[foxboro] AW: HIC's
- From: "Sieling, Marcel" <msieling@xxxxxxxxxxx>
- To: "'foxboro@xxxxxxxxxxxxx'" <foxboro@xxxxxxxxxxxxx>
- Date: Wed, 5 Jun 2002 05:14:56 -0400
Hi,
>What we have been seeing in recent weeks is a weird thing. On one of =
the
>sets of HIC's, [...] the HIC.MEAS and HIC.OUT are constantly
>switching back and forth between the higher and lower value every
>half-second. =20
There are two possible sources for this kind of problem:
1) Two sources of data are writing to one destination (this does not =
apply
to linked connections of course, because it is not possible to link =
more
than one source to a destination, but this may happen in case of =
Sequence
involvment or other software writing to parameters with omset or =
something
like this). This could result in concurrent overwriting of a value and =
then
dependent of block phasing an alternating result.
2) There is some kind of a feedback and due to timing issues resulting =
from
block phasing the value is alternating or oszillating.
This could lead to even more strange results. I once had the case that =
one
single CIN value was shown on two display managers and the first was
constantly TRUE and the second was constantly FALSE. I nearly went mad =
on
this and could not believe my eyes and it turned out, that the CIN was
alternating between TRUE and FALSE each half second and the two =
different
DMs just scanned the block in different moments, so that the first =
always
saw it TRUE and the second FALSE.
Solution:
look carefully in the loop documentation to figure out the information =
flow.
If there is no graphical loop documentation, take a big and empty sheet =
of
paper and create one. Write down all PERIOD and PHASE parameters of the
blocks concerned and carefully think over this drawing. I am sure you =
will
find the problem.
Good luck!
Best regards -
Marcel Sieling
Systems Technologies
Invensys Systems GmbH
Heerdter Lohweg 53 - 55
40549 Duesseldorf
Germany
Tel.: +49 (0)211 5966-171
Fax: +49-(0)172-50-2673077
Mobile: +49-(0)172-2673077
Email: <mailto:msieling@xxxxxxxxxxx>=20
Homepage: <http://www.invensys-process-systems.de>=20
Visit my private (german) homepage: <http://www.powerslider.de>=20
Anwender-Symposium 2002 - Invensys Systems GmbH
11. - 12. Juni 2002, Essen=20
http://www.foxboro-deutschland.de/symposium2002/index.htm
Invensys Showcase - Worldwide User Conference & Expo
8. - 12. September, Orlando, Florida USA=20
http://www.invensysshowcase.com
-----Urspr=FCngliche Nachricht-----
Von: Lowell, Tim: [mailto:tlowell@xxxxxxxx]
Gesendet: Dienstag, 4. Juni 2002 19:04
An: 'foxboro@xxxxxxxxxxxxx'
Betreff: [foxboro] HIC's
List,
I have a control problem I'm hoping one of you has solved already. =
This
refinery of course used to run on board instruments, and we completed a
cutover to I/A over several years that wrapped up last year. In one of =
the
areas, we have several sets of steam turbine pumps that are alternated =
in
service every few weeks. On the old board, there were pressure =
controllers
that were cascaded to HIC's which actually controlled the steam flow to =
the
turbines. To avoid lots of re-designing and Management-of-Change =
paperwork,
we decided to cut these schemes over as faithfully as we could to the
original design.
The way we ended up configuring these HIC's is as follows:
PIC (PIDA block) -> HIC #1 (PIDA Block) -> HV#1 (AOUT Block) -> =
FBM37 or
05 connected to turbine control valve #1
=3D=3D> HIC#2 (PIDA Block) -> HV#2 (AOUT Block) -> FBM37 or 05 =
connected
to turbine control valve #2
The HIC#1.MEAS is configured to the HV#1.OUT
The HIC#1.INITI is configured to the HV#1.INITO
The HIC#1.BCALCI and .FBK are configured to the HV#1.BCALCO
The HIC#1.RSP is configured to the PIC.OUT
The HIC#1.LR is configured to always be in REMOTE
The HIC#2.MEAS is configured to the HV#2.OUT
The HIC#2.INITI is configured to the HV#2.INITO
The HIC#2.BCALCI and .FBK are configured to the HV#2.BCALCO
The HIC#2.RSP is configured to the PIC.OUT
The HIC#2.LR is configured to always be in REMOTE
The HV#1.MEAS is configured to the HIC#1.OUT
The HV#2.MEAS is configured to the HIC#2.OUT
We couldn't figure out how to make the make the M/A transition on the =
HIC's
bumpless, so they aren't. The Console guys just place the HIC that =
isn't in
service in Manual and line up the HIC output with the PIC remote =
setpoint
before switching turbines, and then put the newly-in-service turbine =
HIC in
AUTO after the outside guy switches pumps. This is essentially what =
they
did on the old board.
What we have been seeing in recent weeks is a weird thing. On one of =
the
sets of HIC's, and only one, the HIC.MEAS and HIC.OUT do not match, by =
about
20% difference, and in fact, the HIC.MEAS and HIC.OUT are constantly
switching back and forth between the higher and lower value every
half-second. The valve actually moves up and down in the field, but =
not as
fast as the HIC is trying to move it, so the pressure stays relatively
stable.
I'm guessing there is some sort of problem with the way we have =
configured
this scheme, having to do with the way the HIC.OUT =3D HV.MEAS and =
HV.OUT =3D
HIC.MEAS such that some sort of bizarre feedback loop has been created. =
The
weird thing is, we have 3 other sets of HIC's configured the same exact =
way,
and they don't exhibit this strange behavior.
Has anybody configured a scheme similar to this? Is there a better way =
to
handle this? We need to send a Pressure Controller output as a =
setpoint to
2 or even 3 different Control Valves, but the valve(s) that isn't in =
service
needs to be able to be controlled manually and shut when it isn't in
service, and then matched up with the Pressure Controller output before =
it's
put back into service. We don't want to use straight AOUT faceplates,
because the Console guys aren't used to these, and will get easily =
confused
when they see a different number in the AOUT.MEAS field (which would be =
the
PIC.OUT) than in the AOUT.OUT field (the actual valve output).
Another question is, what the heck is the deal with this HIC? Why is =
the
HIC.MEAS and HIC.OUT differing by 20% when they are configured to =
follow
each other, and all the other ones like this work fine?
Tim Lowell
Control Systems Engineer
Phillips Petroleum Company, Trainer Refinery
Phone: 610-364-8362
Fax: 610-364-8211
tlowell@xxxxxxxx
=20
=20
_______________________________________________________________________
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
=20
foxboro mailing list: //www.freelists.org/list/foxboro
to subscribe: =
mailto:foxboro-request@xxxxxxxxxxxxx?subject=3Djoin
to unsubscribe: =
mailto:foxboro-request@xxxxxxxxxxxxx?subject=3Dleave
=20
_______________________________________________________________________
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:
