analysis & benchmark of 4 years of hipo events and events...

Analysis & Benchmark of 4 years of HiPo Events and Events with Real Major Consequences in TOTAL RC

D. Roosendans / D. Robillard Department Major Risks TOTAL Refining & Chemicals

TOTAL Refining & Chemicals

2 Analysis & benchmark of 4 years of HiPo events and events with real major consequences at TOTAL RC RC/HSE/SEC, April 2016

Safety Indicators in TOTAL RC


Safety Performance Indicators in TOTAL RC

5

o HSE performance indicators

Leading Indicators

Lagging Indicators

Monthly follow-up of these indicators

o Leading indicators

Number of failures of safety critical systems encountered during inspection, testing or operations

Number of reported substandard conditions/acts

Ratio of reported substandard conditions/acts per employee

Degree of realization of tests of safety critical systems

Number of out of date implementation of works following recommendations of inspection service

Number of excursions of safe operating window

Number of permit violations related to air emissions

Number of permit violations related to water emissions

Analysis & benchmark of 4 years of HiPo events and events with real major consequences at TOTAL RC RC/HSE/SEC, April 2016


6

o Occupational Safety Lagging Indicators

Total Recordable Injury Rate (own employees + contractors)

Total Loss Time Injury Rate (own employees + contractors)

TOTAL number of injuries (with and without loss time cases)

o Process Safety Lagging Indicators

Number of LOPC of hazardous products (Tier1 + Tier2, API754)

Number of fires and explosions

Process Safety Event Rate (Tier 1) (using API754 classification)

Process Safety Event Rate (Tier 2) (using API754 classification)




0

1

2

3

4

5

6

7

8

9

10

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Year

ConocoPhillips

Dow

BP

Shell

ChevronTexaco

TOTAL

DuPont

ExxonMobil


TRIR

(o

wn

+ c

on

trac

tors

, p

er m

illio

n h

ou

rs w

ork

ed)

LOPC

8

o A loss of primary containment (LOPC) = accidental, sudden and uncontrolled loss of combustible, flammable, explosive, corrosive, toxic products or products dangerous to the environment whatever steps are taken to confine or collect them or to protect the environment.

o The products to which this indicator relates are those identified in API 754.


Benchmark with CONCAWE, AFPM & GBG data


CONCAWE

10

o Benchmark study in 2014 was performed by CONCAWE (downstream oil & gas in Europe) on safety performance of European refineries

o The study includes feedback from 37 CONCAWE members and represents about 98% of European refining capacity (EU-28, Norway and Switzerland)


AFPM


2.30

1.44

1.23

1.74

1.66

1.91

1.62

1.87

1.46

0.0

0.5

1.0

1.5

2.0

2.5

2013 2014 2015

PSE

Rat

e (

Tie

r 1

+ T

ier

2)

TOTAL RP

AFPM Refining

CONCAWE

AFMP Petrochemicals

TOTAL Refining Petrochemicals vs CONCAWE / AFPM


o Better performance of Petrochemicals compared to Refining

o Number of Tier 1 + Tier 2 Process Safety Events in TOTAL RP :

149 in 2013

83 in 2014

67 in 2015

CO

NC

AW

E

AFP

M P

etro

chem

ical

s

AFP

M R

efin

ing

CO

NC

AW

E

AFP

M P

etro

chem

ical

s

AFP

M R

efin

ing

Benchmark with GBG/API data


TOTAL Refining & Chemicals vs GBG (Global Benchmarking Group)


Process Safety Event Rate (Tier 1 + Tier 2) :

o Significant improvement in 2014 compared to 2013

o Tier1 rate in 2015 still above GBG range

o Tier1 + Tier2 rate in 2015 within GBG range

Analysis of Events


Type of Events

16

o Events in the period 2012 -2015 with high potential consequences (HIPO) and events with real significant consequences (Real) were analyzed

o In total 296 such events were recorded in 2012-2015


o 28% of the 296 studied events are LOPC events

o 89% of these events are HIPO events

A « High Potential Event » (HIPO) is an accident or incident which might have given rise to at least a level 4 accident that was just avoided

Characterization of Events (domain)


TOTAL RC data : Analysis of 296 HIPO events and events with real major consequences in the period 2012-2015

o 54.7% of the 296 studied HIPO events are in the domain of Occupational Safety

42.836.2

49.753.4

7.5 10.3

0

10

20

30

40

50

60

70

80

90

100

Refining Petrochemicals

% o

f Ev

en

ts

Transport Safety Events

Occupational Safety Events

Process Safety Events

Type of Events per Sector



The distribution of different types of events in Refining and Petrochemicals is similar

11

103

64

17

8

93

0

20

40

60

80

100

120

Startup related

Works related

Operations related

Driving related

Transport related

Integrity related

Nu

mb

er

of E

ven

ts

Characterization of Events (nature of activity)


TOTAL RC data : Analysis of 296 HIPO events and events with real major consequences the period 2012-2015

66% of the 296 studied events are related to works and integrity issues

29.7

35.5

24.7

20.2

45.744.3

0

5

10

15

20

25

30

35

40

45

50

GBG 2012, Tier 1 LOPC events 2012-2015

Co

ntr

ibu

tio

n in

per

cen

t

TechnicalOrganizationalHuman

Benchmarking with GBG data (LOPC)

20

GBG data : Refining Industry 2012 Analysis of 70 LOPC events

TOTAL RC data : Analysis of 82 LOPC HIPO events (excl. fires/explosions) in 2012-2015

Profile of LOPC events in TOTAL RC is similar to the profile of LOPC events in GBG (2012 data)


26.0

40.1

17.1

35.5

18.8 19.8 19.4 20.2

55.2

40.1

63.6

44.3

0

10

20

30

40

50

60

70

All events Process Safety events Occupational Safety events LOPC events

Co

ntr

ibu

tio

n in

per

cen

t



Causes for Categories of Events o Organizational causes have the same contribution for all types of events (LOPC, Process Safety, Occupational Safety)

o Human related causes are the most important, irrespective of the category of events

o Technical related causes are more important for LOPC events and Process Safety events

5.0

9.0

9.0

2.0

0.0

7.0

13.0

10.0

12.0

2.0

5.0

12.0

3.2

7.9

7.6

2.1

5.4

4.7

14.0

18.0

9.6

4.1

10.1

13.4

0 5 10 15 20

Fixed Equipment Inspection

Equipment Reliability

Design

MOC

Operating Limits

Procedure

Risk Assessment

Safe Work Practices

Work Monitoring

Communication

Knowledge and Skills

Human Factors

Contribution in Percent

Analysis of TOTAL RC data (LOPC, 2012-2015)

22

Human related

Organization

Technical

GBG 2012 data

More than 50% of all LOPC are related to: • Human Factors • Knowledge & Skills • Safe Work Practices • Work Monitoring • Risk Assessment


GBG 2012 data

3.2

7.9

7.6

2.1

5.4

4.7

14.0

18.0

9.6

4.1

10.1

13.4

0 5 10 15 20



Design

MOC

Operating Limits

Procedure

Risk Assessment

Safe Work Practices

Work Monitoring

Communication


Human Factors




Knowledge & Skills o Lack of experience o Lack of skills

Human Factors o Complacency (lack of focus) o Normalization of deviation (“bypass is

tolerated”) o Lack of presence of management in the

field o Expectations of management are sub-

standard o Inadequate application of work permit

process o Inadequate application of Golden Rules

Work Monitoring o Insufficient oversight of contractor work o Lack of competence of the supervisor

Safe Work Practices o Lack of information or wrong

information o Improper delegation of high risk work o Poor communication with contractor o Lack of clear instruction o No visit in the field o Poor preparation of equipment for

handover

Risk Assessment o Insufficient quality of risk analysis and

risk assessment o Inadequate application of outcome of

risk assessment

More than 50% of all LOPC are related to:

1. Human Factors 2. Knowledge & Skills 3. Safe Work Practices 4. Work Monitoring 5. Risk Assessment

3.2

7.9

7.6

2.1

5.4

4.7

14.0

18.0

9.6

4.1

10.1

13.4

0 5 10 15 20



Design

MOC

Operating Limits

Procedure

Risk Assessment

Safe Work Practices

Work Monitoring

Communication


Human Factors




MOC o An efficient and well documented MOC

program is not in place and not well monitored

Operating Limits o Operating window is not clearly and

formally defined o Operating window is not known by all

operational people o An excursion outside the operating

window is not immediately and formally reported to site management

Design / Equipment Reliability o Lack of use of applicable specifications in

the design o No full compliance with standards and

good practices

Fixed Equipment Inspection o Insufficient experienced and well trained

technical staff o Safety critical items not well defined nor

well managed o No comprehensive documentation of

degradation mechanisms o Insufficient monitoring of the inspection

program o Findings and associated action plan of the

inspection programs are not accepted

Action Plans


Action Plans

26

o The analysis on the previous slides indicate the importance of HIPO and REX knowledge

o Factual evidence of efforts that need to be done to decrease HIPO and REX : human factor , knowledge, skills, risk assessment quality, etc …

o More specifically, a number of desired behaviors need to be developed in order to avoid the majority of HIPO events and event with real major consequences

o This study was helpful to prioritize safety actions


Def

ine

Ver

ify

Sup

po

rt

12. Implement REACH regulation and integrate product safety management

5. Verify the use of mandatory rules and methods in our LBUs

11. Continue the development and implementation of TOTAL referential, guides, audits, training

10. Continue the development of the technical integrity program (method, network, audits)

9. Develop the Safety Culture program in RC

6. Improve the efficiency of the RC HSE REX process (COREX, follow up of important REX, KPI,…)

1. Develop and organize process safety training of process safety teams on sites

2. Improve process safety culture at operational level (operators, engineers, …)

7. Continue the development & follow-up of conformity programs (storage of LPG/Flam. Liq, … )

3. Assure assistance to new projects according to TOTAL HSE/INDUSTRY guidelines

4. Assure quality of analysis of HIPO events, LOPC events and important accidents/incidents

8. Prepare a mapping of major process safety risks and verify that these risks are managed

27

Action Plan for HSE Corporate (Safety Division) in Line with Findings of Analysis


BACK UP


LOPC

29

o A loss of primary containment (LOPC) = accidental, sudden and uncontrolled loss of combustible, flammable, explosive, corrosive, toxic products or products dangerous to the environment whatever steps are taken to confine or collect them or to protect the environment.

o The products to which this indicator relates are those identified in API 754.

o Examples:

Leaks or releases of liquid and/or gas hydrocarbons or flammable or toxic materials occurring on site from a closed circuit:

• Either directly into the atmosphere (Ex: burst of a pipe, of a seal, rupture of pump seals)

• Or in a collection network (drain network or gutters).

• Or in a secondary containment (storage tank, service station separator).

Opening of pressure relief valves (PRV) to the atmosphere giving way either to liquid carry over, or to discharge to a potential unsafe location, or to a shelter in place or to a public protective measure (road closure)

Break of rupture disks


Thresholds for reporting

30

o Values in kg/h o The thresholds in Kg/h are corresponding to the quantity released in any one hour period


CONCAWE

31

o Benchmark study in 2014 was performed by CONCAWE (downstream oil & gas in Europe) on safety performance of European refineries

o The study includes feedback from 37 CONCAWE members and represents about 98% of European refining capacity (EU-28, Norway and Switzerland)


TOTAL RP sites – Process Safety Event Rate (Tier1+2)


Bars indicate range of annual PSE Rates (Tier 1 + Tier 2) for each of the RP sites in the period 2013 to 2015

TOTAL Refining – Process Safety Event Rate (Tier1+2)


Average CONCAWE (2014) : 1.62

Bars indicate range of annual PSE Rates (Tier 1 + Tier 2) for each of the refineries in the period 2013 to 2015

TOTAL RP – Number of Tier1+Tier2 Events


o Bars indicate range of annual percentage of PSE cases (Tier 1 + Tier 2) for each of the sites in the period 2013 to 2015

o About 60 percent of all Tier1 + Tier 2 PSE cases (period 2013 to 2015) are generated by 5 sites

TOTAL RC data : Analysis of 308 Tier 1 + Tier 2 events in the period 2013 to 2015

35

ACTUAL LOSS of less or equal than 3

POTENTIAL LOSS of 4 to 6


A « High Potential Event » (HIPO) is an accident or incident which might have given rise to at least a level 4 accident that was just avoided: o Either through the existence of

particular favorable circumstances oOr by the successful use of an

ultimately mitigating emergency device

oOr through an unexpected or unforeseen recovery action

Level Label Health & Safety Environment

4 MAJOR Permanent disability, death or injuries to the population

Significant external pollution. Implementation of external emergency plan. Emissions into the environment of 10 t of toxic product.

5 CATASTROPHIC Several deaths Major pollution with long-term environmental consequences outside the site

1 Minor

2 Moderate

3 Serious

4 Major

5 Catastrophic

6 Disastrous

Events per Sector



Better performance of Petrochemicals and Chemicals sites in comparison to Refining

81

51

2320

13 12 11 10 85 4 3 3 2

0

10

20

30

40

50

60

70

80

90

Nu

mb

er

of e

ven

ts

Characterization of Events (category of consequences)



The consequences of 52.4 % of the 296 studied events are (not ignited ) loss containment, falling of objects and falling op people

52.4%

Benchmarking with GBG data (2012)

38

A - Mechanical failure 1 - Fixed Equipment Inspection

Technical cause (sudden failure) (failure while operating within design enveloppe) 2 - Equipment Reliability

3 - Design

B - Process failure 4 - Change Management

(failure because of operation out of design enveloppe) 5 - Operating Limits

Organisational cause (inadequate operation) C - Inadequate operation preparation 6 - Procedures

7 - Risk Assessment

8 - Safe Work Practices

D -Inadequate practices 9 - Work Monitoring

10 - Communication

Human related cause

E - Other human related cause 11 - Knowledge and Skills

12 - Human Factors

Other F - Other 13 -Other

Classification of accident causes according to API & GBG


26.0 25.726.9

28.9

21.118.8 19.1

17.3

20.218.8

55.2 55.1 55.8

50.9

60.2

0

10

20

30

40

50

60

70

All events (2012-2015) All Events 2012 All Events 2013 All Events 2014 All Events 2015

Co

ntr

ibu

tio

n in

pe

rce

nt



Causes for All Events (Process Safety + Occupational Safety +Transport Safety)


o Profile of events is stable over the years

o Compared to LOPC events: Higher contribution of

human related causes Lower contribution of

technical related issues

Focus on Integrity


Degradation Types


14.7

12.7

12.7

10.8

10.8

7.8

2.0

0 2 4 6 8 10 12 14 16

Fatigue

Other Corrosion Type

Vibration

Atmospheric Corrosion

Thermal Effects

Mechanical Impact

Erosion


TOTAL RC data : Analysis of 93 integrity related HIPO events and events with real major consequences the period 2012-2015

Corrosion, fatigue and vibration are responsible for more than 50% of all integrity related events

Primary Causes


21.0

20.2

12.9

12.9

8.9

7.3

0 5 10 15 20 25

Poor Maintenance

Operating Procedures not Followed

Specifications Not Followed

Out of Operating Window Issues

Supporting Issues

Management of Change Issues



Poor maintenance and not following operating procedures are the most important primary causes

Equipment Involved


19.4

13.6

13.6

9.7

9.7

6.8

6.8

3.9

1.9

1.0

0 5 10 15 20 25

Line

Valve

Rotating Equipment

Support Structure

Small Bore Connection

Flanged Connection

Process Equipment (Exchangers,...)

Flare / Pressure Relief Systems

Weld

Storage Equipment



Lines and associated equipment (valves, small bore connections, flanged connections,...) are responsible for half of the equipment involved in integrity related events

analysis & benchmark of 4 years of hipo events and events...

Documents