Tips & Tricks: Characterization of Oils with the Minimum Amount of Data

The objective of this section is to show how to obtain a comprehensive oil characterization with a reduced amount of oil pseudo components when only experimental transport properties of a crude oil are available.

From a Gamma Distribution to a TBP Distillation Curve

Sometimes Process engineers face the problem of characterizing crude oils when distillation data and bulk experimental properties are missing and the only information available is transport properties data.

VMGSim has integrated correlations to calculate a complete oil characterization based on transport properties data using the Gamma distribution. The results of this type of characterization are very interesting from an academic point of view since they can provide lots of information due to the large quantity of pseudo components that can be created. However; a large number of pseudo components can increase several times the convergence time of a flowsheet.

The following example will show how to characterize an oil when minimum quantity of experimental data are provided and how to obtain a reduced number of pseudo components that can define the TBP distillation curve of such oil. In this case, a heavy oil, for which the only experimental information available is the bulk kinematic viscosities at 100 and 210 °F, will be characterized.

Oil Summary

Kinematic Visc @ 100 °F = 1.424 e-03 m2/s

Kinematic Visc @ 210 °F = 3.458 e-05 m2/s

To start the characterization, open a new case in VMGSim and select Advanced Peng-Robinson as the Thermodynamic Model. Then, open the oil environment and select Gamma MW distribution as the C7+ Characterization Mode and Oil as the characterization Fluid.

Now, it is necessary to change the default values for the Whitson’s Gamma model in the Gamma Distribution form in order to have the recommended values for a heavy oil.

For a heavy oil, the recommended Whitson’s parameters are:

α = 25.0

η = 90 g/gmol

Gamma Max MW = 1210 g/gmol Number of Pseudo Components = 80

Delta MW = 14 g/gmol

The values of Gamma Max MW and Delta MW were selected in order to have 80 pseudo components distributed in intervals of 14 g/gmol. 80 pseudo components will guarantee a smooth molecular weight distribution.

Now add the kinematic viscosity values in the Bulk Information grid in the right hand side of the oil environment.

The assay is now ready to cut, click on the Cut button to see the results. Click the Install button to install the characterized pseudo components.

Go to the Result curves and select the Fraction vs Cn curve in Mole basis.

It can be observed that the characterization follows a modal distribution based on the entered Whitson’s parameters. This oil is characterized based on just two kinematic viscosity points and contains 80 pseudo components; however, using this oil with this large amount of pseudo components could slow down the solution of some unit operations.

The first reaction would be to change the number of pseudo components to be created in the Gamma Distribution, but in this case the fine resolution achieved with 80 pseudo components will be lost, as it is shown in the following figure. In this figure the oil was re-cut with the Gamma Distribution using the same parameters as before but with only 8 pseudo components and Delta MW = 140 g/gmol.

The oil environment of VMGSim allows characterizing this oil with less pseudo components without losing the achieved resolution with the Gamma distribution. To do this, follow the next procedure.

Return to the original Gamma Distribution assay with 80 pseudo components and Clone the assay as Assay 2. Now, select Experimental information as the characterization type and go to the Distillation Curve tab.

It can be seen that the TBP curve that was created during the Gamma distribution characterization is available in the Distillation Curve tab; this curve will be used as the TBP for the new oil. Click on the Liquid Properties tab an it will be seen that the Molecular Weight and Standard Liquid Density curves are also available.

It can also be noticed that the kinematic viscosities that were added in the previous assay are also available as specifications for the new one.

Before cutting the oil, let’s use the “Heavy_Oil” Range Set to specify the different cuts for this oil, by inspecting the TBP, it can be seen that most of the pseudo components are present in the Distillates and VGO ranges of the oil. Then, set the number of cuts of the “Heavy_Oil” Range as: Naphtha = 1, Distillates = 2, VGO = 4 and Residue = 1.

Assay 2 is now ready to cut, click on the Cut button to see the results. Click the Install button to install the characterized pseudo components.

Note that the resulting TBP is similar to the one provided by the Gamma Characterization; however, in this case we have only 7 pseudo components defining the curve as compared with the 80 obtained from the previous assay.

To compare both assays, close the oil characterization environment and go to the VMGSim flowsheet. Create two material streams, S1 and S2; connect S1 to Assay 1 and S2 to Assay 2 and assign a pressure of 101.325 kPa to each stream.

Open a Case Study and connect the Temperature of each stream as the Independent Variable and the Kinematic Viscosity of each stream as the Dependent Variable. Set the minimum and maximum values of the Temperature to 0 and 200 °C, respectively with 20 points. Be sure that the Run All Combination box is not checked to avoid running all the different combination of the stream temperatures.

Run the Case Study and select the Plot tab to compare the kinematic viscosities predicted for each assay.

As it can be seen in the following figure, both predictions are similar (the pink line represents Assay 1 and the green one Assay 2); meaning that both assays will yield same results for this property; however, Assay2 with 7 pseudo component will give faster answers when used in more complicated flowsheets.

The same procedure can be repeated for different properties and the results will be similar. For instance, the following figure shows the same type of analysis using the Liquid Density.

This exercise exemplified the use of the Gamma Distribution to characterize an oil with the minimum amount of experimental data and, how to covert that characterization into another one with a reduced amount of pseudo components.