5 replies to this topic

[lyla1720]

Hi I’m a first time user. Just a quick question that may be silly. I’m calculating release rates from an isolatable inventory that has a liquid (methanol) tank (50m3) connected to a downstream injection pump. I have to model various hole sizes. For the bigger 50 mm release, I get release rates that are much larger than the downstream injection pumps connected to the tank. Should maximum release be limited to the max discharge rate of the pump? This is for a QRA model. Thanks

[breizh]

Hi Lyla ,

To make sure we understand your query , please issue a simple sketch .

It sounds to me an application of Bernoulli  equation , I may be wrong .

 

Consider the resource attached , you should learn a lot.

 

Good luck.

Breizh

Edited by breizh, 18 June 2019 - 02:47 AM.

[PhilippM]

A sketch would indeed be helpful.

I worked on a similar problem for a QRA a while back, so I think I can make some assumptions on what you want, even without a sketch though:

 

I'll assume you have a vessel B1 (the (atmospheric?) methanol tank), a pump P1 that pumps out of B1, a non-return valve R1 on the discharge side of P1, a short process line L1 after R1, a control valve V1 after L1 and a long process line L2 after V1 which transports the methanol to a downstream equipment B2. So the order is: B1->P1->R1->L1->V1->L2->B2.

 

The scenario is a rupture (50mm) in either L1 or L2.

 

Let's assume B2 is just an atmospheric tank and there are no other (pressure generating) pieces of equipment connected to L2. In this case the liquid release rate would indeed be limited by the maximum discharge rate of P1, I would probably just use the right end of the pump curve and add maybe 10-20% (there will be no backflow that needs to be considered, as there is no source for this backflow)

 

If B2 however is e.g. a vessel with a (large) pressurized gas volume or there are other systems connected to L2, you will need to take a closer look at the hydraulics of this system, as a rupture on the discharge side will result in a backflow scenario with much bigger release rate.

 

(Analogy: Think about inflating a party balloon using your breath. Suppose it takes you 1 minute to inflate it and then you let go of it: It will be deflated much faster. The release rate out of a pressurized system is only limited by the hydraulics, not by the capacity of the equipment generating the pressure. Another analogy is the different speeds of a cyclist riding up on a hill and back down from it)

 

For a real QRA lots of other things (like the volumes of these systems if you want to know how long the release will last) will be needed to be taken into account of course. 

[Art Montemayor]

Lyla:

 

Please stop posting the same topic in different Forums within our Website.  Take the time to read the user Guidelines found at the front of the Forum to learn what we ask or require of our members in order to maintain some sensible and common sense order within our Forums.

 

I have deleted your other, duplicated post in our Industrial Professional Forum since I have assumed you are a student who doesn't take the time to read our posting instructions.  If you are not a student, tell me and I'll move this thread to the Forum where it belongs.

[lyla1720]

Thanks and apologies to Admin for posting in two forums. I am in industry, and not a student. 

Thanks for sending through the Yellow Book which I am familiar with as PHAST bases its algorithms primarily on models described in this book.i have attached the diagram for the rupture case I am talking about. sorry the diagram is very basic. As per one of the replies posted back, it seems like I will have to limit my discharge rate to the downstream pump?

Thanks.

 

     

 

Attached Files

•  Dwg.pdf   40.21KB   10 downloads

[PhilippM]

I thought you meant a rupture in one of the lines. A leak in the tank is something else entirely, the flowrate might be much higher than the pump-in rate.

Why would the pump-in rate even have any effect on the release rate at all?