January 2020

Columns

Digital: Technologies are advancing industry beyond alarm management

Let’s face it, alarm management is a fully mature body of knowledge. Since the mid-1990s, industry has improved thousands of alarm systems, transforming them from overloaded nuisances to valuable operator tools for abnormal situation detection and response.

Hollifield, B., PAS Global

Let’s face it, alarm management is a fully mature body of knowledge. Since the mid-1990s, industry has improved thousands of alarm systems, transforming them from overloaded nuisances to valuable operator tools for abnormal situation detection and response. Tens of millions of alarms have been rationalized. Both the “what to do” and the “how to do it” techniques are well documented in standards, books and papers.

While thousands of alarm systems remain unaddressed, only the will to do so is needed. No “new knowledge” is required, no re-inventing of the wheel is needed, and no magic breakthrough will fix a sub-optimal alarm system with zero effort. Techniques and software are still improving, such as in automating the reduction of nuisance alarms. However, these are enhancements, not fundamental changes in the necessary work process.

Where is alarm management going? Many companies achieve significant improvements in production, efficiency and profitability by leveraging the infrastructure and methods in place for alarm management.

Alarm management infrastructure

The infrastructure of alarm management provides the following capabilities:

  • Connectivity. Alarm system performance monitoring requires a safe, secure, high-speed connection to the control system. Alarm records, operator changes, control system changes and other process events are captured and stored. When open platform communication (OPC) is used as the connection method, other information—such as process values, equipment states and point configuration settings—is easily obtained. A single server can handle multiple control systems of different types.
  • Computing. Most alarm management applications make it simple to create automated alarm performance reports. Just a few years ago, practices such as state-based alarming, alarm shelving and alarm flood suppression were called “advanced.” At present, they are commonplace, created with software that is part of the alarm management system. Operators can add these capabilities to existing control systems, and they can be centrally managed. This is usually far easier than trying to code them into the underlying control system configuration. The alarm server becomes a real-time decision-making addition to the control system.

The master alarm database (MADB)

A required element of the standards, the MADB is a controlled, evergreen list of all alarms, with significant details such as multiple alarm causes, consequences, corrective actions (both console operator and outside operator actions), reasons for setpoint and allowable response time. Operators can easily access alarm data, and the MADB enables automated checking for unauthorized changes to the alarm system.

The combination of process connectivity, computation and the MADB is a powerful one. Automation engineers in forward-looking corporations use these capabilities to do much more than alarm management. The first expanded use of an alarm management application is to automatically monitor, detect and report process excursions that violate pre-defined boundaries (i.e., boundary management).

Going beyond alarm management to boundary management

Typically, important safety, production and quality boundaries of a process reside in a hodgepodge of different procedures, databases and documents, on different servers and managed by different departments at a plant. The data in many of these systems is often contradictory, out of date or even lost. For example, control system interlock settings may not match the correct design documents; or, if the setpoint of an alarm is incorrectly set, it fails to warn the operator before the safety system trips the process.

A well-designed and integrated alarm management application can address this issue. A one-time effort is made to consolidate all boundary information in one place and to resolve the differences, leading to the creation of a “single version of the truth.” This is stored in a new section of the controlled MADB, and the sensor points related to the boundaries are identified and linked.

This augmented alarm management system continuously monitors the process against hundreds of boundaries. Real-time depictions appropriate for operators, engineers, managers and executives bring visibility and awareness to operational and safety risks in real time. Automated reports generate when a boundary is crossed, and the opportunity cost or lost value of each excursion is automatically calculated. Monitoring key performance indicators (KPIs) that identify the most frequent excursions by value, type and location present invaluable insight to operators and management, directing efforts to the best improvement opportunities.

Independent protection layer (IPL) monitoring

The success of boundary management has spawned a related improvement. Companies install independent safety instrumented systems (SIS) to mitigate process risks. Their design requires specialized knowledge. Like process boundaries, there is often not a “single source of the truth.” Design information exists in a variety of uncontrolled, disparate documents. The many different operating and maintenance procedures referring to the SIS have the potential for inconsistency with their design. Uncontrolled systems often accumulate errors.

Once an SIS is operational, complex record-keeping and design verification calculations should occur. To report that certain tasks are accomplished, plants use a variety of methods, most often inconsistent, error-prone and potentially unreliable. These include uncontrolled spreadsheets, notes, homegrown applications and manually marked-up drawings and sketches. These tasks can also be overlooked, as operations does not always have people with the knowledge to perform these specialized tasks. A controlled and automated solution to continuously monitor safety systems addresses these risks.

A new section of the MADB consolidates the correct settings for all SIS functions, which are then continuously monitored. Automated reports of every SIS activation show detailed performance of sub-components. Demand and operations rates are calculated and updated. Response times are calculated. Bypasses are monitored and shown. Relevant configuration changes are automatically reported to ensure proper management of change (MOC). When any of the IPLs are unavailable, out of service, overdue, bypassed or compromised, immediate reports and notifications are generated, and current risk profiles are recalculated.

Potential cost savings are possible for SIS testing. Activating an SIS function—if fully documented—can often be used as a full or partial test, allowing for reduced testing costs.

Risk visualization

All levels in the organization want succinct and accurate visuals of plant, region and fleet performance showing risk exposure. Control system, alarm system, boundary and safety system performance directly affect production, financial and safety performance. Continuous monitoring of these functions can supply KPI-based dashboards with aggregated and drill-down data. Dynamic changes in the operational risk level can be shown, based on combinations of factors.

Takeaway

The key to success is to use an alarm management application that integrates safety-critical functions in a single environment. Boundary management and IPL monitoring are examples of effectively using this infrastructure to solve plant problems. As a result, they are gaining widespread adoption, often as part of operational excellence initiatives. More ideas are surfacing and are taking the industry far beyond alarm management. HP

The Author

Related Articles

From the Archive

Comments

Comments

{{ error }}
{{ comment.comment.Name }} • {{ comment.timeAgo }}
{{ comment.comment.Text }}