PDA

View Full Version : Parallel rack compressor leak



racknovice
01-09-2007, 02:43 PM
I am working right now on a supermarket parallel compressor med-low temp rack.
After noticing that the only way to get one particular low temp circuit down to its -15 C setting was by adding quite some refrigerant to the system (about 50 kg), I began suspecting some sort of leak (main liquid line sight glass was also half empty at all times). I looked for it unsuccesfully in the mechanical room and condenser with a propane torch detector. Now I must confront the fact that it must be either in the piping or evaporators.
In such a case, where there are so many common manifolds and piping shared by the entire system of circuits, what should be the most efficient way to locate the leak: nitrogen, UV dyes or regular dyes ?
Thank you for any useful advice.

DEVITG
01-09-2007, 03:18 PM
I am working right now on a supermarket parallel compressor med-low temp rack.
After noticing that the only way to get one particular low temp circuit down to its -15 C setting was by adding quite some refrigerant to the system (about 50 kg), I began suspecting some sort of leak (main liquid line sight glass was also half empty at all times). I looked for it unsuccesfully in the mechanical room and condenser with a propane torch detector. Now I must confront the fact that it must be either in the piping or evaporators.
In such a case, where there are so many common manifolds and piping shared by the entire system of circuits, what should be the most efficient way to locate the leak: nitrogen, UV dyes or regular dyes ?
Thank you for any useful advice.

The best leak detetector is OIL mark , and in the high pressure side better .

As it is a supermarket , and sure you can no stop it , I think any DYE can be the best option.

Nitrogen is the best and only option when the system is out of service , and empty.

Lowrider
01-09-2007, 03:38 PM
The best leak detetector is OIL mark , and in the high pressure side better .


Indeed, look for oil! If there's a leak under the isolation, it will eventually come out trough the isolation.

Dan
02-09-2007, 04:41 AM
First, you need a really good electronic leak detector. You turn it on outside the store and then see what happens. If you have a Tif or Detek and understand how they work, you lower the sensitivity and progress your way toward the leak. When you get confused, go back outside and start over. It is not easy finding leaks.

If you have what we call a screamer when you first step foot into the store, you might have to deal with it differently than if you have a localized leak. Here are some drastic moves you should make in such situations.

1. Turn off everything. If you are in a machine room, turn all the equipment off. Turn the air handlers off. If you have glass door cases, turn them off. Then listen and look. You will be surprised how audible leaks can be even to us machine room deaf guys. Turn lights off and fans off. If you have an ultrasonic, use it.

2. Look for oil. Best advice already provided by
DEVITG and Sanderh.

3. Stay in the zone. Forget everything except that you are looking for a refrigerant leak.

4. Use all of your senses. Hearing, seeing, touching, tasting, smelling. One of my techs found a leak simply by shining his flashlight above the ceiling grid finding an ice ball on a liquid line. He didn't know it was an ice ball, but it was something 75 feet away that looked out of place.

5. Use histories and observations from people who work at the supermarket. History might tell you that certain lines have rubbed together before in the glass door cases.

6. Always go back to the previous repair.

Smaller leaks are just as difficult. If you are losing 100 lbs every 3 months, the same rules apply, but you might look at my list of things and work backwards. Nahh, that's too simplistic. When you get into the area that you know you have a significant leak, but it is not easy to find, you have to really go into the zone.

We have had success developing a leak check team of young fellows who are finding leaks that my top service people are missing. Even they express frustration as they chase mirage leaks, etc. But they are in the zone, focused... and that is the most important thing.

racknovice
02-09-2007, 04:00 PM
First of all, thank you for your kind advice.
Regarding the use of oil as evidence of a leak, being this particular rack one including an helical oil separator after the compressorsīs discharge and an oil return system back to the crankcases, would it be possible to find noticeable oil quantities up in the pipes and evaporators, seeping out through the leak location?

Lowrider
03-09-2007, 03:45 PM
Even with an oil separator there's some oil going trough the system. So unless it's a recent leak you'll have to find oil.

An oil separator usually only removes 90% off the oil from the discharge.

Mark
03-09-2007, 07:11 PM
First, you need a really good electronic leak detector. You turn it on outside the store and then see what happens. If you have a Tif or Detek and understand how they work, you lower the sensitivity and progress your way toward the leak. When you get confused, go back outside and start over. It is not easy finding leaks.

If you have what we call a screamer when you first step foot into the store, you might have to deal with it differently than if you have a localized leak. Here are some drastic moves you should make in such situations.

1. Turn off everything. If you are in a machine room, turn all the equipment off. Turn the air handlers off. If you have glass door cases, turn them off. Then listen and look. You will be surprised how audible leaks can be even to us machine room deaf guys. Turn lights off and fans off. If you have an ultrasonic, use it.

2. Look for oil. Best advice already provided by
DEVITG and Sanderh.

3. Stay in the zone. Forget everything except that you are looking for a refrigerant leak.

4. Use all of your senses. Hearing, seeing, touching, tasting, smelling. One of my techs found a leak simply by shining his flashlight above the ceiling grid finding an ice ball on a liquid line. He didn't know it was an ice ball, but it was something 75 feet away that looked out of place.

5. Use histories and observations from people who work at the supermarket. History might tell you that certain lines have rubbed together before in the glass door cases.

6. Always go back to the previous repair.

Smaller leaks are just as difficult. If you are losing 100 lbs every 3 months, the same rules apply, but you might look at my list of things and work backwards. Nahh, that's too simplistic. When you get into the area that you know you have a significant leak, but it is not easy to find, you have to really go into the zone.

We have had success developing a leak check team of young fellows who are finding leaks that my top service people are missing. Even they express frustration as they chase mirage leaks, etc. But they are in the zone, focused... and that is the most important thing.

Sound advice Dan. I have found the D-Tek tekmate likes the insulation/lagging glue in which bonds it together!.And i would advise having the detector(Tekmate) calibrated / checked anually with the substance in which the manufactures can supply in a sample pot.
I have also found many leaks on the fixtures etc, high and low side with the Packs turned off. And taking ceiling tiles out and removing services boxing if necessary.
With Kind regards
Mark

Dan
03-09-2007, 07:41 PM
I have found the D-Tek tekmate likes the insulation/lagging glue in which bonds it together!.And i would advise having the detector(Tekmate) calibrated / checked anually as it senses the R134A in R404A.

Mark, I have 5 extra sensors in my right hand drawer for Tekmates. When I have a technician missing a leak that another tech finds, I give him a replacement sensor. You really should try the Tif ZX-1. It surpasses the Deteks when looking for HFC's. But you bring up some good points:

Foamed assemblies or any foam will mislead you into thinking you have a leak. Hot water tanks such as Mueller or Thermastore must be cut loose and pressurized to determine if there is a legitimate issue, same with cold wall cases. I am told that the Detek sensors deteriorate or lose sensitivity when chasing HFC leaks as compared to CFC or HCFC leaks. I don't put much faith in that argument. It is my feeling that the software incorporated into the leak detector makes the difference. Resetability, etc.

I have never liked a single product from Tif until somebody insisted I try the ZX-1. We now have 10 of them.