Condensate Stabilizer Vacuum Condition

[maddocks]

We've been asked to look into the possibility of a condensate stabilizer dropping into a vacuum condition during a shutdown in a very cold Canadian winter. My gut reaction is that this won't happen, but I'm struggling with how to perform this calculation. The volume of the tower and reboiler is fixed, the mass at the time of the shutdown is fixed, and the system simply starts to cool down. The vapour condenses into condensate and the pressure begins to drop due to vapour collapse - much like a amine regenerator. Can anyone provide tips on how to calculate the final pressure during a cooldown? Is it as simple as looking at the vapour pressure of the tower in the equilibrium fluids at an ambient of -30°C?

 

[dcasto]

hint mass is conserved

 

[maddocks]

Yeah, I got that part. But, each stage has a different composition, different temperature, and a different dewpoint. So, once the tower cools off, a lot of the vapour collapses into condensate and the tower liquid level is raised slightly. But, the real question is, how much does the pressure drop on the way down?

 

[dcasto]

ok, look at material balance, unless you are removing more C5's than you are putting in and building up a batch, why would you expect to have a lopsided average composition?

If you want, just get the process simulator to give you the tray by tray composition and density.

 

[EmmanuelTop]

The answer to your question should be rather simple. If the whole system is shutdown (SDV's at the inlet and product outlets, reboiler shut down) while being in normal operating conditions prior to shutdown, the resulting pressure can be calculated by simulating feed stream at the final cooldown temperature.

In case there is a different procedure for shutdown, you might need to consider different overall composition. This again can be calculated by mixing all the product streams at given conditions and flash the resultant stream at the final cooldown temperature.

My guess is - as long as you have a few % of Methane in the column, you won't see any vacuum. Methane boils at -161C and it needs something more harsh than Canadian winter to condense.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[maddocks]

It doesn't seem to be that simple. I've tried just a straight forward JT valve with the outlet set at -30°C and the pressure drops into a vacuum. This is still with almost 5% methane in the stabilzier column feed mixture. It almost seems like I have to do a stage by stage flash where I manually cool each stage to -30°C and then see how much of the vapour has condensed and then calculate the loss of pressure as a result of condensing. Simple bubble point calculations don't to represent the overall column performance when things start to cool.

 

[maddocks]

Yeesh, fast fingers but slow brain. I was trying to say that simple bubble point calc's don't seem to represent column performance because each stage is at a diferent temperature.

 

[EmmanuelTop]

Can you upload the stabilizer feed composition? I would like to have a look.

I don't think that detailed stage-to-stage calculations will bring more accurate results. After sufficient time, all liquid phase will drain through the trays and end up in the bottoms anyway, so the column will contain 2 distinct phases - vapor and liquid.

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[maddocks]

Haven't done the image thing before -

 

[maddocks]

ok, see below. Feed is at 109°F and 60 Psig and the reboiler will run at about 267°F to make a 12 RVP stable condensate. Tower will run at pretty much 60 Psig.

 

[maddocks]

Hang on, I'll put this in excel format so you can copy/paste. Here it is in mole %

 

[EmmanuelTop]

Thank you.

It was a wrong assumption to use the flash operation across a valve (J-T). Ambient cooldown is not an isenthalpic process because a part of the process heat is consumed by the ambient surroundings. Also, different things happen within a flowing system than in an isolated (V=const) system.

The most accurate results can be obtained by builiding a dynamic model of the tower (with accurate system volume = the tower + associated piping), then close all SDV's and introduce a negative heat stream that will cause the entire system to cooldown e.g. 5 degC/hour till it reaches -30 degC. The resulting pressure will be automatically calculated by the model. HYSYS can do this for you in a few minutes.

In the attachment, I presented a simplified manual procedure that I have used on a few occasions when resolving similar subjects. The result will not be within +/-0.1 accuracy but it should be in the same ballpark (order of magnitude) as the real number.

The key concept is to follow the sequence of events, and to consider conservation of mass inside the system, and volume of the system.

I hope this helps.
Best regards

http://antwrp.gsfc.nasa.gov/apod/astropix.html

 

[RPilch]

Keep a bottle or two of nitrogen around as a backup plan. If you pull nearly full vacuum on a vessel not designed for it you'll likely be very, very sorry. If you're taking shortcuts you're probably going to get burned. Maybe not the first time, but down the road after you've gained a false sense of security.

 

[MortenA]

I think that your own suggestion to calculate the vapour pressure at -30ºC sound like a good starting point.

You should consider upset situations (those often cause the units to shut down), and what about steam-out? If they use steam out then vacuum is certain.

Best regards,

Morten