Underlying causes of offshore incidents

Harpur Hill, Buxton Derbyshire, SK17 9JN T: +44 (0)1298 218000 F: +44 (0)1298 218590 W: www.hsl.gov.uk

Underlying Causes of Offshore Incidents

FP/09/21

Project Leader: John Hare, PhD

Author(s): John Hare, Michael Johnson

Science Group: Hazard Reduction Group

DISTRIBUTION Tom McLaren Head OSD3, Customer Authorising Officer Ian Whewell Head of Division, OSD

Julie Voce OSD5 Rob Miles OSD3 Alan Thompson OSD3 Peter Mills Head OSD1 Steve Wing Head OSD2 Howard Harte Head OSD4 Steve Walker Head OSD5

Jill Wilday Fire and Process Safety Unit, HSL Laurence Cusco Head, Fire and Process Safety Unit, HSL John Hare Fire and Process Safety Unit, HSL Michael Johnson Fire and Process Safety Unit, HSL Peter Davies Head, Hazard Reduction Group, HSL (circulation only)

PRIVACY MARKING:

Available to the public.

HSL report approval: Mr Peter Davies Date of issue: 15 May 2009 Job number: JN0004209 Registry file: FP/08/28586 Electronic file name: FP09ir21_offshore incidents

© Crown copyright (2009)

ACKNOWLEDGEMENTS The authors would like to thank Donna Irons, Elaine Kirwan and Pauline Knight for providing a list of offshore incidents involving fatalities and major injuries (via the COIN data mining team) and then supplying the inspectors’ investigation reports for these incidents.

iii

CONTENTS

1 INTRODUCTION......................................................................................... 1

2 RIDDOR DATA ........................................................................................... 3 2.1 Injury Types ............................................................................................. 3 2.2 Injury Location ......................................................................................... 4 2.3 Kind of Accident....................................................................................... 5 2.4 Operating Companies.............................................................................. 6 2.5 Grouped Agents ...................................................................................... 7 2.6 Work Process – Environment .................................................................. 8

3 INVESTIGATION REPORT DATA ............................................................. 9 3.1 Investigation Reports ............................................................................... 9 3.2 Direct Causes .......................................................................................... 9 3.3 Underlying Causes ................................................................................ 10 3.4 Health and Safety Management Failings ............................................... 13 3.5 Regulation and Guidance ...................................................................... 15 3.6 Design Faults......................................................................................... 18

4 FURTHER ANALYSIS .............................................................................. 21 4.1 Kind of Accident..................................................................................... 21 4.2 Injury Type............................................................................................. 21 4.3 Direct Cause.......................................................................................... 21 4.4 Underlying Causes ................................................................................ 21 4.5 Work Process Environment ................................................................... 22 4.6 Grouped Agents .................................................................................... 22 4.7 Health and Safety Management Failings ............................................... 23 4.8 Regulations and Guidance .................................................................... 23

5 COMPARISONS WITH PREVIOUS STUDIES ......................................... 25 5.1 Offshore Injury Ill Health and Incident Statistics .................................... 25 5.2 Previous Offshore Incident Causal Analysis .......................................... 29 5.3 Previous Analysis Work Done by HSL................................................... 33

6 GUIDANCE ............................................................................................... 37

7 DISCUSSION............................................................................................ 43 7.1 RIDDOR Data........................................................................................ 43 7.2 Investigation Report Data ...................................................................... 43 7.3 Further Analysis..................................................................................... 43 7.4 Comparisons With Previous Studies...................................................... 45 7.5 Guidance ............................................................................................... 46

8 CONCLUSIONS AND RECOMMENDATIONS......................................... 47 8.1 Conclusions ........................................................................................... 47 8.2 Recommendations................................................................................. 48

iv

9 APPENDIX A – CROSS REFERENCING TABLES ................................. 49 9.1 Kind of accident ..................................................................................... 50 9.2 Injury type .............................................................................................. 51 9.3 Direct cause........................................................................................... 52 9.4 Underlying Cause .................................................................................. 53 9.5 Work Process Environment ................................................................... 55 9.6 Agent ..................................................................................................... 57 9.7 Health and Safety Management System Failing.................................... 60 9.8 Regulations and Guidance Mentioned................................................... 61

10 APPENDIX B - HYDROCARBON RELEASES CAUSAL TABLES...... 63

11 APPENDIX C – GROUPED AGENTS................................................... 65

12 APPENDIX D - INJURY TYPE AND AGE TABLES ............................. 67

13 REFERENCES ...................................................................................... 69

14 NOMENCLATURE ................................................................................ 71

v

EXECUTIVE SUMMARY Objectives

Offshore Safety Division (OSD) is committed to providing the Offshore Industry Advisory Committee (OIAC) with an analysis of the causes of offshore incidents. To achieve this, the underlying causes of offshore incidents (fatalities, major injuries and major dangerous occurrences) were identified using OSD Inspection Management Teams (IMT) inspectors’ investigation reports supplied to the Health and Safety Laboratory (HSL) by OSD. The project objectives were to:

• identify as far as possible the underlying causes of offshore incidents resulting in fatalities, major injuries and (time permitting) major dangerous occurrences - over the period 2004 to 2008 and determine any trends that may emerge; and

• use the existing work done by HSL for Offshore Division on Hydrocarbon Releases (HCR) and Key Programme 3 (Asset Integrity Programme) to see if they can provide any other information on underlying causes.

Main Findings

1) A sample of 67 offshore incidents involving fatalities and major injuries has been analysed using both investigation reports and RIDDOR (Reporting of Injuries, Diseases and Dangerous Occurrences Regulations) reports. These incidents resulted in 5 fatalities and 62 major injuries. By far the most predominant injury type was fractures. Younger workers seem to be involved in fewer incidents than older age groups. The injured party was more likely to work for a contractor (42 individuals) rather than the installation operator (25 individuals).

2) The most common kinds of accidents were (in order of importance): • Hit by moving, flying or falling object (mostly impact from dropped object); • Injured while handling, lifting or carrying; • Falls from height; and • Slips trips and falls on the same level.

3) The most important underlying causes of accidents were (in order of importance): • Inadequate hazard analysis / risk assessment; • Inadequate supervision; • Lack of / inadequate operating procedures; and • Inadequacies in permit-to-work.

4) As might be expected, the work process environment and grouping of RIDDOR defined agents involved in the incidents correlated with the kind of accident: • Hit by moving, flying or falling object was associated with offshore production, construction and maintenance and with the agents of lifting and storage (hit by object free falling from lifting machinery). • Falls from height and slips, trips and falls at the same level were associated with offshore construction, maintenance, deck operations and transport and the main agent was surfaces at different levels. • Injured while handling, lifting or carrying was mainly associated with offshore drilling / workover and the agent of building materials and components.

vi

5) The main underlying causes associated with the main kinds of accident were: • Operating procedures for ‘hit by moving, flying or falling object’; • Permit-to-work for ‘fall from height’; • Permit-to-work for ‘injured while handling, lifting or carrying’; and • ‘Unknown’ for ‘slipped tripped or fell at the same level’.

In addition supervision was implicated for the agents ‘surfaces at different levels’ and ‘lifting and storage’, which were associated with slips, trips and falls and with being hit by falling objects.

6) The main identified safety management failings were in monitoring, audit and review and planning and implementation. In terms of failings associated with the main kinds of accident: • Hit by moving, flying or falling object was due to poor monitoring; and planning and implementation; • Falls from height were due to poor monitoring; and • For slips, trips and falls at the same level and injured while handling, lifting or carrying, the management failing was usually unknown.

7) The main health and safety regulations referred to in investigation of incidents were LOLER (Lifting Operations and Lifting Equipment Regulations) and MHSWR (Management of Health and Safety at Work Regulations).

8) Consideration of previous causal analysis of offshore incidents confirmed the current analysis and adds the additional underlying causes and Safety Management System (SMS) failings: • Human factors including distractions, stress and culture for slips, trips and falls; • Operator error and poor positioning for lifting incidents; • Poor communication was an underlying cause of hydrocarbon releases in addition to all the underlying causes found in the current analysis; and • Backlogs and deferrals in the maintenance of Safety Critical Elements (SCEs) were highlighted by the analysis of the Key Programme 3 (Asset Integrity Programme) and indicate underlying management system failures.

Recommendations

1) Targeted inspection activity is needed on the following issues:

• Falls from height – Permit to work, monitoring;

• Injured whilst handling, lifting or carrying – Permit to work, supervision, audit and review, reference to MHSWR (Management of Health and Safety at Work Regulations); and

• Hit by moving, flying or falling object. – Operating procedures, monitoring, planning and implementation, use of LOLER (Lifting Operations and Lifting Equipment Regulations) and MHSWR.

2) Further work could usefully be undertaken to improve understanding in the following areas:

• Slipped, tripped or fell - Underlying causes are often unknown, health and safety management failings are often unknown, there is little reference to regulations and guidance.

vii

• Injured while handling, lifting or carrying – Sometimes no reference to guidance or regulations.

3) Underlying causal analysis using both RIDDOR (Reporting of Injuries, Diseases and Dangerous Occurrences Regulations) and HSE (Health and Safety Executive) Inspectors’ reports could usefully be undertaken for major offshore dangerous occurrences.

viii

1 INTRODUCTION

Offshore Safety Division (OSD) is committed to providing the Offshore Industry Advisory Committee (OIAC) with an analysis of the causes of offshore incidents. To achieve this, the underlying causes of offshore incidents (fatalities, major injuries and major dangerous occurrences) were identified using OSD Inspection Management Teams (IMT) inspectors investigation reports supplied to the Health and Safety Laboratory (HSL) by OSD. The project objectives were to:

• identify as far as possible the underlying causes of offshore incidents resulting in fatalities, major injuries and (time permitting) major dangerous occurrences - over the period 2004 to 2008 and determine any trends that may emerge; and

• use the existing work done by HSL for Offshore Division on Hydrocarbon Releases (HCR) and Key Programme 3 (Asset Integrity Programme) to see if they can provide any other information on underlying causes.

Given the time and resources available, only offshore fatalities and major injuries have been investigated. Underlying causes of major dangerous occurrences (including HCRs) can be done separately if required.

Investigation reports were available for 67 of the offshore incidents, involving fatalities and major injuries, which occurred over the period 2004-2008. Information on these 67 incidents was available from two sources: the RIDDOR (31) reports (supplied by the offshore operators) and HSE Inspectors’ Investigation Reports. The RIDDOR reports include records of: Injury Type, Injury Location, Kind of Accident, Operating Company, Agent and Work Process Environment. The inspectors reports were searched and any information on Direct Causes, Underlying Causes, Regulation / Guidance mentioned and Health and Safety Management System Failing was recorded. Therefore as well as highlighting major issues in offshore incidents identified from RIDDOR and inspection reports, cross–referencing of data was possible.

The report also includes comparison with KP3, HCR and HSE Offshore Annual Accidents Reports. Relevant Guidance is discussed: OSD-specific, LOLER, PUWER, MHSWR and HSWA. Human factors issues are also commented on.

2 RIDDOR DATA

Appendix D gives the following information on the 67 incidents – number of major injuries, number of fatalities, calendar year of incident, age of injured party, and injured party - employee status (worked for a contractor or worked for the installation operator). These incidents resulted in 5 fatalities and 62 major injuries. Younger workers seem to be involved in fewer incidents than older age groups (11 individuals came from the 21-30 age group, 18 from the 31-40 age group, 17 from the 41-50 age group and 15 from the 51-60 age group). The injured party was more likely to work for a contractor (42 individuals) rather than the installation operator (25 individuals).

2.1 INJURY TYPES

Figure 1 shows the injury types for the 67 incidents considered. Fracture is the largest injury type with over 40 cases. None of the other injury types show more than six cases. The incident data on fracture injury types will go forward to a detailed comparison with other data in section 4.

45

40

35

30

25

Number 20

15

10

5

0 AMPUTATION BURN CONTUSION DISLOCATION FRACTURE LACERATION MULTIPLE OTHER KNOWN OTHER N/K STRAIN Injury type

Figure 1 Injury types

2.2 INJURY LOCATION

Figure 2 shows the injury location for the 67 incidents considered. There is more variation in the injury location than the injury type. Injury location was not selected to go forward for the detailed comparisons in section 4.

12

10

8

Number 6

4

2

0 AN BA ER T CS D EA LI EC B CS K CS B R KL C FO AN D M K /L UN LIM IS K FI O LO H H B N IM LO LO T E N W TR U W G AL ER L SE SP PP G SE VE EC ER EN LO VE R IF ER R AL IE AL D UN Injury location

Figure 2 Injury location

2.3 KIND OF ACCIDENT

Information is recorded in RIDDOR on the kind of accident, using a large number of categories. The kind of accident can easily be grouped into a number of condensed categories, which are generally similar to the actual response provided. Figure 3 shows the kind of accident, using these condensed categories for the 67 incidents considered. Hit by moving, flying or falling object is the dominant kind of accident with over 20 cases. Falling from a height, Injured while handling lifting or carrying and Slipped, tripped or fell on the same level also show some prominence. The four kinds of accidents data mentioned here will go forward to a detailed comparison with other data in section 4.

25

20

15

Number

10

5

0 Contact with Exposed to Fall from a Hit by a Hit by Hit by Injured Injured while N/K Other known Slipped, moving fire height moving, material object(s) free through cuts handling, kind of tripped or fell machinery or flying or under falling from from lifting or accident not on the same material falling object pressure lifting sharp/coarse carrying listed above. level being during machinery, material or machined normal vehicles and equipment or operation of other from trapped machinery or equipment. fingers. Injury other Include is not through equipment. components the weight of of machinery material, which may objects or Kind of accident fall but still equipment attached. being handled.

Figure 3 Kind of accident

2.4 OPERATING COMPANIES

Figure 4 shows in anonymous way, the offshore operating company involved in the 67 incidents. Three companies seem to have many incidents (FL, EY and HS with seven to eight incidents). Operating companies were not selected to go forward for the detailed comparisons in section 4. Normalisation of this data is necessary to take account of the number of installations each company has and the total number of people (direct staff and contractors) working on all the company’s installations. 9

8

7

6

5

Number 4

3

2

1

0 BD BZ EJ EU EY FL FZ HC HS HY JF KE LB LS PV PX RR RW SK SX TN VU VX WG YP ZD ZK

Anonymous Operating Company Codes

Figure 4 Operating companies

2.5 GROUPED AGENTS

Information is recorded in RIDDOR on the agent involved in the incident, using a large number of categories. The agent can easily be grouped into a number of categories using the first number in the classification. Figure 21 in Appendix C (section 11) shows the agents and how they are grouped together. Figure 5 shows the various agent groupings for the 67 incidents. Five of the agent groupings shown seem to be significant and this data will go forward to a detailed comparison with other data in section 4. The significant five agent groupings are: • Surfaces at different levels: floors ladders, scaffolding; • Pipes and pipe Fittings; • Lifting and storage; • Tools and machines; and • Building materials and components.

18

16

14

12

10

Number 8

6

4

2

0 1.xx Surfaces at different 3.xx Pipes and pipe 5.xx Electrical equipment 6.xx Lifting and storage 8.xx Tools and machines 9.xx Building materials 14.xx Person levels, floors, ladders, fittings and components scaffolding

Grouped agent

Figure 5 Grouped agents

2.6 WORK PROCESS – ENVIRONMENT

Figure 6 shows the work process carried out in the environment where the incident occurred. Offshore construction and maintenance (code 1814) and Offshore production (code 1811) with 20 and over cases would seem to the worst working environment. Offshore drilling (code 1812) and Offshore deck operations (code 1818) also significant cases. Data for the four working environments mentioned here will go forward to a detailed comparison with other data in section 4. Normalisation of this data is necessary to take account of the number of people working in each work process environment.

25

20

15

Number

10

5

0 1811 - Offshore production 1812 - Offshore 1813 - Onshore drilling 1814 - Offshore construction 1815 - Offshore diving 1818 - Offshore deck drilling/workover operations and maintenance operations and offshore transport, including all lifting and crane work, and all helideck and sea transport activities Work process - Environment

Figure 6 Work process – Environment

3 INVESTIGATION REPORT DATA

3.1 INVESTIGATION REPORTS

The HSE Inspectors’ investigation reports for the 67 incidents were studied carefully and information on direct causes, underlying causes, health and safety management failings and regulations/guidance identified by the inspector was noted. The categories used for both sets of causes and the failings were the same as SPC/Enforcement/132 (30). If more than one cause, failing or regulation was noted, it was recorded. One limitation of some incident analysis recording systems is that only one cause or failing may be selected. Examples of incidents are provided with the same incident sometimes used to illustrate more than issue.

3.2 DIRECT CAUSES

Figure 7 shows the direct causes contributing to the 67 incidents. Impact/dropped objects is the most significant direct cause with nearly 30 cases. The incident data on impact/dropped objects will go forward to a detailed comparison with other data in section 4. Examples of incidents involving this direct cause are provided below.

35

30

25

20

Number 15

10

5

0

Procedural violation Unsuitable PPE Corrosion - External Inadequate isolation Inadequate procedures Incorrect installation Slip/fall

Corrosion - Internal Unknown

Inadequate control Inadequate equipment Unsuitable equipment Natural Causes Human error

Impact/dropped object Degradation of material

properties Direct cause

Figure 7 Direct causes

3.2.1 Examples of incidents involving impact/dropped object as a direct cause

• The IP was guiding a suspended load when part of it slipped and hit him on the foot, the resulting injury required parts of two toes to be amputated.

• A suspended load was being manoeuvred through a gap between handrails and the IP pushed the load with his foot. The load moved suddenly and trapped the IP’s ankle against the handrail.

• A scaffolder knocked an old bolt through the grating deck while he was working in an area directly above the IP. The bolt hit the