8

A concern that comes up regularly when troubleshooting A/C problems is whether an A/C system has too little refrigerant (hello leaks) or too much of it (hello overcharging).

The traditional way to do this involves drawing out all the refrigerant (often with expensive, specialist equipment), which adds considerable overhead to a DIY-er like me as I then have to worry about sourcing compressor oil or recycling the existing refrigerant.

So, is there an environmentally-conscious way to determine the amount of refrigerant present inside an A/C system that does not rely on complete evacuation?

Zaid
  • 39,111
  • 48
  • 142
  • 290

1 Answers1

2

You can draw out all refrigerant from a system with no expensive equipment. All you need is a hose (preferably with manifold gauge), a suitable storage tank, a decently insulated container (like a large cooler) that it fits mostly inside, and dry ice to pack around it, around 10 pounds, or less if you have a liquid acetone or alcohol bath to transfer the heat more efficiently out of the tank. Dry ice is around -70°C, and automotive refrigerants liquefy around -25-30°C, so all (modulo some extremely low vapor pressure) the refrigerant will liquefy in the tank; you can confirm this by observing vacuum on your manifold gauges, which should match the vapor pressure of your refrigerant at whatever temperature you can get the tank down to.

Dry ice will run you about $2/lb and can be purchased at Kroger and some other grocery stores.

You can get a 30 lb tank for around $80-100. I got one off Amazon. In theory, if you have an empty refill tank that can hold at least as much as your system, you could use it. This is of questionable legality, and dangerous if there's any risk you might overfill it (assuming it might return to a gaseous state before you move it back, in which case an overfilled tank could explode), but it's not as bad as actually pumping hot pressurized gas into such a tank.

Here's a picture of the setup I've used, in progress:

enter image description here

R.. GitHub STOP HELPING ICE
  • 5,219
  • 6
  • 36
  • 46
  • It seems like you've gone through this exercise before. Am I right in saying that you will be left with some residual refrigerant inside the system with this technique? – Zaid Sep 15 '18 at 06:33
  • @Zaid: It depends on what you mean by "some". There's always some vapor pressure, but it's very low. You should see at least 25 in of vacuum, ideally closer to 29. If you don't, it probably means there's something (air) that doesn't condense to a liquid contaminating the system. – R.. GitHub STOP HELPING ICE Sep 15 '18 at 06:57
  • If you figure the system is normally something like 60psig, 25 in of vacuum would mean you've evacuated all but about 1/30 of it, and 29 in would mean all but about 1/150 of it (<1%). – R.. GitHub STOP HELPING ICE Sep 15 '18 at 07:08
  • This has results that are much better ie 1% compared to about the 10% of the other answer. – Solar Mike Sep 15 '18 at 08:42
  • What I'm wondering is, how does this answer the question? The question specifically asks "... without evacuating the entire system ..." While your method may be accurate and can be done by the DIY-er, you are evacuating the system to make it happen. – Pᴀᴜʟsᴛᴇʀ2 Sep 15 '18 at 12:29
  • @Pᴀᴜʟsᴛᴇʀ2: Technically that's right, but I read it as an XY problem for "how do I do this without expensive equipment?" and I think such a reading is supported by the text of the question. – R.. GitHub STOP HELPING ICE Sep 15 '18 at 15:01
  • Realistically, with just evacuating the system I'll know the amount of refrigerant inside the system: zero (or near zero). The whole point of the question @Zaid poses is to figure out the amount refrigerant without the evacuation. I'm not saying I disagree with your methodry or with the posed outcome, as it will still give you the amount of refrigerant in the system. This method may actually be easier and ultimately more complete for ensuring the proper charge in the system. – Pᴀᴜʟsᴛᴇʀ2 Sep 15 '18 at 15:08
  • @Pᴀᴜʟsᴛᴇʀ2: The assumption is that, after you measure it, you put it back. You'll be left with some stuck in the tank though, which is a motivation for using the "reuse a small tank" approach I discouraged. There may be approaches with returning it in liquid form that allow transferring back almost all of it, though. I could write some of them up but I haven't tried them personally since my application wasn't measurement and I had excess (to leave in tanks, of course). – R.. GitHub STOP HELPING ICE Sep 15 '18 at 15:14
  • And of course if your goal of measuring is to reach a known final mass in the system, you can just subtract what remains in the tank after putting it back, then top off with a known amount of new refrigerant. – R.. GitHub STOP HELPING ICE Sep 15 '18 at 15:17