View Full Version : Help me, I'm leaking

02-03-2012, 04:49 AM
Greetings Ammonia Gents:

Say, I have performed a leak test and have discovered a PRV that is leaking in one of our small plants (750 lbs flooded chiller). The PRV is on a de-superheater. The de-superheater has an isolation valve on the high side HP gas inlet. And, it has an isolation valve on the condensate outlet. The condensate drain line leads from there to a float drain that allows high pressure condensate through to the low-side so it can return to the flooded chiller through 8 feet of 1/2" tubing. There is an isolation valve where the liquid enters the chiller to valve of the drain tubing. A bench test on an identical float drain indicates that approximately 10 psi will be all that is needed to allow the liquid in the tubing to boil off and back into the de-superheater if it were depressurized.

Unfortunately, there is no service valve on this assembly. Due to the physical location of the de-superheater, the only place to install a service valve is on the bottom of the desuperheater and this would require moving the float down about 4 inches, and that requires opening the tubing.

Our main tech is unavailable to help with this, and while the rest of us have training in the use of safety equipment, PPG and the like, I would appreciate your thoughts on how best to vent this assembly so I can install a new PRV and a service valve and then vac the unit down. One idea is to turn on the pump and fan to the evaporative condensor and then to valve off the liquid return lines so as to pump out the chiller and therefore this little return tubing. Another thought would be to gear up and unscrew a fitting at the top of the de-superheater enough to allow gas to vent - after isolating the tubing from the chiller of course.

If any of you can share your techniques and safety tips, warnings, encouragement... with a newbie, I would greatly appreciate it.

Kind regards,


02-03-2012, 06:33 PM
First check the pressure just in case the valve is about to burst open. If so follow emergency procedures immediately! (Just to make sure this is not the case).

I suppose you don’t have dual relief valves with the appropriate fitting to stop one valve while servicing the other as you should have.

If this is the case talk to your boss, stop production, evacuate and fix! Before anybody gets hurt!

... my best advice.

02-03-2012, 07:15 PM
By the way, you may get more intelligent suggestions if you post this under the "industrial/NH3" forum.

03-03-2012, 06:50 PM
I moved your posts to the right section: Industrial NH3

03-03-2012, 11:39 PM
Thank you for the help Peter_1 in moving me to the right area in the forum. In reply to Aramis above, the PRV vented when a tech overpressured the system a month ago. It didn't seat fully, and has been weeping into the red-lines to the diffusion tank since. So, that is why I want to replace it. But, rather than introduce non-condensables into the system, I would like to take the time to install a service valve so I can evacuate this little de-superheater with my vac-pump, before putting it back on-line.

This is one of the places in this plant that doesn't have a 3-way (due to physical constraints, the area is tight), so the vessel has to be vented to plant with exhaust on, or, I need to draw it down using the recips (setting the low-suction to just above zero... and/or both.

I can shut down this system, and I can fully isolate the vessel based on my description above. So, based on that, how would you old-timers proceed?

Kind regards,


12-03-2012, 07:26 AM
Hi Mate, I could be wrong but it sounds to me like the line you describe is a high side float SV that controls a PMFL make up to this chiller? otherwise a it could well be the liquid drainer you describe ,Its hard to diagnose without siting a P&ID. From how most de-superheaters are piped they usually have a bypass to keep the plant running without the use of heat recovary, knowing designers they have not installed isolation valves or service connection to make the job simple an quick. The safest option is by far stopping production and evacuating the chiller in its entirety. Install a 3 way valve and a service connection if space is an obsticule move it! raise it on a stub etc, and install an isolation valve in the half inch bleed, there is more than one way to skin a cat. If you need to install a 2 or three piece SS hydraulic ball valve in this line they are rated at huge pressures and have PTFE seats and glands, I have never had one leak in 12 years +. If the leak is not causing any harm to anyone isolate the area and wait for someone who knows exactly what there up to, this is how intelligent people like yourself do stupid things.
If you are truly stuck, for the right price I can leave in the morning

13-03-2012, 03:47 AM
Hello NH34Life:

It is very kind of you to reply. This system does have a pilot operated valve operated by a low-side float. But, the de-superheater that I speak of has it's own tiny condensate return line from the bottom of the de-superheater down to the bottom of the chiller. That return line is controlled by a float of the kind that is commonly found on the bottom of an oil seperator. This de-superheater is on a branch line end-run, so it can be completely isolated from the main system. The HS gas doesn't flow through the de-superheater, but simply to it as gas collapses and condenses on to the internal tubing. Then the liquid flows down out of the de-superheater to the float.

I fully intend to wait to have a more experienced hand replace the PRV in a few weeks. There is no pressing danger as this PRV is piped to a diffusion tank full of water and it is only weeping. And, I am sure our tech will have to vent the ammonia gas to the plant to open and service the vessel, and I will be certain to be there to learn. I am just curious how you all would approach doing that safely, and if you think installing services valves when a vessel like this is opened makes sense??

Thank you very much for your warning. I have complete respect for NH3, and am always careful to wear my PPG , monitor, and to NEVER trap liquid into a hydrostatic loc that could cause plant or worse people harm.



13-03-2012, 11:05 AM
A few things come to mind for an experienced qualified person only to do the job.

If another part of the plant connected to it up/downstream of desuperheater, that has a purge valve that you can vent into water/atmoshere etc. This can be isolated & used purpose of emtying desuperheater.

If you empty liquid in the desuperheater & could run the plant into a vacuum (to allow for your 10 psi pressure drop). Then completely isolate, crack a fitting carefully to check if vacuum etc.

If relief valve leaks, it may empty itself if isolated & given time. Maybe even break seal on relief adjustment cap & take pressure of spring to purge into your diffusion tank.

Let us know how you go about it when time comes