R32 Mini split/heatpump lineset leak detection.

[Luk88]

Tldr: Nitrogen at 25bar/360psi shows no leak but vacuum seems to leak a but.

First let me say thanks for the great video by Will about a mini split install

I'm having a possible problem with leak detection on my no-name chinese 8kW R32 air to water heatpump.

The manual says to pressure test with nitrogen at 45Bar (650 psi) ! I have a regulator that goes up to 50 bar so thats not a problem but pressure testing lines with such pressure seems completely bonkers. So i tested with 25bar (360psi) as i found r32 pressure is in the region of 20bar.

No leaks found. So then I wanted to pump down vacuum. I have a gauge built into my vacuum pump (and an automatic cutoff valve. Some 6mm ptfe line and 1/8th sae to barb on pump side and 6mm push in to 1/4 sae on pump side.

But the pump has schrader valves. High pressure can push through, but to pump down a stem has to be pushed in. I've inserted a 3in long piece of solid copper wire and made a loop in the hose to immobilise it. So when I push in the push in connector it presses the valve.

I pumped it down until the gauge needle sits on the stop and i left it for a bit. Upon return i discovered some tiny amount if air must be leaking through, but I'm unsure where from. In a span of 5 minutes the needle moves about 1mm up (where it initially says -1000mbar it moves to maybe -990mbar).

The question I have is. Is it likely i indeed have a leak in my lineset despite pressure testing with nitrogen? Or is the leak likely between the pump and its fittings and nitrogen testing at 25bar/360psi is better than vacuum?

 

[Plum Crazy Rob]

I had this exact thing happen, and so left Nitrogen on for a couple days (be aware Nitrogen PSI is affected something like 0.9 PSI per degree F change). Nitrogen pressure was rock solid outside of daily outside temp swings. 8 months later it's working fine. I'd be perplexed how you can prove it's your pump setup not the unit (plug end of pumpset and evacuate / see if same leak condition??)

 

[jpwhit]

Vacuum testing isn't actually a very reliable way to check for leaks. If you have a severe leak, that will show up in vacuum testing, but most small leaks do not. Often, it's because a vacuum pulls components tighter together in the system and will actually stop most small leaks.

You always relay on the high-pressure nitrogen test as the determination for if the system has a leak. And under high pressure nitrogen, don't rely on just the pressure dropping. Use a bubble leak solution on all likely leak points such as all flare joint and brazed connections.

If you want your minisplit systems to stay leak free and to be as reliable as possible, then cut off the flare fitting at the head unit and braze those connections like you do on a normal split HVAC system. That's because for pipe-to-pipe connections, flare joints are much more likely to leak over time than a brazed connection due to flexing of the pipes that comes with thermal expansion and contractions, and the fact that often you have to put pressure and bend the pipes during install and uninstall of the head units. Flare connections at the outdoor condenser do tend to stay well sealed even over a long period of time. That's because one side of the connection is firmly mounted so there is much less stress on the connection due to thermal expansion and contractions. Also, those connection are typically very assessable, so the tightening torque is much easier to get spot on. Always follow the proper two step tightening sequence and use a high-quality torque wrench made specifically for flare connections. I'm fine with using economically priced tools for most HVAC equipment, but the torque wrench for flare connections is one place where I'll only use a name brand like Yellow Jacket. Same for leak detectors. Only a brand like Field Piece or Yellow Jacket.

In your case, I'm 90% sure you have a leak in your pump gauge setup. You really should be using a quality manifold gauge set with proper connections to the vacuum pump. I'd also never use a vacuum pump with an automatic shutoff, because pressure sensor built into that kind of pump tend to leak.

 

[Zwy]

Tldr: Nitrogen at 25bar/360psi shows no leak but vacuum seems to leak a but.

First let me say thanks for the great video by Will about a mini split install

I'm having a possible problem with leak detection on my no-name chinese 8kW R32 air to water heatpump.

The manual says to pressure test with nitrogen at 45Bar (650 psi) ! I have a regulator that goes up to 50 bar so thats not a problem but pressure testing lines with such pressure seems completely bonkers. So i tested with 25bar (360psi) as i found r32 pressure is in the region of 20bar.

No leaks found. So then I wanted to pump down vacuum. I have a gauge built into my vacuum pump (and an automatic cutoff valve. Some 6mm ptfe line and 1/8th sae to barb on pump side and 6mm push in to 1/4 sae on pump side.

But the pump has schrader valves. High pressure can push through, but to pump down a stem has to be pushed in. I've inserted a 3in long piece of solid copper wire and made a loop in the hose to immobilise it. So when I push in the push in connector it presses the valve.

Spend the money for equipment or farm it out. This is not the place to be cheap or rig up something.

I pumped it down until the gauge needle sits on the stop and i left it for a bit. Upon return i discovered some tiny amount if air must be leaking through, but I'm unsure where from. In a span of 5 minutes the needle moves about 1mm up (where it initially says -1000mbar it moves to maybe -990mbar).

Could be many things, even some nitrogen pressure remains in the system.

The question I have is. Is it likely i indeed have a leak in my lineset despite pressure testing with nitrogen? Or is the leak likely between the pump and its fittings and nitrogen testing at 25bar/360psi is better than vacuum?

Only accurate way to use vacuum for leak detection is with a micron gauge. Vacuum pump quality and oil condition play a huge part in achieving good vacuum. My pump oil is changed when vacuum pump can not pull below 200 microns.

A high quality gauge set is needed as cheap manifold hoses can lead you astray. As for the copper wire, it is quite possible it is creating a situation where either the schrader core is not being depressed, core is pressed too far (yes, I've seen it) or restricting the connection so only a tiny amount of flow can be achieved. I'd source a high quality gauge set before proceeding.

 

[Plum Crazy Rob]

You really should be using a quality manifold gauge set with proper connections to the vacuum pump.

Yes, agree. All my stuff Chinese/Amazon vacuum pump gauge set lines etc. Got a much warmer fuzzy feeling from solid Nitrogen test.

 

[Luk88]

Thank you for responses

The pump is high quality and new. I specifically bought a new vacuum pump for this as my old pump is dirty. Oil is new too. I've been using another one of these pumps for years (for various vacuum experiments etc so I know it is high quality - despite the brand literally being called "Value" it is a good local made brand).

I don't really see a problem with my nitrogen testing. I used quality fittings just without the schrader pin, but at 25 bar that seems hardly a problem. The idea some nitrogen was left in the system and somehow it showed up after 5 minutes after the vacuum was lower, I have to say sounds very weird for me. Some materials are known to outgass but copper?

I've ordered a nice gauge and hose set and if there is doubt I'll wait, but haven't you both just said nitrogen is much more solid test than vacuum? So if nitrogen shows fine (at 360 psi) isn't an indication it is fine? Or does it have to pass both a nitrogen and a let's say a 1h vacuum test. And if so, what pressure should it be after an hour? (If it starts at gauge needle bang on at stop - this pump is capable of 2pa or
0.015mmhg while the manual in theory allows 10mmhg (seems a pretty crappy vacuum). Is any increase allowed? Or none at all whatsoever?

Also can you comment about what you think about the idea to test under 45bar (650psi) of pressure? I found few discussions about this on hvac forums and the way I read it, it seems way too much, but I'm not sure.

 

[Luk88]

Guess what. You were right... It was residual nitrogen. I let it pump for half an hour, then I leftvit off for an hour. The needle on the vacuum gauge has not moved even a hair. So it just needed more time tobpump down.

 

[Luk88]

I have an additional question. I watched some additional mini split install videos and I saw this procedure to torque the flare nuts, back them off quarter turn and torque again. Unfortunately my install manual doesn't say so and if Will mentioned it in his install video I must have missed it. So I simply torqued the nuts to the proper value and that's that. I also haven't used any nylock or any-other product. Should I be preparing to redo my lines when it inevitably leaks, or should this hold despite the error?

Just a reminder my vacuum held for an hour, following a nitrogen test at 25bar (350psi) earlier. So I went ahead and let refridgerant in. I run a heating and a cooling test and I went over all fittings with dishsoap water to ensure no leaks again.

 

[Zwy]

Guess what. You were right... It was residual nitrogen. I let it pump for half an hour, then I leftvit off for an hour. The needle on the vacuum gauge has not moved even a hair. So it just needed more time tobpump down.

The issue with the schrader core is what made me bring this up. Commercial refrigeration on large systems, the core is removed to eliminate restrictions. This is done using a special tool that allows for core removal and installation under pressure.

25 years ago I had something similar on a R134A system where the core had been broken off inside and the seat remained. When depressing the core, the new core would push against the old seat and cut off flow. I kept thinking I had a leak as the vacuum kept rising when I turned the pump off.

I have an additional question. I watched some additional mini split install videos and I saw this procedure to torque the flare nuts, back them off quarter turn and torque again. Unfortunately my install manual doesn't say so and if Will mentioned it in his install video I must have missed it. So I simply torqued the nuts to the proper value and that's that. I also haven't used any nylock or any-other product. Should I be preparing to redo my lines when it inevitably leaks, or should this hold despite the error?

I use Nylog on both the flare surfaces and on the backside of the flare so less friction to the nut. As for tightening, backing off and tightening again, this is more a process to relax hoses back when hoses used flare ends. Not needed with hard lines but if there is friction, it can twist the line.

As long as it held high pressure nitrogen and you checked using bubbles, it should not leak.

Just a reminder my vacuum held for an hour, following a nitrogen test at 25bar (350psi) earlier. So I went ahead and let refridgerant in. I run a heating and a cooling test and I went over all fittings with dishsoap water to ensure no leaks again.

I will nitrogen test for several hours. I run electrical and other things while nitrogen testing. One does need to account for any temp change. I've let systems sit overnight with nitrogen and check pressure in the morning and compensate for temp.

As for vacuum, I run a micron gauge and will only charge when the reading stabilizes at rest under 500 microns. If I see under 200 microns with pump running, I know the cake is ready to come out of the oven.

 

[jpwhit]

I use Nylog on both the flare surfaces and on the backside of the flare so less friction to the nut. As for tightening, backing off and tightening again, this is more a process to relax hoses back when hoses used flare ends. Not needed with hard lines but if there is friction, it can twist the line.

On copper lines, the 2-step tighten process is have the copper line flare form to the face of flare joint more completely. Also the correct process is too two different torque values.

 

[Luk88]

The issue with the schrader core is what made me bring this up. Commercial refrigeration on large systems, the core is removed to eliminate restrictions. This is done using a special tool that allows for core removal and installation under pressure.

How is it that despite doing research before, almost always I find out important details after I do my DIY task...

The video I watched demonstrates this process in detail. I wish I found it before I did my install.

25 years ago I had something similar on a R134A system where the core had been broken off inside and the seat remained. When depressing the core, the new core would push against the old seat and cut off flow. I kept thinking I had a leak as the vacuum kept rising when I turned the pump off.

I think it is a very good explanation for what was was happening on my previous tries. One thing I also did slightly differently on the final try was to modify my DIY "schrader valve pusher" by bending the tip of the rod back on itself so it had more surface. This meant it is less likely to slip and close the valve as it is pulling the vacuum or later. One good thing about this heatpump is that it has two service ports and two valves. So when testing my setup I could depress the second valve and see vacuum level shoot up immediately confirming a good connection before.

I hope this description is clear enough, I'm still amazed how many things one has to know to install these devices properly.

As long as it held high pressure nitrogen and you checked using bubbles, it should not leak.

Great. Answering my own question regarding nitrogen pressure the vid I linked above also talks about using 300psi of nitrogen for all his tests. (Instead of manual specified 650psi).

These are soft lines. Although bending the 5/8 by hand was very hard - I used special plastic rods inserted into the lines for hand bending. If anyone does this - don't forget to lubricate the rods and if you use Mineral oil clean the lines and blow them off with nitrogen after cleaning. I read R32 devices use special POE oil. I doubt less than a drop of mineral oil residue in the lines after bending would cause trouble, but better safe than sorry I guess.

I did cut and flare both ends so I haven't done it fully "properly" (with no nyloq, no double tightening), but hopefully it will not leak following the tests.

Still I plan to get spare gas (I already own cylinders). Gauges/adapters are on the way. Why, as my system is already installed? One of the purposes of DIY even more so than price for me is self suffieciency. At this stage I can probably repair almost everything I own and rely upon. I'd like my heat pump to be no different (except certain parts like the compressor itself etc).

I will nitrogen test for several hours. I run electrical and other things while nitrogen testing. One does need to account for any temp change. I've let systems sit overnight with nitrogen and check pressure in the morning and compensate for temp.

As for vacuum, I run a micron gauge and will only charge when the reading stabilizes at rest under 500 microns. If I see under 200 microns with pump running, I know the cake is ready to come out of the oven.

Thank you for the tips.

On copper lines, the 2-step tighten process is have the copper line flare form to the face of flare joint more completely. Also the correct process is too two different torque values.

I wondered why there were two columns with very similar torques in the manual. The second labelled "re tightening".

 

[Zwy]

On copper lines, the 2-step tighten process is have the copper line flare form to the face of flare joint more completely. Also the correct process is too two different torque values.

Refrigeration copper is pretty soft, it should form to the flare on the first try. I occasionally see older equipment that ran steel flares and on those you use a copper flare washer and yes, you will tighten, loosen and re-tighten the connection to get a good seal. Nothing wrong with doing a 2 step on a mini split install, I just don't see a need due to good lubrication on the flare backside and proper torque.

 

[k490]

Vacuum testing isn't actually a very reliable way to check for leaks. If you have a severe leak, that will show up in vacuum testing, but most small leaks do not. Often, it's because a vacuum pulls components tighter together in the system and will actually stop most small leaks.

You always relay on the high-pressure nitrogen test as the determination for if the system has a leak. And under high pressure nitrogen, don't rely on just the pressure dropping. Use a bubble leak solution on all likely leak points such as all flare joint and brazed connections.

If you want your minisplit systems to stay leak free and to be as reliable as possible, then cut off the flare fitting at the head unit and braze those connections like you do on a normal split HVAC system. That's because for pipe-to-pipe connections, flare joints are much more likely to leak over time than a brazed connection due to flexing of the pipes that comes with thermal expansion and contractions, and the fact that often you have to put pressure and bend the pipes during install and uninstall of the head units. Flare connections at the outdoor condenser do tend to stay well sealed even over a long period of time. That's because one side of the connection is firmly mounted so there is much less stress on the connection due to thermal expansion and contractions. Also, those connection are typically very assessable, so the tightening torque is much easier to get spot on. Always follow the proper two step tightening sequence and use a high-quality torque wrench made specifically for flare connections. I'm fine with using economically priced tools for most HVAC equipment, but the torque wrench for flare connections is one place where I'll only use a name brand like Yellow Jacket. Same for leak detectors. Only a brand like Field Piece or Yellow Jacket.

In your case, I'm 90% sure you have a leak in your pump gauge setup. You really should be using a quality manifold gauge set with proper connections to the vacuum pump. I'd also never use a vacuum pump with an automatic shutoff, because pressure sensor built into that kind of pump tend to leak.

Some guys posting on HVAC subreddit claim if you braze a mini-split run the risk of particles getting stuck in the expansion valve because the valve is very small in the air handler. I know when they install a typical ducted split system they flow nitrogen inside the lines while brazing to avoid this issue. I'm assuming with those the valve is not connected during the brazing process where the mini-split it is.

 

[jpwhit]

Some guys posting on HVAC subreddit claim if you braze a mini-split run the risk of particles getting stuck in the expansion valve because the valve is very small in the air handler. I know when they install a typical ducted split system they flow nitrogen inside the lines while brazing to avoid this issue. I'm assuming with those the valve is not connected during the brazing process where the mini-split it is.

Yes, you always flow nitrogen when brazing. There can't be any oxygen inside the lines set when brazing or you'll get oxidation that can clog the expansion valve. If you do it right, there won't be any issues.

 

[Brett V]

I’ve installed 3 mini splits. Used Nylog on all the flare fittings and have had no problems.

 

[k490]

Yes, you always flow nitrogen when brazing. There can't be any oxygen inside the lines set when brazing or you'll get oxidation that can clog the expansion valve. If you do it right, there won't be any issues.

Don't you have to remove the expansion valve when you flow the nitrogen?

 

[jpwhit]

Don't you have to remove the expansion valve when you flow the nitrogen?

From a nitrogen flow perspective, no you don't have to remove the expansion valve. They allow enough flow.

On a minisplit, the expansion valve is in the outside unit. So it doesn't even come into play when brazing the head unit connections.

But even so, I still use a Wetrag heat blocking putty to keep heat from traveling into the head unit.

If you do have braze around the expansion valve in a minisplit, almost all minisplits use EEVs so they aren't as heat sensitive as a traditional expansion valve. But you still need to take measure to limit heat exposure.

On standard split system you alway have to remove the expansion valve temperature bulb but depending on the exact location of the expansion valve, and whether there is a reasonable location for heat blocking putty, you don't always have to remove the valve.

 

[jpwhit]

I’ve installed 3 mini splits. Used Nylog on all the flare fittings and have had no problems.

I'm talking about them staying leak free after 5+ years of use. And if you have good access to the flare connections and can get them torqued accurately for both steps, and don't have to manipulate the lines a lot while installing the head unit, then your odds of staying leak free is higher than in really tight and tricky head unit installs.

Basically, brazing those connections just significantly reduces the odds of having leaks at those joints over the full life of the minisplit.

 

[S Davis]

Guess what. You were right... It was residual nitrogen. I let it pump for half an hour, then I leftvit off for an hour. The needle on the vacuum gauge has not moved even a hair. So it just needed more time tobpump down.

It is not residual nitrogen, what you are doing by pulling a vacuum is removing any moisture by boiling it off with a low vacuum and pulling it out. There was still moisture in the system, I have had systems that required pumping down four or five times with a nitrogen break between to get all the moisture out.

Also these systems will run over 500psi so I wouldn’t be anywhere around the fittings if I didn’t do a proper pressure test, 350psi doesn’t tell you anything. I have seen a fitting blow and it can do some serious damage if anything happens to be in the way.

 

[S Davis]

Some guys posting on HVAC subreddit claim if you braze a mini-split run the risk of particles getting stuck in the expansion valve because the valve is very small in the air handler. I know when they install a typical ducted split system they flow nitrogen inside the lines while brazing to avoid this issue. I'm assuming with those the valve is not connected during the brazing process where the mini-split it is.

You can’t braze with oxygen in the line as it will form oxidation when you heat it up, when you flow nitrogen the oxidation/soot doesn’t form.

 

[k490]

You can’t braze with oxygen in the line as it will form oxidation when you heat it up, when you flow nitrogen the oxidation/soot doesn’t form.

Don't you have to remove the expansion valve to flow nitrogen? Can that be done on a mini-split.

 

[S Davis]

Don't you have to remove the expansion valve to flow nitrogen? Can that be done on a mini-split.

You don’t flow nitrogen through the outdoor unit just the fittings you are brazing on the line set and indoor coil/unit.

 

[Luk88]

I have had systems that required pumping down four or five times with a nitrogen break between to get all the moisture out.

It is interesting such a small volume of just copper can retain stuff.

Also these systems will run over 500psi so I wouldn’t be anywhere around the fittings if I didn’t do a proper pressure test, 350psi doesn’t tell you anything.

There seem to be differing opinions about if pressure test above 350 psi is required (I'm talking about r32 specifically and new precharged equipment). I found few HVAC people online (including install videos) mentioning they don't test pre-charged systems above 350psi for a variety of reasons. One guy said he doesn't trust the refridgerant release valves not to leak nitrogen in at such pressures. This did convince me not to go all the way to the manual specified 650psi. (Despite one guy saying he once has a leak show up only after 400psi).

I have seen a fitting blow and it can do some serious damage if anything happens to be in the way.

My understanding was we nitrogen test to find leaks (with bubbling liquid etc) not to see if it blows up. I kinda assumed if the lines, the flares and the equipment are not visibly damaged and well made the chance of catastrophic failure is extremely low. If it does blow up. Id rather it does so, when I'm not sitting right next to it looking at a manometer, but I'm well clear from it.

However if this was an older system, empty of refridgerant. I think it would be very important to fully pressure test, but for a new precharged system? (Lines where not precharged).

 

[k490]

You don’t flow nitrogen through the outdoor unit just the fittings you are brazing on the line set and indoor coil/unit.

The expansion valve is inside the indoor unit right where the evaporator is that is where it goes from liquid to vapor? The lines on the indoor unit are not removeable so how do you braze and flow nitrogen through the expansion valve which is inside the indoor unit?

 

[Luk88]

The expansion valve is inside the indoor unit right where the evaporator is that is where it goes from liquid to vapor? The lines on the indoor unit are not removeable so how do you braze and flow nitrogen through the expansion valve which is inside the indoor unit?

There is no brazing involved at all. This is a flare screw fittings unit. It is pre-charged with 1.6kg of r32 and the refrigerant is locked inside until manual release valves are opened.

This pressure leak testing is done by connecting to the service ports (with refrigerant valves closed). Pressurising it, then going with a bubbling liquid around the fittings.