Flame arrestor in arbitrary piping

[Tom_H]

A question has been lifted in a HAZID regarding an external fire happening in a plant. The discussion then related to the risk of a line bursting due to this fire and the fire igniting the contents of the line and then the risk of flashback in to the atmospheric vessel to which the line is eventually connected.

The line would only be used during on/offloading operations and even though it is purged, there might be low points with flammable liquid that could potentially flash etc. There were also other lines that similarly would not be in continued use but that could have flammable liquids in them. One "solution" that was lifted to prevent flashback in to the vessel (which has blanket gas) was to put a flame arrestor on the line (close to the vessel). The discussion was kind of strange since this could apply to almost any line, whereas the normal use of a flame arrestor (if used at all) would perhaps be on relief vents or similar. Any experience in the forum on these kinds of issues?

 

[MJCronin]

You state:

"the risk of a (pipe)line bursting due to this fire and the fire igniting the contents of the line and then the risk of flashback in to the atmospheric vessel"

Your proposed scenario fails the common sense test because it assumes that the piping will reach its burst pressure without also pressurizing the connected vessel. The vessel relief devices will actuate OR the vessel will fail before this can happen. The vessel will fail before the piping system.

It seems that you are exploring the fire hazards of a system with long, open interconnected piping. Many similar systems are designed with shutoff valves at the storage vessel

If you are still concerned about the fire effects of a long connected pipelines on your vessel, you could add an inexpensive fire-safety shutoff valve adjacent to the vessel, rather than an expensive flame arrestor.

https://docs.preferred-mfg.com/libr...Safety Oil & Gas Shut-Off Valves Cutsheet.pdf

MJCronin
Sr. Process Engineer

 

[EmmanuelTop]

There are two interesting things here:
- This is actually a fire escalation scenario during which rupture of the subject line occurs, due to exposure to heat/flame (coming from another source). This system is - as you say - normally purged and can be considered fairly free of hydrocarbons, hence unable to sustain fire just by itself.
- The system is at atmospheric pressure. How much flammable inventory can an atmospheric system contain, in real life? We must be speaking of really large volumes then.

I second MJ Cronin's opinion with respect to isolation valves - if the fire scenario can be proven as credible and the consequences are considered severe, whilst no other safeguarding measures exist.

Dejan IVANOVIC
Process Engineer, MSChE

 

[Tom_H]

Thank you very much for your insightful responses. The scenario seemed far-fetched and unrealistic. I see also that API 521 indeed designates this as a double jeopardy scenario (failure to purge segment after filling and simultaneous external fire). That of course relates to relief provisions and not flame arrestors, but still.

Regarding the shutoff valve, the piping does have a ball valve at the vessel, so it is not connected to the vessel when it is not in use (i.e. outside filling operations).

 

[georgeverghese]

General oil/ gas industry standard is to aply a minimum of 7barg design pressure for all piping and equipment for which flashback is possible, so your tank would not survive a flashback event. Your description of this scenario however sounds somewhat vague - how does this line burst in the first place - due to overpressure ?? How would that happen ? - cannnot tell at the moment.

 

[Tom_H]

No, you are right, the tank kan only take a few mbar. This is a case that often occurs, where you get a somewhat strange HAZID finding that someone has dreamed up in a workshop, but you are not entirely sure how to quickly close it out even though it intuitively sounds very strange and not realistic.

The basis is that we have a (fairly long) line used for filling of an atmospheric vessel. The "hazard" that was raised in the HAZID was that, what if we either forget to purge the line, or alternatively that the purging leaves a fair amount of flammable liquid left in low-points, and that subsequently (at a later point when the line is not in use and shut in) a random external fire happens in any vicinity of this line. What if, then, the fire causes escalation and causes this line to burst (either through overpressure or perhaps more likely material fatigue) and then ignites the contents of the pipe. What if, then, the ignition of the flammable liquid and its flashed vapour phase in the line caues flashback into the atmospheric tank. The proposal was then to consider putting a flame arrestor on the line to prevent flashback into the tank. Yes, there is also a normal ball valve at the connection to the tank.

 

[EmmanuelTop]

One of HAZID team members obviously had a vivid imagination and the HAZID Facilitator forgot to wake him up.

Dejan IVANOVIC
Process Engineer, MSChE

 

[georgeverghese]

We only design for firecase events lasting up to 30minutes. Beyond that, you've got a fire thats gone out of control. Within 30minutes, it is not likely that a nominally empty B31.3 pipe will suffer thermal fatigue, resulting in loss od containment - you can check this by applying API521 firecase heat rates to a section of empty pipe that is within a 300m2 fire circle ( ie a 20m length of straight pipe) to see how the yield stress profile compares with required stress.
Your description of this line seems to indicate this is a flash forward event and not a flashback event. The first post seems to indicate this is a bidirectional loading / off loading line - I am not a flame arrestor expert - can a FA work for a high speed flame front that appears on either face of the FA ?