[foxboro] FW: Warning - potential CALC problems (and Simulation Methods)

  • From: <tom.vandewater@xxxxxxxxxxxxxx>
  • To: <foxboro@xxxxxxxxxxxxx>
  • Date: Mon, 18 Sep 2006 11:13:56 -0400

Terry, (and others interested in process simulation)
        We had trouble with commenting the CALC blocks back in 1988
while working on our first IA project initial installation.  We learned
that there was some magic to how many characters the calc block would
process on each line before considering the rest to be a comment.  Later
the ; semicolon was used as a comment separator.  Since that time we
have implemented our logic as Joseph Wu suggested on the list so your
operation would look like this:

CST
IN BI01
IN BI02
IN BI03
IN BI04
OR 4
OUT BO01
END
=09
        Yes, this method uses more steps and that might have been the
reason someone tried to condense the logic.
        As an aside, I never feel comfortable enough to just write
sequence code or calc blocks and install them without testing them.  I
used to do this by dummying the inputs before connecting them to the
final parameters.  For instance, to test your logic I would have written
the steps, and before connecting anything to BI01 through BI04, I would
have typed a "1" into BI01 with zeros in BI02 through BI04, and watched
BO01 transition to TRUE on the block detail display and then worked my
way through BI02 through BI04 in like fashion.  I've gotten used to
doing this because there are a lot of opportunities to make a typo in
the STEP logic such as having two BI01's and no BI02 or other brain
farts that would compile but wouldn't work as I intended. =20
        Before we started using Esscor's FSIM CP Emulator coupled with
their SIM4ME simulator package I used to do the same thing to test
sequence logic.  This gave me the best comfort factor that my code not
only compiled, but functioned exactly like I wanted it to.
        (ESSCOR sales pitch follows): =20
(Janet Parker, please send my commission checks to my home address ;<)

        IN ESSCOR's FSIM "CP" we are able to leave all of the block
connections intact and then use SIM4ME to dummy the signal all the way
through the FBM's ECB's just as if the field I/O point had transitioned
from an open contact to a closed one.  In addition, by creating
simulation logic based on DCS outputs we can drive the associated inputs
to create the same feedback that the live process would normally
provide. =20
        For discrete loops, if we open a block valve that has an open
and closed limit switch we can make the CIN points transition to their
expected state.  For PIDA controllers we can use the output value it
drives through an AOUT point in a SIM4ME calculation that will
dynamically change the AIN point as an appropriate feedback for the
PIDA.MEAS.
        In previous simulations we created, or used, we had to change
.IOMOPT of IO blocks to zero instead of 1 to accomplish the simulation
and then had to make sure that we changed them back before copying the
compound to the live system.  With ESSCOR's offering the Control
application software, (Foxboro's Compounds), remain untouched and run in
the FSIM CP.  The loopback simulation applications you create using
SIM4ME can be built and saved as a separate entity.
        Setting this all up the first time is not for the faint of heart
and we continue to have some technical difficulties keeping it all
running at times, but ESSCOR has provided good service and we have made
a step change in the way we provide new control applications to our
production department.  Production uses our simulations for operator
training while we use it to test and verify that our control logic is
free of typos and or bad connections before we transport it, unchanged,
to the live system by using the same ICC driver tasks used by any other
Foxboro CP.
        Since the FSIM CP is just another station on the Foxboro
nodebus, you can even establish peer-to-peer connections with other
CP's/WP's/Historians, or in our case a TRICONEX TRICON that also
communicates on the nodebus via a FoxGuard module.  The TRICON, (SIS
applications), can then work in conjunction with the Foxboro controllers
or ESSCOR's FSIM CP to provide a complete emulation of your process
control and SIS systems. =20
        This approach requires a certain economy of scale to make it an
economically feasible solution, but for large scale operations it is the
"Cat's Meow"

Cheers,
Tom VandeWater
Control Systems Developer/Analyst
Dow Corning Corporation
Carrollton, KY  USA
 
 
_______________________________________________________________________
This mailing list is neither sponsored nor endorsed by Invensys Process
Systems (formerly The Foxboro Company). Use the info you obtain here at
your own risks. Read http://www.thecassandraproject.org/disclaimer.html
 
foxboro mailing list:             //www.freelists.org/list/foxboro
to subscribe:         mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe:      mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave
 

Other related posts:

  • » [foxboro] FW: Warning - potential CALC problems (and Simulation Methods)