Incomplete rationalization?
Lately, we've been confused when we receive what we are told is a rationalized database. Those of you familiar with TiPS LogMate package (http://www.tipsweb.com/products/logmate/) know that we create a KB (that stands for knowledgebase) that contains all of the configuration information and design parameters surrounding alarms.
This KB is important, in that it captures all of the alarm and tag configuration, including information about cause, consequence, corrective action, trip points, time to resolve, and even links to documentation about standard procedures, etc. It becomes your alarm datamart repository. A datamart is a subset of a data warehouse. It is set aside and optimized specifically for the use of alarm management, and the various methodologies surrounding the professional practice in treatment of alarm management issues.
As such, this rationalized configuration should be rather complete once somebody has spent over $100k with a service company who writes books on alarm management. Unfortunately, that is far from the truth. Our experiences show that on many occasions, less than 10% of the configuration has been completely rationalized and documented. Furthermore, it is usually contained in a spreadsheet that is not current with the prevailing configuration, as it seems to be delivered months after the fact. How does a company stay in business delivering this level of quality workmanship?
The only answer we can come up with is that the general user community is not really aware of what they should expect as a deliverable when they hire a service provider to do this for them. Realizing this, our Chris Wilson undertook the task of writing a white paper a couple of months back on just what you should expect. This whitepaper is available through Control Global white paper releases. It is entitled the "Alarm Rationalization Quality Assurance Guide". You can find it at http://www.controlglobal.com/wp_downloads/TiPS.html.
In numerous conversations with people from the industry, we heard comments of "we've spent lots of money having this done, but we're not sure what we received for our money". When we saw the results, and tried to enter them into our KB, we understood why. Download this paper, and ask for a full quality deliverable when you want a rationalization done on your process units. There are several other companies out there who deliver a complete job. One is Mustang Engineering. They have some of the worlds leading experts on staff. We have also seen impressive work from Invensys Services group. They will supply this work as part of an ongoing service contract. Each of these companies gives a complete delivery of a fully rationalized configuration.
Why is a fully rationalized KB important? The first and most obvious reason is that the documentation is priceless. Not only does it help to have that information available to your operators online, but when regulations come (and they will) their requirement will be more for documentation than anything else. Get things documented now, and you will be ahead of that game.
The second major reason for a fully rationalized system is that not only do you learn a lot in the process, but you identify areas for opportunity. One of those opportunities is dynamic alarming, and operator advisory systems. To see where you can go with this, check out our partners at UReason (www.ureason.com). Their OASYS system adds directly on top of LogMate, and makes use of that rationalized information to dynamically control alarm information that is routed to the operator.
Note that this is more than just a dumb system that sets up a bunch of plant states. It actually advises the operator in times of trouble, and guides his efforts in resolving situations. It does this in three ways. Firstly, it identifies commonly occurring alarm groups, and reduces the overflow of information (thus reducing alarm floods). Secondly, it shows causal connections for events that occur, and reveals information which can help the operator to avoid situations before they ever occur. Thus, avoiding the types of incidents that cause or result from alarm floods. Many who use it see it as the flood avoidance package of the future. Flood avoidance is the holy grail of alarm management.
Lastly, the resultant datamining effort reveals areas of your process that can benefit from additional efforts. Once you are at that point, you are starting to address the full situation awareness issues in your control room. You will soon realize that alarms are only an indicator, not the problem. To learn more on this subject, see www.mycontrolroom.com Alarms are a great place to go to find out where your true automation issues lie. This data yields all of the cost/benefit analysis necessary to justify automation upgrades, should they be called for.
If you think you could benefit from such a system, call us, and we'll have a UReason discussion with you.
In the meantime, be careful to see that you identify what you want for a deliverable for your alarm management efforts. Download Chris's white paper from Control Global to get a good start.
This KB is important, in that it captures all of the alarm and tag configuration, including information about cause, consequence, corrective action, trip points, time to resolve, and even links to documentation about standard procedures, etc. It becomes your alarm datamart repository. A datamart is a subset of a data warehouse. It is set aside and optimized specifically for the use of alarm management, and the various methodologies surrounding the professional practice in treatment of alarm management issues.
As such, this rationalized configuration should be rather complete once somebody has spent over $100k with a service company who writes books on alarm management. Unfortunately, that is far from the truth. Our experiences show that on many occasions, less than 10% of the configuration has been completely rationalized and documented. Furthermore, it is usually contained in a spreadsheet that is not current with the prevailing configuration, as it seems to be delivered months after the fact. How does a company stay in business delivering this level of quality workmanship?
The only answer we can come up with is that the general user community is not really aware of what they should expect as a deliverable when they hire a service provider to do this for them. Realizing this, our Chris Wilson undertook the task of writing a white paper a couple of months back on just what you should expect. This whitepaper is available through Control Global white paper releases. It is entitled the "Alarm Rationalization Quality Assurance Guide". You can find it at http://www.controlglobal.com/wp_downloads/TiPS.html.
In numerous conversations with people from the industry, we heard comments of "we've spent lots of money having this done, but we're not sure what we received for our money". When we saw the results, and tried to enter them into our KB, we understood why. Download this paper, and ask for a full quality deliverable when you want a rationalization done on your process units. There are several other companies out there who deliver a complete job. One is Mustang Engineering. They have some of the worlds leading experts on staff. We have also seen impressive work from Invensys Services group. They will supply this work as part of an ongoing service contract. Each of these companies gives a complete delivery of a fully rationalized configuration.
Why is a fully rationalized KB important? The first and most obvious reason is that the documentation is priceless. Not only does it help to have that information available to your operators online, but when regulations come (and they will) their requirement will be more for documentation than anything else. Get things documented now, and you will be ahead of that game.
The second major reason for a fully rationalized system is that not only do you learn a lot in the process, but you identify areas for opportunity. One of those opportunities is dynamic alarming, and operator advisory systems. To see where you can go with this, check out our partners at UReason (www.ureason.com). Their OASYS system adds directly on top of LogMate, and makes use of that rationalized information to dynamically control alarm information that is routed to the operator.
Note that this is more than just a dumb system that sets up a bunch of plant states. It actually advises the operator in times of trouble, and guides his efforts in resolving situations. It does this in three ways. Firstly, it identifies commonly occurring alarm groups, and reduces the overflow of information (thus reducing alarm floods). Secondly, it shows causal connections for events that occur, and reveals information which can help the operator to avoid situations before they ever occur. Thus, avoiding the types of incidents that cause or result from alarm floods. Many who use it see it as the flood avoidance package of the future. Flood avoidance is the holy grail of alarm management.
Lastly, the resultant datamining effort reveals areas of your process that can benefit from additional efforts. Once you are at that point, you are starting to address the full situation awareness issues in your control room. You will soon realize that alarms are only an indicator, not the problem. To learn more on this subject, see www.mycontrolroom.com Alarms are a great place to go to find out where your true automation issues lie. This data yields all of the cost/benefit analysis necessary to justify automation upgrades, should they be called for.
If you think you could benefit from such a system, call us, and we'll have a UReason discussion with you.
In the meantime, be careful to see that you identify what you want for a deliverable for your alarm management efforts. Download Chris's white paper from Control Global to get a good start.
Comments