Root Cause Incident Investigation IntroOverview

1
Root Cause Incident InvestigationIntro-Overview

• Society of Petroleum Engineers
• June 2005
• Jack Philley

2
A practical definition of root cause per CCPS

• A fundamental, underlying,
• system related
• reason that the event occurred
• and that you have the capability to fix.

3
Root Cause Investigation Concept

• AN EVENT CANNOT BE A ROOT CAUSE

Events are the results (consequences) that follow
root causes. If you identify an event as a root
cause, you have stopped too soon in your
investigation.
4
Useful Definition Management System

• The entire formal administrative system designed
  to accomplish an identified activity,
• along with
• associated procedures, training, resources,
  specific tasks, assignment of authority,
  documentation, monitoring, accommodations for
  changes and deviations, follow-up, and continuous
  improvement components.

5
MANAGEMENT SYSTEM TEMPLATE
Roles/Responsibilities
Performance Assessment
WORKPROCESSDESCRIPTION

• Inputs
• Suppliers

• Outputs
• Customers

Skills/Knowledge
Resources/Funding
Documentation
Goal/Objective
6
Management Systems

• Someone welded a nozzle onto a 55 gal drum.
• What management systems might be involved in
  this situation ?

7
Root Cause Investigation Concept

• Most root causes are associated with
• weaknesses, defects, or breakdowns in
• Management Systems

8
Major Chemical Plant ExplosionEPA Investigation
Published Findings

• Valve unsuitable for service application
• Prior lessons not applied
• Inadequately conducted PHA
• Mechanical Integrity measures inadequate
• Defective operating procedures
• Note Multiple causes, each with ability to be
  traced back to an associated management system

9
Bhopal Incident

• Multiple causes
• Associated with breakdowns in management systems
• 40 tons MIC
• Water Contamination
• Water Reactive

• ? PHA credible reliability of safeguards, LOPA
• ? Emergency Preparedness
• ? Alarms
• ? Flare
• ? Waterspray
• ? Refrigeration System
• ? Scrubber
• ? Resolution of Audit Findings
• ? MOC staffing, training

10
Clarification ?

• Root Cause Incident Investigation
• is not the same thing as
• Root Cause Analysis
• Root Cause Incident Investigation is a
  comprehensive and systematic investigation
  approach that uses Root Cause Identification and
  Root Cause Analysis concepts and techniques as
  integral components of the investigation.

11
Chronic Challenges

• Stopping too soon - multiple root causes
• Stopping too soon - mistaking a symptom for a
  root cause
• The Failure to Follow Procedure root cause trap
• Mistaking an event for a root cause
• Not reaching the system defect level in the
  investigation

12
ROOT CAUSE INVESTIGATION PROCESS
Accident Occurs
Event Occurs
What?
Assess the Situationand EstablishControl Over
theAccident Site
CollectInformation
AnalyzeInformation
IdentifyCritical Events
Search for RootCauses
Why?
Prevent!
Identify Findings andRecommendations
ReportResults
Follow-Up onImplementation
13
Root Cause Concept

• ROOT CAUSE INCIDENT INVESTIGATION IS AN
• Itertiv,
• Iteritave,
• ITERATIVE PROCESS

14
Investigation Activities

• Initial reconnoiter
• Temporary Hazard Control Measures
• Documenting the scene
• Photograph specific evidence
• Witness Interviews
• Physical Evidence Gathering document before
  moving, tag, log, bag
• Information and data gathering
• Liaison with other stakeholders

15
Root Cause Toolbox

• Critical Thinking Skills
• Incident Cause Models
• Time-Line
• Logic Diagrams
• Iterative Loop
• Fact-Hypothesis Matrix
• Stopping Guidelines
• Human Factors Considerations
• Procuring Information from Witnesses

16
Time-line Tool

• Helps identify critical events associated with
  root causes
• Iterative process
• Includes several types of information
• Can clarify apparent conflicts
• Helps identify relationships between events and
  information

17
EXAMPLE TIME-LINE
18
Root Cause Tool LOGIC DIAGRAMS

• Relates information
• Focuses team activity for evidence gathering to
  confirm or refute
• Iterative Process
• Quality assurance in later stages of
  investigation
• Useful in reports

19
or
or
?
?
or
Other
?
Overpressure?
Temp Failure?
20
PARTIAL ROOT CAUSE ANALYSIS DIAGRAM
Hand Carried Oil ContainerLess Than Adequate
(LTA)
Wrong StyleContainer
Proper Type ContainerBut Defective
Proper Type ContainerBut Used Incorrectly
Conscious Decision toUse Wrong Type
These two branches would bedeveloped separately
elsewhere
Use LTA
ProcurementLTA
PotentialRoot CauseLevel
System for inspection
System for repair/replacement
System for defining and recognizing defects
Others
System for establishing and monitoring purchasing
specifications
System for confirming/verifying the vendor
supplied what was ordered
System for managing change
Others
System for ensuring the employee knows (training
and written instructions) which style container
to use
System for marking labeling/identifying multiple
style/contents
System for monitoring actual practices in the
field
Others
21
Stopping GuidelinesorHow do I know I have gone
deep enough ?

• Pipe failed due to
• ?
• corrosion due to
• ?
• wrong pipe material due to
• ?
• LTA instructions for what material to use in
  sulfuric acid service due to
• ?
• no formal, consistent, or effective system for
  updating repair/replacement parts

22
Human Performance Shaping Factors

• Time available to complete the task
• Perceived Threat Level
• Experience Level
• Knowledge Skill Set (competencies)
• Personal Factors
• Procedure System
• Organizational Factors
• Other External Influences

23
Some causes for failure to follow established
procedure

• Procedure not understood as written
• Procedure out of date and does not reflect
  current system or required actions
• Inconsistent enforcement / monitoring
• In a hurry due to task overload (temporary or
  chronic)
• Distraction
• Missing some tool or supply material

24
Some reasons workers Intentionally decide not to
follow established procedure

• Previous deviation successful
• Following personal example set by supervisor/
  manager
• Operating culture or peer pressure
• Perceived that step was not important (no
  benefit)
• Perceived that consequence was not sugnificant
• Perceived low likelihood for adverse consequence
• Perceived supervisor expectation
• My way is safer or better
• Perceived (or actual) excessive administrative or
  documentation requirements that add no value to
  the task

25
Design Traps that contribute to employees
failing to follow established procedures

• Violating conventions for layout or operation
• Fuzzy or ambiguous communication
• Incremental creep changes
• Process modifications that violate initial design
  stipulations

26
Root Cause Investigation Concept

• Failure to follow procedures
• is not a root cause

27
Effective Initial Interviews

• Initial rapport
• Restate purpose/objective
• Neutral location
• Prompt singular
• Un-interrupted narrative
• Reflective listening
• Open-ended questions

28
Piper AlphaJuly 1988, North Sea Platform167
fatalities

• Communications
• Work Permit System
• Lockout/tagout system
• Emergency Preparedness
• Emergency Response
• Fire Protection Philosophy
• Safety auditing/monitoring System