PSV sizing for fire case 2

koshan · Apr 29, 2008

Hi, I want to size a PSV based on fire case scenario on a vessel. The vessel is used as compressor knocks out drum. Sizing information is as following:
Operating Temperature: 47 .C (Tn)
Operating Pressure: 9.4 barg (Pn)
Design Temperature: 165 .C
Design Pressure: 69.4 barg
Vessel Size: ID=2.0 m, T/T=4.0 m
The problem is that the relief temperature base on
API formula (T1= [P1/Pn] Tn) is very high.
Max. Gas temperature in the vessel cannot be more than 1100 .F (API recommendation). So since temperature corresponding to set pressure is about 2400 .C in this case never pressure in vessel goes high up to set Pressure. In my opinion, this vessel does not need any safety valve! Because I can not find any Scenarios which lead to opening of PSV. But on the other it is very odd a pressure vessel without a PSV. Can you help me in this regard?

SeanB · Apr 29, 2008

Generally what I do in a case like this is to simulate the knocked out liquids in HYSYS. Then at your reliving pressure, set your vapor fraction to zero and let HYSYS calculate your bubble point temperature. This is then your relieving temperature.

someguy79 · Apr 29, 2008

As SeanB said, you should consider the effects of your liquid fraction vaporizing.

If liquid vaporization doesn't bring you up to relieving pressure at a temperature below 1100F, it's a rare case indeed. If this is true, I would still consider using a relief device to deal with the issues of CYA and Murphy's Law.

JoeWong88 · Apr 29, 2008

One known fact...PSV does not protect vessel from fire attack. Read more HERE . Nevertheless,it may only "buy time" during fire attack...

SeanB & someguy79 have clearly pointed out that liquid accumulation in suction vessel and leads to liquid vaporisation case.

You may also have to looks at the compressor settle-out follow by external fire which may have different distributed relief load...

One last point, is the suction vessel design pressure is selected base on settle-out pressure with margin ? If not may consider to reduce the suction vessel design pressure...

koshan · Apr 30, 2008

Thank for your comments. SeanB the gas doesn't have any liquid.
JoeWong88, you are right, Design Pressure of compressor suction Drum normally is based on settle-out pressure with some margin, but in my case its design pressure is based on compressor discharge pressure with some margin. The vessel is already available and I want to size PSV only.

SeanB · Apr 30, 2008

Well if you don't have any liquid knock out then you don't have a fire case. With that being said this is usually typical. We usually have a suction knock out drum, but usually no liquid to knock out. So....what you have to do is fake it. I have seen people use an arbitrary mix of 50% i-butane and 50% n-butane and flash that at relieving pressure.

JoeWong88 · Apr 30, 2008

Is your suction fluid is normally saturated ? During turn with recycle mode, mixing feed fluid with recycle flow may see liquid knock out.

You may consider to reduce the Suction drum PSV set pressure to settle out pressure plus margin.

djack77494 · May 1, 2008

You may want to consider potential process upsets. Say you normally have a dry gas because your K.O Drum is downstream of a gas heater. There may be an absorber or contactor upstream of the heater. In the event of a major process upset, some of the liquid in the upstream vessel may carryover into the normally drum vessel. Thus, you could assume that liquid as what will vaporize during a fire. (The "knock out" in Knock Out Drum refers to liquid being knocked out of the vapor stream.)
Doug

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

PSV sizing for fire case 2

koshan

Chemical

Recommended for you

SeanB

Chemical

someguy79

Mechanical

JoeWong88

Chemical

koshan

Chemical

SeanB

Chemical

JoeWong88

Chemical

djack77494

Chemical

Similar threads

Part and Inventory Search

Sponsor