1. Get Control of Your Alarms
David Anderson
Logic, Inc.

2. Agenda Alarm Presentation and Management Challenges
Alarm Standards and Recommendations
Alarms in Wonderware System Platform
Wonderware Alarm Adviser

3. Alarm Presentation and Management Challenges

4. Two Challenges Alarm Proliferation Alarm Display
Two challenges that we’ll cover today, anyway

6. Why did the number of alarms increase?
More information per sensor/actuator
Easily configurable: just a checkbox
Many options: Hi, HiHi, Lo, LoLo, Deviation, Rate of Change, …
Alarm all the things: don’t want to miss anything

7. Operator Overload

8. Operator Overload Over-acknowledgement Inhibiting “noisy alarms”
Happy management
Quieter room
Inhibiting “noisy alarms”
Missing important alarms
Not being able to mitigate in proper time

9. Alarm Mitigation Takes Time
Detect the alarm
Change screens to get context from process
Verify alarm is valid
Acknowledge it, silence it
Analyze, consult, decide proper action
Act, may require leaving the operator console
Continue to monitor to verify action’s result

10. Alarm Displays Too much color
Same color used to convey different information
Too many animations
Too cluttered

13. Impact of Human Error Abnormal Situation
A disturbance or series of disturbances in a process that cause plant operations to deviate from their normal operating state.
 The average percentages shown had the following:
People and work Context Factors: 35% - 58%
Equipment Factors: 30% - 45%
Process Factors: 3% - 35%
Source: ASM Consortium

14. Economical Impact Reduces the operational effectiveness
Economical impact: Unnecessary plant shutdowns (in the USA alone this costs $20 Billion a year on productivity)
Poor alarm management also causes
losses in product quality
danger to population and environment
image loss of a respective company
Source : ASM Consortium
Abnormal Situation Management

15. Safety Impact Piper Alpha North sea 1988 Bunkfield Oil Depot
Texaco Pembroke 1994

16. Texaco Pembroke
“Control panel graphics did not provide necessary process overviews.”
“Warnings of the developing problems were lost in the plethora of instrument alarms…”
“Too many poorly categorized alarms overwhelmed the operators…”
“There was no alarm philosophy for determining what priority an alarm should have and no control was exercised over the number of alarms in the system.”
Refinery in the UK. 26 people injured due to large explosion on July 24, Broke windows up to 10 miles away.
Source:

17. Alarm Standards and Recommendations

18. A Set of Standards and Guidelines
ANSI/ISA-18.2, Management of alarm systems for the process industries
EEMUA 191, Alarm systems a guide to design
Namur NA 102 Worksheet, Alarm Management
NPD YA 711, Principles for alarm design (Norwegian petroleum doctorate slowly adopted throughout Europe as the standard)
VDI/VDE Guideline 3699 (process control using monitors)
API RP-1167 Alarm Management for Pipeline Systems
ANSI/ISA-18.2 Management of Alarm Systems for the Process Industries
API RP-1167 Alarm Management For Pipeline Systems
EEMUA = Engineering Equipment and Material Users’ Association

19. What Is An Alarm?
From ISA-18.2: An alarm is “an audible and/or visible means of indicating to the operator an equipment malfunction, process deviation, or abnormal condition requiring a response.”

20. What Is An Alarm?
Alarms must activate only based on truly abnormal conditions, not expected cases of operation
Alarms must require an operator response
Multiple alarms should not signify the same thing

21. Some Guidance from the Standards
Use broad groups of alarm priorities (severities)
Critical, High, Medium, Low
Use text, color, and shape to indicate alarms
Alarm shelving
State-based alarms
Alarm logging

22. Alarms in Wonderware System Platform

24. Situational Awareness HMI and Alarms Combined
Traditional HMI
Impact
What Happened ?
Critical
Alarm
Grid
Tool
Process
Trends
Tool
Knowledge
Operator
Operational
Limits
SA Graphics
What is Happening ?
Knowledge
Operator
Alarm
Boundaries
Interpretation time
Alarm
Time
- 40 %

25. Alarm Border Animation

26. Global Priority to Severity mapping
One location to change and customizable image…
Can also map severities in InTouch Modern Apps
Default Alarm Border Icons

27. Alarm Styles

28. Alarm Border Animation
Runtime
Global Icons
Global Styles
Auto Configuration for attributes or objects

29. Update Clients - Runtime
Tabbed filtering
Actual alarm indicators on Tabs
Ack buttons
And styles and themes setup as default

30. Widgets for Alarming Area Indicator Object or Device Indicator
Nav Button Indicator

31. Examples

39. Alarm Shelving
“Shelving” an alarm is a manual initiation by an operator to temporarily silence an enabled alarm for a defined period of time after which normal operation is restored.
Not available in InTouch Modern Apps. May look like it is, but it’s not.

40. Alarms Shelving Features
Who Can Shelve?
Shelving from Alarm Client or scripting
Any configured alarm can be shelved
Only enabled alarms can be shelved
Mandatory reason and duration
Audit trail logged to Historian
What can be shelved?
Alarm Border Integration

41. Shelving Audit Trail Logged to Historian

42. Alarms Plant State Based Suppression
Global definition of plant states
Area object based suppression of alarms
Individual state on the area object has a I/O Extension

43. Plant State Example Plant area Two child areas Two grandchild areas
Each area has object with an alarm

44. All Areas Running (Enabled)

45. Grandchild Area Shutdown (Disabled)

46. Parent Area Shutdown (Disabled)

47. Grandchild Area Undeployed

48. Alarm & Event History Database Automatically created by Configurator
Supports mixed & Windows-only security
Includes Alarm DB Purge/Archive Utility
Logging
All Areas deployed to the Engine
Selective based on severity/ event type
Silenced alarms still logged
Based on Engine settings
Licensing: No tag count required for alarms
A2ALMDB
HCAP
WCF
HCAL
Application Server *
* No change for InTouch Alarms/Events

49. Alarm & Event History Super Simple:
1. Enable Historization in Engine… Already doing that!
2. Adjust Priority to Severity Mappings
The old process had 17 – 21 error prone steps.

50. Storage Robustness with Store Forward
Support for:
Redundant Engines
Store forward
Redundant Historians
No configuration required

51. System Platform Recap Situational Awareness Graphics
Enhanced Alarm Client
Alarm Shelving
Plant State
Easy-to-Configure Robust Alarm Logging

52. Wonderware Alarm Adviser

53. A Set of Standards and Guidelines
ANSI/ISA-18.2, Management of alarm systems for the process industries
EEMUA 191, Alarm systems a guide to design
Namur NA 102 Worksheet, Alarm Management
NPD YA 711, Principles for alarm design (Norwegian petroleum doctorate slowly adopted throughout Europe as the standard)
VDI/VDE Guideline 3699 (process control using monitors)
API RP-1167 Alarm Management for Pipeline Systems
ANSI/ISA-18.2 Management of Alarm Systems for the Process Industries
API RP-1167 Alarm Management For Pipeline Systems
EEMUA = Engineering Equipment and Material Users’ Association

54. Alarm Performance Levels
EEMUA = Engineering Equipment and Materials User Association
Alarm Management Handbook disagrees (page 100). For example, average of 60 alarms per hour with peaks of almost 6000 hour is considered stable above, but the book says this would be “very poor alarm performance, and those numbers are quite inappropriate.”
From the Engineering Equipment and Materials User Association (EEMUA)

55. Alarm Distribution and Frequency
Alarm Priority
% Total Alarms
Max Frequency
Critical
< 1%
Very infrequently
High
~ 5%
< 5 per shift
Medium
~ 15%
< 2 per hour
Low
~ 80%
< 10 per hour
These target percentages are for alarm occurrences, not configuration.
The Max Frequency might be too high. It is not generally satisfactory to have 5 high severity alarms in a single shift.
Actually from EEMUA 191, included in the Handbook to point out that it is problematic to actually target these frequencies. “…if alarms have been properly rationalized…then what you get is what you get.” “In fact, in considering the conditions that should require a P1 [High] alarm, most companies would be extremely upset if such conditions occurred even once a shift!”

56. DMAIC ANSI/ISA-18.2 offers a lifecycle model
Define an alarm improvement program
Measure the current situation
Analyze the areas for improvement
Improve the situation
Control and hold the gain with metrics
CLICK for D M A I C
ISA-18.2 Lifecycle model for alarm management.
New site would, or should, start with alarm philosophy. In most cases this is not true, systems have evolved to where they are today with a lack of standards.
Define (Audit): Starting point is an initial audit, or benchmark, of all aspects of alarm management against a set of documented practices, such as those listed in ISA18.2. The results of the initial audit can be used in the development of a philosophy.
Measure: Use Alarm Adviser to create a benchmark of the current system
Analyse: Identify bad actors (areas of improvement), use frequent, standing, fleeting and consequential alarms in Alarm Adviser
Improve: the situation. Ref: Rob Kambach and Alarm Improvements in System Platform Use alarm management tools such as Situation Awareness, Alarm suppression, Alarm Aggregation and the tuning of alarm setpoints.
Control: Set KPI’s and continually measure performance to ensure that the alarm system does not degrade. Define KPI’s and use dashboards in Alarm Adviser to maintain performance

57. It Can Be Done Spectra Energy Control Room 500,000 points
Categorized and rationalized
Now average 6 alarms/hour/operator

58. Wonderware Alarm Adviser
Wonderware Alarm Adviser is a web based tool for discovering nuisance alarms in your process system through interactive visual analysis against standard KPIs
Total, frequent, standing, fleeting and consequential views allow nuisance alarms to be easily identified
Dashboards make it possible to benchmark and maintain your alarm performance in line with industry standards

59. Based on Standard KPIs
Average alarm rates (per day, per hour, per 10 minutes)
Peak alarm rates
Alarm floods
Frequently occurring alarms
Fleeting alarms
Standing alarms
Alarm severity distribution as annunciated

60. Alarm Adviser Walkthrough

61. Dashboard – User Defined KPIs

62. Dashboard

63. Alarm Activity - Time Range

64. Alarm Activity – Severity Distribution

65. Alarm Activity and Filtering

66. Alarm Activity and Filtering

67. Frequent Alarms

68. Frequent Alarms – Detail of Selected

69. Long Standing Alarms – Most Frequent

70. Fleeting Alarms – Most Frequent

71. Consequence/Cascading

72. Consequence/Cascading

73. Architectures / Licenses
Alarm Adviser Demo Mode
Single Node One Collector
Small Systems
Alarm Adviser Standard
1DB, 5000 Analysis points
Single Node One Collector
Small Systems
1DB, 1Milion Analysis points
Alarm Adviser Professional
Single Node Multiple Collectors
Medium Systems
5DB’s, 10Milion Analysis points
Alarm Adviser Premium
Distributed Architecture Multiple Collectors
Large Systems
10DB’s, Unlimited Analysis points
All - Unlimited Clients

75. Wrap Up

76. A Few Resources The Alarm Management Handbook
The High Performance HMI Handbook
Just go to Amazon and search for them
Situational Awareness Actionable Alarm Management
Situational Awareness Actionable Alarming Revisited
Alarm Management Best Practices for Safer Plant Operations
practices.html

77. Thank You