Emerson Exchange 365

Cascade Control - Level/Flow - Slave Loop fails

Hi All,

Condition: Level - Master Loop, Flow - Slave Loop

When Slave Loop fails, let say Transmitter failure, the whole loop goes to Man and the operator have to control the valve output. I had a requirement before that when this happen, Master Loop should have a direct control to the valve, meaning from Cascade Control to simple Level Control.

There are ways to do this:

1. Use of tracking, TRK_VAL = Level controller output. TRK_IN_D = True when Slave Loop fails

2. Use of two controllers, 1 for Cascade Control and 1 for Direct Control. AO_BKCAL is connected to both for bumpless transfer and an input selector to do the switching when Slave Loop fails

3. Use of Selector for the switching and CALC block for the switching of BKCAL output from AO for bumpless transfer

but is there any Standard way of doing this? i prepare to use #1 option?

Thanks,

Gelo

GeloCortez

• 12 Oct 2015 1:09 AM
• Cancel

7 Replies

• DBacker

  

  • 12 Oct 2015 4:04 AM

  Please check books on line for the bypass functionality of the PID function block. In books on line enter "bypass." One of the entries will be for "PID Function Block Faceplate." It has the following text.
  
  "Bypass – This area is visible if the Bypass Enable option is set in CONTROL_OPTS (typically when the PID block is a slave in a cascade pair). Bypass makes the output value equal to the % SP value, which allows the master loop to drive the output if the transmitter is unavailable. When the block is in MAN or OOS mode, click Bypass to turn Bypass on. Click Normal to turn off Bypass."

  • Cancel
• GeloCortez

  

  • 12 Oct 2015 5:07 AM

  In reply to DBacker:

  Hi DBacker,
  
  wow, i've been ignoring "Bypass" functionality and thought it has the same function with Tracking.
  Thanks for the info.

  • Cancel
• Brian Atkinson

  

  • 12 Oct 2015 5:45 AM

  In reply to GeloCortez:

  I'd also look at the control loop webpage (and book) by Terry Blevins and Mark Nixon.
  
  www.controlloopfoundation.com/cascade-control-exercise.aspx

  • Cancel
• James Beall

  

  • 12 Oct 2015 12:00 PM

  In reply to Brian Atkinson:

  The Bypass feature works very nicely. It will account for the control action (reverse or direct) of the Slave loop when it is bypassed. Note that you may have to put different tuning in the Master loop when the Slave loop is being bypassed. For example, a typical flow loop has a process gain of 2%PV/%OUT. Thus, you would need to reduce the gain of the Master loop by 1/2 to have a similiar response from the Master loop.

  • Cancel
• GeloCortez

  

  • 20 Oct 2015 12:24 AM

  In reply to Brian Atkinson:

  Thanks for sharing the site.
  
  Cheers,
  Gelo

  • Cancel
• GeloCortez

  

  • 20 Oct 2015 12:24 AM

  In reply to James Beall:

  Thanks for the info, really appreciate it.
  
  Cheers,
  Gelo

  • Cancel
• MPHymel

  

  • 3 May 2021 7:00 PM

  In reply to GeloCortez:

  James Beall , we created an intermediate PID that we no longer need. Essentially, a slow TIC (0-500°) cascaded to a fast TC (0-500°) cascaded to an FIC (0-27660 BPD). We wish to eliminate the middle TIC. We've been running it this way for some time with the middle TIC in bypass. My question: is there a neat trick to determine a multiplier to use to determine the new gain parameter based on the PV span and Output span of the intermediate controller? Like would 500/27660 = 1.808% work?

  • Cancel