8 replies to this topic

[Diep Thanh]

Dear All,

I need your help, please. 

I am trying to calculate relieving load in case of reflux failure. My column is de-butanizer with total condensing ( 475 kPag, DP=45kPa) and it is already simulated at normal conditions by Pro II. In my opinion, the simplest approach the relieving load is that trying to simulate the column " near" the relieving conditions ( Pset=650kPa) with 2 spec:

- Bottoms temperature: according to the chart in page 149, API 521. ( P relief at bottom = 650*1.1 + 45= 760 kPag)

- Minimum reflux rate to converge my column.

 Then the relieving load, properties = vapor out of top tray.

 

What do you think about my approach? Can it be acceptable?

 Thank you for your time!!!

Edited by Diep Thanh, 11 December 2014 - 11:02 PM.

[Bobby Strain]

Doesn't sound like a good method. You need to search for an acceptable method. Good engineers don't "wing it" like this.

 

Bobby

[ColinR33]

With a reflux failure you dry out the trays above the feed, so you can simulate it as a reboiled absorber (only using trays below the feed) assuming constant feed and maintaining reboiler temperature at the relieving pressure (assuming the reboiler can achieve this - there may be a pinch point that will limit the achievable temperature/duty) to get your vapour (relief)load.

Cheers,

[latexman]

In my opinion, the simplest approach is the relieving load = condenser duty of your normal, base case run.  No need to simulate anything else.  Then, size the relief device in a conservative way to make up for any compostional changes during the boil-off.  For example, use the pure component, that is not a minor component concentration-wise, that gives the largest relief.  You can check how conservative this is by comparing to mixture properties from your normal, base case run.

[shekhar dhuri]

Diep Thanh,

I am attaching one article which may help you while sizing Column PSVs.

Regards,
Shekhar

Attached Files

•  Avoid Mistake Distillation Safety Valve Bradford ChE Mag 1984.pdf   984.8KB   200 downloads

Edited by shekhar dhuri, 13 December 2014 - 11:27 PM.

[Bobby Strain]

I don't recognize the authors as authorities. I would continue to search for accepted practices. This is not one of them.

 

Bobby

[AlertO]

Hi Diep

 

The column relief detemination normally come from 2 methods: 1. Conventional method (hand calculation) 2. Simulation

 

 

For conventional method, you can get the relief load from doing mass and energy balance. I've attached a literature for you below.

 

For simulation, dynamic simulation is more accurate and able to get the optimum relief load. However, you need to well understand about the boundary specification otherwise, your simulation will be useless. 

 

 

 

Attached Files

•  Column Relief Load Cal Part II.pdf   1.03MB   170 downloads

[Diep Thanh]

Thank you all,

In ColinR33's opinion, we also have to determine the flowrate/ compositon ...etc for the liquid come into top of the column. It may be back to my approach ( minimum reflux rate) ?

Latexman, I belive you said about " normal reboiler duty" right? OK, It may be a conservative way to calculate the relieving load, however, if i try to calculate ( base on the increasing temp. on the bottom of the column) , duty of reboiler decrease remarkably ( it seems to be 5-10 times less than duty in normal).

I was trying to calculate by unbalanced heat and material method ( load= unbalnced heat/ latent heat). But I don't know how the distillate and the bottom product  flowrate change? Can I assume that they are the same as normal?

[latexman]

I did not say "normal reboiler duty".