Alarm Management 101

Omniflex has more than 40 Years experience in the design and manufacture of annunciators and has been instrumental in development from relay logic to microprocessor controlled SIL Rated Annunciators.

Alarm definition

An alarm is classified as an event to which an operator must react, respond and acknowledge (not simply acknowledge and ignore) and no plant should have more than 6 such alarm occurrences an hour. [EEMUA 191 guidelines]


If you have more than that you got it wrong!


Alarm Objective

The objective of an alarm system is to minimize or prevent physical and economic loss through operator intervention.


Note Safety takes precedence over economics


1         Introduction

The aim of an alarm system is to prevent, or at the very least minimize, physical and economic loss to plant or people through operator intervention in response to a plant condition that has occurred thus making the necessary corrective action optimizing production.

Alarm management has new meaning in major process plant disasters of recent years, such as Buntsfield, Chernobyl, BP Texas City, Bhopal and 3 Mile Island to name but a few.

Collaborative work by large process companies, vendors and industry organisations like the Electrical Equipment Manufacturers and Users Association (EEMUA) and the Instrumentation Society of America (ISA); accepting the need for an alarm management strategies; and applying new technology and standards to the process, companies can achieve multiple benefits. These include better productivity, quality improvement,  a more motivated and responsible staff, improved profitability and efficiency while minimizing the potential liability of management.


2         The Alarm System where does it fit in on a plant

The three layer plant model is the ideal for plant control, efficiency and safety however in practice and for economic reasons this is blended into one single system. For example a small plant may have PLCs doing control, Alarms and Shutdown are added into the same devices. At the same time the risk to the plant and personnel rises.


3         The Driving Forces

The driving forces can be classified as forced or voluntary. Forced motivation may be compliance (or regulatory) or risk mitigation. Voluntary motivation is that which is driven by perceived business benefits, profit, quality, efficiency. Alarm management strategy is driven by both forced and voluntary forces.

3.1      Regulatory Tools and Standards

IEC61508 Safety Instrumented Systems

EEMUA Guide Engineering Equipment and Material Users Association Publication 191

ISA standards and practices committee ISA SP18-02




4         What is an alarm?

The generally accepted definition of an alarm is: "An alarm is an event to which an operator must react, respond and acknowledge". Therefore the purpose of an alarm system is to alert an operator(s) to a potential problem, that if not addressed will cause some production, process, quality or safety compromise. To put an economic slant on this: the alarm system is there to prevent, or at least minimize, physical and economic loss through operator intervention.


5         What governs the operators response?

General Human Psychology

• As a rule, people do not process all available information, but rather generally look for patterns in the world.

• Once a pattern is recognized, then it becomes "reality".

Operator Intervention

• As soon as a pattern is selected, focus will shift from understanding the upset to monitoring intervention activities.

• Focus will not shift back to a new search for the root cause until it becomes clear that the intervention is not improving the situation or several new alarms do not match the pattern selected.


6         Current Situation

Studies have shown that there are several common problems experienced in alarm systems:

• There are too many alarms

• Alarms are ignored by operators because they associate them as a nuisance

• Alarm lists are overcrowded with data not alarms

• Operators perform mass alarm acknowledgments I.e. cannot cope

• Alarm systems fail to provide real insight to support operator decision making



Rationalise and Process

• Think of the HMI (Human Machine Interface) as the funnel through which data flows to the Operator, processed it becomes information

• And the alarm system and alarm priority is a tool to direct the operators attention as needed.

• What you do not show is as important as what you do!


6.1      Too many alarms

Most SCADA and DCS installations are configured creating too many alarms.. According to the EEMUA 191 guidelines, an alarm is classified as an event to which an operator must react, respond and acknowledge and no plant should have more than 6 such alarm occurrences an hour.

How many installations have more than this? Most do!

6.2      Alarms get ignored

Most alarms get ignored by operators because so many are either inconsequential and/or irritating. Data information only

Operators miss or ignore genuine alarms when overwhelmed by trivial information.

6.3      Mass alarm acknowledgement

Alarms get globally acknowledged without review.

6.4      Lack of real information

Masses of data is nothing without some form of analysis and rationalization.

7         Key Performance Indicators (KPIs)

Two good reference standards used currently are EEMUA 191 and ISA 18.02. alarm system and they focus on the following:

Four core principles of the guide are:

• Alarm system should be designed to meet user needs

• The alarm system purpose is to protect the safety of people, the environment and the plant equipment

• The performance of the alarm system should be assessed

• Every alarm is justified and properly engineered


EEMUA System metrics:

• 3 alarms/control valve

• 1 alarm/analog measurement

• 0.4 alarms/digital measurement

• 10% high - 20% medium - 70 % low priority alarm distribution


EEMUA Usability metrics:

• average alarm rate <1/10 minutes

• <10 alarms in 10 minutes after upset (peak rate)

• <10 standing alarms

• <30 disabled or inhibited alarms


The ISA Alarm Management requires (extract)

• Purpose of alarm system,

• Motivation for improving the alarm system,

• Definitions,

• References,

• Responsibilities for alarm management,

• Alarm management work processes,

• Operator responsibilities related to alarms,

• Alarm design principles,

• Alarm presentation,

• Priority assignment method,

• Alarm system performance targets,

• Alarm system maintenance,

• Advanced alarm management techniques,

• Required alarm documentation,

• Periodic testing of alarms,

• Alarm system's relationship to other site procedures,

• Implementation,

• Management of change,

• Training,

• Alarm history preservation,

• Recommended process for developing and maintaining the philosophy.




Bench Mark against the best in practice


8         Alarm strategy

To get started, at a minimum, you need:

• A philosophy that sets the basic concepts.

• It can be a working draft!

• Lots of good resources from EEMUA and ISA.

• A project plan:

• Goals,

• Metrics (with a base case),

• Schedule,

• Budget,

• Resources.


9         The Role of the Annunciator

The traditional Alarm Annunciator has enjoyed a revival of its status in Industry over the last few years having been deposed by PLC and SCADA system implementations over many years. Strings of spectacular and catastrophic failures as a result of the lack of clearly defined alarm strategies has redefined the thinking of plant designers worldwide. Annunciator implementation forces the designer at design stage to think about the alarm strategy for the system not at implementation where SCADA is configured. This eliminates the unnecessary alarms being added. Only those deemed to be necessary to plant function and safety are designed in.


A typical 16 Point Annunciator from Omniflex

10   The Role of an Event Recorder

Event Recorders have always been specialized equipment but there is no other way to tell for sure what happened at the last plant trip in sequence (down to better than 1millisecond resolution) The Control PLC cannot provide this nor can the SCADA or DCS they just do not have the speed to collect this information. Sequence of Events recorders can vary from 32 Points to 4800 Points. This Data can be presented in a PC based Management package or integrated directly into the SCADA system database.

Typical Rack Based Events recorder from Omniflex

Omniflex Automation Products - Solutions By Design 

Back to PR Listing

For further information on this press release please contact