Overcooling problem during depressuring

[Richler]

Hi,
I'm performing a blowdown calculation in adiabatic conditions for a gas header (mainly methane).
Initial conditions are P=460 bara and T=60°C. The header has to be fully depressurized.
The problem is that overcooling over -10 °C has to be avoided during depressuring (pipeline design temperature).
The flare header backpressure is nearly atmospheric.

Is there a way to perform the blowdown (using several time intervals or whatever) in a way to avoid overcooling?

Thanks for the patience,

best regards,

Francesco

 

[TangoCleveland]

We've done similar blowdown work on pressure vessels. You'll need to have the volume of your header. In a spreadsheet, figure the mass and thermo properties of gas in the header. Reduce pressure (maybe by 10%), volume stays constant, and use constant entropy (isentropic) properties to determine the new temperature. Keep doing this in small increments until you get to your temperature of concern. At that point, you'll need to stop blowing down and let the header heat up again.

If you're really adiabatic, there's no way that you can heat up the header. Please check my conversion - 460 bara = 6671 psia? Where does such a high pressure come from?

Larry

 

[Latexman]

Richler,

Anyone with Ch4 at that high of a pressure must have access to a dynamic simulator like Aspen, right?


Good luck,
Latexman

 

[Richler]

Hi all,

of course I'm using a dynamic simulator (OLGA 2000 and Hysys, actually).
Right now my problem is performing a correct simulation of the fluid entering the flare header, where very low temperature can be reached.

The main problem with OLGA seems to come from the PVT table values (P & T range is huge).

Anyone with a good hint?

Tnanks again,

Regards,

FR

 

[Latexman]

Richler,

IOWs, the equations of state (EoS) used by OLGA are not predicting PVT accuratly according to a reliable reference table, right? If not, please rephrase, I'm not familiar with these two software.


Good luck,
Latexman

 

[TangoCleveland]

Can't tell whether you're having a modeling problem or a properties problem. I'd suggest contacting the OLGA people for tech support. We use AFT products for flow modeling, but only single phase. AFT Arrow does compressible, AFT Fathom does incompressible flow. I believe OLGA is a multi-phase modeling product.

Larry

 

[jay165]

If I'm reading your post right, the minimum design temperature of the pipe material is -10 C (+15 F). I would challenge this (this sounds like the 106 Gr B lower temp limit for the design minimum yield strength). ANSI B31.3 has flexibility for lower temps at lower pressures. Most carbon steels have min temps of -20 F (-29 C) w/o impact testing. B31.3 allows carbon steels to be used at even lower temperatures (down to and including -50 F) w/o impact testing provided that the maximum operating pressure of the components will not exceed 25% of the max allowable design pressure at ambient conditions and the combined longitudinal stress due to pressure, dead weight and displacement strain does not exceed 6 ksi. In other words your pressures may be low enough at the time the pipe cools down to be within the "safe" zone during blowdown. We have used this rule successfully in assesing blowdown systems in the past.

Not sure if the applicable code in your area allows this, but I would check into it before designing a complicated blowdown logic system to limit the rate of depressurization.

 

[Richler]

Hi,
I answer some questions: to use OLGA 2000 without the compositional model you need to supply a property table. I do it via PVTSim.
In the case I'm facing the P&T range, as I said, is really large, so you could get results that are not so accurate.
By the way, Latexman, Hysys 3.1 is from AspenTech. I used its depressuring utility to model what happens to the fluid before the blow down valve while I intended to use OLGA to model the whole scenario.

Kind regards,
Francesco

 

[Latexman]

Richler,

I'd call the dP you are looking at gigantic! Seriously, I recommend looking for some real PVT and H data on CH4. A header good for those pressures, must have been expensive, and I wouldn't risk it undergoing brittle fractures as it depressurizes to an EoS without testing it. Either that, or I'd rig up several temperature measurements to stay safe during the depressurization.


Good luck,
Latexman

 

[Latexman]

Richler,

I stumbled across a good reference for you. Check this out:

http://chumba.che.ncsu.edu/publications/predictions of the joule-thomson inversion curve.pdf

Good luck,
Latexman