3 replies to this topic

[Platonicus]

Hello!

 

There is a plate heat exchanger.

 

Hot side: inlet - liquid, outlet - liquid.

Cold side: inlet - liquid, outlet - gas-liquid mixture (with a small vapor fraction by mass).

 

A PSV is installed at the outlet from the cold side.

 

We assume the following triggering scenario:

If the cold flow is blocked, abnormal heat flow may occur.

 

How to correctly calculate the flow rate when discharging through PSV?

 

How is the actual geometry of the heat exchanger (heat transfer coefficient, heat exchange surface) taken into account in the calculation?

 

As I understand it, if the cold side is blocked, then the heat transfer coefficient will be lower than in normal operation of the heat exchanger.

 

The task is further complicated by the fact that the cold side is a gas-liquid mixture.

 

 

On the Internet I was able to find only one more or less clear article on this topic:

https://kevindorma.c...-design-margin/

 

But it is not clear how the flow graph was constructed depending on temperature and, in general, how the calculations were made.

 

Below is a picture for understanding:

Attached Files

•  PSV_heat_exchanger.PNG   27.31KB   0 downloads

Edited by Platonicus, 27 May 2024 - 01:02 PM.

[Platonicus]

Continuation:

API 521 clause 4.4.12.3 has a formula for calculating liquid discharge:

q=αv x φ / (1000d x c).

 

However, API 521 itself states:

"If the blocked-in liquid has a vapor

pressure higher than the relief design pressure, then the PRD should be capable of handling the vapor-generation

rate. If discovery and correction before liquid boiling is expected, then it is not necessary to account for vaporization

in sizing the PRD."

 

It is also doubtful that the dimensions of the heat exchanger are not involved in the calculation, and also that the maximum heat exchanger duty during operation is taken to calculate the total heat transfer rate φ.

[breizh]

Hi,

Google : Thermal expansion design margin

 

You will find a video from Kevin Dorma on the topic you mentioned.

I tend to agree that API 521 is conservative. I agree also that the heat transfer coefficient cannot be the one used to size the equipment because no convection is taking place meaning the quantity vaporized is much less.

my 2 cents

Breizh 

[Platonicus]

Hi,

Google : Thermal expansion design margin

 

You will find a video from Kevin Dorma on the topic you mentioned.

I tend to agree that API 521 is conservative. I agree also that the heat transfer coefficient cannot be the one used to size the equipment because no convection is taking place meaning the quantity vaporized is much less.

my 2 cents

Breizh 

Thank you, Breizh!

Edited by Platonicus, 28 May 2024 - 07:43 AM.