PRV Icing Issue at liquid ammonia service 1

[JZ59]

A PRV is installed at our site on Liquid Ammonia Service with operating temperature of 90F and Operating Pressure of 300 Psig. CDTP of PRV as mentioned on nameplate is 600 Psig and same is used during its calibration in shop.

Icing was reported on PRV body at disc area in past month (picture attached). This PRV was brought to our shop facility for calibration and pressure testing. No issue was found in Pressure and Leakage test in “as found condition”. PRV calibration was also found Okay and we installed it back.

Same issue of icing was reported after a couple of days of its installation. PRV was again brought to shop facility for pressure and leakage tests and no issue was found. Inspection of PRV internals was also carried out this time and no anomaly was found.

In the light of your experience, what do you people think should be the probable causes of this recurring issue?

Thanks,
JZ59

 

[The Obturator]

You have a large gap between operating and set pressure, so that rules out the obvious problem of leakage.

The leakage is probably being caused by a build up of icing caused by escape of ammonia, but can also be attributable to a variety of reasons including;-
1. Seat design - is it metal to metal or elastomer (material?)
2. If metal seats, have they been lapped properly? Have OEM parts been used?
3. Are the seats still meeting manufacturers minimum dimensions?
4. Are the other internals in good condition? - Is the spindle true?
5. Has the spring been checked - does it have flat ends - is it the right spring for the stated CDTP?
6. Is there anything at the valve installation that may make the valve leak? - eg, is there any outlet piping 'hanging off' the valve causing the seats to distort? Your picture suggests heavy outlet piping 'bending' against the valve inlet possibly distorting the seats - seen this happen a lot and most users don't realise.
7. How are you testing the valve? with ait, liquid?
8. Is there any back pressure?
9. Have low temperature materials been selected for the construction of the valve.
etc.

Provide more details about the valve (brand/type) if you can.

Please also note there is a separate safety/relief valve forum.



*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[Compositepro]

I would say that you are observing cavitation in the ammonia exiting the valve due to high velocity and an increase in pipe diameter.

 

[LittleInch]

Can you get a picture of the piping coming off the valve. Point 6 above sounds the most likely to me with the information we have at the moment.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[The Obturator]

@Compositepro Pressure-relief valves are not 'in-line' type valves - they are not seeing constant flow as do control valves. They should only open in event of overpressure. In theory they should never open and are there as the last line of defence against life and property. Hence cavitation is not an issue with PRV's.

*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[Compositepro]

Thank you, that is true for a relief valve. I interpreted PRV as a pressure regulating valve. This is why I often don't bother to reply to posters who do not bother to define their acronyms.

 

[JZ59]

@TheObturator Thank you for your response. Please find answers to your questions below:

1. Seat design - is it metal to metal or elastomer (material?)
Seat Material is SS316 as found in PMI.
2. If metal seats, have they been lapped properly? Have OEM parts been used?
Seats were lapped by experienced workers as per site procedures.
3. Are the seats still meeting manufacturers minimum dimensions?
Actually this PRV dates back to 1990's. Unfortunately we could not found any data sheet or drawings to compare the dimensions. This issue came up in December 2021, and before that same team has been involved in its servicing and calibration.
4. Are the other internals in good condition? - Is the spindle true?
Yes, the spindle is true and all other parts were found in good condition.
5. Has the spring been checked - does it have flat ends - is it the right spring for the stated CDTP?
The spring has flat ends. I searched the spring # online but could not get any details. (Spring # HC1062 ST 35072)
6. Is there anything at the valve installation that may make the valve leak? - eg, is there any outlet piping 'hanging off' the valve causing the seats to distort? Your picture suggests heavy outlet piping 'bending' against the valve inlet possibly distorting the seats - seen this happen a lot and most users don't realise.
No, the outlet piping has a support right after the outlet flange. See attached picture.
7. How are you testing the valve? with ait, liquid?
I am testing the valve on water as per past records. Further, the nameplate mentions the water capacity at 10% overpressure to be 48.6 USGM. This suggests that the valve should be tested on Water not air.
8. Is there any back pressure?
The outlet line is connected to a stack. No back pressure is mentioned on nameplate and CDTP is same as set pressure.
9. Have low temperature materials been selected for the construction of the valve.
etc.
PMI was carried out for all the valve internals. All internals were found to be of SS316. We do not have any vendor provided document to compare it to.

Provide more details about the valve (brand/type) if you can.

Make: Farris
Serial #: EF2001272
Type: 2740 UL
Size: 3/4" x 1"
Service: Liquid Ammonia
Oper. Pressure: 300 psig
Oper. Temp: 90°F

I tried to contact the vendor and asked a couple of questions regarding media for pressure testing, pressure testing procedure, data sheet availability etc. They replied that this model is obsolete and they do not have any information regarding this.

I searched for the PRV type online and found a Farris Manual for it. It has generic diagram for this PRV type (attached) which chalks out it parts.
When we opened this PRV for internal inspection, we did not found any disc retainer ring (picture attached). I asked the vendor whether this item should be part of our PRV or not. He excused that the model is obsolete and no information is available.

The subject PRV was sent to area yesterday for installation. Icing has been again reported on the same part.
Your experience in figuring out the root issue will be of much help.

 

[LittleInch]

Has the seat simply been lapped too many times?

25 years of operation it might just be lifed out.

The support looks a bit rigid.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[The Obturator]

Hi again.

This valve is past its design life. I also found the maintenance manual on-line, one that also tables critical minimum dimensions for the seat (nozzle/base) and disc. Suggest you check these. Manual can be accessed here:

https://www.yumpu.com/en/document/read/38221317/farris-2740-2741-2740ul-2745-manual-boig-hill

Typically, the seats, gaskets (if fitted) and spring should have been changed some 4 to 5 times with OEM parts in a 30 year installation cycle. Especially on an application protecting Ammonia.

Other than that, I recommend you change the valve simply due to its age. It is installed to protect life and property so don't compromise its function. I would also discuss seat tightness with a potential vendor and make known the relieving temperature causing icing.

*** Per ISO-4126, the generic term 'Safety Valve' is used regardless of application or design ***

*** 'Pressure-relief Valve' is the equivalent ASME/API term ***

 

[MortenA]

I assume that the pressure side/inlet is "down"?

If so why is the ice forming in this direction? I would think that it should be in the direction of the discharge?



--- Best regards, Morten Andersen

 

[JZ59]

@The Obturator, I thank you again.

The seat and disc dimensions matched the ones given in the manual.

 

[JZ59]

@MortenA, you are right.
Actually the valve is installed on Ammonia Service. In case of slight passing, Ammonia would flash right after passing through the seat causing temperatures to drop in that region. The icing that we can see in the picture is at the same area where seat and disc are located inside the body.

But the question is if the valve has cleared the pressure test in shop at 600 psig on water, why is it passing at ammonia at 300 psig?

 

[SPDL310]

It may be a difference in hold time vs leak rate. The initial leak rate for the ammonia may be low enough that it is undetectable in the water/time you are holding on the hydro test. You could try a longer hold period or gas testing to see how resilient your seat is. Once you are in line and holding for long periods of time on the ammonia the leak could cause uneven seating due to the temperature drop from the leak flashing off. Another possibility is the piping loads @The Obturator mentioned earlier causing seat distortion. Either way I agree with his recommendation to replace the vale.

 

[LittleInch]

Do you ever get pressure spikes at this location, even for 1/2 second?

otherwise if it only starts a few days after installation SPDL310 may well be correct and that yoou have a very small amount of passing which is difficult to detect.

Ask the shop to test it in static conditions for a few hours and see what happens.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.