A good alarm system will give operators actionable information, in time for them to take action to avoid upsets, off-spec product, shutdowns, and similar situations. A proven approach to creating such an alarm system exists and has resulted in thousands of vastly improved alarm systems.
This approach is grounded in the widely adopted 7‐step methodology for alarm system improvement developed by PAS. This methodology has been proven in the field and aligns with the ISA 18.2 Alarm Management Standard. PAS participated in developing this standard, as well as the seven related ISA Technical Reports. The steps are:
- Alarm Philosophy Development
- Data Collection and Benchmarking
- Bad Actor Alarm Resolution
- Alarm Documentation and Rationalization
- Alarm Audit and Enforcement
- Real-Time Alarm Management
- Alarm System Monitoring and Change Management
When the control loops are not working correctly, the operator must intervene in the process. There are two important tools that they use to do that, the operator’s Human Machine Interface (HMI) and the alarm system contained in that HMI. (Future blogs will look at the operations risk associated with our safety systems, control system management of change, and boundary management.)
When the control system can no longer keep performance inside the proper boundaries, operators must intervene. We get alarms that require the operator to take some sort of action to get the process back on track. When that action is successful, we minimize the upset and increase our production and efficiency. We avoid loss of production efficiency, environmental releases, and undesirable trips from our safety shutdown systems.
Even though a trip of the safety system will take us to a safe state, it then requires us to do a startup. Besides being costly, we all know that startups are the most hazardous process operation that we do in our plants. If your alarm summary shows page after page of alarms, it will not help the operator detect a burgeoning abnormal situation. Poorly performing alarm systems have been cited time and time again as contributing factors to major accidents.
The alarm system must not be a collection of miscellaneous status notifications – things that are really alarms, things that are not alarms – all mixed together in the same viewer. We want the alarm system to contain only events that meet the definition of an alarm, showing abnormal conditions that require the operator to make a timely action to avoid a consequence. An existing alarm system that is not designed with such principles in mind will need some work to achieve this.
Alarm management is a mature body of knowledge. All the techniques and knowledge needed to fix our alarm systems are well‐known and documented. We've calculated that more than 40 million alarms have been rationalized per the 7-step methodology. To discover more and see examples, watch the on-demand webinar or read the transcript that goes through the seven‐step methodology.
Additionally, the Alarm Management Handbook, published in 2006, contains a work process for methodology, people, and technology. It includes detailed explanations of the problems, along with practical solutions. The “what and the how” are both addressed, as are the people-related aspects discovered through hundreds of successful alarm improvement projects. This book contains full details for fixing your alarm system.
This blog is an excerpt from the PAS On-Demand Webinar, Alarm Management: 7 Steps to Maximizing Operator Effectiveness.