Re: High Purity Tower DMC control - effect of operating state variation on controller performance

Name    : 
E-Mail  : 
Subject : 

Comments:
: Naidu
: I suspect that a significant amount of error could be attributed to shifts in feed composition.  Small (e.g. 1% shift from propylene to propane could cause significant errors in overhead composition).    Before you automatically take ln of both sides(since you really haven't changed operating targets and you are not varying over a broad range) you should check to see if it improves you prediction of the plant data (i.e. prediction utility within the Model).   Sometimes the ln can actually make the prediction worse.

: I wouldn't worry about pressure compensating steam flow as it is automatically getting direct feedback from the level.  That's why implementing the regulatory scheme you have is good.


Optional Link URL:  
Link Title:         
Optional Image URL: