Lethal Service: To categorize a process as lethal service (or not), to use a U-stamped vessel in th 1

[jmedclay]

A client wishes (to their credit only if totally acceptable from H&S perspectives) to re-purpose a 6000 gallon, U-Stamped pressure vessel for use in a process which will generate about 4 lbs/min of H2S in the head space. IDLH is 100 ppm; a small leak would have serious H&S consequences. We have been asked to help navigate the issue. In searching for guidance, and in spite of the lethal service language in the code, it appears that some industries don’t, by default, consider vessels containing substantial concentrations/quantities/pressures of H2S as constituting lethal service while others do. Quantitative breakpoints are not apparent to me. Neither I nor my firm are pressure vessel experts or are calibrated to this part of the CPI world, hence my kicking this around, here.

Steam jacket pressure will be approximately 40 psig, process liquor pressure about 4 psig. The process area is within a building and neighborhoods aren’t far away. My interpretation is that this is indeed a pressure vessel as defined by B31.3 and that a lethal service verdict must be formally made wrt H2S release. The client would like to avoid that classification and use the existing vessel. Are there methods whereby that be done in good conciseness? Are there engineering and administrative approaches which are professionally sound and well protective of human health in which this service could be considered non-lethal? Say, H2S gas monitors at all flanges, manways, seals etc. with alarms in the op-area and CR, SCBA + radio for all in the op-area, vacuum conditions in the transport of headspace gas as soon as possible in the line, burst disks ducted to the scrubber (or flare), immediate feed SD on alarm (which quickly stops the reaction), induced building ventilation, and so on. I assume many of those safety items would be used regardless but, again, I am not calibrated to the CPI world; I do not know what is considered reasonable & customary in this context. If the answer to the question is NO, that it’s absolutely contrary to accepted practice, there is no way around it, then so be it. If the answer is “maybe, sometimes one can reasonably do that sort of thing” then a next question seems to be: Is this U-Stamped vessel close enough in the details that matter to be reliably used for this process? Because time is short (surprise!) I requested the U-1, any R-1s and 4s, all inspection reports, mfg fab drawings & materials used, and anything else in the vessel file, with the suggestion that we put them in front of a U, R & L-Stamp shop for an opinion as to how close it is to complying with L-Stamp requirements. I also suggested consulting the client’s insurer straight away.

Any reactions? I’d greatly value experienced feedback.

Thanks much,
John

 

[georgeverghese]

I have not come across a plant or equipment operating with H2S classified as being in lethal service - all were classed as toxic service(and surely not at 100ppm H2S)- this is in SE Asia.
Maybe it is more severe in your location.

 

[SJones]

See if you can distill anything from all of this:

http://www.eng-tips.com/search.cfm?q=lethal service&action=search

If you want to develop some form of rational basis with the client regarding a "lethal service" definition, try this:

http://www.unece.org/trans/danger/publi/ghs/ghs_rev05/05files_c.html

I'm not a technical safety expert, but I would expect that the recovery barriers of monitors, alarms, BA sets etc result from the hazard rather than serving to reclassify it.

Steve Jones
Corrosion Management Consultant

http://www.linkedin.com/pub/8/83b/b04

All answers are personal opinions only and are in no way connected with any employer.

 

[chemjock]

My experience is in oil refining where there are streams at high pressure and temperature with high concentrations of H2S and none of the equipment was classified as lethal service. The only time I have ever seen it applied was in a plant for destroying chemical weapon mustard agent and when the agent was in concentrated form. It was not applied After dilution. While I found no precise definition I got the impression that if the escape of very small quantities of the chemical was a hazard to nearby personnel then lethal service was justified.

 

[EmmanuelTop]

If there is nothing specific written on service classification in the codes of the country where this vessel/system will be operated, the asset owner still needs to demonstrate that the risk of injuries/fatalities has been reduced to an ALARP level, as a minimum. This is also the requirement in standards of most of the operator companies worldwide.

You also need to look at what are the legal requirements for obtaining the operational permit for this facility, and perhaps there you can find some further insight. Given the presumably high level of consequences in cases of loss of containment, this facility might qualify for Safety Case.

The classification of a vessel itself for lethal service will certainly build additional inherent safety in the design, but you (or the asset owner) as an end-user should have a systematic and documented risk assessment done by certified 3rd party. This study will quantify the risks of operating the facility, and there you can see if additional and which safeguards are required to reduce the risk to a tolerable level.

Dejan IVANOVIC
Process Engineer, MSChE

 

[georgeverghese]

a)What would the vapor space H2S concentration in this vessel be in the proposed service? Cannnot tell from this statement that 4 lb/min of H2S is generated.
b)You say process liquor pressure is 4psig, but you also say that vapor space will be operated under partial vacuum ?
c)Are the relief valves on the vessel good for steam jacket rupture into the vessel?
d)When was the last time this vessel was inspected for corrosion
e)Do these new operating conditions with H2S require compliance to NACE on materials and construction methods?
f)Are there any new corrodants that would increase the likelihood of loss of containment due to corrosion?

 

[georgeverghese]

Topic unrelated to this discussion:
If the emergency relief collection header is routed to a regular flare, this may not be acceptable given this unit is in a populated area - full combustion of H2S will most likely not be possible, especially on turndown at the low pressure flare burner tip - what would the corrresponding ground level concentration of H2S be in this neighborhood ? - how would you deal with fugitive H2S rich emissions from RVs - is an incinerator or thermal oxidiser required for this location?

 

[georgeverghese]

At the very least, it does look like this vessel needs to comply with NACE MR 01-03 / MR 01-75 for sour service operating conditions.

 

[jmedclay]

Thanks for the input everybody. You've given me a number of specific areas to investigate (which I shall pursue). The overarching take-away for me, though, is that we need the guidance of a firm experienced this sort of evaluation. There are many angles to consider. I had a little bit of HAZOP experience with Risk Management Professionals many years ago so I'll ping them to see if this is fully within their experience base. If you know of any others please sing out. Any other technical info appreciated as well; at the least I'll review that material enough to be able to ask halfway reasonable questions.

Thanks again everybody,
John

 

[jmedclay]

Thank you George!

a) What would the vapor space H2S concentration in this vessel be in the proposed service? Cannot tell from this statement that 4 lb/min of H2S is generated.
- Process engineer noted that 4 lb/min of H2S is generated but didn’t note the concentration. Will ask.

b) You say process liquor pressure is 4psig, but you also say that vapor space will be operated under partial vacuum ?
- The vessel and headspace piping up to the PCV is at 3 to 4 psig. Beyond that it’s feasible to operate at a vacuum.

c) Are the relief valves on the vessel good for steam jacket rupture into the vessel?
- I mentioned the possible injection of steam into the process cavity (corrosion, whatever type of failure) and noted the need for adequate RV capacity. Am I right in supposing that evaluation would be based on the maximum steam supply capability of the delivery loop under operating conditions, rather than assuming a fixed size hole in the vessel?

d) When was the last time this vessel was inspected for corrosion
e) Do these new operating conditions with H2S require compliance to NACE on materials and construction methods?
f) Are there any new corrodants that would increase the likelihood of loss of containment due to corrosion?
- Don’t yet know date of last inspection. Have requested the U1, R1, R4, all inspection, repair & testing documents and original engineering & fab documents. Ammonia and CS2 are present. Have requested concentration data.

g) If the emergency relief collection header is routed to a regular flare, this may not be acceptable given this unit is in a populated area - full combustion of H2S will most likely not be possible, especially on turndown at the low pressure flare burner tip - what would the corresponding ground level concentration of H2S be in this neighborhood ? - how would you deal with fugitive H2S rich emissions from RVs - is an incinerator or thermal oxidiser required for this location?
- Thanks for that info/perspective. Current notion is to route relief flow thru the wet scrubber. Have noted the need to check for adequate capy, performance & general suitability. Use of a flare was just noted as a consideration – your questions are helpful there. Therm ox – Don’t know. I wonder if they use one for anything else at that location.

h) At the very least, it does look like this vessel needs to comply with NACE MR 01-03 / MR 01-75 for sour service operating conditions.
- I ran into those specifications – will look to start educating myself as to rqmts/recommendations.

I also requested pH & temp data so as to figure where, in the sour service continuum, this system is (low, moderate, severe). I think this would be considered wet sour service.

John

 

[georgeverghese]

c)You could size the RV on the basis of max normal steam pressure leaking through an orifice equivalent in cross section area to that for a 1inch tube into the vessel which would be at relieving pressure at this time.
f)Ammmonia is another contaminant ( other than wet H2S) that may require PWHT, if I recall correctly.
h)It should be possible to get a theoretical value for pH given the H2S concentration in the vapor space and the pKa values for H2S at the corresponding operating temp / pressure.