Alternative relief path for tube rupture - API 521 interpretation 1

[chemter]

I am evaluating a tube rupture scenario in which both sides contain liquid. There is no vapor or flashing involved with the tube rupture. The tube side is the high pressure side and the maximum operating pressure of ~700 psig exceeds the low-pressure shell side hydrotest pressure of 390 psig. I am trying to reduce the relieving requirements for a relief device. API 521, Section 5.19.5 in the 5th edition (or section 4.4.14.2.4 of the 6th edition) says “…if a detailed analysis is performed, a capacity credit may be taken for acceleration of the low-pressure side liquid.” In this case, is a “detailed analysis” referring to a dynamic analysis to simulate the pressure profile as discussed in API 521 Section 5.19.3 (5th edition) or can a hydraulic analysis be performed to show the low-pressure side can accept the increased liquid flow without unacceptable pressure increases? I am curious how others interpret this and if this approached has been used, what type of analysis was performed? Thank you.

 

[don1980]

The meaning of "detailed analysis" in 6th ed 4.4.14.2.4 is completely different from "dynamic analysis" which is discussed in 5th ed 5.19.3 (4.4.14.2.2 in the 6th ed.). Dynamic tube rupture analysis refers to the potentially catastrophic shell failure that can occur when the tube side contains a very high pressure gas (or liquified gas) while the low-pressure shell side is liquid-filled. In such cases, a tube failure can cause the shell to "explode" due to the sudden shock force applied by the escaping tubeside fluid. Due to its inertia, the shellside liquid can't move out of the way fast enough, resulting in catastrophic shell failure.

In the application which you describe, this dynamic (or inertial) type failure isn't a risk. Your case is a simple one in which the tubeside fluid will flow out of the exchanger along with the shellside liquid. As long as that flow path is open, a tube failure won't cause overpressure.

 

[chemter]

don1980,
Thank you for your response.