Finding an air conditioner refrigerant leak that's completely empty

Question

I have a central air conditioner that worked briefly when recharged professionally, but the refrigerant all leaked out within about a month, so there's a pretty substantial leak somewhere. It's a "York" branded system of unknown age.

A HVAC company come out and diagnose the system, they could not find the leak and just used "stop leak". Needless to say, one month and $1100 later, the system is dead again.

I have worked on automotive air conditioners. Normally to find a leak I would charge the system with refrigerant and dye, then either use a refrigerant sniffer or UV light to find the leak.

However, I understand that R-410A requires a license so I can't just buy some and shoot some in there, so I was considering using compressed air from a portable air compressor, then soap solution.

Is this a bad idea or what would be a better way to diagnose the system?

Edit/Add: A small 20-lb tank of nitrogen would be about $200 filled around here. Is there some other reason not to use shop air besides moisture? I have inline water separators and desiccants I can use with the compressor. The system is already full of air since it has a leak.

Also will the dye show anything if the system is just statically pressurized? Will it just settle to a low point in the system unless the system is turned on?

Answer 1

The gold standard for that kind of dry pressurization is nitrogen. You could use argon, but your contractor won't be very happy about it and may refuse to issue a warranty on their work. Yes, I know argon is a noble gas (that doesn't necessarily mean argon welding gas is pure argon, either).

If it's that old, think about replacing it. I just saw a video by a contractor in the corrupt old air conditioning business, the one that refuses to modernize, and they showed how if you tore out a recent 13 SEER system (coarsely, 13 BTU/hr per watt) and replaced with the current bottom-end 14 SEER (minimum legal to sell these days) it would take 30 years to earn back the energy savings.

And I thought to myself "wait a minute. If that's the payback on such a miniscule upgrade, the payback on a wider upgrade would be vastly better." So what was your system at its best, maybe 8-9 SEER? Suppose you go even to 14 SEER you're talking maybe 5 year payback.

If you can get into some of the 25+ SEER Asian stuff your payback would be a lot better than that! Unfortunately most contractors have their livelihoods married to the corrupt old industry, and they will not help you do that. The industry does have knock-offs of the Asian stuff, but they make them grudgingly and in very low production, so they're not very good. And the prices are over the moon. Zero effort from a typical contractor to up-sell you into them -- they seem to only exist in the catalog to get you to decide "it's not worth it" and buy the same old "That 70’s A/C".

It's madness, look at any Ukrainian apartment block, every other apartment has a modern system from Asia. Yet to get a comparable system from the old domestic makers costs 250% of the annual wage in Ukraine? Give me a break.

Answer 2

If pros filled and ran the system and could not find a leak, that means the leak is very small. If it's very small the only way to find it is to charge the system with added dye and let it run for a long time ... that could mean weeks or months depending on the weather .... long enough for sufficient amounts of dye to leak out through the tiny holes that the cumulative mess it makes on your coils becomes visible. You cannot do this with nitrogen.

Also ... if the leak is very small it's probably in the evaporator coils. There is a certain amount of gambling you have to do here, and if you have a trusted technician you should go with their instinct. If the system is more than 10 years old, you know it's leaking but you don't know where, and you are 90%+ confident (not you, but your experienced and trusted technician) that the leak is in the evaporator ... there is no point proving it with dye. That test costs a lot of money.

If the system is in warranty you HAVE to do the dye test. If it's out of warranty, and if you are paying typical labour costs, it's not worth doing the dye test OR replacing the evaporator coils. Just replace the entire indoor unit so you get a new (hopefully 5 year) warranty.

Some therapy for you: This sucks. It really, really sucks. Get over it. I'm sorry. Don't try to save money by doing mickey mouse tests of your own design with substitute gases.

Answer 3

A leak of this severity should have left a wet oily mess in its wake. You don't need UV dye to spot that. Inspect the condenser coils, look at the compressor, the area around the service valves, the line set, etc. If there's no door that allows viewing/cleaning the evaporator coil then cut a hole in the ductwork so you can see into there (beware of the evaporator's tubing just behind the ductwork; you don't want to damage those!). Line sets can leak too unfortunately, especially if there's a joint somewhere, and if it crosses finished spaces then it's pretty hard check for this.

It's really unfortunate that service call cost so much -- before spending that much again I'd think really hard about system replacement. If you find the right contractor the next $1100 might be 25% of a new system installed.

Regarding shop air: it will contain moisture, could contain oil carry-over from the compressor, could contain debris degrading from the interior wall of a hose, could contain dust or whatever happened to be collected in the blow nozzle or other hardware/fittings you might use for connecting to the A/C system.

Then again, your leading alternative to fixing the leak is buying a new system. So if in the effort to find the leak you happen to do something that shortens its life a little or a lot, "nothing of value was lost." It's an opportunity to learn something anyway, and that might or might not have personal value to you.

I think you're correct that adding dye without refrigerant will be unproductive because the system cannot be run to circulate the dye.

One thing to be aware of is that normally an A/C system is charged with the refrigerant bottle inverted so that both oil and refrigerant are dispensed into the system. Too much re-charging can leave you with a system that doesn't work as well as it should because of the excess of oil (because the oil consumes volume that ought to be occupied with refrigerant gas/liquid). It's a bit of work to open a system sufficiently to remove some of that oil. I don't have any hard numbers to share, but I can say that most of the time when I replace a system I cut the tubing open and nothing drips, but I've had a few "we've had some leaks" systems that, when cut open, drained out a few ounces of oil and I had to work quickly to mop up the mess.

You could ask the service person to put dye into the system and recharge it with reclaimed R-410A. This is the stuff they capture out of other systems they're upgrading. Normally they would turn it in for recycling and receive some credit for it, but the value is small compared with the cost of new refrigerant. There is a risk that the reclaimed refrigerant may be contaminated and, like testing with shop air, this means a risk of damaging your system. But perhaps it's less expensive and worth the risk in your situation.