Connecting Dynamic Aspen Custom Modeler (ACM) with FOQUS

[boverhof]

Description

One of the main issues with running dynamic ACM models with FOQUS is the requirement to specify a ‘TimeSeries’ vector which determines how long the simulation will run for and when to record dynamic data. This is a problem when working with cycle models which may operate using event driven conditions or switching (ex: switching to desorption when so much CO2 is captured) and the total cycle time or time you need to simulate for isn’t known before starting the simulation.

Test 1:

A simpler version of a model to test this connection was developed using the BFB example model. The cycles are set up using ACM’s Task tool, and here a Task is used to just change the inlet gas flow rate and simulate until the CO2 capture reaches a certain value. From running the simulation in ACM, it is found that this takes 375 seconds for the model to achieve this capture and finish running. The first test is to just only set a single value in the ‘TimeSeries’ vector which is sufficiently larger than the time required from the task. In this test I just used 1000 seconds.

When running this through FOQUS with the 1000 second time series entry, the simulation stops at 375 seconds (when the task completes) and FOQUS gives the ‘Run Completed Successfully’ message and everything reads fine. So, this method works for this smaller simulation, and it can be tested on the full cycle model.

The task tool in ACM is like a structure that sits over top of the simulation that can change boundary conditions within the model based on certain criteria (time, model variables, etc.). Now with this, there are tons of different ways to set these up. For test 1, I write one myself that activates after the model starts. It changes the flue gas flow rate and then will simply pause/finish the simulation when the capture reaches a certain value. This is really simple, only 5 lines of code total need to be written. The other tests use a task which has been used to simulate the entire temperature swing CO2 adsorption cycle. This still uses the ACM task tool but contains much more code (234 lines that is actually generated using an interface from ACM).

[boverhof]

@Ryan-Hughes-8 can you provide the simulation files so I can attempt to run test1.

[Ryan-Hughes-8]

@boverhof this zip file contains the BFB ACM file, simsinter .json file, FOQUS file, and a word doc explaining how to create the test 1 task in ACM.

simple model.zip