5 replies to this topic

[ziron]

Hello dear friends,

 

I have a extra heavy crude that has a 9 API degree and I want to simulate a topping unit in Pro/II. The main probleme is that the crude assay characterization is made from 1% to 50% volume with D2887 not the usual TBP or D86.

 

I tried to input the data in pro/II but pro/II says that for the method D2887 i can only insert %wt values not %volume values, I tried in hysys to input the values but the same. I read on the internet that the D2887 method is similar with the TBP method so I inputed in pro/II the values with TBP method and %vol. At 55% I have 560 degree C, and Pro/II says that @98%vol I have a temp of aprox 1000 degree C, I crosschecked with Hysys and I got similar results.

 

But here is the problem.. that I can't undersand. I found some info on the crude from another partner that this crude at 98%vol TBP has a temperature of 820 degree C. From 1000 degree C to 820 degree C is a long way.

 

So what I'm trying to find out is how to input the D2887 crude assay data into Pro/II in %mass (i have in %vol). Is there any coreletions between D2887 and TBP ? 

 

In Pro/II i used Peng-Robinson, and some heavy oil option which affect the final viscosity of the simulated crude, The values between 50 - 98% didn't differ to much.

 

Any advice regarding simulating heavy crude oils and how to convert D2887 %vol to TBP %vol are welcome.

 

Thank you very much.

[PingPong]

ASTM D2887 results are always in wt%.

You should ask the person who provided the data to you to confirm that it is wt%, not vol%.

 

As you only have D2887 data up to 50 % the rest of the TBP curve is no more than an educated guess, irrespective whether it is done by PRO/II or Hysys or you do it yourself by hand.

Nobody knows what the TBP curve looks like at the end as it is impossible to determine. I was taught in the past that if a crude contains X wt% asphaltenes, then its (100 - X) wt% point is at 800 ^oC. The endpoint can be taken as 850 ^oC or whatever number one likes. It is all virtual reality. It makes no difference for the design of the CDU.

[ziron]

Thank you PingPong for the quick response.

Just an observation, if the client want's a specific API in the bottom of the column, the curve with the high temperature(1000 C) end points will have a smaller API and the curve with 800 C end points will have a bigger API, correct ?

 

Another thing that I want to ask everyone is it ok to use Peng-Robinson, Critical, Ideal-Gas Enthalpy: heavy oil, Molecular weight: heavy oil, Liquid viscosity: Simsci Heavy Oil, Liquid Thermal Conductivity: Simsci Heavy Oil.. for Crude with low API.

 

Thank you very much.

[Chellani]

Ziron,

 

Why are you worried about 98% boiling point?

 

Try to look at your assay definition. Define in form of cuts rather than bulk values/properties.

 

End point will not have any impact on API as API should be defined in assay. If you define API of 550+ or 650+ cut as XX, you will get that value in simulation (or somewhere close to that because of actual cuts rather than absolute / TBP cuts defined in assay).

 

In most of the cases, heavies cut (VR / AR) is defined till 50% boiling range and simulator should extrapolate that to whatever it wants.

[PingPong]

if the client want's a specific API in the bottom of the column, the curve with the high temperature(1000 C) end points will have a smaller API and the curve with 800 C end points will have a bigger API, correct ?

No, provided you also enter the API of the crude.

The crude is cut into many pseudofractions which, when lumped together, form the crude that you entered. Choice of Endpoint has a minor impact on the shape and gravities of the part of the TBP curve that the simulator has to extrapolate. In your case you only have data points up to 50 %, so half the TBP curve generated by the simulator (from 50 to 100 %) is imaginary anyway. Shocking really, but that is reality.

 

Another thing that I want to ask everyone is it ok to use Peng-Robinson, Critical, Ideal-Gas Enthalpy: heavy oil, Molecular weight: heavy oil, Liquid viscosity: Simsci Heavy Oil, Liquid Thermal Conductivity: Simsci Heavy Oil.. for Crude with low API.

The Heavy Oil option is the same as the standard SIMSCI option, except that it is (supposed to be) more accurate for fractions above 1000 K or 727 ^oC.

Note however that that part of the TBP curve is never available in any assay or analysis because it is impossible to determine. In your case everything above 50 %, say above 560 oC is simply pulled out of thin air by the "extrapolation" done by the simulator. I find it very funny that  a lot of people really think that they can determine any property above 600 ^oC with any accuracy. They blindly put their trust in a piece of software like Hysys or PRO/II and really believe that the simulator knows everything that is missing in the crude assay.

 

I am not going to comment or advise on the methods for K-values, or enthalpy, of critical properties, or MW, because I do not want to get involved in a discussion on what is the "best" method. None of the methods are very accurate. Selection is not based on that, but on what the purpose of the simulation is.

Is it for a design that your company has to guarantee, then you best make sure you choose those methods that produce a design that is on the conservative side. Your company manuals should guide you in that.

If it is to model an existing operation then obviously you try several untill you have those methods (and tray efficiencies) that result in the best approximation of reality.

Edited by PingPong, 30 April 2014 - 10:58 AM.

[ziron]

Tank you PingPong, I like your approach