17 replies to this topic

[Adi.putra]

Hello All,

I'm just going to start explaining the problem.

Consider two two-phase separators in series, one on the upstream (V-01) with normal operating pressure of 18 barg and the downstream vessel (V-02) operating at 3.5 barg. On a normal basis, liquid will be flowing through the line connecting these two vessels, controlled through level control valve.

Consider the following event; the two vessels were working on a normal condition, liquid level on each vessel was at normal liquid level. Then suddenly level reading was going wild and the level control valve was ended up being opened 100%.

The liquid level from V-01 went down very quickly, and within 2 minutes the liquid level was all loss. The liquid was then in V-02 which was able to contain all liquid from V-01. However the inlet nozzle in V-02 was immersed in liquid after containing all liquid from V-01. Assuming only vapor flow coming in from V-01 and because of the level control valve the vapor breakthrough is choked. But the vapor breakthrough is going to mix with the liquid in V-02 since the nozzle is immersed in liquid. In some articles, this is called as liquid swell relief or two phase relief.

During the relief these are considered:

• V-01 level control valve is 100% open (control valve in consideration)
• V-02 level control valve is at turndown position (as per API 521)
• V-02 vapor outlet has an “LO” manual valve.
• The fluid is non-reactive and non-foamy.

Some calculated figures:

• Vapor breakthrough (choked flow) thru V-01 LCV is 11,000 kg/hr or about 2,300 m3/hr at relieving condition.
• V-02 liquid level right before the vapor breakthrough is 1 m above the inlet nozzle. (If it is necessary, say 1 m liquid level in V-02 is 1m3).
• V-02 dedicated nozzle for relief line is on the top of the vessel, which is 2 m higher from the inlet nozzle.

My concern is whether this scenario needs to be considered? If yes, then how do you calculate the relieving two phase flow? (Yes, in this case some will go to the vapor outlet line of V-02).

I hope I have the problem explained clearly.

Thanks,
Adi

[fallah]

Adi,

Please upload the relevant sketch.

Fallah

[Adi.putra]

Fallah,

The sketch of the system.

 Sketch.pdf   12.03KB   137 downloads

Adi

[fallah]

Adi,

Vapor breakthrough due to full open position of LCV-01 can be a credible scenario in your system, but the information to do the calculation of relevant relief load isn't adequate. Some of the main information which are needed for system evaluation to do so are as follows:

-Design conditions/volume of V-01/V02
-How much is the liquid flowrate the LCV-02 can handle in turndown conditions
-Liquid portion of the main incoming fluid
-Stability of flowrate/pressure of the main incoming fluid at the beginning of the vapor breakthrough

Actually due to full opening of the LCV-01 the liquid level rises in V-02, and the inlet nozzle in V-02 Is immersed in liquid after a while. In this situation both liquid and vapor may enter the vapor discharge line.

The main point is to determine when the V-02 reaches to set pressure and then to relieving conditions. At this time, incoming two phase flow from V-01 in combination with liquid outlet from LCV-02 will establish a liquid level in V-02, may be higher or lower than 1 m above inlet nozzle, and based on corresponding outgoing flowrates from LCV-2 and vapor discharge line the two phase relief load through the PSV can be evaluated.

Fallah

[Adi.putra]

Fallah,

Thanks for your reply.

Some of the necessary information:
Design condition for V-01 is 24 barg, volume = 4 m3 and V-02 is 6.5 barg, volume = 3.5 m3.
LCV-02 is at 15,000 kg/h at turndown valve position and relief condition.
The incoming fluid to V-01 can be assumed as 100% vapor during relief.
Pressure of supply to V-01 will not be affected during vapor breakthrough as it is capable of supplying much higher than the vapor breakthrough rate.

Let's say that after analysing, we still find the V-02 inlet nozzle immersed in liquid and vapor is coming through from LCV-01, how do we determine the two-phase flow rate (regardless whether or not it is going to relief) because of this event? Frankly I do not know how to as the information on the liquid side is available as liquid level above the inlet nozzle, and the information on the vapor side is choked flow rate... or do you find out whether the vapor disengage from the liquid based on the velocity and available vapor space? or something else...

Many thanks

Adi

[fallah]

Adi,

Actually this assumption: "The incoming fluid to V-01 can be assumed as 100% vapor during relief." cannot be a right one because the main incoming fluid is received in order to be processed for two phase flow separation.

On the other hand, in your first post you mentioned: "Vapor breakthrough (choked flow) thru V-01 LCV is 11,000 kg/hr...at relieving condition" while in your recent post you mentioned: "LCV-02 is at 15,000 kg/h at turndown valve position and relief condition". It cannot be acceptable from material balance point of view.

Fallah

Edited by fallah, 13 October 2012 - 03:06 AM.

[Adi.putra]

Fallah,

Sorry for trying to oversimplify things. Actually the incoming fluid contains very little condensation like < 1%-w of total flow, at max.

The vapor breakthrough flow rate (choked) is calculated based on the LCV-01 Cv and the normal upstream pressure and downstream @relieving cond.

The flowrate through LCV-02 is calculated based on the Cv at turndown valve position and relief condition, which gives 15,000 kg/h (liquid rate).
Well in this case, if the pressure in V-02 never reach the relief pressure, liquid rate through LCV-02 will not reach the 15,000 kg/h mark.
Thanks for pointing this out.

The vapor volume rate into V-02 is definitely higher than the liquid volume rate through LCV-02, will this still be a potential cause of overpressure in V-02?

My question still remains, if the inlet nozzle in V-02 still immersed when the vapor breakthrough, what will happen? I hope this is not a silly question...

Many thanks,
Adi

[fallah]

My question still remains, if the inlet nozzle in V-02 still immersed when the vapor breakthrough, what will happen? I hope this is not a silly question...

Adi,

It is not only isn't a silly question but also is an interesting one...

When the liquid level just approaches the V-2 inlet nozzle during vapor breakthrough, a mixture of both vapor and liquid may enter the vapor discharge line of this vessel even before the V-2 set pressure would be reached. The capability of liquid level such that can immerse the inlet nozzle depends on the fact that if there is sufficient space for liquid and vapor disengagement and if so how much.

Fallah

[Adi.putra]

Fallah,

Before the vapor breakthrough the liquid level is at 1m above the inlet nozzle and the inlet nozzle is 2 m below the top of the vessel, as provided in the previous sketch.

With the given properties, I tried to calculate the settling velocity of a vertical separator (available in Coulson&Richardson) and with the given geometry of the vessel, I found that the volumetric flow to have a full disengagement without a demister pad is only 520 m3/h of vapor load, this is not sufficient as the vapor breakthrough is 11,000 kg/hr (2,600 m3/h). I was quite doubtful when using this approach, as the separator described in Coulson&Richardson does not have the liquid level above the inlet nozzle.

Any other approach, in your experience, you can suggest here?

Oh by the way, can I still continue since this may not be a relief-related topic?

Thanks,

Adi

[fallah]

Adi,

IMO with lack of accurate and adequate information, you can use simplified methods (assumption) either: 1) Volumetric flowrate through vapor discharge line of V-02 remains unchanged; or 2) Mass flowrate through vapor discharge of V-02 remain unchanged. Relief load of the PSV in each two method would be different. The greater value would be final relief load of the PSV.

Fallah

[Adi.putra]

Fallah,

Sorry for being slow, I still do not know how the flow will be (two phase or vapor only) because of the liquid-immersed inlet nozzle. Is considering this scenario not a consideration in common practice? There is a possibility of this happening, right?

Thanks,

Adi

[fallah]

Adi,

As i told in post No.8: " When the liquid level just approaches the V-2 inlet nozzle during vapor breakthrough, a mixture of both vapor and liquid may enter the vapor discharge line of this vessel even before the V-2 set pressure would be reached."

It means when liquid level either approches the inlet nozzle or would immerse it, mostly there would be two phase flow in vapor discharge line especially due to very high breakthrough vapor flowrate.

Fallah

[Adi.putra]

Fallah,

How about the flow rate of this two-phase flow?
Any method or approach to calculating/estimating the rate?

Thanks,

Adi

[fallah]

Adi,

Flowrate in which situation? after beginning of vapor breakthrough or in relieving condition?

Fallah

Edited by fallah, 15 October 2012 - 06:27 AM.

[Adi.putra]

Fallah,

I would like to know the flow rate right after the vapor breakthrough (Yeah, the two phase flow will deplete as the liquid level drops).

Thanks,
Adi

[fallah]

Adi,

Couldn't the explanation of mine in post No.10 help you out in this regard? Of course, i am well aware it is hard to determine the liquid portion of two phase flow passing vapor discharge line of V-02 just after vapor breakthrough.

Fallah

Edited by fallah, 16 October 2012 - 01:44 AM.

[Adi.putra]

Fallah,

Of course, i am well aware it is hard to determine the liquid portion of two phase flow passing vapor discharge line of V-02 just after vapor breakthrough.

Yeah my thought too, as I searched the web for this and have not found anything.

Thanks for the help though....

Adi

[mhs]

Adi,
as per my understanding there will be gas phase (with dispersed droplets of liquid) will emerge from liquid phase.

Edited by mhs, 20 October 2012 - 07:54 PM.