10 replies to this topic

[J_Leo]

Hello,

 

We have a compressor last stage KO drum operate at high pressure (operating P=550 psig) and the condensate is normally routed to another destination. There is a 3" drain attached at the bottom of the vessel with double block vales and it is drained to another low pressure vessel with design P @50 PSIG.

 

For the sizing of the PSV on the low pressure vessel, the vapor blow through case is identified. One engineer estimated the relief rate to be the max flow rate through a 3" orifice, using upstream oper. pressure and downstream relief pressure. The resulting relief rate is very large. Is this a right way to do it? I think we should just estimated the relief rate based on the velocity of sound in the 3" pipe using the down stream relief pressure (55 psig). Beyond this, it is chocked flow, no more mass flow increase.

 

Please shed some light on this.

 

Thank you,

Leo

 

 

[latexman]

Both you and the other engineer are ignoring fL/D + Sum of K's.  This may be valid if the line is short.  Use the compressible, adiabatic flow equation.  If you have compressible flow software, don't forget about the effects of any reduced port valves, if there are any.

Edited by latexman, 20 April 2016 - 01:26 PM.

[fallah]

 

We have a compressor last stage KO drum operate at high pressure (operating P=550 psig) and the condensate is normally routed to another destination. There is a 3" drain attached at the bottom of the vessel with double block vales and it is drained to another low pressure vessel with design P @50 PSIG.

 

For the sizing of the PSV on the low pressure vessel, the vapor blow through case is identified. One engineer estimated the relief rate to be the max flow rate through a 3" orifice, using upstream oper. pressure and downstream relief pressure. The resulting relief rate is very large. Is this a right way to do it? I think we should just estimated the relief rate based on the velocity of sound in the 3" pipe using the down stream relief pressure (55 psig). Beyond this, it is chocked flow, no more mass flow increase.

 

 

Leo,

 

What's the real problem? Do you think your estimated relief rate might be wrong? Or you have problem to handle such large relief load?

 

Please submit the detail of your estimation to be able to submit better assistance...
 

[J_Leo]

Hello Fallah,

 

 The calculated flow rate using orifice (estimated using orifice sizing equation) is much larger than using 3" pipe at sonic velocity (using compressible flow - Conison equation). For the estimation of the max blow though rate for relief, to use 3" orifice or 3" pipe at sonic velocity? In my opinion, we should use the latter. But I am not sure.

 

I just hope somebody who has the knowledge can share some information.

 

Thank you,

Leo

[latexman]

If the pipe and fitting losses are high enough, it may not attain Mach 1.  It just depends on the layout and details of the installation.

[fallah]

 

 The calculated flow rate using orifice (estimated using orifice sizing equation) is much larger than using 3" pipe at sonic velocity (using compressible flow - Conison equation). For the estimation of the max blow though rate for relief, to use 3" orifice or 3" pipe at sonic velocity? In my opinion, we should use the latter. But I am not sure.

 

Leo,

 

In my opinion, you should use the block valve with lower Cv value as flow limiting component...
 

[J_Leo]

Fallah,

 

I think that is a good approach using the gate valve Cv. Can you be a little bit specific about the low Cv value? I know how to use vendor table to get Cv for control valves, but never selected a Cv for a block valve. When you talk about low, do you mean the valve is partially opened?

 

Thank you.

 

Leo

[cea]

Dear J_Leo,

 

As you anticipate, flow cannot be more than value, once Mach No. reaches to 1. At the same time, as indicated by latexman, your system pressure drop cannot be more than available pressure difference.

 

In your case available pressure difference in substantially high & may achieve Mach 1.

 

Now, it is further assumed that there is level control valve on condensate draining line. Please confirm my understanding.

If so, maximum flow through that control valve (at relieving condition) + flow through bypass (if applicable), will drive total relieving rate, even if it doesnot reach Mach 1.

 

If not so, maximum flow through regulating valve will govern relieving flow value. You may find it again may not reaching Mach 1.

 

I feel, considering maximum flow so as to achieve Mach 1, is very conservative approach.

 

Experts can confirm.

[fallah]

 

I think that is a good approach using the gate valve Cv. Can you be a little bit specific about the low Cv value? I know how to use vendor table to get Cv for control valves, but never selected a Cv for a block valve. When you talk about low, do you mean the valve is partially opened?

 

 

Leo,

 

By lower Cv value i mean lower Cv between Cv's of two double block valves. Indeed, a block valve has a rated Cv that vendor can provide the relevant value...

[J_Leo]

Thank you everybody for your inputs.

I have calculated the relief rates by using 3" gate valve Cv, 3" orifice and 3" downstream pipe sonic flow for the conditions mentioned in the original post.

The relief flow rate using gate valve Cv (672 from vendor table) is approximately 3 times of using orifice and 10 times of using downstream (55psig) sonic flow (2500 ft/s).

I just want to give you an update. Still not sure which is right.

Regards,
Leo

Edited by J_Leo, 23 April 2016 - 08:57 AM.

[latexman]

In your case, using 55 psig in the sonic velocity calculation in the 3" pipe is not correct.  This is similar to a recent post and explained very well by fallah here http://www.cheresour...cal-conditions/ .