7 replies to this topic

[kasri]

Dear All,
While doing PSV orifice sizing, we find the relieving rate (based on latent heat,wetted area or nonwetted area) in Kg/hr.Based on the above calculated flow rate and setpressure we size the orifice.Please suggest me how long the PSV release the inventory, it starts to release when its reaches the relieving pressure( set pressure + % margin depends on conventional or fire case) ,then PSV will close when it reaches the pressure below relieving pressure.
Please correct my above understanding is wrong.
How long the PSV will relieve that much kg/hr of inventory.For BDV we size orifice based on the relieving of inventory within 15 min or half of the system pressure and termed it as kg/hr for calculation.
(1)In PSV whts the time for relieving ie how long the valve opens.
(2) We sized orifice and selected the nearest large orifice size.Then during relieving condition it release the rated flow based on the selected orifice , if it is the flare KOD have to be sized for rated flow or required flow.

Please clarify my above doubts.
Still I am not get clear on the above concepts.
Regards,
Kasri

[paulhorth]

Kasri,

it starts to release when its reaches the relieving pressure( set pressure + % margin depends on conventional or fire case) ,then PSV will close when it reaches the pressure below relieving pressure.

You are correct.
The PSV is a constant pressure device. It will stay open so long as the source of overpressure continues to exist. If there is no overpressure, the valve will close.
So in the case of a fire, the valve will tend to stay open until the fire is stopped or the vessel ruptures.Because the sustained fire relief flow is less than the rated capacity of the valve, the valve will either open partially, so that its capacity balances the fire relief load, or will repeatedly close and open again which is called "chattering".
In the case of a relief due to blocked outlet, it will stay open (with possible "chattering") until the source of inflow to the vessel is shut off.
There is no single answer to your question (1).
The PSV is quite different from a blowdown valve. A blowdown valve will reduce pressure because it stays open. The PSV will not reduce pressure in the vessel.

In reply to your question (2), the usual practice is to size the flare header and KO drum for the CALCULATED relief flow, while sizing the tailpipe for the RATED relief flow. This is a compromise based on the idea that the Rated flow will not be sustained, but this transient rated flow will determine the transient back pressure on the PSV, which has to be checked to confirm that the valve will open fully.

Paul

[fallah]

Because the sustained fire relief flow is less than the rated capacity of the valve, the valve will either open partially, so that its capacity balances the fire relief load, or will repeatedly close and open again which is called "chattering".

Paul,

You are explaining generally, or are referring to the situation that kasri described. It generally might not be right and referring to kasri description it seems fire case is governing (sizing) case and also in this situation it couldn't be right because fire relief load (if PSV could find a time for relief before vessel rupture!) would be almost equal to rated capacity of the valve.

Fallah

Edited by fallah, 28 October 2011 - 11:22 AM.

[kasri]

dear fallah and paul,
Thanks for your help.In my system fire case is governing one, if it is whether I have to design the flare header and KOD for rated flow or required flow.
Please guide me and please mention whether API govern this.
Thanks,
Kasri

[paulhorth]

Kasri,
I have already answered this question:

In reply to your question (2), the usual practice is to size the flare header and KO drum for the CALCULATED relief flow, while sizing the tailpipe for the RATED relief flow.

You should check API RP 521 for yourself.

Paul

[fallah]

if it is whether I have to design the flare header and KOD for rated flow or required flow.

Kasri,

In single relief source you should size the flare header and KO drum based on rated flow and in multiple devices installation based on cumulative required capacities.

Fallah

[paulhorth]

Kasri, Fallah,

My reply was based on API RP 521 section 7.2.1 Table 12, which reads:

Table 12 — Design basis for pressure-relief device laterals and disposal system headers
Device Lateral/tail pipe Main header (if applicable)

Pop-action, pilot-operated relief valve: valve rated capacity Required relieving rate
Modulating-action pilot-operated relief valve: Required relieving rate Required relieving rate
Spring-operated relief valve: valve rated capacity Required relieving rate

(sorry if this does not read clearly, I could not arrange the table columns correctly)
This makes the distinction between the Tailpipe sizing (which means meeting velocity and pressure drop limits) and the header sizing. If there is more than one relief flow in the header then the basis is the total of the Calculated simultaneous relief or blowdown rates, as Fallah says. The pressure loss calculation for the header (connected to several valves) done for each single relief flow in turn, should I believe be based on the calculated relief flow.

If there is only one relief source on the plant, which is unusual, then API RP 521 7.2.2. states
The initial design of the disposal system should, as a minimum, be based upon this required relieving rate.
Typically, the required relieving rate is used for the flare header, flare tip and knockout drum design with spring-loaded pressure-relief valves.

Paul.

Edited by paulhorth, 30 October 2011 - 09:57 AM.

[fallah]

If there is only one relief source on the plant, which is unusual, then API RP 521 7.2.2. states
The initial design of the disposal system should, as a minimum, be based upon this required relieving rate.
Typically, the required relieving rate is used for the flare header, flare tip and knockout drum design with spring-loaded pressure-relief valves.

Paul.

Paul,

For one relief source, as per API RP 521 7.2.2 the above statement is just applicable for the cases in which initial flowrates are less than 50% of their rated capacity at set pressure (IMO rare cases). Hence, in most systems with one relief source using rated flow as design basis for downstream equipment would be more probable than required flow.

Fallah