View Full Version : I refilled an R22 window AC with R134a and this is what happened

28-05-2017, 05:16 AM
I did this purely as an experiment, NOT because I thought it would work and I definitely do NOT suggest that anyone try this with a system that they care about. I got a free 8000 BTU Kennmore window AC from the 1990s that I use as a dehumidifier. Over the winter, it developed a slow leak and was empty yesterday when I turned it on and did not cool at all. I installed a Bullet Piercing Valve and filled it with compressed air to find the leak. The condenser had a pinhole, which I soldered closed and vacuumed the system to remove any air or moisture. I turned it on, connected the gauges and started adding R134a refrigerant. It took about 12 ounces of the original 13.6 ounces until it was cooling well and was not developing frost on the evaporator. The performance of the system is identical to with R22 refrigerant. It blows 38 degree air with a room temp of 75 degrees with the fan on high speed. I thought that the air would be warmer because I thought that R22 was better than R134a in terms of efficiency. But what I thought was the most interesting is how much cooler the compressor runs. With R22, the compressor and discharge tube got so hot it would hurt to touch. With R134a, the compressor runs MUCH cooler. It is still hot, but I can hold my hand on it for at least 5 seconds without getting burned. Same with the discharge tube. I have no way of testing it, but I would think that power usage would also be lower. Besides the obvious oil issue, what is the disadvantage to running the system this way? The pressures are MUCH lower, the system runs cooler and even seems quieter and it seems like it vibrates much less, but that makes no sense to me. I have no idea about longevity as I just started using R134a in this system, but it seems like it works great. What are your thoughts?

28-05-2017, 07:51 PM
'filled it with compressed air'...says it all.

28-05-2017, 11:05 PM
A fridge tech mate of mine uses compressed air for pressure/leak testing,it's something I would not recommend.

29-05-2017, 05:03 AM
'filled it with compressed air'...says it all.

I am not a professional tech, so I do not have pure nitrogen to pressure check the system. Besides, the system sat for the winter empty, so it was probably full of air anyway. I vacuumed the system for 3 hours anyway, so any air or moisture that was in the system was removed anyway.

29-05-2017, 02:24 PM
OK to experiment,but in my opinion the system will soon fail because of the oil issue how long it will take not sure but it will fail, unless your lucky?

29-05-2017, 05:18 PM
OK to experiment,but in my opinion the system will soon fail because of the oil issue how long it will take not sure but it will fail, unless your lucky?
I understand the oil issue, but heat kills compressors, right? Since for whatever reason the the compressor runs so much cooler, maybe that will somewhat offset not getting as good of lubrication? The compressor is running the smoothest, coolest and quietest it ever did since I got this system, so I think something must be working right.

05-07-2017, 06:46 AM
If you are lucky. What do you have to lose. you got the unit free. You paid a lot less for 134a. let it run. I think it will last another 10 yrs lol. From my lips to gods ears

09-07-2017, 09:59 PM
I am not sure what kind of oil the system has in it now. I was telling a friend about this test, and he said that the system likely uses alkylbenzene oil instead of mineral oil. How can I find out? Is there a sticker somewhere? I am not sure where I would find the sticker for the oil type, if it has one.

09-07-2017, 10:03 PM
Another interesting thing that I noticed is that the evaporator does not freeze up on cool days anymore since the conversion. The compressor does not seem to be wearing out or anything, it still sounds great and blows 38 degree air with a 75 degree ambient.

12-07-2017, 11:49 AM
As for the oil, that comp is probably hermetic piston compressor, right? They come precharged from the factory according to refrigerant they are meant to be used with. So the way to find the oil type would be to contact comp manufacturer... 27 years old window unit, ouch... But even if you would want to replace the oil, I still think it would be more cost effective to buy brand new window unit, let this run as it is for as long it will run...

15-07-2017, 12:07 AM
How does it compare with electricity running costs?

I have been to jobs where other companies have used R404 a and R 407 a in place of R22, and the system has run but in heat pump mode it has run cooler and the compressor did seem cooler whether it was in cooling or heating mode. The oil inside the original system was mineral and is not an ideal selection to be use with a zeotropic blend such as 404 a, which should have a synthetic oil. They also notice their electricity usage had gone up.

I replaced it with R 422 D and the the overall performance was definately noticeable straight away, and the electricity usage had almost dropped back down to before it was changed form R22.

Any refrigerant will usually work in any system, but its all about horses for courses, and if you work the horse and don't give it clean water or hay it will die.

The type of oil also has a big part to play and will eventually destroy the compressor if it is not compatible with the type of refrigerant that is in the system. Just like an engine using 2, 3 or 4 star, or a high octane fuel only recommended for use with a racing engine, and would probaly destroy a Ford model T engine or any other standard 4 engine within a couple of hours. And the correct oil viscosity also needs to be considered.

Using compressed air in the system for any reason can contaminate the oil and its effect might be noticed later. Did you change the dryer or leave the old one in?

Peter paner
21-08-2017, 03:47 AM
wow. I am also very interested in this matter. (https://www.quangminhhanoi.com/)

22-11-2017, 11:35 PM
I think the reason for the coil frosting up was a refrigerant shortage caused by the leak you have found. Because of the shortage of refrigerant, the suction vapour has not been cool enough to keep the compressor cool, and the heat from the discharge pipe is more the convected heat from the compressor rather than the heat of the refrigerant. That said, it isn't unusual to burn ones fingers on a discharge pipe of any system, if it is running correctly.
One would expect a suction pressure of around 58 psi, for a CAPILLARY based AC system operating on R22.
Because R22 is very good at extracting heat, the condenser coil for R22 would be a lot larger than a system designed for R134a.
I would be concerned about refrigerant velocity through the low side of the system, as it is this that returns the oil to the compressor. I would also be concerned as to whether the oil is mineral or synthetic: Some compressors are marked as R22/ R407C these would have synthetic oil in them. If it is only marked as R22, then it would be a good bet that it is a mineral oil.
You'd be better using R417a....

15-04-2018, 07:22 PM
UPDATE: It's been over a year, and the system is still running great. However, I'm going to be filling it with R290 very soon, just to try it. I'll report back with the results.:)