Leak test at 550PSI per the manual?

NY_Rob

Hopefully I will finish the remaining lineset piping/connects this weekend on my new LG 36K BTU minisplit.

Picked up the rental nitrogen tank last night, so it's on to 24 hour pressure testing...

The LG manual states "test at 551.1PSI with nitrogen"... but then "do not exceed 551.1PSI as it may damage the system".
Seems like a really tight margin for error on a regulator... hit 551 but don't don't exceed 551.

Do you pros run the 24hr test at 550PSI or is it generally done at a lower pressure?

.....FROM THE LG INSTALL MANUAL....
1. The liquid and gas piping in each refrigerant system should be pressurized with nitrogen gas in accordance with the steps listed below.
Step 1: Pressurize to 43.5 psi for three (3) minutes or more. (Will identify if major leaks are present in the system.)
Step 2: Pressurize to 217.6 psi for three (3) minutes or more. (Will identify if major leaks are present in the system.)
Step 3: Pressurize to 551.1 psi for approximate twenty-four (24) hours. (Will identify if minor leaks are present in the system.)
Pressurizing the system to 551.1 psi does not guarantee the identification of minor leaks if the pressure is maintained for only a short time. It is recommended that the system remain pressurized for at least 24 hours.

Piping system should not be pressured to more than 551.1 psi. Pressures greater than 551.1 psi may damage the piping system and cause unit malfunction.

unclejohn

I will state for the record that I have never installed a mini split and never will. But it seems to me that 250# is plenty to test for leaks over 24 hr.

Jamie Hall

Somebody was translating from kiloPascals using a calculator. That's enough to make me weep... they started off at 3800 kPa and should have printed 550 psi. (3800.00 kPa is 551.143 psi). If you want to do it by their book, anything close to 550 (but not over) would be what you want -- but I agree with @unclejohn -- I would think that 250 would be fine. What you're looking for is change in pressure, not the exact pressure, after all.

NY_Rob

Thank to both of you!

My gut is just telling me 550psi is too high... I can envision bad things happening at those pressures.

Even the guy at the gas supply where I rented the nitro cylinder was shocked when I told it gets tested at 550psi.

What is the high working pressure of a typical modern minisplit?

ratio

IDK about LG, but Lennox & Mitsubishi both require ±600 lbs N₂ for the extended warranty on their VRF systems. We've had no issues following their procedure during the (admittedly three or so) systems we've installed.

That said, if you don't have any leaks I can't see how they'd know.

JUGHNE

I have done conventional AC at 500 for 12-18 hours.
Mini's at 300 to 500 for hours.
Quite honestly it depends upon your time window and how much is in your tanks! I have at least 2 tanks to cascade with.
I am sure I am not alone is this get er done fashion. But I feel confident that I have done justice to the leak check procedure.
I also triple evacuate everything, this can involve hours.
A leak will show up in vacuum testing by rapid rise with pump off.

However, you want to be sure you have all air purged before high pressure testing. Air will contract overnight and drop your pressure......and likewise expand with warm weather.

EBEBRATT-Ed

How you test is up to you. I think what these manufacturers are requiring is just.....obcene. 200psi is plenty. There is very little gas volume there a leak will show up. In most cases you have a leak our you don't it would show up at almost any test pressure.

I am waiting for someone to tell me that they tested at 200 and didn't have any leaks but at 550 it leaked

All this is is the MFG covering their A--

Tinman

I use to test R22 systems to 300# because that's about as high as they get. I use the same logic for testing R410 systems. Test at the high end of normal operating pressures. Why would you test at less pressure than you'll be operating at? That doesn't making any sense to me.

Mike

I've tested systems at the recommended test pressures of 400 to 500 psi. But after numerous leak searches, and finding the service valves weren't holding back the nitrogen, I only go up to 250#. If it holds for an hour, then pull a vacuum and that holds, I know I'made good.

NY_Rob

Thanks again to all who responded... looks like there's a lot of variation between installers- with some sound logic behind the numbers they test at.

When I look at the manifold gauge and chart for R-410a at 100F ambient it indicates about 317psi. I was reading that you should test at 1.3x your max working pressure. That would put me at about 400psi for the pressure test. I feel better with that vs. 550psi.

As Mike mentioned above- I also saw reports of the service valves not holding and letting nitrogen into the outdoor unit when testing at 500+ psi pressures- don't want to go there.

hvacfreak2

I have seen compromised flare connections with Nylog hold a lot of pressure ( 550 psi ) for a long time. We had an over - torqued 1/4 inch connection that was periodically leaking refrigerant. We were able to valve it off close by and were amazed when we took it apart and saw how messed up it was ( it had almost straightened out through the flare nut ). It was a large system , 52 indoor units , but it held 550 for 2 weeks. I soap every flare connection these days as well as moving the connection slightly before I call it good. Also when the system is operating the refrigerant oil in the system seems to find any flawed connection , almost like a hydro test with oil. I would test at 550 psi if that is what the instructions say. Or you could roll the dice and test at 50 psi or heck don't worry about it at all and see if it pumps down with no pressure test. Or heck for that matter ...

NY_Rob

^ it's definitely getting nitro pressure tested... I'm just not comfortable @ 550psi with the strict warning not to exceed 550psi or it may damage the system.... literally no margin for error.

I did use Nylog on all the flares, I cut off all the factory lineset flares, I deburred every cut, and used the LG supplied nuts on both sides of the linesets which I flared with a R-410a eccentric/clutch type flare tool (which made beautiful flares BTW), and I torqued every connection per the manual based on lineset diameter.

Looking forward to to the pressure test- I'm going to soap all the connections and will take it from there..

unclejohn

Sigh, the good ole halide torch is gone.

EBEBRATT-Ed

I still say a leak is a leak. I am waiting for someone to tell me they tested at 200 and thought it was good but at 500 it leaked. I doubt it. I have had leaks so small that soap bubbles were useless. 99% of the time if 200 won't show a leak 500 won't either

hvacfreak2

I have seen flare connections on systems that held at 300 and leaked at 500. That's when I understood why they want the test so high. If the joints were all brazed connections the test could be less like we did with R22.

Tinman

@hvacfreak2 - To add to your point, when I came up I worked for a couple of Lennox dealers and the equipment was mostly flared fittings. Now we've come full circle with the amount of ductless splits going in with the flared fittings. It's a different animal than brazed joints.

newagedawn

TALKING ABOUT NITROGEN TEST? pull a vacuum to 500 microns if it holds THERE IS NO LEAK, CHARGE SYSTEM, i would recommend some leak lock on the flares for added protection for leaks, but man 550 psi, never heard of such a thing

Tinman

Refrigerant oil on flare threads is all I've ever needed.

EBEBRATT-Ed

@Stephen Minnich & @hvacfreak2 ,

Yeah I wasn't thinking about flares. The newer flaring tools do a great job better than my 45 year old imperial.

ratio

A 500 micron leak test is a 15# leak test.

Jamie Hall

Um... except that there are joints which will leak under vacuum, but not under pressure. And vice versa... Not brazed or soldered (with very rare exceptions), but any mechanical joint may.

NY_Rob

Finally got the last zone piped... started at 7:30am finished at 3pm... only breaks for hydration. It had to run (parallel) through 26' of soffit... with a bay window in the middle for extra aggravation
Did two low pressure tests 75psi and 200psi... soaped every connection on all three indoor units, all the outdoor connections and even the manifold gauge and the connection to the nitro tank... no leaks.
At 3pm I boosted it up to 400psi with the needle between the two zeros in 400. Soaped everything again... no bubbles! It was 87F outdoors (and the sun was partially on the outdoor unit).
The pressure actually went up a hair later on when it hit 90F and I took away the umbrella I was working under. It stayed there until the temp finally dropped to 75F and the sun got low in the sky- now it's reading between the 4 and first 0 in 400psi.

It's not looking good weather wise... rain forecast for tonight and tomorrow.. may have to pull my equipment and the rental nitro tank/regulator tonight.

If the leak test holds.... all that's left is the triple evac!


Re the R-410a eccentric/clutch type flare tool.. I was very careful deburring the cuts (a pita on 1/4" inch pipe)... and I put a few drops of the Nylog on the tool's cone for lubrication... once I figured out how much (about 1mm) the pipe should extend past the top of the die... every flare came out wonderful.

hvacfreak2

EBEBRATT-Ed said:

@Stephen Minnich & @hvacfreak2 ,

Yeah I wasn't thinking about flares. The newer flaring tools do a great job better than my 45 year old imperial.

The eccentric type are a must have for this ( CPS makes a nice one ). 410A and flares are just a pain altogether.

Tinman

@EBEBRATT-Ed and @hvacfreak2 - I have the ancient Imperial that Ed is referring to from trade school days and within the last few years invested in Yellow Jacket's top of the line. What a difference it makes.

hvacfreak2

Stephen Minnich said:

@EBEBRATT-Ed and @hvacfreak2 - I have the ancient Imperial that Ed is referring to from trade school days and within the last few years invested in Yellow Jacket's top of the line. What a difference it makes.

I bought my Ritchie red handle in 1989 or so , I have the blue handle one now. There isn't much difference except the flare cone is slightly wider. The next job I do though I'm getting the CPS.

NY_Rob

Well it held pressure overnight...
400psi at 88F, twenty four hours later 375psi at 57F.

Here's the formula I found for pressure change due to temp.. 1.45psi drop for every 1.8deg below initial fill temp.
31deg temp drop = 25psi drop.

EBEBRATT-Ed

@NY_Rob is that formula for nitrogen or air?? I am assuming it is nitrogen.

@Stephen Minnich & @hvacfreak2
After seeing the new flaring tools work I guess my old Imperial goes to the dust bin.

I bought one of the new ones (don't know the brand) but it's like looking in a mirror the flares are soo nice.

We hired a new kid about a year ago. He did residential burners and ductless splits.

At first I was skeptical about new flaring tools, ratchiting tubing benders and torque wrenches that he insisted on but I was wrong

NY_Rob

EBEBRATT-Ed said:

@NY_Rob is that formula for nitrogen or air?? I am assuming it is nitrogen.

It's for nitrogen, it starts on pg 42 of the manual and ends with the formula on page 43...

RPK

I always feel better about these tests when I have time to wait for more than 24 hours. I like to see the pressure go back up as the temp rises again. Wide temperature swings add uncertainty to the test, especially if part of the piping system is indoors and wasn’t subject to the same temperature change as the outdoor piping.

hvacfreak2

With the majority of the piping being inside the structure I would not use the outside temperature in that formula. If it were me I would leave the test pressure on for awhile longer , consider your lower reading as " stable " and check it again in 6 plus hours.

RPK

I believe the most effective leak test method is adding a trace amount of refrigerant, then pressurizing and checking each joint with an electronic leak detector. On one recent job, I saw lots of tiny leaks on small zoomlock fittings. The pipe fitters said they had bubbled everything and couldn’t figure out why they were losing pressure. We tried this method and found a bunch of 1/4 and a couple 3/8 zoomlock fittings. The leaks were so small that it took a bit of patience to see them with soap bubbles, but all the leak detector hits were eventually confirmed.

On VRF, you can pressure test each branch individually by closing off ball valves to each branch.

ratio

RPK said:

...electronic leak detector.

+1 this. I had a leak that woudn't show up with bubbles but a leak detector located it. It was small enough that I think only oil was leaking out, but the oil was carrying enough refrigerant to set off the leak detector. It was a factory joint in a packaged unit, BTW.

RPK

.