Leak testing by vacuum.

[fxb80]

All the talk going on in the Nitrogen Pros Only thread put me in mind of this.

Who uses vacuum as a leak test, I mean a good vacuum gauge, not your manifold? Does it really work well enough to not do an overnight nitrogen test?

I'm real old school and think nitrogen is the best way to test. It seems to me that a vacuum test is still a pressure test but your test pressure is only 14.7 psi (at sea level).

Vacuum test is the big thing where I work, and if somebody convinces me I should stop preaching nitrogen testing I'm sure my co-workers would appreciate it.

[invisiblecollar]

How would you locate a leak with a vacuum test?
Nitrogen and soap is the way to go.

[jricer2001]

with a digi. micron gauge. yes I have done that, and I tend to think that can be a good way to go, but my thought is that pressureizeing the lines is also a good thing because up here we have had problems with small cracks / fishers in the pipe it's self. typically 5/8 but i have also heard 1/2 as well, my micron gauge is way more accurate than any manifold gauge I have. I may or may not notice +/-1/2 lb with the manifold gauge but will notice 500 up to 100 micron change with the micrometer. now finding the leak is a whole nother story.

[itsiceman]

I thought you were supposed to check both ways
not one or the other :anyone:

[jricer2001]

checking both ways is part of the job. you first pressureize to check for leaks, and then you have to pull a vaccume anyway so, when you are done pulling a good vaccume you should let it sit for a while and check back ... if you cant get below 300 microns you may have a issue somewhere vac oil, leak, ect. or at least that is how I was taught

[fxb80]

Let's say I'm working on a pretty good size system that takes a full bottle of N2 to make 150 PSI. What I think I'm hearing is you pressurize it just long enough to make sure nothing is spewing, since it can take quite some time for a bigger system to lose enough pressure to say for sure there's a leak, and then do the "real" test with vacuum.

Is it impossible to get a system below 300 microns if there's a leak? Is that the magical number?

I'm pretty ignorant about the range of possibilities, but I've noticed techs having difficulties with vacuum testing. More than a few times I have seen techs frustrated because the vacuum stopped dropping at some value like 500 microns. They would blank-off test the manifold and pump and get the manifold down to 50 microns so the possibility of leaking hoses, etc. is eliminated. They would just give up and gas up the system. And time would prove the systems were tight.

My preference would be an overnight test at 150 PSI where it didn't drop at all. Then I don't have to worry about getting a perfect vacuum because pressure showed it tight.

[jricer2001]

the vast majority of what i tend to do is supermarkets. we put up hangers, sling pipe, burn together and pressureize linesets. liq, and suct -we loop then and pressureize together to usually 150# for several days or even weeks. our line sets can be several hundred feet and, 200 ft of line set of 5/8 x 1 1/8 pipe is not uncommon at all. if we think we have a bad bunch of pipe then 400# to "prove the pipe" if a leak is suspected we seperate the suct and liq lines and pressureize ind. I was thinking about the micron thing, I may of been wrong on the # I think it is usually somewhere around 1000 (I have been off since oct, my boss didnt diversify with the companys that we do business for enough and our main one is not doing much lateley) good thing I own a snowplowing and lansdscapeing business on the side. so sorry, 1500-1000 was what i was thinking of. thats right the lower # of microns the lower the vaccume.

[DDDpope]

Pressure test with a 20% mixture of R22 is by far the best way to leak check, but nothing is fool prof, and we all miss leaks once and a while no matter how careful we are. The vacuum leak check does however have its place. If you have a system that you suspect has standing moisture in it, the vacuum test can tell you for sure. Pull the system down to at least 500 microns and let it stand for 24hrs if the microns rise to 1000 to 1200 then there is moisture in the system. If it stays below 800, no leaks or moisture, finish the evacuation process. If it goes above 1200 microns you still have a leak. These values are approximate and there are many variables to factor in. For instance not ever micron gauge reads the same. I hope this helps, and answers your question, please let me know. Contact me anytime.

[VTP99]

I would be hesitant to do a standing vac test. Why would you want to find out there is a leak and you pulled in moisture all night ?

[cmclifton]

There's a lot of numbers on this subject that people throw around arbitrarily. I've heard people say things like "if you pull down to 500 microns, blank off and wait 5 minutes, and it only goes up to 1200 then its moisture, and if it goes above 1400 then its a leak". The reality is that in either case it could be moisture and it could be a leak and it could also be that you just did the blank off test too early. The size of the system, the size of your vac. pump, quality of your hoses, etc all come into play here.

If you want to be sure the best way to do it is to pressure test your system with nitrogen and use the vacuum blank off test as both a leak test and a test to make sure you had it under vacuum long enough to bring all the nitrogen/air/moisture out.

A thermistor vacuum guage reads high if there's moisture in the system. As in, artificially high. The vacuum level could be 100microns but the moisture will make your vacuum guage read higher than that. You should be pulling a vacuum for as long as it takes to bring the system to less than 500 microns and have it hold less than 500 microns in a 15 minute blank off test. If it passes that test then you know it's evacuated to the best of your ability, isn't so full of moisture that your new compressor dies a horrible death in 6 months, and isn't leaking. How long should it be under vacuum? The answer is as long as it takes to pass a proper blank off test. This is where 3/8th vacuum hoses, full flow manifolds, and the biggest vacuum pump you can carry come in handy :-)

[jricer2001]

There's a lot of numbers on this subject that people throw around arbitrarily. I've heard people say things like "if you pull down to 500 microns, blank off and wait 5 minutes, and it only goes up to 1200 then its moisture, and if it goes above 1400 then its a leak". The reality is that in either case it could be moisture and it could be a leak and it could also be that you just did the blank off test too early. The size of the system, the size of your vac. pump, quality of your hoses, etc all come into play here.

If you want to be sure the best way to do it is to pressure test your system with nitrogen and use the vacuum blank off test as both a leak test and a test to make sure you had it under vacuum long enough to bring all the nitrogen/air/moisture out.

A thermistor vacuum guage reads high if there's moisture in the system. As in, artificially high. The vacuum level could be 100microns but the moisture will make your vacuum guage read higher than that. You should be pulling a vacuum for as long as it takes to bring the system to less than 500 microns and have it hold less than 500 microns in a 15 minute blank off test. If it passes that test then you know it's evacuated to the best of your ability, isn't so full of moisture that your new compressor dies a horrible death in 6 months, and isn't leaking. How long should it be under vacuum? The answer is as long as it takes to pass a proper blank off test. This is where 3/8th vacuum hoses, full flow manifolds, and the biggest vacuum pump you can carry come in handy :-)

I would agreee with this, niceley stated

[VTP99]

CM Quote (and the biggest vacuum pump you can carry come in handy :-)


CM,
Bigger is not necessarily better when it comes to pumps.

[mmmdave]

Nitrogen to 150 psi to test for leaks (with a trace of r22 if using a detector). I only pull a vacuum after getting past the pressure test. Vacuum is only for removing moisture/non-condensables. Ideally, a rise in microns should only indicate moisture boiling off inside the system since the pressure test already checked for leaks... although sometimes it can indicate gases bubbling out of the walls of charging hoses if your micron guage is sensitive enough.

14.7 psi doesn't seem like much of a leak test to me.

[cmclifton]

Bigger is not necessarily better when it comes to pumps.

The only reason I've ever had anyone give me for that is because of the danger of freezing any moisture in place in the system. Is there another cause for concern? With the small amounts of moisture we're typically dealing with that's mostly snake oil, and the schraders are going to slow down the evacuation process anyway. Not to mention that frozen moisture in the system can still be removed via vacuum. Sublimation...

I've seen it happen in freeze driers when someone leaves a drain valve closed accidentally so the water doesn't drain and then it all boils then freezes when the vacuum pump is turned on (and then it snows inside the chamber. very cool thing to see) but I've never seen it cause a problem when evacuating a small self contained system.

I used to use a 10cfm pump to evacuate everything from single door True reach in cases to 32 cubic foot ultra lows, to 300ton chillers.

[Phase Loss]

I use vacuums for leak testing. (sometimes) Most times I am able to chase it down with my H10-G or TIF-ZX1 and bubbles.

(Here is the situation): We have been adding refrigerant to a rack time and time again...guys are "repairing" leaks, yet we keep adding refrigerant.

I will pump 1 loop down at a time and pull it down to a 30" vacuum. I am looking for a leak that ranges from a pinhole leak to a decent size blower. I am not worried about microns because I am not looking for a 12oz a year leak.

I will let the 30" vacuum sit for 30 min...It should hold at 30". I use digi-cools and have found small pinhole leaks this way (and obviously blower leaks).

Once I get to a loop that will not pull down to 30", or is able to pull down to 30", but loose the 30" after 30 min. I know I am on the loop that has a leak. I will focus on tracing out that loop and if need be, pressurize it with N2 and refrigerant to pinpoint it. Rarely is N2 even needed. Most of the time I am able to close the suction ball valve and pump discharge vapor through the suction line and let pressure build up. (if you turn up your ears, most leaks can be heard) If I still can not find that leak in the loop...its time to cut the line sets loose from the header and from the cases and pump them up individually because the leak is going to be in either the undergrounds or overheads.

Again, this method works wonders for me when I leak check. When I hunt for leaks, they are usually substantial leaks. I'm not chasing down 12oz a year leaks on a rack.

[nchvac]

Do you guys ckeck your pumps/gauges from time to time to see if they are working ok on holding vacuums? If so, what is the proper method of doing this?

[CoolWine]

seems like bad practice to pull a vac w/o pressure testing with nitro. if you have a leak you will suck moisture into the system.

I pressure test with nitro to find my leaks, if its problematic i'll blast in some r-22 (with the nitro) as a tracer gas.

[zzonko]

If you test with vacuum, and there is a leak on the suction side, the armaflex will sometimes get sucked into the defect, sealing the leak temporarily.

Test it like it will be used, under recommended label pressure with nitro. Anything less is a shortcut and will possibly create future problems

[fxb80]

Thanks for the responses. Thanks cmclifton and Phase Loss for showing the value of vacuum testing as supplementary to pressure testing and giving specifics.

[VTP99]

The only reason I've ever had anyone give me for that is because of the danger of freezing any moisture in place in the system. Is there another cause for concern? With the small amounts of moisture we're typically dealing with that's mostly snake oil, and the schraders are going to slow down the evacuation process anyway. Not to mention that frozen moisture in the system can still be removed via vacuum. Sublimation...

I've seen it happen in freeze driers when someone leaves a drain valve closed accidentally so the water doesn't drain and then it all boils then freezes when the vacuum pump is turned on (and then it snows inside the chamber. very cool thing to see) but I've never seen it cause a problem when evacuating a small self contained system.

I used to use a 10cfm pump to evacuate everything from single door True reach in cases to 32 cubic foot ultra lows, to 300ton chillers.

CM,
Using a vacuum pump on a/c systems from 1 ton to 10 tons, you will not see the difference between a 3 CFM pump and 10 CFM pump. For example,if you pump on a system and you notice in 2 minutes the pump quiets down and you cannot really feel any air leaving the exhaust. This means that there is no more CFM left in the system and you are now working with molecules. Thus,at this point if you replaced a 3 CFM pump with a 10 CFM pump,there is no change in vacuum or time.

[cmclifton]

CM,
Using a vacuum pump on a/c systems from 1 ton to 10 tons, you will not see the difference between a 3 CFM pump and 10 CFM pump. For example,if you pump on a system and you notice in 2 minutes the pump quiets down and you cannot really feel any air leaving the exhaust. This means that there is no more CFM left in the system and you are now working with molecules. Thus,at this point if you replaced a 3 CFM pump with a 10 CFM pump,there is no change in vacuum or time.

Clearly the difference in that case is all in that theoretical 2 minutes. It might be 2:25 on a 3 cfm pump and 1:50 on a 10 cfm pump. Not really noticeable on small stuff. It would certainly be handy to have a large pump when you're changing out a grounded compressor on a big tonnage chiller though.

I missed this post earlier-

Let's say I'm working on a pretty good size system that takes a full bottle of N2 to make 150 PSI. What I think I'm hearing is you pressurize it just long enough to make sure nothing is spewing, since it can take quite some time for a bigger system to lose enough pressure to say for sure there's a leak, and then do the "real" test with vacuum.

Pressure is pressure, it goes both ways. It can take a long time to notice a leak on a large system with a vacuum blank off test too. Since there's no real standard operating procedure, its tough to know when you've given it enough time to make sure the pressure isn't rising in the system. It's one of those things that comes with experience I guess. I pressure test big systems the same way I do small ones, but I give them more time to lose or gain pressure. 150 psi with nitrogen in a system that holds 6 oz of gas, I'll notice a leak fairly quickly. In a system that holds 40 lbs of gas, I might let it sit for an hour or two. I've let them sit overnight or for several days also. The vacuum test is the same way. If I'm on a small system I'll do the vacuum blank off test for 5 minutes or so. On a big system it might be 20 minutes, half hour, just whatever point I start feeling comfortable that the system is tight.


Is it impossible to get a system below 300 microns if there's a leak? Is that the magical number?

I've pulled systems to under 100 microns and then found leaks before. It doesn't happen often, but it can happen. One that comes to mind was a Revco -80*C freezer. System holds roughly 7 ounces of gas (2nd stage). I pressure tested it overnight and lost 1-1.5 psi. Left it for another night and it lost 1-1.5 psi again. I started looking for leaks and couldn't find anything. I was still a little green so I decided to evacuate it and see what I would get. It pulled down to 80 microns or so and wouldn't go any lower. I eventually found the leak but it took forever. Turned out to be a tiny little pinhole that I couldn't even see in my 56% silver braze joint where the oil return comes out of the oil separator.


I'm pretty ignorant about the range of possibilities, but I've noticed techs having difficulties with vacuum testing. More than a few times I have seen techs frustrated because the vacuum stopped dropping at some value like 500 microns. They would blank-off test the manifold and pump and get the manifold down to 50 microns so the possibility of leaking hoses, etc. is eliminated. They would just give up and gas up the system. And time would prove the systems were tight.

You have to remember that the longer the vacuum is running, the slower it's going. The vacuum pump moves the most amount of air/nitro/whatever right at startup and it only goes downhill from there. It's exponential. The more gas you remove from the system, the less there is to be pulled into the pump. I've seen it take hours to go from 550 microns or so to 450ish on a larger system. If the system is tight, it will eventually make it down there. (unless there's something wrong with the pump/hoses/etc)


My preference would be an overnight test at 150 PSI where it didn't drop at all. Then I don't have to worry about getting a perfect vacuum because pressure showed it tight.

I love having a nitrogen test overnight too. If I could do that every day, I would. You still need a good vacuum test though, just to make sure you got all the moisture and nitrogen out.

[Phase Loss]

Who uses vacuum as a leak test, I mean a good vacuum gauge, not your manifold? Does it really work well enough to not do an overnight nitrogen test?



If your question is (to prove the absence of a leak)

Lets keep it simple :) and leave the proof in the numbers.

1 PISG=1 PSIG

Now lets look at microns.

0 PSIG=762,000 microns. every additional 1" of a vacuum subtracts 25,400 microns from 762,000.

Now when proving the absence of a leak...what is going to be more accurate? working with a Standing pressure in the range of 0 to 400 PSIG?

Or

working with the absence of pressure that is in the range of 50 to 762,000 microns?

Nitrogen has it's place. I use it for flowing while brazing, pressurize components for strength testing, and pressurizing for leak detection
*NOTE* I said detection, after proving a leak exists with a high vacuum.

If you manage to find a leak during the pressurization for the strength test because your pressure dropped, that's good. But the final test is always always always a high vacuum.

[MikeySq]

I'm not chasing down 12oz a year leaks on a rack.

so you have that factory option on your racks too ! haha

[rayr]

I will pump 1 loop down at a time and pull it down to a 30" vacuum. I am looking for a leak that ranges from a pinhole leak to a decent size blower. I am not worried about microns because I am not looking for a 12oz a year leak.

I will let the 30" vacuum sit for 30 min...It should hold at 30".

Does not exist on this Planet!!

Soooooooooo at a 29.975" vacuum, the tech is depending on 14.7 Psig external force to indicate a leak. Hmmmm methinks a piece of dirt could plug that leak enough to indicate a non lead situation. Think I will stick to proven methods or get me one of them thar Freezoneeeee sniffin' dogs.

[Dowadudda]

the vast majority of what i tend to do is supermarkets. we put up hangers, sling pipe, burn together and pressureize linesets. liq, and suct -we loop then and pressureize together to usually 150# for several days or even weeks. our line sets can be several hundred feet and, 200 ft of line set of 5/8 x 1 1/8 pipe is not uncommon at all. if we think we have a bad bunch of pipe then 400# to "prove the pipe" if a leak is suspected we seperate the suct and liq lines and pressureize ind. I was thinking about the micron thing, I may of been wrong on the # I think it is usually somewhere around 1000 (I have been off since oct, my boss didnt diversify with the companys that we do business for enough and our main one is not doing much lateley) good thing I own a snowplowing and lansdscapeing business on the side. so sorry, 1500-1000 was what i was thinking of. thats right the lower # of microns the lower the vaccume.

hey pm me. I think I might know you. Seriously.

[VTP99]

Does not exist on this Planet!!

Soooooooooo at a 29.975" vacuum, the tech is depending on 14.7 Psig external force to indicate a leak. Hmmmm methinks a piece of dirt could plug that leak enough to indicate a non lead situation. Think I will stick to proven methods or get me one of them thar Freezoneeeee sniffin' dogs.

rayr,
I understand Phase as saying with a micron measurement scale, he can see any movement more easier using 762,000 increments verses 400. Of course this is only good to determine a leak not find one.

[Phase Loss]

Think I will stick to proven methods.

pulling a vacuum is a proven method.

And if you read that entire post that you pulled the quote from, it would have been very obvious that I mentioned pressure testing and sniffer testing.

[VTP99]

pulling a vacuum is a proven method.

And if you read that entire post that you pulled the quote from, it would have been very obvious that I mentioned pressure testing and sniffer testing.

True but he did get ya on the 30" vacuum. Perfectionist :)
We no you were rounding off.

[Greg M]

If you test with vacuum, and there is a leak on the suction side, the armaflex will sometimes get sucked into the defect, sealing the leak temporarily.


How is something as porous as Armaflex going to plug a leak? The vacuum will suck air through armaflex with no problems at all.

[cboe]

Iam old schooled also. pressureize with notro mabey put a little reff. chaser with the nitro. check for leaks. repair if any. then evac. down to 300 miccrons. do something for a few hours. if the vac. holds you are good to go. good luck.
another thought! I wonder how many of use nitro when braseing reff. lines. good question.

[surenuff]

I like to blow those lines just a little after I finish soldering on a new system and have the schraders out. You already are hooked up. So when you crank down that regulator for a second or two, just **** your gauges and let it build a little pressure. Then I soap up connections I just made. If no bubbles, time for the pump. I have two different articals I will try to bring back to the table here when I find them.(I don't know how most of you guys reference info so fast. You must have a library right by you ). But one artical was talking about how you can actually find a leak with a smaller amount of positive pressure, I think in the range of 30# to 40#. The other artical was from JB Industies and showed that even with 3/8 hoses and a 8CFM pump that even with the schraders out, the best you get is 3CFM. If I had to say which way I trus the most for leaks, I would have to agree with the guys that like the vaccuum. Any micron gauge will show the existance of a leak faster than anything.(except oil and freon spraying in your face) I still use my old analog with the thermister on top of the old valve. I do have a couple of digitals, but I have a hard time learning to trust them.

[Ti Llaves]

If you test with vacuum, and there is a leak on the suction side, the armaflex will sometimes get sucked into the defect, sealing the leak temporarily.

Test it like it will be used, under recommended label pressure with nitro. Anything less is a shortcut and will possibly create future problems

I agree. Always test for leaks using a pressure method. Think about this, checking for leaks using the vacuum method is not a good idea. If a leak does exists you are pulling air into the system.

[itsiceman]

FWIW You can locate a leak in a system under vacuum using an Ultrasonic Leak Detector.

[nchvac]

Which ultrasonic are you using? Seems Accutrak and Amprobe are the hits around here.

[itsiceman]

I have an old Amprobe
I want a new AccuTrak ;)

[nchvac]

Can you give a reason why? This topic comes up a lot and it is hard to get an accurate evaluation of these two brands. It sounds like you have some exprience that would help you form your opinion.

[itsiceman]

I haven't used any newer Amprobe units so take this how you would like.
The AccuTrac VPE is much better at not amplifying unwanted background noise.
I have not used the VPE-1000 but hear it is superior to the VPE.
I will probably just save for the VPE-1000 but the VPE is a very good unit IMO.
Also for the price you can't beat the Amprobe
HTH

[vacdaddyt]

I have always instructed pressurize with nitro and check for leaks with a bubble solution first, if that fails pressurize with nitro and a trace of R22 checking with a good leak detector second, and then condem low side parts, usually the evap first. I have personally had equipment that would hold 150 psi over night and it would not pull a 1000 micron vacuum let alone the 500 I normally look for. We had checked for leaks using different methods. The sytem would leak at the evap when the compressor was on, there was a small flap of epoxy closing the hole during the pressurization( aluminum evap to copper). Because the low side was not leaking under pressure the detector would not go off or bubbles show. This is easier on the small stuff we handle most of the time now. The larger equipment has so many places to leak and air currents influence the detectors. I hate leak searching cars and the few tractors and combines the farmers bring around occasionally,they are a pain.

[gsxrsquid]

I would be hesitant to do a standing vac test. Why would you want to find out there is a leak and you pulled in moisture all night ?

+1

[Green_Tech]

I would be hesitant to do a standing vac test. Why would you want to find out there is a leak and you pulled in moisture all night ?

Somebody please correct me if I am wrong. But doesn't pulling vacuum boil moisture off at a lower temperature? Also, if the pressure holds that means no leak and there wouldn't be any moisture introduced into the system. If there was a change in pressure that means either there was a leak and that we would have to vacuum the system down to boil moisture off anyway.

I'm new to this so please correct me if I'm wrong.

[gsxrsquid]

Well I use pressure to find the leak. Once it is repaired pull the vacuum which removes non condensables and moisture. as long as thee is moisture boiling off under vacuum your level won;t hold. once you get the moisture out then yo should be able to pull it down and hold it down

[Texas-Tech]

I think using the two together is the best way. There have been times when the pressure test seemed to be good but it would vary a tenth of a psi and you just had that nagging doubt. With a good vac gauge like the YJ29075 you're not gonna get below 100,000 microns with any leak in the system.

[nchvac]

100,000 microns? I have a YJ and it starts at 25,000. If you can't pull down below 100,000 on your micron gauge you need a new vacuum pump and a hearing aid, because you should be able to hear that leak.

[Texas-Tech]

My micron gauge starts out at 760,000 and goes to 1. I don't need a new pump.

Reread the post, I said with a "leak" your not gonna get below that point.

I also got the number wrong it's a 69075 and No I wouldn't be able to hear the leak either, I'm deef.

[roadrunnerderek]

This is a great thread on using micron gauges for finding leaks. I've been using micron gauges for about the last 8 years now. It definately depends on the quality of your hoses (I'm always putting new rubbers/gaskets into my hoses to help keep the system tight and sometimes buying new black heavy duty vacuum hoses). Just one cracked gasket can ruin your micron reading. Of course after pressure testing with nitrogen and some trace gas I like to get that micron gauge on for a final check on system tightness. Depending on the size of the system and the kind of compressor (hermetic, scroll, or semihermetic) you can expect different levels of deepness of your vacuum. I like to see on smaller systems between 250-500 microns initially, larger systems the vacuum might only get down to 500-1000 microns especially if a semihermetic compressor has very small leaks from discharge valve gasket. After isolating system from the vacuum pump the vacuum micron level will rise (especially if you have your gauge set [4 hose set] hooked up to you vacuum gauge due to the permeability of the hoses) usually about 200-700 microns. It will then "level off" if you have a tight system. If the vacuum level remains steady between 800-1250 microns for a few minutes (staying steady at say 1100 microns [it should stay within approximately 30 microns of where it "levels" off) for a good 3 minutes you most likely have a tight system. If it skyrockets quickly above 2000 microns there is something wrong (leak, or oil stuck in hose set). This week I had to clean some oil out of my hose set and afterwards my vacuum was holding good. Once again it all depends on the quality of your equipment and how you maintain it. I don't know yet if anyone has got this down pat with a great vacuum leak testing method. But you can sure tell if there is a leak fast by doing a vacuum test while you are vacuuming the system anyway.

[nchvac]

My micron gauge starts out at 760,000 and goes to 1. I don't need a new pump.

Reread the post, I said with a "leak" your not gonna get below that point.

I also got the number wrong it's a 69075 and No I wouldn't be able to hear the leak either, I'm deef.

I have the 69086, which I think is a step down from yours. I can pull a vacuum with it if there is a leak, but when the pump is taken off is when you will notice it. A leak big enough to keep you from going below 100,000 should be easily noticed with a very light pressure test.