PDA

View Full Version : Phantom leak help needed







cryotec
06-06-2015, 01:45 PM
System tested at 10 bar for 30 minutes each part
System break down as fallows
1. Compressor discharge to receiver out let
2. Receiver to cabinet before TEV
3. TEV to cabinet suction out
4. Suction line
Each section has isolation valves all were closed and started with rest Line connected to compressor for 30 minutes then tev iso valve and last suction out to condensing unit but no drop?
But system looses gas

GHAZ
06-06-2015, 02:05 PM
Hi it might leak when the joints get hot, or at higher pressure,

cadwaladr
06-06-2015, 02:20 PM
Is this a commercial cabinet? Sometimes putting a little refrigerant in the system pressurised with nitrogen get the sniffer out and check it,or throw it in a swimming pool!

hookster
06-06-2015, 07:50 PM
For really difficult leaks a tracer gas is used which has a low molecule gas to pin point small holes. (Helium/Hydrogen).

Do not to forget to check your schrader valves after removing gauge sets, many an engineer has been caught with these.

cryotec
07-06-2015, 01:10 AM
Hi it might leak when the joints get hot, or at higher pressure,

Cant be higher pressure , heat is possiable how do i put some heat in ?

cryotec
07-06-2015, 01:18 AM
no schrader it has back seat service valve ( 6 hp condensing unit feeding 2x 3m dairy cabinets ) shutting down in 50 - 70 hours on lp

martinw58
07-06-2015, 06:58 AM
have you tried putting a UV dye into the system

cryotec
08-06-2015, 12:16 AM
have you tried putting a UV dye into the system

yes no show in obese places

Rob White
08-06-2015, 12:52 AM
Cant be higher pressure , heat is possiable how do i put some heat in ?

10 bar is quite low for pressure testing / leak finding.

You say it is a dairy with a 6hp comp, I can't see where
you mention the type of refrigerant?

If it is 404a then low side would be 32degsC = 14barg and the high side
at 55degsC = 25barg so if you can isolate the system you can pressure
test to higher pressures than 10barg.

What is on the system that limits the pressure to 10barg?

30 mins is no time either, you are wasting your time doing it this way.
Pressure test it correctly up to it's it's safe maximum working pressure
and then leave it for 24 hours if necessary.

Use good quality bubbly water and check every section manually,
if there are 2 cases isolate them and test them separately, play close attention
to evaporator end returns.

Does the system have a pressure release valve on the receiver?
If it does put a party balloon on the end and see if it inflates.

Leak testing is a systematic, boring, time consuming job but it must
be done, leaving something for thirty mins and then putting gas into it
is just going to result in the same way as previous attempts.

Also don't always think it is going to be one big leak, small leaks add up.
It could be more than one very small leak and don't forget the service
valve. Back seating service valves are one of the most common places
for a leak and it can be missed because we fit our gauges to them.

Make sure you test the valve with the gauges off as well as on.

2 x 3mtr cases lasting only 3 to 4 days then going out on LP is some doing.
How much refrigerant? 3 or 4 Kg's in 3 days? Wow......................

Pressurise to the highest safe limit LP switches, evaps and the LP side of the comp
are the weak points but you could still go above 10barg (unless it was on 134a?)

Isolate all sections and wait overnight, check inside case with bubble water and check
all the condenser, compressor and receiver components, then start with the pipe.
Remove section of armaflex insulation and stick a leak detector down each end.

Check for oil, with the amount of gas loss there will be oil everywhere.

Lastly put video cameras on it and see if someone is stealing it, you may laugh at that
but it is becoming more and more common for people to steal refrigerants thinking
they can get high on them (i have seen the video proof if you doubt me) :)


Regards

Rob

.

Rob

.

cadwaladr
08-06-2015, 08:58 PM
10 bar is quite low for pressure testing / leak finding.

You say it is a dairy with a 6hp comp, I can't see where
you mention the type of refrigerant?

If it is 404a then low side would be 32degsC = 14barg and the high side
at 55degsC = 25barg so if you can isolate the system you can pressure
test to higher pressures than 10barg.

What is on the system that limits the pressure to 10barg?

30 mins is no time either, you are wasting your time doing it this way.
Pressure test it correctly up to it's it's safe maximum working pressure
and then leave it for 24 hours if necessary.

Use good quality bubbly water and check every section manually,
if there are 2 cases isolate them and test them separately, play close attention
to evaporator end returns.

Does the system have a pressure release valve on the receiver?
If it does put a party balloon on the end and see if it inflates.

Leak testing is a systematic, boring, time consuming job but it must
be done, leaving something for thirty mins and then putting gas into it
is just going to result in the same way as previous attempts.

Also don't always think it is going to be one big leak, small leaks add up.
It could be more than one very small leak and don't forget the service
valve. Back seating service valves are one of the most common places
for a leak and it can be missed because we fit our gauges to them.

Make sure you test the valve with the gauges off as well as on.

2 x 3mtr cases lasting only 3 to 4 days then going out on LP is some doing.
How much refrigerant? 3 or 4 Kg's in 3 days? Wow......................

Pressurise to the highest safe limit LP switches, evaps and the LP side of the comp
are the weak points but you could still go above 10barg (unless it was on 134a?)

Isolate all sections and wait overnight, check inside case with bubble water and check
all the condenser, compressor and receiver components, then start with the pipe.
Remove section of armaflex insulation and stick a leak detector down each end.

Check for oil, with the amount of gas loss there will be oil everywhere.

Lastly put video cameras on it and see if someone is stealing it, you may laugh at that
but it is becoming more and more common for people to steal refrigerants thinking
they can get high on them (i have seen the video proof if you doubt me) :)


Regards

Rob

.

Rob

.
once again first class advice,when you pointed out what quantity is or may be being lost it certainly has one hell of a leak.

absoair
09-06-2015, 03:09 AM
Had a small walk in freezer that had a leak that took more time to find than it should have. Pressurized to 450 psi and found the leak at the copper to brass factory connection on a solinoid valve. It seemed to leak worse while operating than with 100 psi. Maximum test pressure found it.

cryotec
12-06-2015, 11:52 AM
once again first class advice,when you pointed out what quantity is or may be being lost it certainly has one hell of a leak.
Sorry been away family matter
Late night i could not read my writing test pressure was 20 bar and not 10 bar
it is 404 system

redroge
12-06-2015, 03:13 PM
Have you been round it with a leak detector when there,s gas in it ,make sure you check pressure switch bellows

Tycho
12-06-2015, 11:05 PM
Me, I would have pulled the whole system down to vacuum and then sectioned it of.

-Vacuum system
-Connect vacuum gauge to one of the sections
-leave over night

next day, observe reading on vacuum gauge

-open one by one valve, starting with the section closest to vacuum gauge
-observe change in vacuum between each time you open a new part of the system.

If I know the system is loosing refrigerant and I can't find the leak with nitrogen, I always go for vacuum, and use a 0-100mbar gauge, because you get a bigger reading when you loose the vacuum than leaving the system overnight with 10 bar in it. leaving 10-20 bar in a system light leave you with the needle on the pressure gage moving 0.01 mm.

With a 0-100 mbar gauge, you have a span of 0.1 bar, or a span of 0-1.4 psi and that's not a lot, so you can detect the tiniest of leaks
----------------------------------

Also, as Rob White said, "30 mins is no time either, you are wasting your time doing it this way.
Pressure test it correctly up to it's it's safe maximum working pressure
and then leave it for 24 hours if necessary"

and

"Leak testing is a systematic, boring, time consuming job but it must
be done, leaving something for thirty mins and then putting gas into it
is just going to result in the same way as previous attempts."

Both of those quotes are words to live by, it's boring as hell, but it has to be done...

The worst part is standing around waiting, but that is a whole lot better than having to come back every 3-4 days for the same problem...
The client will forget that you spent 15 hours trying to find the leak but fixing it, but he will never forget that you had to come back every three days for a 2 week period, making him pay for the ***** and your time, before you managed to fix it.

cryotec
07-07-2015, 12:54 AM
Found it !!!!!!
32 bar no drop in 24 hours so i got sick and got a big propane torch and heated the condenser and got to about
80 C before i saw drop
Thanks for all your help

Rob White
07-07-2015, 03:14 PM
.

:)

Well done

Rob

.

cadwaladr
07-07-2015, 10:11 PM
Wish I had been paid in full for finding leaks and faults,but nothing gives me greater pleasure in the job than sorting a system the ones that are easy compensate,an old retired fridgy mate said to me once your like a scrap yard dog you won't let go!

hvac01453
06-10-2015, 02:04 AM
And the pecker head connection on the compressor, and the fuse able plug on the receiver tank.
On split systems with long line sets, I cut and cap them with a Schrader on one end, cap and pressurize the two end pieces ( condenser and air handler ) and pump up all four, to try isolate where to look. Some are vibration leakers and dyes are usually the only way to find them.

wooleybull
11-10-2015, 12:26 AM
yes if you have systematically checked all components at max test pressure u can almost guarantee its going to be in the condenser, i would be checking oil level cause there so
should have been a **** load of oil, soaking the condenser with bubbles once you put dye in and shining you uv light on the liquid running into the condensate pan can work as well

pie man
02-11-2015, 11:10 PM
Most common leaks for me porous evap coils and receivers plugs (usually on a Friday evening)!