View Full Version : Painful safety lessons

13-04-2002, 04:46 AM
This just happened last week. It truly belongs in the Trail of Tears.

A simple removal and replacement of a back room freezer case with a remote condensing unit. The freezer was in a commercial kitchen. The technician soldered up the liquid and suction piping and pressure tested the loop by putting liquid 404A into the liquid line. After a short time somebody ran out to him and said the freezer was on fire.

The tech valved it all off and rushed to the kitchen. When he arrived at the kitchen, he couldn't see any fog or vapor, but there was a kitchen employee who was clearly distressed and coughing well after being evacuated from the kitchen.

There was a bad solder joint in the liquid line. The tech repaired it, evacuated the system and put it back into service.

The affected employee gasped to the technician "That was really bad." I have to think there was an open flame somewhere.... but I don't really know.

It is an ugly situation. But other than using liquid to pressure test, I really cannot fault the technician. But we hurt somebody, and that is really bad.

And I am in a quandary. How much safety can we afford to instill? This was a $300.00 job.

Any advice, anyone?

13-04-2002, 10:53 AM
Hi, Dan these things happen, but to minimise events like this and their harm to people and property we should employ a risk assesment and a method statement system to each and every job. This envolves considering the likley safety issues and writting out a preplaned way of dealing with these events or minimising their effects.
One of the main things for a pressure test method statement is to restrict entry to the area being pressure tested to a safe distance, another is to provide adequate ventilation during the test.
We have generic method statements which are wrote out for work such as this. The service guy is issued with these, he gives a copy to the customer and keeps a copy himself. This statement is signed by both parties, both parties should be then clear in their duties. In the case of the above incident if a risk assesment had have been undertaken and the method statement given the fault would have been on the customers side for not removing staff whilst the pressure test was undertaken.
I know it's a pain, but so is a court appearance. A few copys printed off on the computer for standard proceedures is time saved in the end and it looks more professional when you hand one to a customer, it also puts a lot of the responsibility with the customer.
Regards. Andy.

13-04-2002, 01:53 PM
Thanks Andy. Would it be too much a bother for you to e-mail me what you have or direct me to a URL? I guess hindsight is better than no sight at all.

13-04-2002, 05:49 PM
Andy... That is really great. I'm taking that under advisement, also.:)

15-04-2002, 04:58 AM
Why didn't the tech use R-22 & Dry Nitrogen as a leak test gas (I believe the EPA would be much happier with this method than the one the tech employed)?
There is less danger with this method regarding phosgene gas, and he could have recycled the test gas and used it again on the next job.


15-04-2002, 12:48 PM
In Uk you are not allowed to leak test with ANY refrigerant.

All leak tests must be carried out using OFN. (Not even a trace of refrigerant is allowed.)

15-04-2002, 04:36 PM
If you're not allowed to leak test with any refrigerant, then why was the tech using R404a???

In the US, we are allowed to use a mixture of 10% R22/90% dry nitrogen as a leak test gas. The test gas is required to be recovered when finished.

27-09-2004, 10:53 PM
R 404 a for leak testing?!! Try nitrogen...As for the US being able to use 10% R22,I believe they signed on to the Montreal Protocol which forbids this.Then again U.S. seems to make their own rules with most things!

27-09-2004, 11:28 PM
Pressure and strength test on new installations are conducted with OFN, Leak testing of systems with OFN with a trace of R22 is still allowed as leak detectors cannot detect OFN. Technically speaking R134a and R404a (Hfc's with an ODP of 0) are not included in the Montreal protocol but are in the Kyoto protocol (limits on GWP)(and F-gas) which the aussies and yanks wont sign up too. :rolleyes:

28-09-2004, 04:38 PM
I can honestly say I could not believe what I have just read. Leak testing with any refrigerant is dangerous particularly in a kitchen. I have only leak tested with nitrogen where the system is new or devoid of gas for the last 15 years and I can honestly say I don't know any one who would use a refrigerant in any mix or quantity.

Good Brazing Good Practice and proper pressure testing with OFN is the only way it should be done unless the system is just short of gas and then a good leak test with an electronic locates the leak. If this doesn't work reclaim the gas and pressure testing using OFN. If that fails I have used florescent die, you do find the leaks although on open screw or recip compressors you can get a lot of mess around the shaft and the clean up can take forever.

28-09-2004, 09:07 PM
My first post was a long time ago. The technician is no longer with us and upon further analysis and questioning, he was only saying he was leak testing. He was, in fact, charging the system after having failed to perform a leak test. Which also speaks poorly of any vacuum he claimed to pull.

As far as the UK not allowing the use of a refrigerant for leak testing, that is a stringency that goes beyond the Montreal Protocol. Montreal protocol permits using trace amounts of a refrigerant combined with nitrogen in order to effectively use electronic leak detectors. Using a trace amount of refrigerant in the example I cited at the beginning of the discussion would not have caused any harm. The technician was not being truthful.

28-09-2004, 09:13 PM
Hi Guys,

Surely its about time that someone (BOC,Rhone,Gasco ect) came up with some kind of gas that can be added to OFN when pressure testing so any leak can be picked up with an electronic detector and that doesn't contravine any legislation on emmisions



29-09-2004, 09:56 AM
Hi Guys,

For me, I always prefer using N2 with trace refrigerants. As for safety and economic measures I do gradual increase of test pressures by using the regulator and perform leak test in between.

Big leaks can be easily identified at low pressure test. Contamination of the surrounding can be minimized, saving of the pressure test gas can be considered depending on the equipment size and above all-surprises can be elliminated if ever there is a pressure break.

The best formula for safety = observation+common sense+precautionary action/measures.

But, sad to say, we sometimes forget.


29-09-2004, 12:31 PM
With regard to producing a gas with a substance in it which can be picked up by electronic detectors, I believe that one of the manufacturers produce one with a trace of helium in it. (I would guess it is OFN with a trace of helium).

So you can get electronic leak detectors which pick up helium, although they are quite expensive. :rolleyes:

30-09-2005, 03:07 PM
does anyone remember the pipework guys who leak tested a multisplit system in the north east of uk with water?

seemed like a good idea at the time----LOL

30-09-2005, 03:52 PM
Using refrigerant to do a pressure test actally cost higher than using dry nitrogen gas, especially 404a. But to carry the bottle of dry nitrogen cylinder around makes the job tougher.

30-09-2005, 08:23 PM
I'm quite amused when I see people saying that watching a vacuum gauge rise will indicate whether or not a system is leak tight.

Bear in mind that a system under absolute vacuum will only be subjected to 1 bar atmospheric pressure (14.9psi) and is mainly used to indicate the presence of water vapour.

If you apply nitrogen (OFN) at multiples of atmospheric pressure then you stand a much greater chance of seeing if your system is leak tight and capable of withstanding working pressures. :confused:

US Iceman
01-10-2005, 12:43 AM
does anyone remember the pipework guys who leak tested a multisplit system in the north east of uk with water?

This is not too surprising. I heard of a large centrifugal chiller that was hydrotested. Not the water piping mind you, the chiller package...:eek:

The compressor and refrigerant piping were completely filled with water and then pressurized to find the leaks.

Someone always has to get creative and have a "better" way of doing things. Similar to the crimped tubing used as an expansion device in another post for the propane refrigerator.

Who knew refrigeration could be so difficult?:p