20 replies to this topic

[liecdre97]

I just started with a rupture disc company and we were approached to come up with a pressure relief solution for a High Pressure Separator.  This is a relatively new application for my company, so I was hoping to get input on what solutions others have used for this application and their experiences with those solutions (good or bad).

[fallah]

Hi,

 

Please exactly specify what you are hoping to get about pressure relief in a high pressure separator...

[liecdre97]

Sorry.  I thought I had included the general requirements (280 bar, up to 250 degrees Celsius and 6" diameter outlet).  We are looking into the application and what solutions we can provide, but it would be helpful to have some feedback as to what is currently used by others under similar applications (e.g. other options) and what experiences they have had, so we can best know what attributes our solution might need to have (or would be nice for it to have).  

 

I would assume a PSV is the standard for many applications, but this seems to be relatively large pressure, diameter  and temperature requirements, so I would like to know if the requirements might make a PSV less practical.  In other words, is there a diameter, pressure or temperature when designers tend to go away from PSV's?  Also, if PSV's are used under similar requirements, is leakage and/or chattering an issue.  

 

In addition, if anyone has used a rupture disc in the past for a similar application (or any application), I would like to know what issues, if any, were had with it.  We want to have the best solution for our customer, so any information we can gather is helpful.

 

I hope this is clearer.  Sorry again for the vagueness of my original post.  Thank you for asking for a clarification.

[fallah]

Hi,

 

Appears there is no reason using a PSV for your application; nevertheless at first find the relieving rate at governing scenario and reflect it along with the other info in a data sheet, then send it to reputable vendors to get the proposal(s). If there would be any limitation in vendor manufacturing method in pressure, temperature, material,...standpoints due to which can't provide a PSV they will inform you along with the alternative(s) relieving device(s) for the application...

[liecdre97]

Thank you for your response.  That sounds like a good idea, but I would also like to possibly hear from end users or designers of the systems to get a better feel of what is desired and what might cause issues.

 

Thank you again for your help.

[shvet]

Sorry.  I thought I had included the general requirements (280 bar, up to 250 degrees Celsius and 6" diameter outlet).  We are looking into the application and what solutions we can provide

 

Special applications: high corrosive fluids (e.g. acids, some organics), fluids with solids particles (e.g. catalizator, adsorbent), fluids with polimerization agent (many chemicals, aromatics and organics), extra toxic chemiclas (e.g. warfare agent - HCN), special cases of protection from chemical reactions in reactors (i haven't seen yet).

 

General rule of thumb for conventional spring and balanced bellows PSV - the higher pressure is the better. The most expensive, large and heaviest PSV are for low pressure application with discharge to close collecting system.

 

Rupture disks are very inaccurate in Pset (in your case app. -3/+6% of Pset). Have your disks good accuracy?

 

 

In other words, is there a diameter, pressure or temperature when designers tend to go away from PSV's?  

 

Only case of "dry vessels" when protected vessel is filled with gas and no liquid.

 

Also, if PSV's are used under similar requirements, is leakage and/or chattering an issue.  

 

as described higher

[liecdre97]

Shvet, thank you very much for your response.  That is very helpful information.

[shan]

You may invent a relief device that one-size-fitting-all scenarios to save tons of engineering hours for PSV sizing.  Hopefully, I am retired before your invention so that no worry of layoff.

[shvet]

 In addition, if anyone has used a rupture disc in the past for a similar application (or any application), I would like to know what issues, if any, were had with it.  We want to have the best solution for our customer, so any information we can gather is helpful.

 

I've worked 9 years in a chemacil plant. We had rupture disks for overpressure protection of reactors. They were rather small in size - only 2m3 and RD were 4". Reactors work on 110°C and after 116°C they explode. ∆t=6°C is rather small value and from time to time they realy explode. RD worked and stok discharged but in spite of it ~50 bolts cut and the lid of reactor penetrated the ceiling and laid on the next floor.

[fallah]

 

RD were 4". Reactors work on 110°C and after 116°C they explode. ∆t=6°C is rather small value and from time to time they realy explode.

 

 

shvet,

 

What was the corresponding pressure increase of temperature increment by 6 C? It might due to such relatively small temperature increment, the pressure, say following to a runaway reaction, would reach near the RD set pressure... 

[curious_cat]

 

 

In other words, is there a diameter, pressure or temperature when designers tend to go away from PSV's?  

 

Only case of "dry vessels" when protected vessel is filled with gas and no liquid.

 

 

I am curious why. Why are PSVs not very good for dry vessels? 

[curious_cat]

Reactors work on 110°C and after 116°C they explode. ∆t=6°C 

That sounds like a really small margin for error. How did you control the reaction so tightly under normal operating scenarios? What was the process control strategy?

 

When T reaches 116 what exactly happens? Gas evolution? An accelerated runaway? Why couldn't it be run at (say) 100 C under normal operations? Would reaction rate be too low? 

Edited by curious_cat, 10 July 2015 - 01:13 AM.

[shvet]

 

 

RD were 4". Reactors work on 110°C and after 116°C they explode. ∆t=6°C is rather small value and from time to time they realy explode.

 

 

shvet,

 

What was the corresponding pressure increase of temperature increment by 6 C? It might due to such relatively small temperature increment, the pressure, say following to a runaway reaction, would reach near the RD set pressure... 

 

 

Runaway reaction. In reactor pressure is slightly higher atmospheric, ~0.1 barg.

 

 

 

 

In other words, is there a diameter, pressure or temperature when designers tend to go away from PSV's?  

 

Only case of "dry vessels" when protected vessel is filled with gas and no liquid.

 

 

I am curious why. Why are PSVs not very good for dry vessels? 

 

 

Because of temperature rise. Just imagine: normally "dry" vessel with hydrogen inside, 40 °C, Pwork.=5 barg, Pdesign=12 barg, Pset=12 barg, allowable overpressure = 10%, case - firecase.

 

Releaving conditions of gas are T = (40+273) * 13 *1.1 / 6 = 473 °C. Approximately after 300 °C gasket rupture accurs and PSV never relieves, but hydrogen feeds fire.

 

 

Reactors work on 110°C and after 116°C they explode. ∆t=6°C 

That sounds like a really small margin for error. How did you control the reaction so tightly under normal operating scenarios? What was the process control strategy?

 

When T reaches 116 what exactly happens? Gas evolution? An accelerated runaway? Why couldn't it be run at (say) 100 C under normal operations? Would reaction rate be too low? 

 

 

Reaction of acrylonitrile hydratase to acrylamide. Real temperature in reactor was ~108°C, if temperature <100°C then acrylamide has very high viscosity and plugs pipes, after 116°C uncontrollable reactions (unresearched) start. When temperature 116 °C reaches operators had ~30 sec and 50/50 luck to stop reactor, if not - temperature rises >120 °C, stock starts "to swell", some gas appears and reactor explodes.

Temp in reactors is controled by cooling water flowrate. If you interested I can share P&IDs but they are bad quality.

Edited by shvet, 10 July 2015 - 07:07 AM.

[curious_cat]

 

 

Temp in reactors controls cooling water flowrate. If you interested I can share P&IDs but they are bad quality.

 

 

Very interesting! Many thanks for elaborating. 

 

Yes, I'd love to see the P&IDs.  I'll message you my email address if that makes sending them easier.

[breizh]

Shvet,

 

" When temperature 116 °C reaches operators had ~30 sec and 50/50 luck to stop reactor, if not - temperature rises >120 °C, stock starts "to swell", some gas appears and reactor explodes." 

 

Something is wrong , this operation is dangerous and gambling with the life of people is not acceptable .

 

External cooling or other cooling material should help to control the reaction , attached an example .

 

 

My view.

 

Breizh 

Edited by breizh, 12 July 2015 - 04:29 AM.

[shvet]

External cooling or other cooling material should help to control the reaction , attached an example . 

Thanks for link, but it's extremely special and expensive. Unreasonable capital costs. New bolts and window glasses are much cheaper.

 

 

Something is wrong , this operation is dangerous and gambling with the life of people is not acceptable .

 

It works only in countries called "gold billion". Life of other 8 billions is much cheaper. Yes there's some risk for personel to get injuries from reactor explosion, but everybody understand that if alarm is yelling (there's general alarm system in industrial building of unit) you should move away reactor instead of approaching. And there wasn't any hero to try stop reactor manually if emergency system didn't work.

 

Routine risks and everyday harmful hazards much more dangerous than reactor explosion.

[Zauberberg]

I seriously doubt that each and every employee in that company is aware of potential hazards, including loss of life, immanent to the place of work. I doubt even more that he or she accepted this as a normal way of performing work duties.

 

What you are describing - and please apologize if I got things in a wrong way - is just a plain ignorance or corruption of people who own and/or operate the facilities. There is absolutely no excuse for that.

[shvet]

Zauberberg:

 

Sorry for my English, I'm not so good in "not engineer" english.

 

But guys, hey... mmm...  Get this XXXX out of your heads.  Money rules the world, and the cheaper technology is the more effective it is.  Personel are just an item of expenses and everybody's life has it's price. In any country in the world.  Even in your country insurance companies define the prices of citizens and gasterbeiters.  Just in your country price of (life of) citizen is higher than price (life of) gaster.  In my country this prices are =.

 

If you have any "codes" or "principles" lets move to China or Bangladesh, work 1-2 year...  Sometimes there were Wounderland too and life of citizens was priceless.  Do you realy think it'll never happen in your country?  Or it has never happened in you country.

 

Question is only "when", not "where".

Edited by Art Montemayor, 13 July 2015 - 02:50 PM.
Filthy, garbage vocabulary

[Zauberberg]

I'm just going to repeat what I discussed with my good friend on this topic, in a personal conversation.

 

You are describing this situation in a manner that permanent proximity to explosion is like an evening stroll in the park... One really has to wonder about what causes people to have such attitude.

 

There are cases in 3rd world countries where safety is really at unacceptably low levels, but the point is that this happens not because somebody wants to have the workers killed (and easily replaced by another cheap workforce), but because there is ignorance, negligence, and corruption, in the whole system - from plant owners to governmental institutions, who all help this system survive and exist in the way it exists.

 

If you don't care about safety in the place where you work, how can you expect somebody else will do that for you?

[Art Montemayor]

shvet:

 

I not only reinforce everything Zauberberg has stated, but I also wonder what garbage can or pig sty you learned your English from.  I won't tolerate nor allow filthy language on the forums.  If you don't have a proper, decent manner of expressing your thoughts, then perhaps you don't belong here.  I deleted your word describing your state of mind; next time, I'll just delete the post.

 

Our Forums are composed of serious, professional engineers and dedicated students.  Please show some respect for them as well as for their goals and work effort.

[curious_cat]

@Zauerberg:

 

That's a very interesting point, in general and has often caused me much consternation. (not in the specific case of the risk taken in shvet's example )

 

e.g. I think we all agree that there is nothing that can be labelled absolute safety? There is a risk-vs-cost tradeoff and we choose to operate at certain points on that curve. The question is whether every country in the world today ought to be operating at the same point on this curve. 

 

e.g. Even in a country like USA the amount of risk we take today is far less than say the way someone might have taken an Engineering Safety Tradeoff decision 50 or 100 years ago. So if we do accept that the risk appetite of a society changes with time then must we also accept that different nations extant today might have a different level of acceptable risk? 

 

I'm no way condoning an attitude treating "proximity to explosion is like an evening stroll " but does this not raise a more nuanced question about  when "gambling with the life of people is not acceptable"? 

 

The actuarial cost of a life is different in various nations. Is it right or wrong for this to impact safety decisions?