Tips & Tricks - Getting the Most from your VMGSim Case Studies
The Case Study in VMGSim is a powerful tool that enables you to perform sensitivity analyses of simulations allowing you to inspect the behaviour of models across any number of variables. This article explores some of the most important features in the Case Study such as adding feed stage location in distillation columns, and specifying points back into the simulation model.
The VMGSim Case Study utility runs sensitivity analyses by manipulating any number of independent variables and reporting the effect on any number of dependent variables. The Case Study can be accessed through the Tools menu.
A Case Study runs in a separate memory space from the current simulation case. The entire simulation case is cloned in memory and then run on this copy, which leaves the main case untouched.
- The Clone button allows you to copy the configuration of a Case Study into a new Case Study
- You can see details on the progress of the Case Study in the Status tab
- Many of the features described in this article apply also to Model Regression and the Optimizer
- The Case Study points can be configured to run in parallel processors in the Settings tab - this option is most effective for large case studies or large simulation cases as small cases do not see the benefit because of the overhead to set up the multiple processes
Selection of Variables
Virtually any variable from the simulator can be added to the Case Study, including any data type such as numbers, booleans (check boxes), integers, strings and items from vectors. It also includes variables from any unit operation and other objects such as process calculator cells, interaction parameters (kij) and physical properties from hypothetical components. The only restriction is that an independent variable must be a specification within the simulation case.
Variables can be added through the Add button from the Case Study. A more flexible way of adding them is through the right click menu Add to Case Study which is supported by almost every variable in the simulator. This right click menu is only available if the Case Study form is already opened.
Compositions as Independent Variables
The composition of any component can be added as an independent variable. The Case Study contains a special algorithm to monitor composition specifications to handle normalization. The input values selected for these components are preserved while the other components are normalized around them to add up to 1. Mole, Mass and Volume compositions are supported.
Feed stages to distillation columns can be added to the Case Study with the right click menu in the Configuration tab.
The same also applies to draw stages and energy feeds.
Any cell of the process calculator can also be used in the Case Study using the right click menu. The only thing to keep in mind is that the cell must have a value before it is added to the Case Study. Variables from the thermodynamic environment can also be added to the Case Study including interaction parameters and pure component properties from hypothetical components; all you have to do is open the form and select the variable.
Data can be input in two main modes: all combinations or as a list of data. The first mode runs all the combinations of the variables selected. For example, the Case Study will run a total of twenty points for the two independent variables selected below where the first variable has four points and the second variable has five points.
If you uncheck Run All Combinations there is a user specified total number of points defined in each row of the Input Data grid.
When in doubt, you can always check the Results tab before running the Case Study to check the exact order of the point that will run.
The points within every independent variable can be specified directly or auto segmented between two values. The number of points or the delta can be specified, while the labels are Min and Max, the limits can be any values and not necessarily Min < Max.
Once the Case Study is running you can see the progress in the Status tab or inspect the calculated values as they become available in the Results tab. There are a number of different plots available, including 2D plots, 3D plots and density, to inspect the results one the run is complete.
A very important feature when inspecting the Case Study is the ability to specify any point directly on the live simulation case. This can be accomplished with the Spec. in Flowsheet button from the Results tab.
The Case Study in VMGSim is a very powerful tool to perform sensitivity analyses. This article described some key features when configuring and running it but it is also worth noting that many of these tipsalso apply to the Regression and Optimizer utilities. The power and flexibility of VMGSim does not end here. There are a variety of tools beyond the Case Study that can be used to analyze and extract results from the simulator such as the Excel Add In, Scripting or COM Programming.
Please contact your local VMG office for more information.
Raul Cota, Ph.D., P.Eng., VP Research & Development