What is a good alarm management philosophy? Is there an accepted standard in the HPI?
Specific questions:
1)number of alarms/total number of loops
2)alarm rate (eg # of alarms/10 minutes)
3)industry benchmark
4)the tools for alarm analysis
5)are we concerned with number of alarm points or are we concerned with the rate of alarm anunciated to the boardmen?
6)when we take out alarms from the system, do we need to re-hazop? Taking them out is actually un-hazop-ping the hazop study.
7)any tools that are highly recommended for alarm analysis?

Thanks for the info shared.

 

There is a guideline... EEMUA 191 which I believe covers a lot of the issues. I am not aware of any particular general purpose tools, but I
expect there may be some. Some systems vendors probably have tools for alarm analysis. DeltaV, in particular has some tools. In my opinion,
digital systems have reduced the "cost" of alarms to a point where they are treated as free. The problem is you get what you pay for. If the true cost of engineering and maintaining alarms were factored, there would probably be fewer, but more effective alarms. Of course, you may want to bear in mind what you paid for this opinion.

Regards,

Lou Heavner
Emerson Process Management

 

Many people get tripped up by the questions that you are asking. Most of the solutions you find in the industry look at these various statistics.
This entirely ignores the true problem of alarm systems. Is the operator getting the information he needs? Of course, it is true that an operator
can only handle a certain number of alarms in a given time frame, but is it logical to just reduce the number of alarms configured in that operator's area in order to meet that goal? Simply, rationalization to reduce the number of alarms configured in the system is not the solution. Static rationalization masks the symptoms while the underlying problem grows worse.

ProSys will be presenting a paper on "The Pitfalls of Alarm Rationalization and Benchmark Analysis" during the Texas A&M University, Mary Kay O'Connor Process Safety Center 2001 Symposium. This is on the very subject that you
are inquiring about. Additionally, if you want to find more about our unique solution (Dynamic Alarm Management) that addresses the true problem
facing alarm systems you might want to check out our Alarm Management Resource Center at:
http://www.prosysinc.com/solutions/amrc
We also have a seminar series:
http://www.prosysinc.com/seminars/index.php
If you have any additional question you can email me, or contact us via the website. We are proud to be the largest provider of real-time alarm
management solutions with over 100 process units in operation worldwide. We offered the first commercially available alarm management software and services in 1990.

Dustin Beebe
dsb<@NOSPAM>prosysinc.com

 

The EEMUA guide can be obtained from www.eemua.co.uk/publications/index.htm
There is also includes a summary of the guide (in the control category.

 

Regarding the tool for Alarm Management kindly refer www.ics-ltd.co.uk

Thanks

Ranga
Contact for addtional info:
[email protected]

 

In order to reduce the number of alarms, a rescheduling need to be done. A good practice include:

1.- Prioritize alarms
2.- Segregate alarms in operational units
3.- Autoinhibit low priority alarms depending on process conditions
4.- establish procedures
5.- Identify alarms demanding actions from operators, just indications, troubleshooting, ...