HYSYS viscosity error for diluted bitumen

[PetroBob]

Hi folks

A HYSYS simulation for blending diluent and bitumen is showing lower kinematic viscosity in the dilbit than that actually measured in the field. Is this a typical error for HYSYS? (Ie is it normal for HYSYS to under-predict viscosity when blending light and heavy hydrocarbon liquids?)

Does anyone have any suggestions as to corrections I could make to the simulation to improve the accuracy of viscosity prediction?

Thanks!

 

[jmw]

Where there is disagreement, it is necessary to question each element and assumption.
One ought to suspect, as a starting assumption, that the actual calculation is good because if it is a standard calculation any defects ought to have been discovered.
If all other aspects check out then the calculation can be cross checked by using independent methods. It would be feasible to use a fuel oil blend calculator (Shell has one and DNV PS produce one also.)These also calculate the density.
It is worth while also checking if the predicted density is as expected.

But if we also presume the calculation is correct is the process measurement correct? Has it been validated by lab sample and analysis?
Are you satisfied the process measurement is reliable?
What type of viscometer is it, what are the measurement conditions and how does it correlate measured viscosity to reference conditions?
Does the density measurement correspond with expected? and has it too been tested in the laboratory?
If the process measurements are verified, and if independent calculations agree with each other (more than they agree with the process measurement) then the next step is to suspect the the original data for the density and viscosity of the distillate and raw bitumen. In a great many cases these values may be in question due to storage and management.
Finally, were the original viscosities (and densities) of the blend components verified?

JMW

http://www.ViscoAnalyser.com

 

[PetroBob]

Thanks JMW for the response. I have limited information ie I don't know the viscosity and density measurement methods. Good points re completing calculations independant of HYSYS to cross-check, and to validate the original assays and density/viscosity information for the bitumen and diluent in the simulation.

If the Shell and DNV PS fuel oil blend calculators are available free online could you please send the link? All my google search could find was refernce to old links that no longer work.

Thanks for your time and useful suggestions, much appreciated.

As I haven't had any response to my main question (does HYSYS typical show viscosity errors when blending different density hydrocarbons) I will re-post this question in the Aspen forum.

 

[jmw]

Shell Bunker Calc is here:

http://www.shell.com/home/content/marine_products/products/fuels/bunker_calculator/

Of course, they don't say if they use ASTM D341 for temperature corrections (to bring the entered data inline) or their own Shell V50 equation but if you enter the viscosity data for both components at the ssame temperature and make that the temperature ffor the calculation that will get you round that problem. Also, such calculators use fuel oil references curves for the temp correction - another reason to do it this way.

JMW

http://www.ViscoAnalyser.com

 

[PetroBob]

For interest, I used Shell BunkerCalc to cross check HYSYS viscosity results when blending diluent (condensate and synthetic crude oil mix) with bitumen. I found that the results are very different (HYSYS gives lower synbit viscosity than BunkerCalc) at low temperatures. The HYSYS and BunkerCalc viscosity results become more consistent at high temperatures, and at higher diluent/lower bitumen concentration.

Thanks JMW for the info and link.