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

Posted by Don Snowden on November 29, 1999 at 13:19:36:

In Reply to: High Purity Tower DMC control - effect of operating state variation on controller performance posted by S. R. Naidu on November 22, 1999 at 01:19:56:

Before making too many comments, I would ask a few questions. First is this a refinery application or ethylene plant application? There is a difference in how one approaches setting up the regulatory controllers. This leads into my second question, how is the regulatory control setup? In other words, is the reboiler tied to bottoms level or is the top reflux tied to the reflux drum level, etc.. Are there any temperature controllers active (namely in refinery applications).

If the controller was initially designed and built correctly it should be able to handle the 30-35% turndown. Some difficulty could be experienced if the product spec is changed (e.g. 99.5% propylene purity verses 97%).

In anycase, setting up multiple models is not trivial since you need to test the plant in both conditions. How are your prediction errors (or your accumulated prediction errors?

Follow Ups:

• Re: High Purity Tower DMC control - effect of operating state variation on controller performance Naidu 13:53:36 11/30/99 (1)
  • Re: High Purity Tower DMC control - effect of operating state variation on controller performance Don Snowden 20:57:04 12/02/99 (0)

Post a Followup

Name    : 
E-Mail  : 
Subject : 

Comments:
: Before making too many comments, I would ask a few questions.  First is this a refinery application or ethylene plant application?  There is a difference in how one approaches setting up the regulatory controllers.  This leads into my second question, how is the regulatory control setup?  In other words, is the reboiler tied to bottoms level or is the top reflux tied to the reflux drum level, etc.. Are there any temperature controllers active (namely in refinery applications).

: If the controller was initially designed and built correctly it should be able to handle the 30-35% turndown.  Some difficulty could be experienced if the product spec is changed (e.g. 99.5% propylene purity verses 97%).

: In anycase, setting up multiple models is not trivial since you need to test the plant in both conditions. How are your prediction errors (or your accumulated prediction errors?


Optional Link URL:  
Link Title:         
Optional Image URL: