Valve Pulsing During CIP

Usually when there are discrete valves associated with most process unit operations, there’s a need to pulse those valves during CIP. 

So from a DeltaV perspective, I’d like to ask the community their thoughts on the best way to accomplish the pulsing.  Should it be handled in the valve control module as a separate setpoint?  Is this something better handled at an equipment module level to allow some sequencing in addition to the pulsing?  Should it be done in the CIP phase logic? 

Along the same line, are there any rules of thumb related to how fast the pulsing needs to occur?  How critical are the pulsing speeds and sequencing synchronizations?

  • Hi Bruce

    Our team in AP have worked on several biotech projects which uses CIP and SIP extensively throughout the plant. We not only have to pulsing (or toggle is the word we use) the valves, we had class based ways to handle similar tank/reactor/skid unit classes.

    Leveraging on concepts of S88 for CIP we split the CIP procedures into 3 units, one the CIP unit which handles the CIP skid, the second a distribition unit and lastly the unit to be cleaned. The toggling is then handled by the unit phase running on the unit to be cleaned. we kept with the EM concept pulsing or toggling the valves groups, I have seen methods of using composites to individuallly drive the valves position as well.

    Well this is putting everything simple in a paragraph. The implementation is much more involved and takes understanding the customer's process and involved your piping design and process design function as well. But the modular and class based concepts served us very well in reusuablilty and ease of validation.

    Hope this helps

    Robin

  • In reply to robin.toh:

    Robin,

    Thanks for the feedback.  Was speed ever an issue or concern?  How precise / repeatable did the timing have to be (open/close of a valve or one valve compared to another valve)?

    Bruce

  • I have always found it better to push the control as low as possible in the stack because that means there is less for the upper levels to manage.  If coordination between valves is not required then the pulsing can be done at the CM level.  If coordination of multiple valves within a unit is required then the function has to be implemented at the EM level.  But, creating CM classes that have the capability means that it is available for the code to use if and when they need it.