serious injury and fatality (sif) precursor customization ... s/2019spring/eeisif.pdf · specific...

Serious Injury and Fatality (SIF) Precursor Customization Project

2

Motivation

0.00000.01000.02000.03000.04000.05000.06000.07000.08000.09000.1000

2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018

Fata

lity

Rate

(per

100

0 w

orke

rs)

Year

Fatality Trends in Electrical Power Generation and Delivery(moving 3-year average)

3

Gregg Slintak, Consolidated Edison Tom Dyson, Ameren ServicesMatthew Hallowell, University of ColoradoTodd Gallaher, Southern California EdisonTerry Halford, ClecoDave Flener, Quanta ServicesDean Larson, KCPLJenny Bailey, Southern CompanyKathy Wilmer, Duke EnergyJoe Armatys, Bonneville Power

Bill Messner, Portland General ElectricEric Bauman, Electric Power Research InstitutePaul Mackintire, Eversource EnergyIan Wenzel, AlleteScott Lange, WEC Energy GroupMarguerite Porsch, CenterPoint EnergyRick Hoffman, American Electric PowerJames Goodnite, American Electric PowerPatrick Winkel, Consumers EnergyBob Spencer, Tennessee Valley AuthorityLen Colvin, Tennessee Valley Authority

EEI SIF Research Team

4

Key Definition

Precursor: Reasonably detectable event, condition, or action that serves as a warning sign of a serious incident or fatality

All precursors are causal factors, but not all causal factors are precursors

Use of the SIF precursor analysis does not guarantee that a SIF event will not occur

5

Objective

Customize precursor analysis for electrical generation, transmission, and distribution

What factors best distinguish success from failure?

6

Mindset

7

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases

Process

Customized Method

8

Brainstorm New Precursors

Specific to EEISelect Top Tier

Rate Predictive Ability of New

Precursors

Rate Generalizability

Validated General Industry

Precursors

Our Investigation Set

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

(43) (13) (16)

(29)

9

1. Safe Work Procedure2. Working Alone3. Hazard Recognition4. Control Barriers5. Plan for Change6. Safety Attitudes7. Schedule Pressure8. Improvisation

Original Precursors from General Industry Study

9. Significant Overtime10. Fatigue11. Distraction12. Prior Safety Performance13. Safety Supervision14. Front-Line Supervisor15. Pre-Task Plan16. Congestion

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

10

1. Departure from Routine2. Adherence to Rules3. Familiarity with Task4. Worker Assumptions5. Multitasking6. Risk Normalization7. Use of PPE8. Equipment ID and Steps

9. Communication Barriers10. Safety Culture11. Stop Work Execution12. Worker Engagement13. Safety Devices

New Precursors to Test

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

11Create ‘case’ template

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

List of 29 Factors

• X• X• X• X• X• X

Case Template

• ???????????• ???????????• ???????????• ???????????• ???????????• ???????????

12

SIF No-SIFTD 12 12

GEN 8 8

n = 40 cases

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

13

SIF No-SIF TotalTD 12 12 24 60%

GEN 8 8 16 40%

n = 40 cases

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

14

SIF No-SIF TotalTD 12 12 24 60%

GEN 8 8 16 40%

Total 20 2050% 50%

n = 40 cases

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

15

High-energy near miss

Fatal or DisablingHigh-Energy

Success

Screen, Scrub, Randomize

Case 1 Case 2 Case 3 Case 4 Case 5 Case 6 … Case 40

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

16

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

72% of cases correctly predicted by the team

17

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

Goal: Find the precursors that best distinguish no-SIF cases from SIF cases

With this information we can shorten the engagement to something reasonable.

18

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

Simple data structure

Principal Components

Analysis (PCA)

Reduces the number of variables

Generalized Linear Modeling

Predictive equation

0 1 0 … 11 1 0 … 01 1 0 … 01 1 0 … 0

1 1 0 … 0

Case 1Case 2Case 3Case 4

Case 40

Precursors

X1 X2 X3 X29

1000

1

Y

OutcomesCases

19

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

0

2

4

6

8

10

12

14

Presence in SIF v Success

20

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

0

2

4

6

8

10

12

14

Presence in SIF v Success

Total Presence in SIF Total Presence in Success

21

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

0

2

4

6

8

10

12

14

Presence in SIF v Success

Total Presence in SIF Total Presence in Success

22

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What we are trying to predict: Probability of SIF

What is a generalized linear model?

23

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

Depends on (is predicted by)

24

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

What the probability when NO precursors are present (intercept)

25

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

The first precursor (present = 1, absent = 0)

26

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

The first precursor (present = 1, absent = 0)

The weight of that precursor (coefficient)

27

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

The second precursor

28

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

The second precursor

Its weight

29

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

Interaction of two precursors

30

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

P(SIF) ~ B0 + B1 P1 + B2 P2 … Bn Pi Pj

What is a generalized linear model?

Interaction of two precursors

Weight of the interaction

31

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

Precursors Present in SIF Cases

Present in no-SIF Cases

Difference (SIF-no SIF) 𝜷𝜷𝟎𝟎

Rules and Procedures 27% 2% 24% 4.12

Departure from Routine 24% 5% 20% 3.57

Hazard Recognition 22% 5% 17% 0.79

Safety Attitudes 22% 5% 17% 0.02

Workers Inactive in Safety 17% 2% 15% 0.73

Risk Normalization 32% 20% 12% 0.91

Safe Work Procedure 12% 2% 10% 0.92

Familiar with the Task 17% 7% 10% 0.58

Stop Work Execution 12% 2% 10% 0.58

Perceived Safety Culture 15% 7% 7% 1.11

Pre-Task Plan 20% 12% 7% 1.56

Plan to Address Change 22% 17% 5% 0.49

Productivity Pressure 20% 15% 5% 1.06

32

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

Precursors 𝜷𝜷𝟎𝟎 Weights

Rules and Procedures 4.12 3

Departure from Routine 3.57 3

Pre-Task Plan 1.56 3

Perceived Safety Culture 1.11 3

Productivity Pressure 1.06 3

Safe Work Procedure 0.92 2

Risk Normalization 0.91 2

Hazard Recognition 0.79 2

Workers Inactive in Safety 0.73 2

Familiar with the Task 0.58 2

Stop Work Execution 0.58 2

Plan to Address Change 0.49 2

Safety Attitudes 0.02 1

33

11% 14% 17%21%

26%32%

38%45%

52%58%

65%71%

76%81%

85%

97%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

POTE

NTI

AL F

OR

SIF

WEIGHTED PRECURSOR SCORE

SIF Potential

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

34

Identify Precursors

Experiment to Test

Precursors

Objective Statistics

Collect Cases Customized Method

35

What is a field safety engagement? When performing the engagement… Potential pitfalls

Field Safety Engagements to Complete the SIF Precursor Scorecard

36

• 27 Questions and 11

Observations• Validated, Research, and

Customization Process

• List of Questions

• Peel the onion

Field Safety Engagements to Complete the SIF Precursor Scorecard

37

Setting up for Success!For a Strong and Positive Engagement

Déjà vu- Add strategy and Intentional Focus- Framework to engage field workers

Let the engagement begin- Minimize disruption of work- Participate in Job Briefing

Present on not?- Knowledge of the work & workers- Precursors are rare occurrences

Coaching

2 Way Communication

Respect

Comfortable ConversationFeedback

Positive Tone

38

Using the SIF Precursor Scorecard

1

2

3SUM

Check the precursors that were present before any intervention was made.

Find the sum of all the weights for the selected precursors.

Interpret the total weighted score.

40

How did you score?

41

Putting Precursor Analysis into Practice

43

The product of individual and group values, attitudes, competencies, and patterns of behavior that determines the commitment to safetyIt’s the way we do things!

Safety Culture determines: Personal Responsibility Trust Communication Behavior Lessons learned

Safety Culture

44

Empowerment

Empower individuals to successfully fulfill their safety responsibilities to themselves, their family, and their coworkers.

Encourage everyone to: Hold themselves and each other accountable for safety! Exercise authority to stop unsafe behavior without fear of negative

repercussions. Correct unsafe conditions as soon as possible Provide multiple options for your team to report unsafe conditions

and/or behaviors

45

Communication

Build TRUST! Do what you say you will do, when you say you will do it!

Ensure timely and appropriate responses to identified hazards and have an action plan in place to address and remove the hazards.

Reinforce current safety practices through regular coaching

Celebrate the successes along the way!

46

Organization - All Employees“Perceived” Safety Culture, Stop work and Productivity

pressures .

Planning and PreparationRules and procedures, Working Together/Communication

PerceptionHazard Recognition, risk normalization and familiarity with

the work.

EngagementSafety Attitudes and Ownership

IMPLEMENTATION

How it all fits together

47

The presence of precursors like schedule pressure, risk normalization, and poor attitudes compromise readiness and may increase the potential for events.

When strong pre-task planning is performed to manage hazards and precursor analysis is used to check worker readiness, BOTH the demands of the work and readiness of the worker are considered.

Integration with Existing Programs

48

Executive Summary Implementation Guide Project Report Engagement Video Conference Video

Available Resources