Alarms API - ISA 18.2 Standard¶

The Alarms module implements the ISA 18.2 standard for alarm management in industrial automation systems. ISA 18.2 is an international standard published by the International Society of Automation (ISA) that provides best practices for the design, implementation, and management of alarm systems in process industries.

ISA 18.2 Overview¶

ISA 18.2 establishes a framework for effective alarm management throughout the alarm lifecycle, from philosophy development through monitoring and maintenance. The standard addresses common problems in alarm systems such as alarm flooding, nuisance alarms, and poor alarm response.

Key Principles of ISA 18.2¶

Alarm Philosophy: A documented set of principles governing the design, implementation, operation, and management of the alarm system
Alarm Lifecycle: A structured approach covering philosophy, identification, rationalization, detailed design, implementation, operation, maintenance, management of change, and monitoring/assessment
State Management: Standardized alarm states and transitions that ensure consistent behavior
Performance Metrics: Key performance indicators (KPIs) for alarm system effectiveness

Alarm State Machine Diagram¶

The following diagram illustrates the complete alarm state machine as defined by ISA 18.2 and implemented in PyAutomation:

stateDiagram-v2
    [*] --> Normal: Initial State

    Normal --> Unacknowledged: Alarm Condition Detected
    Normal --> Shelved: Shelve Action
    Normal --> SuppressedByDesign: Suppress by Design
    Normal --> OutOfService: Out of Service

    Unacknowledged --> Acknowledged: Acknowledge Action
    Unacknowledged --> RTNUnacknowledged: Process Returns to Normal
    Unacknowledged --> Shelved: Shelve Action
    Unacknowledged --> SuppressedByDesign: Suppress by Design
    Unacknowledged --> OutOfService: Out of Service

    Acknowledged --> Normal: Process Returns to Normal
    Acknowledged --> Shelved: Shelve Action
    Acknowledged --> SuppressedByDesign: Suppress by Design
    Acknowledged --> OutOfService: Out of Service

    RTNUnacknowledged --> Normal: Acknowledge RTN
    RTNUnacknowledged --> Unacknowledged: Alarm Condition Reoccurs
    RTNUnacknowledged --> Shelved: Shelve Action
    RTNUnacknowledged --> SuppressedByDesign: Suppress by Design
    RTNUnacknowledged --> OutOfService: Out of Service

    Shelved --> Normal: Reset (Process Normal)
    Shelved --> Unacknowledged: Reset (Alarm Active)

    SuppressedByDesign --> Normal: Unsuppress (Process Normal)
    SuppressedByDesign --> Unacknowledged: Unsuppress (Alarm Active)

    OutOfService --> Normal: Return to Service (Process Normal)
    OutOfService --> Unacknowledged: Return to Service (Alarm Active)

    Normal --> [*]: Alarm Deleted

Alarm States and Attributes¶

According to ISA 18.2, each alarm state has specific attributes that define its behavior:

1. Normal (NORM)¶

Attributes: - Process Condition: Normal - Alarm Status: Not Active - Annunciate Status: Not Annunciated - Acknowledge Status: Acknowledged (by default)

Description: The alarm is in its normal state, meaning the process variable is within acceptable limits and no alarm condition exists. This is the initial and default state for all alarms.

Available Actions: - Shelve - Suppress by Design - Out of Service - Disable

2. Unacknowledged (UNACK)¶

Attributes: - Process Condition: Abnormal - Alarm Status: Active - Annunciate Status: Annunciated - Acknowledge Status: Unacknowledged

Description: An alarm condition has been detected, but the operator has not yet acknowledged it. The alarm is actively annunciating (sounding, flashing, etc.) to alert the operator.

Available Actions: - Acknowledge - Shelve - Suppress by Design - Out of Service - Silence - Disable

Transition Triggers: - Automatically transitions from Normal when alarm condition is detected - Can transition from RTN Unacknowledged if alarm condition reoccurs

3. Acknowledged (ACKED)¶

Attributes: - Process Condition: Abnormal - Alarm Status: Active - Annunciate Status: Annunciated - Acknowledge Status: Acknowledged

Description: The operator has acknowledged the alarm, indicating awareness of the condition. The alarm remains active because the process condition is still abnormal, but the annunciation may be reduced (e.g., flashing instead of sounding).

Available Actions: - Shelve - Suppress by Design - Out of Service - Disable

Transition Triggers: - Transitions from Unacknowledged when operator acknowledges - Automatically transitions to Normal when process returns to normal

4. Return to Normal Unacknowledged (RTNUN)¶

Attributes: - Process Condition: Normal - Alarm Status: Not Active - Annunciate Status: Annunciated - Acknowledge Status: Unacknowledged

Description: The process has returned to normal, but the operator has not yet acknowledged the return to normal condition. This state ensures that operators are aware that an alarm condition existed and has now cleared.

Available Actions: - Shelve - Suppress by Design - Out of Service - Disable

Transition Triggers: - Automatically transitions from Unacknowledged when process returns to normal - Transitions to Normal when operator acknowledges RTN - Transitions back to Unacknowledged if alarm condition reoccurs

5. Shelved (SHLVD)¶

Attributes: - Process Condition: Normal (typically) - Alarm Status: Active or Not Active - Annunciate Status: Suppressed - Acknowledge Status: Not Applicable

Description: The alarm has been temporarily shelved (archived) by an operator, typically for maintenance or known conditions. Shelved alarms do not annunciate but remain in the system for tracking purposes.

Available Actions: - Reset (returns to Normal or Unacknowledged based on process condition)

Transition Triggers: - Can be entered from any active state (Normal, Unacknowledged, Acknowledged, RTN Unacknowledged) - Returns to appropriate state when reset based on current process condition

6. Suppressed By Design (DSUPR)¶

Attributes: - Process Condition: Normal (typically) - Alarm Status: Active or Not Active - Annunciate Status: Suppressed - Acknowledge Status: Not Applicable

Description: The alarm is suppressed by design, meaning it has been intentionally disabled as part of the system design (e.g., during startup sequences, known process conditions). This is different from shelving, which is a temporary operator action.

Available Actions: - Unsuppress by Design

Transition Triggers: - Can be entered from any state - Returns to appropriate state when unsuppressed based on current process condition

7. Out Of Service (OOSRV)¶

Attributes: - Process Condition: Normal (typically) - Alarm Status: Active or Not Active - Annunciate Status: Suppressed - Acknowledge Status: Not Applicable

Description: The alarm is out of service, typically when the associated equipment or process is not in operation. This state prevents false alarms during shutdown or maintenance periods.

Available Actions: - Return to Service

Transition Triggers: - Can be entered from any state - Returns to appropriate state when returned to service based on current process condition

State Transitions and Logic¶

Main Operational Flow¶

Normal → Unacknowledged: When the monitored process variable exceeds the alarm threshold
Unacknowledged → Acknowledged: Operator acknowledges the alarm
Acknowledged → Normal: Process variable returns to normal range
Unacknowledged → RTN Unacknowledged: Process returns to normal before acknowledgment
RTN Unacknowledged → Normal: Operator acknowledges the return to normal
RTN Unacknowledged → Unacknowledged: Alarm condition reoccurs before RTN acknowledgment

Suppression States Flow¶

All operational states (Normal, Unacknowledged, Acknowledged, RTN Unacknowledged) can transition to suppression states: - Shelved: Temporary operator action - Suppressed By Design: Design-based suppression - Out Of Service: Equipment/service not available

When reset from suppression states, the alarm returns to the appropriate operational state based on the current process condition: - If process is normal → Normal - If alarm condition exists → Unacknowledged

ISA 18.2 Compliance in PyAutomation¶

✅ Implemented Features¶

Standard Alarm States: All seven standard states from ISA 18.2 are implemented
Normal, Unacknowledged, Acknowledged, RTN Unacknowledged, Shelved, Suppressed By Design, Out Of Service
State Attributes: Each state correctly implements ISA 18.2 attributes:
Process Condition (Normal/Abnormal)
Alarm Status (Active/Not Active)
Annunciate Status (Annunciated/Not Annunciated/Suppressed)
Acknowledge Status
State Transitions: Complete state machine with all standard transitions
Alarm Types: Support for standard trigger types:
HIGH (High alarm)
LOW (Low alarm)
BOOL (Boolean alarm)
Alarm Configuration:
Deadband (hysteresis) to prevent chattering
On/Off delays for transient filtering
Setpoint configuration
Alarm History:
AlarmSummary model tracks alarm occurrences
Timestamp tracking for alarm and acknowledgment times
User Actions:
Acknowledge functionality
Shelve/Unshelve operations
State transitions with user tracking

⚠️ Partially Implemented Features¶

Alarm Philosophy Documentation:
Framework exists but may need enhanced documentation tools
Recommendation: Add API endpoints for alarm philosophy management
Alarm Rationalization:
Basic alarm creation exists
Recommendation: Add tools for systematic alarm review and rationalization
Performance Metrics:
Alarm history is tracked
Recommendation: Add KPIs such as:
- Average alarms per operator per hour
- Percentage of time in alarm
- Alarm response time
- Alarm chattering metrics

❌ Not Yet Implemented (Future Development)¶

Alarm Flooding Prevention:
Status: Not implemented
Description: Mechanisms to prevent overwhelming operators with too many alarms
ISA 18.2 Requirement: Maximum recommended alarms per operator per 10 minutes
Alarm Priority Management:
Status: Basic priority exists in database model
Description: Enhanced priority-based filtering and display
ISA 18.2 Requirement: Priority-based alarm handling
Alarm Grouping and Correlation:
Status: Not implemented
Description: Group related alarms to reduce operator cognitive load
ISA 18.2 Recommendation: Related alarm grouping
Dynamic Alarm Suppression:
Status: Basic suppression exists
Description: Advanced suppression based on process conditions or other alarms
ISA 18.2 Recommendation: Context-aware suppression
Alarm Response Procedures:
Status: Not implemented
Description: Link alarms to response procedures or operating instructions
ISA 18.2 Recommendation: Procedure-based alarm response
Alarm System Monitoring Dashboard:
Status: Not implemented
Description: Real-time dashboard showing alarm system health and performance
ISA 18.2 Requirement: Continuous monitoring and assessment
Alarm Audit Trail:
Status: Basic logging exists
Description: Comprehensive audit trail of all alarm state changes and user actions
ISA 18.2 Requirement: Complete auditability
Alarm Testing and Validation:
Status: Not implemented
Description: Tools for testing alarm logic and validating alarm system behavior
ISA 18.2 Requirement: Testing procedures

Components¶

Alarm: The main alarm class implementing the ISA 18.2 state machine.
States: Definitions of standard alarm states with ISA 18.2 attributes.
Triggers: Logic for detecting alarm conditions (High, Low, Boolean).

Usage¶

Alarms are typically created via the PyAutomation core instance, but they can be interacted with directly for advanced use cases.

# Example of creating an alarm via PyAutomation
app.create_alarm(
    name="HighPressure",
    tag="PressureTag",
    alarm_type="HIGH",
    trigger_value=100.0,
    description="Tank pressure too high"
)

# Acknowledge an alarm
alarm = app.get_alarm_by_name("HighPressure")
alarm.acknowledge(user=current_user)

# Shelve an alarm
alarm.shelve(user=current_user)

# Return from shelved state
alarm.unshelve(user=current_user)

Best Practices¶

When implementing alarms following ISA 18.2:

Develop an Alarm Philosophy: Document your alarm management principles before creating alarms
Rationalize Alarms: Review and justify each alarm - not every process deviation needs an alarm
Set Appropriate Priorities: Use priority levels to help operators focus on critical issues
Implement Deadbands: Use deadbands to prevent alarm chattering
Use Delays Appropriately: On/off delays help filter transient conditions
Monitor Performance: Regularly review alarm system performance metrics
Maintain Documentation: Keep alarm documentation up to date
Train Operators: Ensure operators understand alarm states and appropriate responses

References¶

ISA 18.2 Standard: "Management of Alarm Systems for the Process Industries"
ISA 18.2 Lifecycle: Philosophy → Identification → Rationalization → Design → Implementation → Operation → Maintenance → Management of Change → Monitoring