UK HQ Your time

Prioritisation & Classification

Alarm Management & Escalation — in depth

In an upset, priority is everything. Each alarm is assigned a priority — typically low, high and critical/emergency — from the severity of its consequence and the time available to act, with a sensible distribution so that critical alarms are rare and unmistakable, and the operator is never faced with a wall of equal-priority noise.

Prioritisation

What matters in practice

Consequence-Based

Severity of the outcome sets priority.

Response Time

Time available to act considered.

Priority Levels

Low / high / critical distribution.

Sensible Mix

Critical alarms kept rare.

Priority Levels

PriorityMeaningTarget share
LowAwareness~80%
HighPrompt action~15%
CriticalImmediate~5%
BasisConsequence/timePer ISA-18.2

Related Topics

Continue across this series

Talk to our engineers

Reynolds & Bauhm designs and delivers alarm management & escalation solutions backed by process engineering and performance guarantees.

Prioritisation & Classification: Engineering Detail

Fundamentals, design drivers and practical guidance

Alarm prioritisation — classifying each alarm by consequence and the time available to respond, so operators always know what to tackle first.

Key performance metrics benchmark the system against EEMUA 191 targets: an average of fewer than one alarm per operator every ten minutes in steady state, fewer than ten alarms in the ten minutes following a major upset, and standing/stale alarms driven to near zero. Chattering and fleeting alarms are eliminated with deadbands, on/off delays and hysteresis, while alarm shelving and state-based suppression keep the displayed set relevant to the current plant mode.

Reynolds & Bauhm rationalises alarm databases against a written philosophy, prioritises by consequence severity and operator response time, and tunes setpoints, deadbands and delays to suppress nuisance activations. We deliver the master alarm database, priority distribution, suppression logic and a monitoring regime so the system stays compliant through plant changes rather than degrading back into a flood.

Effective alarm management is governed by ANSI/ISA-18.2 and EEMUA 191, which define an alarm as an audible and/or visible means of indicating to the operator an equipment malfunction, process deviation or abnormal condition requiring a timely response. The discipline exists because poorly configured systems flood operators with nuisance and standing alarms, masking the genuinely safety-critical events during an upset — the failure mode implicated in several major process incidents.

Design & Specification Considerations

What our engineers assess on every scope of this type

  • Deadbands, on-delays and hysteresis to kill chattering alarms
  • Priority distribution targets (e.g. ~80% low / 15% medium / 5% high)
  • State-based and mode-based alarm suppression logic
  • Master alarm database with unique operator action per alarm
  • EEMUA 191 KPI monitoring: rate, standing, stale, flood
  • Management-of-change control so the database stays rationalised
ParameterTypical basisWhy it matters
StandardISA-18.2 / EEMUA 191Defines lifecycle, metrics and good practice
Steady-state rate<1 alarm / 10 min / operatorKeeps the operator able to respond, not react
Post-upset flood<10 alarms / 10 minLimits the burst after a major trip
Priority split~80 / 15 / 5 (L/M/H)A healthy distribution avoids priority inflation
Standing alarmsNear zeroStanding alarms desensitise operators
Operator actionOne clear, unique actionAn alarm with no action is removed

Frequently Asked Questions

Common questions on alarm management

How do you reduce nuisance alarms?

Chattering and fleeting alarms are removed with deadbands, on/off time delays and hysteresis; duplicate and consequential alarms are suppressed with state-based logic; and standing/stale alarms are reviewed out. The aim is the EEMUA 191 target of fewer than one alarm per operator per ten minutes.

What is alarm rationalisation?

Rationalisation is the structured review where each candidate alarm is tested against the philosophy: does it indicate an abnormal condition, is there a unique operator action, and what is the consequence and time available to respond? Only alarms that pass are retained, each with a justified priority.

How is Prioritisation & Classification kept healthy over time?

Through monitoring and management of change. Alarm rate, standing, stale and flood metrics are trended; any new or modified alarm passes through rationalisation before implementation, so the rationalised state does not erode as the plant evolves.

What is the difference between an alarm and an alert?

An alarm requires a timely operator response to a defined action; an alert is for information only. Conflating the two is a common cause of overload, so the rationalisation process re-classes action-less alarms as alerts or removes them.

Industries We Serve

Our expertise spans multiple industries with sector-specific water treatment solutions.

Related Pages

Explore closely-related topics, equipment and guides