2 May 2008
Contents
1
INTRODUCTION .................................................................................................................................. 4
3.1
3.2
3.3
4.2
4.3
Injuries............................................................................................................................................ 27
4.4
4.5
Conclusions ........................................................................................................................................... 36
References ............................................................................................................................................. 37
PAGE | 2
1 LIST OF ANNEXES
1. Rules for scenario modelling, - Linda Bellamy, Martijn Mud, Andrew Hale, Hans Baksteen, Ben
Ale 2005
2. Modelling management deliveries and tasks - Linda Bellamy, Martijn Mud, Andrew Hale, Ben
Ale 2005
3. Rules for the construction of the right hand side, - Hans Baksteen 2005
4 Storybuilding injury classification rules, - Hans Baksteen 2004
5. Table of All Storybuild Data exported from Superfile 9b, - Linda Bellamy 2007
6 Accidents per Storybuild, - Linda Bellamy 2007
7 Accident consequence ratios - Linda Bellamy & Martin Damen 2007
8 Injuries per year.- Linda Bellamy 2007
9 Storybuilder information sheets: Definitions Linda Bellamy 2007
10 Facts and Figures Sheets Linda Bellamy, Martin Damen & Martijn Mud 2007
11 Sector specific accident frequencies Linda Bellamy 2007
12 Ad hoc questions - Martijn Mud & Marc Blauuw 2007
PAGE | 3
2 INTRODUCTION
In 2004-2006 a team was formed for the sole purpose of analysing occupational accident
investigation reports made available by the Dutch Labour Inspectorate. In this period the software
tool StoryBuilder was developed to support the detailed analysis of the accident data. These
horrible stories were analysed in order to build cause and effect structures according to a set of
building rules. The history of these developments is outlined in a number of papers [ Bellamy et al
2006, 2007, 2009 ].
The accidents analysed are reportable occupational
accidents1.
Employers are obliged to report serious
occupational accidents to the Dutch Labour Inspectorate
(Arbeidsinspectie). Sometimes this does not happen and the
accident is either not notified at all or brought to the attention
of the Labour Inspectorate by police, insurance companies or
victims. Accidents are reportable according to article 9 of the
Dutch Working Conditions Act (Arbowet 1998) if they are
occupational accidents resulting in serious physical or mental
injury or death within one year.
A physical injury is
considered to be serious if the victim is hospitalised within 24
hours and for at least 24 hours or the injury is permanent
whether or not the victim is hospitalised. A reportable
accident has to be reported within 24 hours. Then there are
also criteria concerning whether an injury is permanent or not
(physically or mentally).
GISAI (Gentegreerd InformatieSysteem ArbeidsInspectie) is
the Dutch Labour Inspectorate management system for
occupational accidents. The Dutch Labour Inspectorate
Figure 1 Splash screen stores all correspondence about occupational
accidents in GISAI. Data was available on 22,892
from the software
occupational accidents that were reported between 1
January 1998 and end February 2004. 10,237 of these had no offence or investigation report and
were not analysed. The main reason why there was no report was that they were not reportable
(82% of the accidents without report). The other cases were waiting to be investigated or were
under investigation or too sensitive to be made available.
Only accidents with reports could be used for detailed analysis of causes. There are different kinds
of reports and only if a breach has been found is the report complete with respect to witness
statements and injury classes. If there is no breach report then there is a summary of the
A few cases that were investigated turned out not to be reportable but they still give a lot of information about how
accident happen so they were included
PAGE | 4
investigation findings and the reason why it is not a breach. The latter were also analysed but
contain less information. If the conclusion is that the accident was not an occupational accident
e.g. natural death or suicide then these were not included in the analysis.
There are also limitations on which occupations appear in GISAI e.g. self employed are excluded
unless they are working under the authority of another company. Also excluded are occupational
accidents occurring in air transport for flying aircraft, for cabin personnel on stationery aircraft.
Loading/unloading and other non-cabin staffs are included as are cabin personnel outside the
aircraft.
Oil and gas exploration is also excluded. All other drilling for scientific research,
geothermal energy, groundwater accidents are covered by the Labour Inspectorate. Shipping all
accidents while building repairing maintaining or cleaning ships are included, and loading and
unloading is only in the Labour Inspectorate database if the crew is not involved. Railway
accidents are in principle incuded, but really serious incidents with collision and electrocution are
done together with another inspectorate which means the investigation is not in the GISAI
database If a pupil of any school/university has an accident it is considered a labour accident
where the teacher or school can be held responsible after the investigation. Military is included
other than in wartime. Passers by or trespassers on worksites during working hours are included.
Accidents to illegal workers, foreigners etc working on Dutch soil are included the company is
obliged to report them.
An accident is an occupational accident if it occurs at the workplace working during work.
Underreporting for serious occupational accidents is considered to reach 50%.
The
underreporting % cannot be distributed evenly across the 3 categories of consequences (death,
permanent injury, recoverable injury).
9142 reported investigated accidents of the Dutch Labour Inspectorate have been analysed.
These accidents are distributed across 36 Storybuilds, graphical structures in the software
StoryBuilder each representing a type of occupational accident, characterised and named in each
case by the release of the hazard agent or centre event of a bowtie of causes and effects.
The analysed accidents occurred between 1998 and end February 2004 inclusive except for 03.1
Contact with falling objects cranes, 03 Contact with falling objects not cranes, and 08.1 Contact
with moving parts of machine, where only accidents between 2002 to 2003 were analysed.
PAGE | 5
3 OVERVIEW OF ANALYSES
3.1
The 36 Storybuilds
There are 36 Storybuilds. A Storybuild is a structure of events which has a bowtie shape and a
centre event (see Figure 2 and Figure 3). The Storybuild is named according to the centre event.
On the left hand side are events leading up to this event, the release of the hazard agent.
The original building rules for Storybuilding are given in Annexes 1-4. Annex 1 provides the basic
rules for barriers and loss of control events which are a consequence of barrier failures, Annex 2
gives the important delivery systems-PUMM tasks which attach to barriers. Annex 3 and 4 are
concerned with storybuilding to the right of the centre event. These ae notably dose determining
factors like the height of a fall, the weight of a falling object, the speed of a vehicle or the length of
an exposure to a hazardous substance and injured body parts and type of injury. Ultimately the 3
types of consequence death, permanent or recoverable injury - must be recorded and if
available the time of absence from work.
The 36 storybuilds were imported into a single file called a Superfile. The first complete set of
bowtie quantifications for the risk model were based on Superfile version 9b which was frozen at
the start of the quantification period.
An export of all the data in this Superfile 9b is given in Annex 5. This is a very extensive document
of several hundred pages since it covers 24559 data points (storybuild boxes) for the 9142
accidents. The first column gives the Storybuild number and name, the next column shows the box
numbers for each storybuild in sequence followed by box code, event name and description. The
total accident pathways going through each box are shown in the next column and finally this is
shown as a per year average.
PAGE | 6
6 (2)
DDF
1m < Height of
fall <= 2m
83
BSM
Barrier Suc cess
Mode
82
SSB
Safety Barrier
Group
7 (2)
DDF
2m < Height of
fall <= 3m
84
BSU
Barrier State
Unknown
8 (2)
DDF
3m < Height of
fall <= 5m
118 1_BSM
Right c ondition
of ladder
113
SSBWrong/
s ubstandard
ladder
120
G
Management
Delivery
Sy stem s GROUP
119
G
Tas ks Group
117
1_B
Condition of
ladder
114 1_BFM
Subs tandard
condition or ty pe
of equipm ent
115
LCE
Ladder fails
116
PSBLadder Strength
10 (2) DDF
Height of fall >
10m
141 2_BSM
Correct
placement of
ladder
102 (3) REG
Ens ure s afe
procedures for
work ing with
electric ity are
followed
143
G
Management
Delivery
Sy stem s GROUP
142
G
Tas ks Group
88
A
Moving up/ down
a s tep/ s tairs /
ladder
3
G
Dose
determining
fac tors
140
2_B
Placement of
ladder
138 2_BFM
Wrong
placement of
equipment
166
G
Management
Delivery
Sy stem s GROUP
165
G
Tas ks Group
96 (4)
ET
ESAW 0202.01
Fix ed ladders
90
A
ins pecting/adjus
ting/meas uring
on ladder/steps /
stairs
87
Activity
91
A
(de-)installing/
repairing on
ladder/ s tairs/
steps
97 (2)
ET
ESAW 0203.01
M obile ladders,
stepladders
86
G
Equipment Type
98 (2)
ET
s tep ladder or
s teps
92
A
c leaning/
painting on
ladder/s teps/
stairs
100 (4) ET
Rope ladder
93
A
c onstruc tion
related ac tiv ities
on ladder/
s teps / s tairs
101
ET
Unk nown
109 (3) MT
Prov ide s afe
work ing
loc ations at the
c onstruc tion s ite
94
A
other ac tiv ity on
ladder/ s teps/
stairs
95
A
Unk nown
13 (2) DDF
Hard s urfac e
(c oncrete, tiles ,
frozen s oil)
162
PSBLadder Stability
1
CE
Fall from height ladder & s teps
190
G
Tas ks Group
15 (2) DDF
Person
underneath
187
LCE
Support/ surface
fails
16 (2) DDF
Objec ts in line of
fall
210 5_BSM
Correc t pos ition
of person on
ladder
211
G
Tas ks Group
207 5_BFM
Subs tandard
pos ition of
pers on on
equipment
79 8_BSM
Us e of adequate
fall arres tors/
safety nets
208
LCE
Ov erreaching
leads to loss of
balanc e
78
8_B
Us e of fall
protection
(arres tors, nets )
231 6_BSM
(Phys ically ) fit
pers on on ladder
206
SSBUs er ability to
stay on ladder
failure
233
G
Management
Delivery
Sy stem s GROUP
253
G
Management
Delivery
Sy stem s GROUP
232
G
Tas ks Group
252
G
Tas ks Group
73 9_BFM
No (adequate)
emergency
res pons e
18
G
Number of
casualties
56 (5) INJ P
53 Hand
71
INJN
N of C = 1
54 (5) INJ P
50 Upper
Ex tremities , not
further spec ified
228
LCE
Pers on gets
unwell
248 7_BFM
Los s of control
over body
m ov ement
249
LCE
Uncontrolled
m ov ement/
miss tep
38
FOP
(Probably)
permanently
injured
25
G
Hospitalis ation
42
HOSP
NHS = Unknown
27
G
Consequence
clas s
39
FOI
(Probably ) Non
permanently
injured body part
29
G
Abs ence from
work
34 (5) ABS
> 1 week
36 (5) ABS
1 day < abs ence
<= 1 week
40
FO
Unknown
37 (5) ABS
<= 1 day
48 (5) INJ T
120 Multiple
injuries
58 (5) INJ P
55 Wrist
3309 BWT
SEE 1.1.2
227 6_BFM
Subs tandard
c ondition/
fitness of pers on
229
PSBUs er Stability
251 7_BSM
Correct care/
attention /
conc entration of
pers on on ladder
28
FOD
Death
41
HOSP
NH = Not
Hospitalis ed
23
G
Ty pe of injury
47 (5) INJ T
050 Conc us sion
and internal
injuries
57 (5) INJ P
54 Fingers
3308 BWT
SEE BOWTIE 1.1.
3
45 (5) INJ T
030
Dis locations ,
sprains and
strains
46 (5) INJ T
040 Traum atic
am putations
(Loss of body
parts)
72
BWT
SEE BOWTIE 3.2
77 8_BFM
No or inadequate
use of fall
arres tors or
safety nets
26
HOSP
H = Hospitalis ed
31 (5) ABS
> 1 month
21
G
Part of body
injured
19
INJN
N of C = 0
17
G
Dos e-Response
Factors
30 (5) ABS
unk nown
44 (5) INJ T
020 Bone
frac tures
52 (5) INJ P
40 Torso and
organs, not
further spec ified
55 (5) INJ P
51 Shoulder and
shoulder joints
20
BWT
SEE BOWTIE 14.1
80
8_IF
Inadequate fall
arrestor
230
6_B
Fitness /
condition of
person on ladder
250
7_B
Care/ attention/
c oncentration
76 8_BSU
Unk nown
2
LCE
Impac t by fall
209
5_B
Position on
ladder
53 (5) INJ P
41 Rib c age, ribs
including joints
and s houlder
blades
75 9_BSM
Adequate
emergency
res pons e
74
9_B
Emergenc y
Res pons e
81
G
Barrier Failure
Group
212
G
Management
Delivery
Sy stem s GROUP
43 (5) INJ T
010 Wounds and
s uperfic ial
injuries
51 (5) INJ P
30 Bac k,
including s pine
and vertebra in
the back
14 (2) DDF
Soft s urfac e
(soil, sand)
188
4_B
Condition of
support/ s urfac e
186 4_BFM
Subs tandard
surface/ support
85
G
Regulations
violated
161
LCE
Ladder/ person
is hit
189 4_BSM
Right surface/
s upport for the
ladder
191
G
Management
Delivery
Sy stem s GROUP
24 (5) INJ T
000 Ty pe of
injury unknown
or uns pec ified
50 (5) INJ P
20 neck
163
3_B
Location of
ladder
99 (2)
ET
M obile ladders
49 (5) INJ P
10 Head, not
further spec ified
12 (2) DDF
Unknown s urfac e
ty pe
160 3_BFM
Wrong location
of ladder
89
A
Moving up/down
ladder/s tairs/
s teps carry ing
loads
22 (5) INJ P
00 Part of body
injured, not
s pecified
11 (2) DDF
Height of fall
unknown
139
LCE
Ladder/ steps
m ov es
164 3_BSM
Safe location for
the ladder
4 (2)
DDF
Height of fall
9 (2)
DDF
5m < Height of
fall <= 10m
59 (5) INJ P
52 Arm,
including elbow
60 (5) INJ P
60 Lower
Ex tremities , not
further spec ified
67 (5) INJ P
70 Whole body
and m ultiple
sites , not further
s pecified
70 (5) INJ P
99 Other Parts of
body injured
PAGE | 7
3.1 Contact with falling object - cranes, part of cranes or crane loads
...it is NOT to do with cranes:
3 Contact with falling object - other than cranes, part of cranes or crane loads
The person is struck by a missile:
4 Contact with flying object
The object is rolling or sliding (not a moving vehicle)
5 Hit by rolling/sliding object
The object is hanging or swinging when it makes contact
8.2 Contact with hanging/ swinging objects
STRUCK BY AN OBJECT BEING USED OR CARRIED BY A PERSON
The object of contact is carried by another person or it is carried or being used by the victim but it
is not a handheld tool
6 Contact with object person is carrying or using
The object of contact is a hand held tool operated by the victim
7 Contact with hand held tools operated by self
MOVING OR JUMPING INTO AN OBJECT
The victim moves into a stationary object or bumps into a person
9 Moving into an object
CONTACT WITH OR TRAPPING BY MOVING PARTS
The machine has moving parts and the victim contacts those moving parts
8.1 Contact with moving parts of a machine
An object moves against a person and traps them against a static obect
8.3 Trapped between/against something
BURIED OR IMMERSED
Victim is buried by a large amount of a physical substance like earth or grain or sand due to
collapse or downpour
10 Buried by bulk mass
Victim is in a liquid and they lose buoyancy or are trapped under the liquid
23 Impact by immersion in liquid
IN OR ON A VEHICLE WHICH LOSES CONTROL
Victim is in or on a vehicle which loses control
11 In or on moving vehicle with loss of control
PAGE | 9
CONTACT ELECTRICITY
The victim contacts active electrical parts directly, indirectly or by short circuit
12 Contact with electricity
CONTACT EXTREME HOT OR COLD SURFACE OR FIRE
The victim contacts an extremely hot surface or an extremely cold one or an open flame (which is
not a fire burning out of control) because of movement of either the victim or the damage source
13 Contact with extreme hot or cold surfaces or open flame
A fire is caused by a loss of control, including the victims clothes catching fire, and the victim is
exposed to the effects.
17 Fire
HAZARDOUS SUBSTANCE OR ATMOSPHERE
Contact is due to a LOC (LOSS OF CONTAINMENT) of a substance which carries a hazard, such
as by spill or overflow, and it is normal for the containment to be open
14.1 Release of hazardous substance out of Open containment
Contact is due to a LOC (LOSS OF CONTAINMENT) of a substance which carries a hazard
and the containment is normally closed
15 Release of a hazardous substance out of Closed containment
Contact due to explosion - substances are explosive, reactive (solid, liquid or gaseous):
27 Explosion
There is contact with a substance which carries a hazard through ingestion, inhalation or skin or
eye contact but there was not a LOC
14.2 Exposure to hazardous substance without Loss of Containment
There is contact with a hazardous atmosphere and the victim is confined
22.1 Exposure to hazardous atmosphere in confined space
There is contact with a hazardous atmosphere through breathing apparatus
22.2 Exposure to hazardous atmosphere through breathing apparatus
VICITM OF HUMAN AGGRESSION OR ANIMAL
The victim is caused harm by an aggressor who loses control
20.1 Victim of human aggression
PAGE | 10
The full set of Storybuilds which were used for quantification purposes in the risk model are
available in Superfile:
01 D d09 Storybuilds Superfile 070329 LJBMM.sb
The storybuild went through a number of changes due to quality issues. In particular:
1. Pathways were dragged into group boxes, originally meant to be path free, for counting check
purposes. These group boxes, coded G, are not events but structurally hold a set of boxes of the
same typ (eg Activities) together.
2. Mismatches between the accident names and the original names in GISAI were corrected and
these appeared fixed in version 9 of the superfile. A record was kept of the corrections
3.2
A Storybuild is built up from a number of boxes in the software Storybuilder (see Technical Report
11 Storybuilder Software User Manual and Help). These boxes are coded according to the type of
event.
PAGE | 11
PAGE | 12
Figure 4 Examples of box coding using shape and colour and letter codes
The idea is to record an accident by identifying its passage through the Storybuilld
structure which in turn is built up as accidents are analysed
PAGE | 13
Figure 5 Accident scenarios appear as pathways through the Storybuild The lines show the passage of
scenarios through the structure (a green line means some of the boxes have been popped away and
are not being viewed)
Figure 6 Accident and victim counts. The number count below the box indicates number of accident
scenarios with number of victims in parenthesis.
Figure 7 More detailed description visible through a hint. The hint shows box number, name of the box,
the box code, whether it is a popup box and the further description
PAGE | 15
PAGE | 16
|
Figure 8 Typical left to right structure of a storybuild shown in sequence (in reality all the boxes are a single row): Activity - information about
equipment and other characteristics of the situation regulations barrier failure groups and loss of control events, centre event, dose
determining factors, mitigating factors, number of casualties, injuries, hospitalization and consequences
Full details of specific Storybuild structures are given in Technical Report 8 Bowtie models and
quantification.
3.3
The following summarises the main event codings in alphabetical order with examples. This is
illustrative of the content of a storybuild in terms of cause/consequence events and other relevant
information which may be important for understanding how and why accident occur.
PAGE | 17
3.3.1
3.3.2
3.3.3
Also has a numerical coding associated with a barrier number and barrier related events e.g.
1_BFM (Barrier failure mode 1 which is linked to 1_DS, 1_T)
PAGE | 18
3.3.4
3.3.5
PAGE | 19
3.3.6
3.3.7
CE (Centre event)
PAGE | 20
3.3.8
579
DDF
3.3.9
PAGE | 21
3.3.10
3.3.11
PAGE | 22
3.3.12
440
FOI
(Probably) Non
permanently injured
body part
3.3.13
438
FOP
(Probably)
permanently injured
3.3.14
409
Mitigating Factors
434
G
Hospitalisation/
Medical Treatement
PAGE | 23
3.3.15
435
HOSP
H = Hospitalised
437
HOSP
NH = Not Hospitalised
3.3.16
IF (Incidental factor)
3.3.17
470 (5)
INJP
30 Back, including
spine and vertebra in
the back
PAGE | 24
3.3.18
455 (5)
INJT
040 Traumatic
amputations (Loss of
body parts)
3.3.19
294
LCE
Control over vehicle
and/or its route failure
3.3.20
295
PSBVehicle collision
prevention failure
PAGE | 25
3.3.21
REG (Regulations)
140 (3)
REG2
Artikel 3.14.
Connecting routes/
pathw ays
3.3.22
137
SSBPlacement &
protection failure
3.3.23
T (Barrier task)
1_T (Barrier task which is linked to 1_DS, 1_BFM)
PAGE | 26
A summary of the number of accidents per Storybuild is given in Annex 6. The accident
frequencies were obtained by exporting all the paths from the storybuild and calculating the total
and per year frequencies, bearing in mind the number of years spanned by the data. The accident
frequencies are per storybuild, are also given as a % of the total number of accidents per year.
The storybuilds are shown in sequence from the most frequent accident per year to least frequent.
The top scorer is contact with moving parts of a machine accounting for 21% of accidents per
year. Fall from height - roof/platform/floor, Fall from height ladders and Contact with falling objects
(NOT cranes) each account for around 9%. So together these 4 hazards account for 50% of
accidents each year.
4.2
As a further breakdown of these data, Annex 7 shows accident consequence ratios for death:
permanent injury: recoverable injury frequencies. Death is always given the value of 1. The ratios
are dependent on the type of hazard. The accident triangles of Heinrich (1931) who indicated
ratios of 1:29:300 for major: :minor:: no injury inspired this formulation.
Contact with moving parts of a machine, for example, produces a high ratio of permanent injuries
1:57:11 whereas fall on same level has ratios of 1:17:64 indicating the high frequency of
recoverable injuries in relation to number of deaths and permanent injuries. When more deaths
are occurring the ratios get smaller as with contact with electricity which is approximately 1:1:5
(similar for falls from roofs).
4.3
Injuries
Injuries were also evaluated at a more detailed level as shown in Annex 8. This shows the most
frequent hazard per body part injured. For example the head is most frequently injured by flying
objects whereas the lower extremeties are most frequently injured by falling objects not crane
related. Most frequent injury type per hazard is also given, such as amputations for contact with
moving parts of machines and bone fractures for falls from height. Finally a body picture shows the
highlights for each body part.
4.4
A facts and figures sheet was provided for each storybuild giving the causal highlights on 3 sheets
per storybuild. An information sheet giving background and definitions is provided in Annex 9.
The facts and figures sheets are given in Annex 10 for each Storybuild.
The 3 page sheets provide the following information:
PAGE | 27
4.5
Ad hoc questions
The Dutch Labour Inspectorate were particularly interested in industry sector specific and other
(demographic) information concerning the accidents analysed in Storybuilder such as age of the
victim. These data which in combination with the accident analyses could help in formulating
sector specific inspection plans were not directly available in the storybuilds. However, they could
be found for each accident in the GISAI database.
The linkup between GISAI data fields and the storybuild superfile was made in a tool called Story
Filter (see Technical Report 11). Story Filter uses the accident record i.d. to make this link, a a
method of filtering the data in a sequence of filters that can be defined by the user. The filters
include:
Storybuild (1-36)
BIK code - this defines the industry branch according to the Bedrijfsindeling Kamers van
Koophandel The choice can be 2, 3, 4 or 6 digit codes
Job e.g.nurse
Consequences this is the final outcome according to GISAI (not necessarily the same as the
storybuilds)
Sex
Permanent injury according to GISAI (not necessarily the same as the storybuilds)
Accident i.d
PAGE | 29
Figure 9 Story Filter. Here the user has picked the wood industry (20) using 2 figure BIK and analysed
for storybuilds
PAGE | 30
Figure 10 Story Filter. Here the user has picked a storybuild (Contact moving parts machine) and
analysed for 2 figure BIKs
PAGE | 31
Figure 11 Story Filter. Analysis of a single accident using multiple analysis filters
By exporting from Story Filter valuable industry sector specific information can be derived
which can be useful for focusing on dominant accident themes for that sector. A summary
of 2 code BIKs and storybuild frequencies in each sector are given in Annex 11 for
overview.
Using the Story Filter methodology combined with in depth analyses in the Storybuilds
themselves, a number of questions from the Labour Inspectorate were responded to in
depth by RPS Advies. The answers to these ad hoc questions are given in Annex 12.
These reports address the following areas.
Accidents in construction
PAGE | 32
All reports are in Dutch. Each analysis concludes with suggested inspection points. For
example for the accidents in the potatoes, vegetables and fruit sector suggested areas for
inspection, based on the most frequently failing barriers in that sector, are:
1. Guarding of moving parts of machines
2. Marking of danger areas of machines
3. Signaling danger zones, and supervision of work with machines.
4. Information concerning safe working with/ machines, especially when clearing jams
5. Emergency stops for machines
6. Provision and use of personal protective equipment when working with dangerous
substances
7. Storage and labelling of dangerous substances
8. Information about working with dangerous substances
9. First aid
10. Lay-out/infrastructure of the logistics especially separation of pedestrians and vehicles
like forklift trucks
PAGE | 33
PAGE | 34
5 ORGANISATIONAL FACTORS
In total there were 420 safety barriers identified. There were a total of 16,314 identified and 8,293
unknown barrier failures in total. In addition there were 16,007 known management delivery
failures across 8 management delivery systems and 15,632 barrier task failures including 2047
unknowns. The analysis of delivery system and barrier task data was carried out for presentations
to industry and at conferences (e.g. OECD conference on Human Factors in 2007). These data
were totally new and provided insight into the fact that different storybuilds have different
management failures. These data are still being investigated. So, for example, while motivation is
the highest contributor DS (28%) as shown in Figure 12, ergonomics is the highest contributor
(34%) for 7 Contact with handheld tools, and procedures is the highest contributor (26%) for 15
Loss of containment normally closed. Regarding barrier tasks provide and use failures are around
45% each with maintain and monitor around 5% each of known failures.
The implications are that different hazards require different forms of management. This also has
implications for interpreting safety performance indicators.
ALL
30%
28%
25%
20%
16%
15%
14%
13%
11%
10%
5%
9%
8%
2%
0%
Figure 12 Percentage failure contribution of management delivery systems across all storybuilds
PAGE | 35
6 CONCLUSIONS
The analysis of facts and figures data has not yet been exhausted as there are still very many
aspects to consider. Data export enhancements of Storybuilder have provided extensive
possibilities from quick analysis of data in the form of percentages using tree view to exports of a
complete superfile to excel for example where further analyses can be carried out.
The use of Story Filter and its enhancement in terms of speed has greatly expanded the
possibilities for analysis which combine the accident analyses in Storybuilder with sector and
victim specific data of GISAI. Investigations that were carried out on using data cubes to analyse
these data more quickly indicated problems with dealing with the hierarchical nature of storybuilder
data. In the end Story Filter proved to be more useful and is easy to train.
Currently management delivery system data are being further explored. This area is still a big
unknown. Further work on classifying barrier types and examining modes of failure is also
underway.
PAGE | 36
7 REFERENCES
Bellamy L.J. Oh J.I.H., Ale B.J.M., Whiston J.Y., Mud. M.L, Baksteen H., Hale, A.R.,
Papazoglou, I.A., 2006. Storybuilder: The new interface for accident analysis, International
Conference on Probabilistic Safety Assessment and Management, May 13-19, 2006, New
Orleans, ASME, New York, ISBN 0-7918-0244
Bellamy L.J., Ale B.J.M., Geyer T.A.W., Goossens L.H.J., Hale A.R., Oh J.I.H., Mud M.L.,
Bloemhoff A, Papazoglou I.A., Whiston J.Y., 2007. StorybuilderA tool for the analysis of
accident reports,Reliability Engineering and System Safety 92 (2007) 735744
Bellamy L.J., Ale B.J.M., Whiston, J.Y., Mud M.L., Baksteen H., Hale A.R., Papazoglou
I.A., Bloemhoff A., Damen M. and Oh J.I.H., 2008. The software tool storybuilder and the
analysis of the horrible stories of occupational accidents. Safety Science 46 (2008) 186197
Heinrich, H. W. (1931). Industrial Accident Prevention. New York: McGraw Hill.
PAGE | 37