Managing alarms effectively

Many SCADA systems have built-in features that can be used to improve alarm handling, operator response times, and root cause analysis
By Fabio Terezinho, Indusoft May 20, 2013

Figure 1: Avoid alarm flooding by using an alarm management system to prioritize alarms. Courtesy: IndusoftHas managing alarms at your facility become a problem? Are operators missing critical alarms that could lead to unscheduled downtime or incidents because they’re inundated with too many low-priority alarms? Once alarms are acknowledged, does it take too long to find the root cause?

Conditions change continually in a manufacturing or process plant. Some changes are routine, such as when a motor stops or starts in response to normal operations. Others can be mission critical, such as a temperature rising above maximum limit. Operators are constantly monitoring conditions to prevent problems and achieve maximum efficiency and productivity.

Benefits of using an alarm management system include:

  • Ensure critical alarms get a timely response
  • Avoid operator fatigue from too many low-level alarms
  • Send real-time data to operators for immediate response
  • Get detailed information from alarm worksheets
  • Identify root causes quickly through alarm filters with string tags. 

The introduction of computer-based SCADA systems and the great technological leaps that followed have significantly reduced the cost of alarming, making it almost free to add an alarm point. Most plants are controlled by one or more automation systems, and these systems provide a wealth of data to the SCADA system, data that can be used to monitor nearly every aspect of a plant’s operation. With modern automation and SCADA systems, alarms can be added to a screen without incurring the expense of adding a hardwired alarm point and a panel-mounted indicating device.

While nearly free alarms can provide great benefits, care must be taken in SCADA system design each time an alarm point is added. For example, adding too many alarms can create alarm fatigue for operators who may be distracted by low-priority items at the expense of identifying the root cause problem.

A SCADA system should clearly show appropriate alarms indicating possible problems, but it shouldn’t cry wolf for every routine change in plant operations. It’s easy to blame the SCADA system for alarm overload and other alarm-related issues, but it’s only a tool. Plant personnel need to use this tool correctly by selecting and implementing systems that avoid alarm overload while making sure critical events are addressed in a timely fashion.

Creating an alarm management system

The key to preventing alarm flooding from distracting operators is a proper design of the SCADA system, specifically its alarm handling features. A well-designed system will prioritize alarms to prevent serious alarms from being overlooked as a result of an alarm overload involving a large number of low-priority alarms. It will also offer other advantages, such as the ability to more easily diagnose the root cause of a problem.

The first step in designing a SCADA alarm system is developing a comprehensive plan that determines alarm criteria. This design step should be accomplished with a team consisting of plant operators, plant engineers, and others who have a deep understanding of how the plant actually works. This team will determine and document:

  • How alarms should be prioritized
  • What type of operator intervention is required for each alarm
  • How much time should be allowed for an operator to respond to an alarm
  • Operator alarm handling and escalation procedures
  • How to structure the system to handle various routine alarms. 

The importance of alarm limits in configuration

Figure 2: Operators can get immediate real-time data on a variety of devices, such as Smartphones and other handheld devices. Courtesy: IndusoftMost modern SCADA systems are PC-based, and the selected system should include built-in features to allow low-cost and straightforward implementation of an effective alarm management regime. A properly designed SCADA-based alarm management system should help operators distinguish between high-priority and low-priority alarms, respond properly to each alarm, and quickly determine alarm root cause. The alarm management system should contain tools to implement these features—including, but not limited to, configurable alarm limits, alarm displays, and most importantly, filters.

When selecting a new SCADA system it’s important to look for features that will make it possible to prioritize alarms and messages by setting limits and providing sorting capabilities. Creating worksheets with alarm limits is one of the best ways to begin the alarm management process.  Alarm limits define high and low values along with deviation set points, creating betting alarm management by helping operators to quickly determine the alarm priority with a glance at the screen.

These alarm limits can be properties of tags, and shouldn’t count against the total number of tags in the SCADA system. This is important as SCADA software is often priced based on total tags.

For example, a tag can be created called “Temperature” that has a setpoint value of 100 degrees. The HiLimit tag property might be set at 110 degrees, and the LoLimit tag property might be set at 90 degrees.

When creating alarm limits, some or all of the following items are typically included:

  • HiLimit, the first high alarm point reached on an increasing value
  • HiHiLimit, the second high alarm point reached on an increasing value, indicating an escalating condition
  • LoLimit, the first low alarm point reached on a decreasing value
  • LoLoLimit, the second low alarm point reached on a decreasing value, indicating an escalating condition
  • DevSetpoint (Deviation), the allowable deviation or variation above and below the desired setpoint
  • Dev+Limit, the allowable deviation plus the alarm limit
  • Dev-Limit, the allowable deviation minus the alarm limit
  • DeviationDeadband, the range through which a deviation can occur without initiating an alarm.

There are many ways to set alarm limits and display them. Alarm limits can be set for each of the items listed above, and more. Alarm limits can be displayed on SCADA screens using numerical values, or via graphical representations such as sliders. With many SCADA systems, alarm limits can also be saved in separate worksheets with tag names, making it easy to view and adjust settings as necessary.

Alarm worksheets set a good foundation

Alarm worksheets, also known as configuration sheets or pages, are used to configure the setpoints and limits for the tag names, such as a HiLimit of 110 degrees for the “Temperature” tag with a nominal setpoint of 100 degrees. By spending time to set the alarm limits and deviations correctly, fewer nuisance alarms will be generated.

Figure 3: By filling in one field in the Filters Dialog, the user can either trigger the same dialog during runtime, or change the dialog in response to operating conditions. Courtesy: Indusoft

The alarm worksheets also contain message sections that correlate to the alarm, the priority, and the selection field—all to provide the operator with more information on the alarm. For example, for a HiLimit temperature alarm, within the worksheet users can set a message (Too Hot), a priority (0-255), and a “selection” that is a user-defined field such as “Tank2.”

All the items in the alarm worksheets can be set as well as changed. Later on, we’ll see how by including values in the messages, the worksheets enable easy filtering.

Alerting operators 

The information on the alarm worksheets should appear on the screen as an alarm object. Once imported into the SCADA system, the worksheets should provide functionality to automatically e-mail alarm messages. Furthermore, the SCADA system should provide settings to control exactly when and how these alarm messages are sent. Some common settings are frequency, number of alarms, message format, response times, and escalation procedures.

Typically, the SCADA system default setting shows all active alarms, but these settings can be changed to only show certain alarms. The ability to have the “online alarm object” as the default setting, meaning it doesn’t need to pull information from a database, is vital. This is important when operators only need to view current alarms as opposed to all active alarms over a longer time period 

The SCADA system should also provide functionality to allow alarm data to be pulled from a database to get a view over a selected period of time, as this is often useful for troubleshooting by maintenance personnel and others. Characteristically, this functionality permits viewing of all alarms that have been logged. With many SCADA systems, alarms can be logged using a built-in PDF writer that turns alarms into a PDF document. This document can then be saved, e-mailed as an attachment, or printed as a report.

Using filters to manage alarms 

Filters are one of the most important features for preventing alarm overload and insufficient prioritization of alarms. Most modern SCADA systems provide easy configuration of features that offer a wide variety of sorting and filtering mechanisms.

One of the biggest alarm management issues is that operators are inundated with low-priority alarms and therefore miss critical high-priority alarms through alarm fatigue and the inability to quickly distinguish an alarm’s urgency. This is where a SCADA system with superior filtering mechanisms becomes highly valuable.

The first step in filtering was discussed in the worksheet section in which information can be prioritized to have only certain alarm limits appear. For example, the screen might display only Priority 10 and above alarms, or only the HiHiLimit alarms.

In addition to setting alarm priorities, plant personnel should consider SCADA systems that offer a filter dialog that includes an expression to let users customize the filter. For example, only HiHiLimit alarms that have the word “Hot” in the message are displayed on the screen in response to certain conditions. This expression can be hard coded, but to increase effectiveness it’s often better to create a string tag that changes values during runtime depending on specific operating conditions.

When an incident occurs, hundreds of related alarm messages can be quickly generated. Usually, only one alarm is the root cause, and the rest are symptoms of the problem. This is where many SCADA systems fail because the operators have no way to distinguish the root cause from the symptoms. But, the ability to build strings based on key criteria can help operators locate the root cause of the problem quickly.

For example, temperature alarms may start appearing with messages saying, “Tank 1 temperature too high” or “Pipe 1 pressure too low.” Operators must now quickly determine the root cause of why the temperature in Tank 1 is too high. The ability to filter all messages with a single word such as “hot,” or with a phrase like “valve closed,” can help them discover why the tank temperature is soaring. Wild card expressions can also be used to sort through words or individual phrases in the messages.

Unfortunately, many operators and even some plant managers have found managing alarms so burdensome they have almost considered them nuisances. This attitude can lead to important alarms being missed or ignored, a dangerous situation. It doesn’t have to be this way, as alarms are a key part in monitoring plant operations and can be managed effectively using off-the-shelf software tools.
One of the keys to effective alarm management is selecting the right SCADA system, one that offers advanced and easily configurable alarm management features. These features should include alarm prioritization to alert operators of conditions that require quick responses. Filtering features are also needed to enable operators to quickly locate root causes.

Implementing these and other features of modern SCADA systems as part of an effective alarm management program will improve response to alarms. This will minimize downtime, increase safety, and keep minor alarms from escalating into major incidents.

Fabio Terezinho is vice president of consulting services and product manager for InduSoft. www.InduSoft.com.