[foxboro] Alarm detection policies (MANALM/INHOPT) (was: how to prevent IOBAD during PT online calibration)

New thread time :)

Tom, interesting topic.  For inputs (AIN/CIN), I like to "lock" these 
blocks in Auto with a loopback connection (i.e., :BLOCK.MA = :BLOCK.MA.1). 
 We almost always leave INHOPT at 0, as it is the default, and we almost 
always want alarm detection (i.e., flags like HHAIND) responding to alarm 
conditions even if the block in inhibited.  MANALM gets left at the 
default of 0 with no adverse effects.


The reason for this approach is that interlocks are defeated when 
operators put input blocks in manual.  This has been an issue in several 
plants at my site, and the locking scheme above seems to be the best, most 
"foolproof" resolution to this problem.  Getting there of course is 
another matter, as some plants/operators rely on the ability to put blocks 
in manual to get around certain conditions in the process.  This latter 
bit is not the right approach IMHO, but it is what we have, and it will 
take some case-by-case engineering evaluation to determine what is the 
best approach.  It's not an impossible task, but it takes time and 
prevents the simple running of a script to do the locking (much as I would 
like to do that :).


On controllers we set MANALM to 1 to get alarms when the controller is in 
manual.


If we truly need to prevent alarm detection or interlock action, we write 
an MOC and do it.  But the need to do that tends to be a rare occurrence, 
usually only a result of field device malfunction, or some unusual (less 
then once a year frequency) plant state.


I wish Foxboro would implement a shortand for loopback connections, 
something along the lines of COMPOUND:BLOCK.MA = :..1, in addition to 
making all settable parameters connectable.  Or better yet have a 
system-wide security/ACL mechanism (in the CPs, not the graphics) to 
control access to each parameter, so we didn't have to make all these 
loopback connections.  But I don't expect to see that until the CP380 at 
the earliest.


Corey Clingo
BASF





"Tom Badura" <tbadura@xxxxxxxxxxx> 
Sent by: foxboro-bounce@xxxxxxxxxxxxx
11/07/2009 01:58 PM
Please respond to
foxboro@xxxxxxxxxxxxx


To
<foxboro@xxxxxxxxxxxxx>
cc

Subject
Re: [foxboro] how to prevent IOBAD during PT online calibration






Well, I still can't seem to post anything form work, and the bad boys in 
IT
don't seem to care. I'll just have to post from home again and get the 
usual
"auto-scold" message about unauthorized email address, but so far it seems
to work.  My apologies to the list administrators.
Hi Ainuddin,
Is your Transmitter ECB is connected directly to an AIN block (IOMOPT =
1), and do you use the PNT value from the AIN for all of your control
logic?  If so you may be able to get away with just putting the AIN
block to Manual and possibly inhibiting the IOBAD alarm.  This is
something you can do from the Detail (Foxselect) displays without any
program download. 
Two things to look at first are the settings of the AIN Block MANALM and
INHOPT parameters. 
If MANALM = 0 (False) the AIN block will not generate any alarms when it
is put into Manual.  This is the default for AIN blocks.
INHOPT defines how inhibited alarms are handled.  The default setting of
0 only inhibits the alarm message - not the alarm detection and boolean
output (BAD). This could still cause your control logic to take
exception action if you use the BAD bit elsewhere in your logic.
Setting INHOPT to 1 or 3 will also inhibit the alarm detection and thus
not set the alarm boolean output (BAD).

I had a chance to test this on our system this morning.
If MANALM = 0 - Alarm Detection will be disabled and the BAD Boolean
will not be set when the AIN block is in Manual 
If MANALM = 1 
 - INHOPT must be set to 1 or 3 to disable alarm detection and setting
of the BAD output Boolean 
 - you must also toggle the Bad Alarm to Disable.

* You MUST put the AIN block in Manual and Disable the Bad Alarm (if
MANALM = 1) before disconnecting the transmitter.  The AIN block will
not reset the BAD output Boolean once it has been set until the
transmitter is reconnected.

In general, we have standardized on setting our blocks up with MANALM =
1 and INHOPT = 1 or 3 (depending on need for auto acknowledgement).
This allows us to still monitor (or simulate) alarms with a block in
Manual, but also completely inhibit alarm messages and control actions
when necessary.  I would love to get some feedback from other users on
this idea.

I hope this may be helpful.  It was worth my time to verify the exact
operation.

Tom Badura
Plastics Engineering Company
920-458-2121 x3366
tbadura@xxxxxxxxxx





 
 
_______________________________________________________________________
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:             http://www.freelists.org/list/foxboro
to subscribe:         mailto:foxboro-request@xxxxxxxxxxxxx?subject=join
to unsubscribe:      mailto:foxboro-request@xxxxxxxxxxxxx?subject=leave

Other related posts: