target one par for the evaporator pressure when the high pressure liquid propane expands and boils off to this pressure it absorbs heat until it cools to -42c to demonstrate this concept I've got a bottle of propane the bottle is brand new so it should be almost entirely liquid inside the bottle in liquid weigh 806 grams right now now I'll hook up a thermocouple to it it's the same temperature as their surroundings which is 19c meaning the pressure inside should be around 8.2 bar or 106 PSI gauge now I'm going to open the valve to release the propane note that this stuff is extremely flammable and you shouldn't do this inside near any ignition sources dry grass Etc you can see that the temperature starts to drop and before long a layer of frost forms on the bottle in theory we should be seeing -42c but since the environment is warming up the bottle we only get down to around -20 or so if I had a very thick insulation jacket around the bottle we probably would see around -40c and here's our weight after discharging we lost 153 out of the 400 grams in this bottle now you may be wondering why the entire bottle didn't empty out this is because to leave the bottle the liquid propane had to boil to a gas which took energy from the surroundings mostly the remaining liquid as more and more liquid boil to gas and left the bottle the remaining liquid got colder and colder and its vapor pressure got lower and lower as its temperature dropped until it finally got cold enough that its vapor pressure was equal to one atmosphere so there was no more of it flowing outward at this point we'd have to wait for the liquid to warm back up to discharge more gas I should point out that this is different from an air conditioning system or the target temperature of the evaporator coil might be somewhere between 5 and 10c which would correspond to an evaporator pressure of around five to seven bar for propane which would be much less work for the compressor because the pressure ratio would be so much smaller now of course there's got to be some way to drop the pressure so we need to add some flow resistance to the system between the output of the condenser and the inlet of the evaporator large air conditioning systems use a valve to do this automatically and it adjusts its opening based on temperature but you can also use a needle valve however in experiments I found the needle valve to be a real pain to deal with because it required super fine adjustment and the opening was so small that even the slightest amount of debris would pretty quickly clog it so the simplest solution is what's known as a capillary tube which is just the length of really small diameter tubing that drops the pressure of the flow with friction this is simple and reliable the only downside is that it creates a fixed pressure drop for a given flow rate so it's only optimal for one particular load this is a little bit like if you only had one gear ratio in your car's transmission also the system is going to have to be charged with more than one atmosphere of refrigerant if we just left our refrigerant gas in there at one atmosphere when the compressor turned on it would pull the evaporator way below one atmosphere and the condenser might be two or three atmospheres and practically nothing would happen in terms of temperature drop because there would be no face change happening the best way to add the correct amount of refrigerant is to turn on the compressor and then start adding refrigerant until the evaporator pressure is up to your target pressure in my case that's one bar absolute which would read a zero PSI gauge ideally enough refrigerant will be in the system that when it's turned off it's a saturated mix of liquid and gas meaning that your static pressure will correspond to a boiling point at ambient temperature which would be something in the neighborhood of 125 psi gauge for propane however for a system with a really low power compressor it might be all in a gas phase when the system is off which would mean it would have a much lower pressure than that that's enough Theory let's start actually building the heart of any refrigeration system is the compressor I got this one off of eBay but you can just as easily Harvest them from old fridges window AC units water chillers Etc the first thing I noticed was that it didn't have any oil in it and that's required for any Refrigeration compressor fortunately the label showed the exact amount required so I bought some Ester oil and filled it up I've heard you can also use mineral oil with these but as far as I know most of them run with Ester oil because it dissolves in most types of refrigerant which helps return it back to the compressor as it Cycles around the system so you don't need an oil separator next I start adding some quarter inch copper tubing and fittings to the discharge side for the condenser coil and discharge pressure gauge I made the condenser coil by winding 25 feet of quarter inch copper tubing around a four inch PVC pipe as a former and then added a filter dryer to the outlet this will filter out any traces of moisture in the system and also remove debris and the horrible smelling odorant that's added to the propane this was easy to make but I think it might not be very effective at removing heat from a larger compressor in the future I added a flare fitting to the outlet of the filter dryer and hooked up a pressure gauge to make sure the compressor was actually building pressure and there weren't any leaks in the system so far everything seemed okay next I started adding fittings to the compressor Inlet known as the suction Port this port used 3 8 inch tubing presumably since the low pressure refrigerant would be at a much lower density then I added a third line for filling refrigerant this was also a 3 8 inch tube now for the capillary tube this was one millimeter inner diameter copper tubing figuring out a good length took a little research though too long of a tube and you'll have a relatively high pressure drop but excessively low flow rate resulting in low cooling power and low efficiency conversely if the tube is too short you'll have a very low pressure drop and high flow rate but the temperature drop will be so low that very little cooling is actually happening I figured I'd be leaning a little more toward the long side since my goal with this project is to get the lowest possible temperature and not necessarily the best cooling power or efficiency I found a few tables online that suggested lengths based on power refrigerant type and evaporator temperature and then a calculator that determines equivalent diameters in length since the difference in pressure drop for a given length is extremely non-linear with respect to diameter I also found the simple little software program made by danfoss that allowed me to put in refrigerant type high and low temperatures and power and it spit out a result that seemed to roughly agree with the tables I found online ultimately I ended up using 10 feet or about 3 meters of the one millimeter diameter tubing for my half horsepower compressor there don't seem to be any adapter fittings available for such tiny diameters so what I did was crush the tip of a quarter inch piece of tubing down onto the one millimeter tubing until the gaps were minimized and then backfilled the gaps with solder the wall thickness relative to the diameter of the small tubing was so large that there's not really any risk of pinching it when you're crushing the larger tubing with pliers so you can put lots of force into this I turned on the compressor and tested it real quickly to make sure that there was air flowing through the capillary tube and no blockages in general you should avoid using these compressors with air but for a quick test it shouldn't really be a big deal the compressor pulls about 150 watts with the load from the capillary tube which is around 1.2 amps so pretty well within the locked rotor amperage of 11.7 next we're going to need an evaporator coil this is where all the refrigeration actually happens because liquid refrigerant which is at close to ambient temperature suddenly boils off and expands absorbing tons of heat and dropping the temperature as it does so in a simple air conditioner or freezer the evaporator coil is literally just a simple coil of wire with a fan blowing over it to cool off a room or a box but this refrigerator is actually going to be used in the future as a first stage for a Cascade system where it pre-cools the refrigerant of a second much colder stage so we need something a little bit fancier than that my evaporator coil is actually going to have a second coil of smaller diameter tubing running through the inside of it this is known as a coaxial heat exchanger to do this I took some elbow fittings and very carefully drilled a hole through the elbow so that the smaller diameter tubing would fit through without much Gap here's a shortened version to demonstrate what I'll be building the full-size version is actually 18 feet of coil if you're going to try this yourself I highly advise getting the outer tubing as straight as humanly possible beforehand because otherwise it takes a ton of strength to push the inner tubing through this took me the better part of an hour and I was sweating by the end of it my inner tubing was 3 16 of an inch and my outer tubing was 3 8 but if I did this again I'd definitely use at least one half or 5 8 inch tubing for the outer anyway once that's done I slide my specially drilled elbow fittings back on and solder up the gaps the trick to filling these large gaps is to heat the material just enough to get the solder mushy but not so hot that it becomes runny like water otherwise it'll run right through any gaps getting this right took a little bit of patience here's what the joint looked like when it was finished it wasn't very pretty but it held pressure and here's the finished evaporator coil winding this was even more difficult than threading the inner tube though because the double tube Arrangement made it much stiffer and harder to bend even around a relatively large radius like with the condenser I created my own jury rigged adapter to the capillary tube by crushing a piece of 3 8 inch tubing down onto a one millimeter tubing and soldering up the Gap now let's try to hook it up so far so good the only things left are the pressure gauges and the fill valve a one-quarter to 3 8 female npt adapter fits snugly on the 3 8 inch tubing and allowed me to easily seal up the gaps with solder the fill valve will be a Schrader valve just like what's on a bike or car tire I use silicone caulk to seal the threads which I've had far better success with than Teflon tape or plumber's putty the pressure gauge fittings are one quarter inch female npt caps that I drilled out to fit snugly on the tubing and then soldered the gaps I'm pretty happy with the result now let's fill it up with some propane for this I've got an adapter for propane bottles that connects to a needle valve and flare fitting for a refrigerant hose on the fill side of the refrigerant hose I've got an adapter with a flare fitting on one end and a female Schrader valve on the other everything seems to seal up very nicely now let's crack the valve and try letting some propane in note that I made a mistake when I filmed this assuming the bottle has no siphon tube it should be standing upright so that it releases propane into the system as a gas and not as a liquid I pressurized this to over 60 PSI and then put the entire assembly in the bathtub even the slightest leak would show itself as small bubbles which would tell me immediately what I needed to fix looks like the evaporator pressure gauge had a small leak fortunately it was an easy fix okay so now I ended up being pressurized to about 55 PSI after the temperature settled and I'm going to add this fan to move the heat out of the condenser coil I plug in the compressor and get about 1.4 amps which seems pretty reasonable one of my favorite Parts about this machine is watching the frost form on the evaporator coil this is about the only way you're going to get snow in Florida I've got the condenser running at close to 200 PSI and the evaporator close to zero so the cold end should definitely be close to -42c but it's only showing about minus seven my guess is that this is because of the lack of insulation on the evaporator coil to remedy this I added foam around the tubing and when I ran out of that I finished off the rest of the coil with some fiberglass insulation and then covered that up with some tape to keep it from crumbling away which it has a tendency to do sometimes this time I got much closer to my goal and settled at around -32c this actually makes sense since I ended up averaging a little more than one bar on my evaporator pressure and I'd usually sit at about 5 PSI gauge which corresponds to 1.34 bar at this pressure the boiling point of propane is -35c and we're within about three degrees of that on the outside of the tube so that seems about right now that we know this works let's try cranking up the power a little bit I'm going to put aside the modest little fridge compressor and replace it with this rotary compressor rated at 5000 BTU rotary compressors use a rolling piston instead of a reciprocating one and this geometry allows them to run at a much higher Inlet pressure their geometry also means that the entire shell is occupied by the motor so they have more power for their size than a reciprocating fridge compressor speaking of power let's break down that 5000 BTU number this is a familiar unit for HVAC technicians in America but from an engineering standpoint it's kind of an annoying one to use one BTU is about 0.29 Watts so 5000 BTU translates to 1450 Watts that's still not quite the whole picture though see it's pretty common for a well-designed air conditioning unit to have a coefficient of performance or cop of about three meaning that you pump three watts of heat out for every watt of power you put into the system and no that's not a violation of thermodynamics because you're just moving energy somewhere else you're not generating it that Assumption of a cop of about three is baked into the power rating of most compressors so if it was expected to produce 14 850 watts of heat pumping the nominal power of the compressor would be around 450 Watts or 4.1 amps at 110 volts of course this number will vary a good amount based on the load and the temperature but it gives us a ballpark idea of what to expect another metric I found for estimating a safe power consumption on your compressor is the maximum operating amperage is typically right around 20 percent of the locked rotor amps so for example the compressor I'm going to use has an lra of 27 meaning if we're seeing more than about 5.4 amps there's too much load on it this makes sense since as you'll see later in the video the compressor started to become alarmingly hot at around 5 amps here's what the system looks like after adding the bigger compressor I added this convenient switch box fancy 3D printed feet for the compressor a nice printed cover for the start capacitor so I don't electrocute myself on the terminals and I've covered the entire evaporator coil assembly and shiny aluminum tape to insulate it just a little bit better by reflecting any infrared heat that tries to get in from the environment by the way I forgot to mention before that when you're charging one of these systems for the first time you have to make sure to suck the air out with a vacuum pump first this is because the air is taking up space that more refrigerant could have occupied and carrying some moisture in it that could damage the compressor also for a flammable refrigerant having air mixed in could form an explosive mixture this time when I fill with propane I'm going to weigh the bottle as I'm doing so this is because the pressure alone isn't a reliable indicator of refrigerant Mass when the refrigerant is in a saturated state it could be 100 liquid or 100 gas or any combination in between I zero out my gram scale and start filling this video was cut short before I could finish the fill but I finished at about 49 grams of refrigerant I estimate that the internal volume of the system is around 400 cc meaning the average density of the refrigerant in the off state is about 122 kilograms per meters cubed at 25c liquid propane has a density of about 490 kilograms per meter cubed and gaseous propane is about 20. using some simple math I can determine that we're at about 22 percent liquid by volume when the system is at rest when I first fire up the system we start off right at around zero PSI gauge on the evaporator and 180 PSI gauge on the condenser in theory this means the evaporator should be right at -42c and the condenser should be at around 39c both these numbers look really good within seconds ice forms on the evaporator another good sign but when I probe the evaporator Inlet with a thermocouple we're only seeing about -27c a bit of a ways off from the target -42c then I notice the condenser pressure is rising progressively higher passing 250 psi and approaching 300 this means the temperature is getting close to around 60c I tape a thermocouple to the condenser Inlet and another one to the outlet the temperature readings confirm my suspicions the inlet has risen to over 60c and the outlet goes over 40c this means the coil isn't getting rid of enough heat the temperature drops a little bit between the inlet of the filter dryer and the start of the capillary tube but this is still going up as well then I measured the compressor which exceeds ADC and was still climbing when I shut it off it's pulling a decent amount of current but well within the rating of the motor this has to be heat from the condensation this explains why the coldest part of the system reached only about -27c rather than the target of minus 42c unfortunately my copper coil just isn't going to cut it for the condenser I need something that can get rid of more heat and what better to use than an actual purpose-built condenser coil from an air conditioning unit you can buy these Standalone but I found the best thing to do is just grab them from used air conditioners worst case they just need a little cleaning before use you can find these pretty easily at local flea markets junkyards or crack houses or you could just Harvest them directly out of people's windows or one of those weird floor mounted units that hotels have not advocating theft but taking other people's stuff is probably okay if it's for scientific reasons so I got this window AC unit and stripped it down this got me two decent sized coils uh 35 microfarad capacitor and even a new compressor of course this begs the question why didn't I just start off this project by tearing down a window AC unit instead of putting together the individual Parts the answers that I have an incredibly deep junk drawer so I already had most of the parts from other projects if your whole goal was to just get a really low temperature as cheaply and simply as possible with the window AC unit a bit of copper tubing and brass fittings is really all you'd need the window unit came with a 35 watt fan but I wanted a little more assurance that my condenser would be kept cool so I strapped a 65 watt radiator fan to it the only downside was that the inrush current at startup was over 15 amps so it kept resetting my 10 amp dc Supply I guess I'll just run it off my bench Supply until I get a suitable alternative I kept a smaller fan from earlier to cool the compressor casing and so both fans pull 1.2 amps at 115 volts we'll keep this in mind for power readings later now I turn on the compressor and wash the condenser pressure climb it settles out just a hair over 150 psi meaning the average temperature is much lower than before I put thermocouples on the condenser Inlet and Outlet the inlet is almost 64 degrees but the outlet is sitting right at ambient temperature this is exactly what I want to see with proper cooling ice forms all the way up the evaporator line and even the casing of the compressor's liquid receiver my meter reads 5.4 amps subtracting the 1.2 for the fans means the compressor is running at 4.2 which is perfect with the additional cooling the compressor shell never exceeded 70c it's still toasty but well within limits now the layer of frost on the exposed portion of the evaporator is pretty impressive here's a little icicle forming on the fill line the left side isn't cold enough to freeze so the liquid water flows down around the elbow until it hits the colder pipe section and freezes solid the frost layer is several millimeters thick after just a couple minutes the evaporator Inlet settles right around -37 there's probably one or two degrees of difference between the inside and outside of the pipe and I'm slightly over one atmosphere of pressure on the evaporator so this number looks about exactly right the evaporator Outlet sits a few degrees warmer at around -34c this makes sense because the insulation isn't 100 effective as you can probably tell by the condensation on the aluminum tape here the fun part about turning off the machine is watching the ice melt and recede back up the evaporator line the last thing I want to do is measure the actual cooling power I got this 12 volt water pump and I'm going to hook it up to the inner tube of the evaporator first I check the flow rate I want to make sure there's enough water flow that it won't immediately freeze up and stop up the tube with ice I measured 1000 grams in 74 seconds or 13.5 grams per second it would take more than a kilowatt of cooling power to drop the temperature 20 degrees from the inlet to the outlet and we definitely don't have that much cooling power so there shouldn't be any risk of immediate freezing now I run my return line back to the same Reservoir this is a glass Beaker filled with two liters of water and I'm going to measure the temperature drop with respect to time this isn't a perfect test because heat from the surroundings will warm the beaker and the tubing somewhat but the temperature difference will be relatively small and the materials involved have poor thermal conductivity so the error caused by warming shouldn't be that big I have a thermocouple in the reservoir but also one in the return line to tell me the differential generated across the coil we start off with the reservoir and ambient temperature 24.8 C it quickly begins to fall when the compressor is turned on here you can see the condensation on the beaker and the tubing causing them to look cloudy as they both drop below the ambient dew point the lines were between 3.2 to 3.6 degrees below the reservoir temperature throughout the test here's a graph showing the numbers I recorded the data yielded a very clean straight line and the kerfit gave an average drop of 0.83 C per minute multiplying that by the heat capacity of my reservoir gave me a cooling power of 116 Watts my input amperage settled at 5.8 and voltage was 115 meaning average input power was 667 Watts resulting in a coefficient of performance of about 0.173 this is kind of disappointing considering that for this temperature the cop should be really closer to around 1.0 but there's a few factors influencing this first off 138 Watts out of those 667 are just fans and I probably have way more fan power than what's actually necessary second imperfect heat transfer from the test water to the refrigerant in the evaporator this one was pretty obvious Because by the end of the test the evaporator Outlet was starting to ice up a little bit while the test water was well above freezing temperature third losses in The Test Tubing and reservoir from the environment which I mentioned earlier unless these elements were perfectly insulated some heat would leak in after cooling resulting in a measurement that's below the actual cooling power the fourth and by far most important factor is a load mismatch remember how I said before that a capillary tube is kind of like having a single gear ratio on a car transmission and there's only one condition where it would be optimal yeah we're really far from that condition because my capillary tube is sized for Peak efficiency at the lowest possible temperature but my load never went below 10c if I wanted Optimal Performance at that temperature I'd need a much much shorter capillary tube for a smaller pressure drop which would in turn reduce compressor power draw by lowering the pressure ratio as it's currently set up my condenser runs at around 11 bar and my evaporator is just over one bar this is a very aggressive pressure ratio meant to maximize the temperature drop in a future video I'll be using my evaporator coil to cool the condenser of a second Refrigeration stage this second stage refrigerant will then go on to evaporate to a much much colder temperature closer to -100c I'll be using ethylene for the second stage refrigerant which is very hard to buy so in the next video I'll show how to make it thanks for watching

Channel: Hyperspace Pirate

Views: 543,415

Keywords:

Id: 7QZrHzd3RA8

Length: 28min 5sec (1685 seconds)

Published: Mon Apr 10 2023