7 replies to this topic

[kkimic]

Hello everybody,

I have got a question that may have been asked already although I could not find a clear answer. Hopefully someone could point me in the right direction.

I have sized few PSVs already for fire case located at the side of vessels, so I normally work out the API orifice required in two parts:

1.- First I calculate the steam rate as per API 520.
2.- Since the PSV is located at one side of the vessel, the PSV needs to relieve the liquid displaced by the steam in a fire case scenario. This is done by calculating the density of the steam at relieving temperature.
3.- Calculate the required orifice in acc. with API for liquid relief (PSV with capacity certification).

Problem is that I have been taught to do it this way but would like someone to agree with me this is the correct way to do it.

Thanks beforehand

[pleckner]

Please help me out here. Your PSV is located on the side of the vessel, in the liquid space?

Please describe your vessel and the contents. Better yet, can you produce a quick sketch (EXCEL) with some information and post it so we can see it. This is higly unusual.

If I'm picturing this right, you probably have a two-phase relief to deal with.

[kkimic]

Thanks for the quick reply.

Imagine for instance a vessel with a quick opening closure on top (for instance some of the cartridge filter vessels I deal with) so the PSV cannot be located on top of the vessel. Instead the PSV is located on the side of the vessel, as high as possible but still in the liquid space.

Normally I calculate the relief gas rate from a fire case scenario (steam relief). Since I know the relieving conditions (relieving temperature from water tables @ relieving pressure) I can calculate the density of the gas and hence the liquid displace by this gas in a fire case scenario. Then I simply size the PSV for liquid relief as if it were blocked discharge.

As I said I was taught to do it this way but last week I came accross a PSV vendor that could not understand why I was doing it this way. This obviously worried me. A PSV sized for steam relief in a fire case scenario is way smaller than a PSV for liquid relief, hence price and so on... but from a process point of view, how would you deal with this?

Forgot to say, vessel contains mainly water (produced water)

Thanks.

[pleckner]

Again, highly unusual.

Can you give me some dimensions of this vessel (curiosity)?

Recheck your calculations; a PSV sized for steam relief in a fire case scenario is typically way larger than a PSV for liquid relief. Perhaps you can provide your calculations or provide the parameters and reference the equations you used so I can review them. Nevertheless, I feel you will have two-phase flow so you aren't calculating the PSV size properly anyways.

Looking forward to reading your response.

[JoeWong]

I believe this is an interest post.

Phil,
Sorry to cut in. I am total agree with you kkimic may have missed some scenarios.

kkimic,
For a vessel complete filled with produced water (in subcooled) region and expose to external fire, i can imagine there are main events occur one and another.

i) Liquid* (sub-cooled) expansion until saturation point - Liquid relief
ii) Vapor generation forcing saturate liquid** passing PSV - Boiling liquid relief
iii) Vapor-liquid mixtures (two phase relief - clearly indicated by Phil and many events shown that this is the governing case) until vapor-liquid disengaged - Two phase relief
iv) Liquid** boiling & vapor expansion until no liquid present - Vapor relief
v) vapor*** expansion until it fail - Vapor relief

# Jet fire or Pool fire
* Ignore present of non-condensable /dissolved gas
** Ignore liquid expansion
*** Ignore localized heating / perfect heat transfer

JoeWong

[pleckner]

No problem Joe. That's what this Forum is all about, multiple ideas/thoughts/comments.

I suspect this is a small vessel and when dealing with small vessels where the surface area to liquid volume ratio is large, the thermal relief many talk about as the initiating event to me is basically meaningless as it will not govern anything. Sure it may make for interesting conversation on theory. The rapid boil-up will cause two-phase relief and if the vessel is small enough, you will never get to the point of good vapor-liquid disengagement where the event can change to all vapor venting.

[JoeWong]

Phil,
I agree with your thoughts. This really came from many years of experience and hundred even thousand time of similar calculations.

As a young engineer (includes myself), i believe we really need to go through same calculation for many times and get the "feel" and wisdom from the calculations.

Thanks Phil.

JoeWong

[kkimic]

Thanks for the replies, you are really helpful.

I do not have here the calcs but tomorrow I will copy them in my pendrive and send them over to you.

Any ideas welcome.

As you say this is in small vessels. I will send you some of the last calcs over the weekend.

Regarding your comment ''a PSV sized for steam relief in a fire case scenario is typically way larger than a PSV for liquid relief''. In my experience it is the other way around... but I have just started so may be doing something wrong.

Thanks.