Valves for Lethal Service 1

[Giuss]

I am specifying valves for a oil centre.

In this specific case, the fluid is hydrocarbon with a high presence of hydrogen sulfide H2S (up to 160000 ppm).
The Owner of the plant defines "Lethal Service" each H2S concentration greater than 1000 ppm.
The piping class is designed according to Cat.M of ASME B31.3.

What are the precautions you would take for preventing stem/packing leakage to the environment?

If ISO 15848 is the chosen standard for fugitive emission testing, would you specify Tightness Class A?

In case you specify bellow seal valves, Would you in any case require the above mentioned emission test?

Thank you all.

 

[bcd]

For lethal gas, the ASME standard simply mentions that the valve should have enhanced stem seals. No further definition or reference to any standards.

My opinion, ISO 15848 does not prove that the shaft seals will not leak once the valve is in the plant. It is just an expensive, time consuming and overly complex lab. experiment that at best measures the permeability of the seal material. It does let you know if the seal will remain tight after a thermal cycle. But thats about it.

Once the valve is in the plant, outside factors will detemine if the shaft seal leaks. Is the operator mounted properly or are there side loads on the valve shaft? Is anyone monitoring for leaks and properly adjusting the packing load? etc. Do not rely on certificates from a lab. experiment as proof that a seal is going to remian tight in service. Especially with lethal media. You need to rely on experience and ensure you have the best seal possible for the application as proven through actual use in a plant.

 

[JimCasey]

Bellows seals are the traditional answer, but they are VERY expensive and have a limited cycle life. ALso when they fail it tends to be a catastrophic failure.

Contemporary solution to the problem is to use a double packing set with a lantern ring between the packings, Leakoff from the Lantern ring is piped to a detector and then to safe disposal. THe second packing set has no differential pressure so it is assured that nothing escapes to atmosphere. Additionally such packing is customarily live-loaded to compensate for wear or thermal effects.

Rotary valves avoid a major packing wear mechanism since the stem does not drag in and out of the packing box. Any micro-roughness in a rotary valve only moves packing particles circumferentially inside the packing stack. Packing wear is drastically decreased so the packing stays tighter and lasts longer. Compared to a rising-stem valve where the same shaft surface finish drags particles out of the packing box, and also drags any contamination that lands on the stem back IN to the packing. This accelerates packing AND stem wear, so leakage occurs much, much sooner with the rising stem valve.

You say "High pressure" but that's a relative term.
Some of the tightest valves are rotary valves designed for ASME class 150 and 300. Example: The Durco TSG4 has a primary seal of a tapered plug within a sleeve. There is a secondary seal of a PFA diaphragm with a pressure-energized reverse-lip stem seal. Then they stack on a double TFE Live-Loaded packing set. That valve is frequently used for really nasty chemicals such as phosgene or HF. Above class 300, a ball valve with double live-loaded packing will still outperform a rising-stem valve for lowest stem leakage.

 

[JLSeagull]

Review articles related to fugitive emissions. Most companies tag the stem seals for non-lethal valves as part of their environmental program. Methods to control fugitive emissions include extensive stem leak detection/monitoring. Search for acronyms like LDAR (leak detection and repair), VOC (volatile organic compounds), etc.

Once the H2S level exceeds some nominal percentage, some companies require welded end valves instead of flanged end valves. Even in these cases the instrumentation is usually threaded.

 

[unclesyd]

Here is a link to Vogt's standard style forged valves some of which are designed for H2S. If you go to the Vogt page at Flowserve there are other style valves for your service.
The question is whether your client will accept a standard figure valve for his "M" classification since they are used by others under the same service conditions but other classifications.

http://www.flowserve.com/vgnfiles/F...tLiterature/FlowControl/Vogt/VVACT0000-01.pdf

 

[NGiLuzzu]

Giuss,
I agree with bcd; in particular, I believe that in no way new valve test results may replace good maintenance practices and LDAR programs!!!

First of all, you should choose valves from experienced and reliable Manufacturers: to do that, the availability of type test certificates and/or the proven capability to production test according to the most common fugitive emission test standards (ISO 15848, TA-Luft regulation and VDI 2440 procedure, EPA 21 method, Shell MESC SPE 77/312 specification, etc... *) may represent a positive reference, but should not be the only one to drive your choice!

In addition, as mentioned by JimCasey, quarter turn design with lantern rings within the stem packing should be preferred whenever possible: the relevant ports should be used to pressurize with inert gas or to suck possible leakage away from the stuffing box internal volume, and this should be done continuously.
Moreover, the rotating stem design may also allow the application of elastomeric o-rings (e.g.: in Viton, Kalrez or FEP materials) on the gland bushing and/or the stem bush bearings, in order to provide an additional barrier to the fugitive emissions.


In any case, as I said before, it is mandatory to monitor the valve external tightness performance along its service lifetime and to put in place, whenever necessary, the proper maintenance interventions as soon as possible.



Hope this helps, 'NGL


- - - - - - - - - - - - - - - - -

* See also:

- thread408-143739 ;

- thread408-93545 ;

- thread408-191910 ;

- thread408-40826 ;

- thread408-150132 ;

and other discussions about the same issues within this Forum.

- - - - - - - - - - - - - - - - -

 

[DrJim]

In addition to the above, NACE MR0175 is a must read if H2S is involved.
Also, API Spec 6A (a well head publication) gives an informative overview with specific detail on how the manufacturing requirements are increased as as seriousness of the application (PSL - product specification level) is increased. There is a lot more involved than stem seals.

 

[bmoorthy]

The original posting sounds like Maersk application.

Is it?