Expert Alarm Management and Guidance System Trends Harnessing Newest Technology Introduction
K. Jayaprakash
Discipline Head (C& I)
Tata Consulting Engineers Ltd

In Process Automation, an advanced alarm management could be harnessed to improve the plant performance & enhance safety by quicker and appropriate responses while preventing spurious plant trips and preventive maintenance.The challenges in the alarm management continue even with Modern Automation Systems, number of alarm configurations and display are virtually unlimited leading to poor alarm management. Inefficient alarm management can lead to serious consequences in operation of plants.

Today's modern control systems, mostly standard software components, enable us to create alarms that are flexible and with no incremental cost. The combination of flexibility and economy has resulted in one unintentional & undesired by-product-alarm overload. In fact, the mindset "if it costs nothing, why not alarm it?" becomes an easy trap to fall into. However, excessive alarms could deprive the operator's attention to the really serious alarms while he is distracted by "not so critical" type of alarms. Poor alarm management can lead to serious consequences in operation of plants.

Expert alarm management system(EAMS) is user friendly and, provides opportunity to improve currently available alarm systems. It is preferred as guidance tool to identify the root cause of an alarm condition by traversing the decision/fault trees. These can be particularly useful in developing countries and plants where the operators may not be well versed with complete operation processes and in the process plants with latest technology in which operation skill set is still not much available.The ease of plant operation in the presence of Expert alarm management and guidance system provides opportunity to improve currently available plant alarm systems.

Also, considering the importance of increasing standardization and consolidation activities across the globe, various bodies like ISA, EEMUA, IEC have prepared standardized approach to alarm management system.

Need for EAMS


Figure 1: Block of Alarm management system

Today thrust for operational excellence in process plants is driving the need for effective alarm management. Plants are now looking always for reducing down time, increasing productivity and using plant performance system for achieving desired results


Figure 2: Typical standalone EAMS

In process plants, alarm systems are used to alert the Operators and plant personal of the abnormal process conditions or equipment defects. In conventional alarms system before adoption of Distributed control system (DCS), visual and audible indication on Panel board were available and restricted because of space limitation and cost.

With growing flexibility and increased facility available in DCS, various tools are developed to aid personnel operating and maintaining the plants .

In this context, now-a-days to aid the operator and also build intelligence into alarm system, an expert guidance system is introduced as a part of alarm management system. This shall facilitate the operator to play a pivotal role in operator decision making and requires the information empowerment. The EAMS shall free the operator from tedious or repetitive tasks and shall give more time to focus and make intelligent decisions that improve productivity and plant performance.

Objectives of EAMS
  • Root cause analysis of processdisturbances
  • Recommend operator actions
  • Operator guidance messages
Features of EAMS

Main
Alarm analysis application tracks different alarms generated in control systems and keeps track of all the process values that are required to perform alarm analysis. Only configured alarms & process points are tracked in the system, hence non critical alarms & process points are not tracked. Once the critical configured alarm is detected, root cause analysis program is activated. The alarm analysis decision tree is traversed and possible cause of the alarm is identified and same is displayed on the operator work station. Diagram of EAMS in Figure1.

Decision support on the corrective actions for process operation and maintenance:
System has four Modules
  • Logic Tree Module
  • This comprises of various fault tree logics to help analysing root cause. These trees are scanned/executed continuously.
  • Graphic Tree Module
    These trees provide the operator graphical representation of fault Trees & the traversal of tree logic for an Alarm with cause / probability.
  • Probability Computation Engine
    Provides the probability for various causes triggered by alarm.
  • Reporting Engine
    Responsible for annunciating the cause
Alarm diagnostic functions follow above steps for Root Cause analysis in case of process abnormality.

EAMS are implemented in 2 options:
  • Standalone EAMS - All process points & alarms are fetched through OPC and are configured in application package as standalone system since few DCS vendors do not have standard package for above functions. Figure 2.
  • EAMS Integral system - All process data and application software reside in DCS systems. Few DCS vendors have developed these functions in DCS. Figure 3. Alarm decision trees shall be available in Operator work station.1
Special features:

Critical alarms apart from displaying on OWS in CCR are sent as an SMS to identified plant personnel mobiles. So personnel shall know the major events ,

Justification & Conclusion

Alarm management shall be viewed as investment for risk management. The effectiveness of Automation is governed and improved by Expert effective alarm management system. This system shall play a pivotal role in operation's decision making and requires information empowerment. The system shall transform the role of operator from doing tedious & repetitive to well guided task.
Figure 3: AMS integral to DCS

In process plant, number of alarms will be triggered. However only critical alarms with root cause are to be maintained and displayed for operator attention on priority when alarm is detected. Alarm management system will guide operator to take appropriate decisions/actions by viewing the root cause analysis program triggered and diagnosis tree traversed and possible cause for the alarm is identified .These Fault trees for Root cause analysis are to be customised for each type of process industry and are prepared by group of Domain experts.

However some of the DCS vendors have implemented these expert alarm functions in DCS These shall not affected main control functions and have brought value addition to the EAMS. Seamless integration of database with plant control system adds value to the system without any additional external interface. Future trend & demand may push all DCS Vendors to implement as standard modules

Effective alarm management system is crucial for efficient plant operation. Relevant intelligence built into the "Alarm management system" design with flexibility to incorporate plant best practices for continuous improvisation. As the right action is likely to be taken more quickly, this alarm management system often leads to bottom line benefits such as less equipment damage and improvements in process up time, & quality. Nuisance -free intelligent alarms provide the operator with the time window to implement a corrective action prior to escalation.

With implementation of these packages in all the future projects by all Utilities, we can presume that benefits are seen by the Operation & Maintenance personal for providing guidance to them for attending to problems and avoiding spurious trips, prevention of loss of life and property, environment protection, etc.

However, this Expert Alarm management system is in the infant stage of implementation of complete required alarms with a Faults decision trees. All the stake holders', consultant , clients, OEM’s and DCS vendors have to develop these data base of fault decision trees for all alarm to cover all process plants and take this initiative & concept forward. The cost involved is less, but long term benefits are more.

All DCS suppliers provide an effective foundation for implementation of alarm management system confirming to ISA 18.2. The latest advances in dynamic alarm conditioning methods, like eclipsing logics, out of service conditions, alarm shelving, alarm prioritization, alarm grouping etc can be implemented in present DCS. These alarm practices are developed under a common management framework; these remove bad actor alarms, rationalisation of alarms, alarm sorting, etc. Hence alarm flooding is avoided and plant safety & reliability is improved. However, the intelligent root cause analysis functions are not covered as a present alarm module and ISA 18.2 alarm model.

Drive towards Excellency in operations shall promote the need for effective alarm management .Plant users are operating close to limits and always look for ways to increase the operation efficiency by reducing the down time, increasing productivity and implementing real time performance management strategies These are gaining importance due to the growing void in experienced operators and increased level of complexity and sophistication in control system.