Principles of forensic engineering applied to industrial accidents

Table of Contents
Cover
Foreword by Giomi
Foreword by Chiaia
Foreword by Tee
Preface
Acknowledgement
List of Acronyms
1 Introduction
Who Should Read This Book?
1.2 Going Beyond the Widget!
1.3 Forensic Engineering as a Discipline
References
Further Reading
2 Industrial Accidents
2.1 Accidents
2.2 Near Misses
2.3 Process Safety
2.4 The Importance of Accidents
2.5 Performance Indicators
2.6 The Role of ‘Uncertainty’ and ‘Risk’
References
Further reading
3 What is Accident Investigation? What is Forensic Engineering? What is Risk
Assessment? Who is the Forensic Engineer and what is his Role?
3.1 Investigation
3.2 Forensic Engineering
3.3 Legal Aspects
3.4 Ethic Issues
3.5 Insurance Aspects

3.6 Accident Prevention and Risk Assessment
3.7 Technical Standards
References
Further Reading


4 The Forensic Engineering Workflow
4.1 The Workflow
4.2 Team and Planning
4.3 Preliminary and Onsite Investigation (Collecting the Evidence)
4.4 Sources and Type of Evidence to be Considered
4.5 Recognise the Evidence
4.6 Organize the Evidence
4.7 Conducting the Investigation and the Analysis
4.8 Reporting and Communication
References
Further Reading
5 Investigation Methods
5.1 Causes and Causal Mechanism Analysis
5.2 Time and Events Sequence
5.3 Human Factor
5.4 Methods
References
Further Reading
6 Derive Lessons
6.1 Pre and Post Accident Management
6.2 Develop Recommendations
6.3 Communication
6.4 Safety (and Risk) Management and Training
6.5 Organization Systems and Safety Culture

6.6 Behavior based Safety (BBS)
6.7 Understanding Near misses and Treat Them
References
Further Reading
7 Case Studies
7.1 Jet Fire at a Steel Plant
References
Further readings
7.2 Fire on Board a Ferryboat
References
Further Readings


7.3 LOPC of Toxic Substance at a Chemical Plant
7.4 Refinery's Pipeway Fire
References
Further Readings
7.5 Flash Fire at a Lime Furnace Fuel Storage Silo
Further readings
7.6 Explosion of a Rotisserie Van Oven Fueled by an LPG System
Further Readings
7.7 Fragment Projection Inside a Congested Process Area
Reference
Further Readings
7.8 Refinery Process Unit Fire
Reference
Further readings
7.9 Crack in an Oil Pipeline
References
Further Reading

7.10 Storage Building on Fire
Further Readings
8 Conclusions and Recommendations
References
9 A Look Into the Future
References
Appendix A: Principles on Probability
A.1 Basic Notions on Probability
Index
End User License Agreement

List of Tables
Chapter 02
Table 2.1 Incident typologies and correlated potentiality and magnitude.
Table 2.2 Flammability limits of some gas and vapors.
Table 2.3 MOC values (volume percent oxygen concentration above which
combustion can occur).


Table 2.4 Approximate values of the Auto Ignition Temperature for some
substances.
Table 2.5 Storage pressure of some compressed gasses.
Table 2.6 Classification of flammable liquids according to CLP Rule (EU Directive
1272/08).
Table 2.7 Classification and FPT of some common flammable liquids.
Table 2.8 Extinguishers and their actions.
Table 2.9 Categories of growth velocity of fire.
Table 2.10 Values of t 1 for some materials commonly used.
Table 2.11 Characteristic explosion indexes for gasses and vapors.
Table 2.12 Characteristic explosion indexes for powders.

Chapter 03
Table 3.1 Example of “what if” analysis [23].
Table 3.2 Guide words for HAZOP analysis.
Table 3.3 Extract of example of HAZOP analysis.
Table 3.4 Subdivision of the analysed system into areas.
Table 3.5 Subdivision of the analysed system into areas.
Table 3.6 List of typical consequences.
Table 3.7 HAZID worksheet.
Table 3.8 Relations between discrete values of SIL and continuous range of PFD
and PFH.
Chapter 04
Table 4.1 Possible checklist for developing an investigation plan.
Table 4.2 Investigation team members should and should not.
Table 4.3 Some containers for sampling, their main features, pros, and cons.
Table 4.4 Checklists to evidence examination.
Table 4.5 Forms of data fragility.
Table 4.6 Digital evidence and their volatility.
Table 4.7 Example of form to use for the collection of pictures.
Table 4.8 Summary of the evidence and deductions.
Table 4.9 Summary of technical assessments, explosion of wool burrs at
Pettinatura Italiana.


Table 4.10 Sequence of events that led to the explosion.
Table 4.11 Summary of the evidence and deductions.
Table 4.12 Summary of the evidence and deductions
Table 4.13 Summary of the evidence and deductions.
Chapter 05
Table 5.1 Examples of unsafe acts and conditions.
Table 5.2 Example of spreadsheet event timeline.

Table 5.3 Example of Gantt chart investigation timeline.
Table 5.4 Example of human factors in process operations.
Table 5.5 Human and management errors.
Table 5.6 Definition of BRFs in Tripod.
Table 5.7 Causal factor types and problem categories.
Chapter 06
Table 6.1 PIF (current configuration).
Table 6.2 PIF (A configuration).
Table 6.3 PIF (POST configuration).
Table 6.4 Frequency of the considered incidental hypotheses
Table 6.5 Comparative table for teaching differences between incidents and
nonincidents.
Chapter 07
Table 7.1.1 General information about the case study.
Table 7.1.2 Record of the supervisor systems (adapted from Italian).
Table 7.1.3 Threshold values according to Italian regulations.
Table 7.1.4 Summary of the investigation.
Table 7.2.1 General information about the case study.
Table 7.2.2 Some lessons learned from the incident, written so that they can also be
used in other business sectors, such as the process industry.
Table 7.3.1 General information about the case study.
Table 7.4.1 General information about the case study.
Table 7.5.1 General information about the case study.
Table 7.5.2 Chemical substances involved.


Table 7.6.1 General information about the case study.
Table 7.6.2 Reference parameters for scenario b).
Table 7.6.3 Scenario a), release characteristics.
Table 7.6.4 Identification of simulations related to scenario a) indicating the

breaking point and of the released phase.
Table 7.6.5 Results of simulations with C Phast code.
Table 7.7.1 General information about the case study.
Table 7.7.2 Simulation results for steam pressure and temperature variation.
Table 7.7.3 Simulations characterised by a Dynamic Increase Factor.
Table 7.7.4 Results for impacts.
Table 7.8.1 General information about the case study
Table 7.8.2 Tabular timeline of the main events.
Table 7.9.1 General information about the case study.
Table 7.10.1 General information about the case study.

List of Illustrations
Chapter 01
Visual explanation of the addition rule of probability, through Venn diagrams.
Visual explanation of the conditional probability, through Venn diagrams.
Chapter 01
Figure 1.1 The onion like structure between immediate causes and root causes.
Figure 1.2 Galileo Galilei (left) and Roger Bacon (right): two of the brightest
scientists of the world who supported the scientific method.
Chapter 02
Figure 2.1 Causes of industrial accidents in chemical and petrochemical plants in
the United States in 1998.
Figure 2.2 Components related to the industrial accidents in chemical and
petrochemical plants in the United States in 1998.
Figure 2.3 The Fire Triangle.
Figure 2.4 The different mechanisms of heat transfer.
Figure 2.5 The involvement of deck no. 3 of the Norman Atlantic into the fire, due
to radiation: simulation and evidence (plastic boxes, melted at the top).

Figure 2.6 The chromatic scale of the temperatures in a gas fuel.
Figure 2.7 Graphical representation of the concepts of LFL and UFL.
Figure 2.8 Relations among the flammability properties of gas and vapors.
Figure 2.9 Comparison among the MIE of gases and vapors and the energy of
electrostatic sparks. Adapted from [11].
Figure 2.10 Different colors at the access of deck 3 and 4 of the Norman Atlantic,
suggesting two different typologies of fire. The oxygen controlled fire at deck 3 (on
the right) and fuel controlled fire at deck 4 (on the left).
Figure 2.11 Evolution of a fire.
Figure 2.12 Shock front and pressure front in detonations and deflagrations.
Figure 2.13 Primary and secondary dust explosion.
Figure 2.14 Incidental scenarios and their genesis.
Figure 2.15 An example of Flash Fire.
Figure 2.16 On the left, a modelled jet fire for a fire investigation
Figure 2.17 Example of Pool Fire.
Figure 2.18 Schematic representation of a fireball in the stationary stage.
Figure 2.19 A Vapor Cloud Explosion test.
Figure 2.20 Sequence events to BLEVE.
Figure 2.21 Example of BLEVE.
Figure 2.22 Differences between accident (a), near miss (b), and undesired
circumstance (c).
Figure 2.23 Contributing factors in improving loss prevention performance in the
process industry.
Figure 2.24 The evolution of safety culture.
Figure 2.25 Example of BFD for the production of benzene by the
HydroDeAlkylation of toluene (HDA).
Figure 2.26 Example of PFS for the manufacture of benzene by Had.
Figure 2.27 Example of P&ID for the production of benzene by Had.
Figure 2.28 Principles of incident analysis.
Figure 2.29 The importance of incident investigation.

Figure 2.30 Steps of incident analysis.
Figure 2.31 Temperatures at the Seveso reactor.


Figure 2.32 A photograph of the signs used to forbid access into the infected areas
in Seveso.
Figure 2.33 Simplified conceptual Bow Tie of Seveso incident.
Figure 2.34 The chemical plant in Bhopal after the incident.
Figure 2.35 Arrangement of reactors and temporary bypass.
Figure 2.36 The chemical plant in Flixborough after the incident.
Figure 2.37 The Deepwater Horizon drilling rig on fire.
Figure 2.38 Application of the Apollo RCA™ Method using RealityCharting® to
the Deepwater Horizon incident.
Figure 2.39 Application of the Apollo RCA™ Method using RealityCharting® to the
Deepwater Horizon incident. Used by permission. Taken from [43].
Figure 2.40 Application of the Apollo RCA™ Method using RealityCharting® to
the Deepwater Horizon incident.
Figure 2.41 Some LPG spherical tanks during the San Juanico disaster.
Figure 2.42 The IHLS.
Figure 2.43 The site after the incident.
Figure 2.44 Pipe penetrations for the loss of seal between pipes and walls.
Figure 2.45 RCA of the Bouncefield explosion developed by company Governors BV
(NL).
Figure 2.46 Example of a risk matrix.
Chapter 03
Figure 3.1 Phases in accident investigation.
Figure 3.2 The Conclusion Pyramid. Source: Adapted from [10]. .
Figure 3.3 A damaged item under investigation.
Figure 3.4 Handling of an item under investigation.
Figure 3.5 Explosion of flour at the mill of Cordero di Fossano (CN). The damages

caused involved many insurance related consequences.
Figure 3.6 Feed line propane butane separation column. Source: Adapted from
[23]. Reproduced with permission.
Figure 3.7 Top Gates of the Fire Safety Concepts Tree.
Figure 3.8 Use of the Scientific Method according to NFPA 921. Source: Adapted
from [25]. Reproduced with permission.
Chapter 04


Figure 4.1 The forensic engineering workflow.
Figure 4.2 A detailed investigative workflow.
Figure 4.3 During the preliminary and onsite investigation, remember to wear the
PPE.
Figure 4.4 Collection of some portions of metal sheet from the processing tape and
their subsequent enumeration, ThyssenKrupp investigation.
Figure 4.5 Samples in glass cans and in plastic bags with zipping closure.
Figure 4.6Figure 4.6 The collection process of digital data.
Figure 4.7 The sequence of smoke sensors activation. In grey the first group, in
dark grey the following 60 seconds, in dashed circle the first open loop and in
dashed circle and dashed rectangles the residual activation, all in less than 180
seconds
Figure 4.8 The wall collapse a few minutes after the arrival of the fire brigade unit.
Figure 4.9 Rolls of expanded LDPE with flame retardant included invested from
heat.
Figure 4.10 Identification of fire extinguishers by tags (on the left) and
acknowledgement by photography (on the right), ThyssenKrupp investigation.
Figure 4.11 Detail of a small imperfection on the edge of a metal sheet,
ThyssenKrupp investigation
Figure 4.12 Straight graduated ruler, Norman Atlantic fire investigation.
Figure 4.13 Example of metadata related to a photo taken during the ThyssenKrupp

investigation.
Figure 4.14 Example of keywords for filtering the picture of a collection.
Figure 4.15 Example of visualised information when finding a photograph by
keywords.
Figure 4.16 Example of Pareto Chart.
Figure 4.17 Evidence: overpressure damage to a flours repump duct flange.
Figure 4.18 Building (south side) with noticeable damage from excess pressure.
Figure 4.19 Building (north side) with widespread collapse primarily from static
collapse.
Figure 4.20 Explosion of wool burrs, state of places.
Figure 4.21 Explosion of wool burrs, state of the places, card rooms.
Figure 4.22 Explosion of wool burrs, burrs storage boxes.
Figure 4.23 Explosion of wool burrs, state of places, burrs collection boxes corridor


with visible in the foreground signs of material fragment projection on the white
bin.
Figure 4.24 Diagram of the methane and air flow rates (a) during the moments
before the explosion and (b) enlarged detail.
Figure 4.25 Abatement system, detail of exploded fragment.
Figure 4.26 Reduction system, detail of the flue discharge pipe inside the cyclone.
Figure 4.27 State of places and damage to the abatement system.
Figure 4.28 Remains of the bag filter.
Figure 4.29 Sample Chain of custody form. Taken from [1].
Figure 4.30 Front view of the conic spiral.
Chapter 05
Figure 5.1 Fishbone diagram. Step 1: Identify the problem.
Figure 5.2 Fishbone diagram. Step 2: categorise the causes.
Figure 5.3 Fishbone diagram. Step 3: identify possible causes.
Figure 5.4 Example of event and causal factor diagram.

Figure 5.5 Domino theory by Heinrich (1931) [6].
Figure 5.6 Loss Causation Model by Bird [7].
Figure 5.7 Sequence of dominos.
Figure 5.8 Events and causal factors analysis.
Figure 5.9 The different nature of human and technical systems.
Figure 5.10 AND and OR combinations in logic trees.
Figure 5.11 Multiple levels logic tree.
Figure 5.12 Procedure to create a logic tree.
Figure 5.13 Example of timeline developed for the Norman Atlantic investigation
(see Paragraph 7.2 for details).
Figure 5.14 STEP worksheet.
Figure 5.15 An example of STEP diagram for a car accident.
Figure 5.16 Row and column tests for STEP method.
Figure 5.17 STEP worksheet with safety problems.
Figure 5.18 Thought behavior result model.
Figure 5.19 Stimulus response model.


Figure 5.20 Two prongs model.
Figure 5.21 Two pronged model – accident analysis.
Figure 5.22 Categorization of human factors in petroleum refinery incidents.
Figure 5.23 Method to determine the type of human error.
Figure 5.24 Reason's classification of human errors.
Figure 5.25 Causes of human error.
Figure 5.26 Self correcting process step.
Figure 5.27 MTO worksheet.
Figure 5.28 Swiss cheese model by Reason.
Figure 5.29 Workflow of structured methods.
Figure 5.30 Workflow of pre structured methods.
Figure 5.31 The deductive logic process.

Figure 5.32 The inductive logic process.
Figure 5.33 The morphological process.
Figure 5.34 Example of root causes arranged hierarchically within a section of a
predefined tree.
Figure 5.35 Top portion of the generic MORT tree.
Figure 5.36 MORT Maintenance Example.
Figure 5.37 Difference between SCAT and BSCAT™ (Courtesy of CGE Risk
Management Solutions (NL)).
Figure 5.38 Events types in a BSCAT™ diagram (Courtesy of CGE Risk
Management Solutions (NL)).
Figure 5.39 Incident barrier states (Courtesy of CGE Risk Management Solutions
(NL)).
Figure 5.40 Relation between barrier state and barrier lifecycle (Courtesy of CGE
Risk Management Solutions (NL)).
Figure 5.41 Example BSCAT™ diagram (Courtesy of CGE Risk Management
Solutions (NL)).
Figure 5.42 Example BSCAT™ diagram (Courtesy of CGE Risk Management
Solutions (NL)).
Figure 5.43 The bowtie diagram (Courtesy of CGE Risk Management Solutions
(NL)).
Figure 5.44 Bowtie risk assessment & incident analysis (Courtesy of CGE Risk


Management Solutions (NL)).
Figure 5.45 Example a Tripod Beta diagram (Courtesy of CGE Risk Management
Solutions (NL)).
Figure 5.46 Possible Tripod Beta appearances (Courtesy of CGE Risk Management
Solutions (NL)).
Figure 5.47 Accident mechanism according to HEMP method.
Figure 5.48 Example of a BFA diagram (Courtesy of CGE Risk Management

Solutions (NL)).
Figure 5.49 Example of a BFA diagram (Courtesy of CGE Risk Management
Solutions (NL)).
Figure 5.50 BFA core elements (Courtesy of CGE Risk Management Solutions
(NL)).
Figure 5.51 General structure of a BFA diagram (Courtesy of CGE Risk
Management Solutions (NL)).
Figure 5.52 Event chaining in BFA (Courtesy of CGE Risk Management Solutions
(NL)).
Figure 5.53 Defeated barriers are not BFA events (Courtesy of CGE Risk
Management Solutions (NL)).
Figure 5.54 Barrier identification in BFA (Courtesy of CGE Risk Management
Solutions (NL)).
Figure 5.55 Correct and incorrect barrier identification in BFA (Courtesy of CGE
Risk Management Solutions (NL)).
Figure 5.56 BFA analysis (Courtesy of CGE Risk Management Solutions (NL)).
Figure 5.57 Levels of analysis.
Figure 5.58 TapRooT® 7 Step Major Investigation Process.
Figure 5.59 The TapRooT® Basic Investigation Process.
Figure 5.60 Example of SnapCharT®.
Figure 5.61 The Corrective Action Helper Module.
Figure 5.62 Apollo RCA™ diagram (it continues in Figure 5.63). Used by
permission from “The RealityCharting® Team”.
Figure 5.63 Apollo RCA™ diagram (it continues from Figure 5.62). Used by
permission from “The RealityCharting® Team.
Figure 5.64 Example of Reason© RCA screenshot.
Figure 5.65 Numerical simulations in CFD to support the incident investigation of


the Norman Atlantic Fire.

Figure 5.66 Basic structure of a Fault Tree.
Figure 5.67 Example of fault tree, taking inspiration from Åsta railway incident.
Figure 5.68 Flammable liquid storage system.
Figure 5.69 Example of FTA for a flammable liquid storage system.
Figure 5.70 The structure of a typical ETA diagram.
Figure 5.71 Event Tree Analysis for the Åsta railway accident.
Figure 5.72 Pipe connected to a vessel.
Figure 5.73 Example of Event Tree for the pipe rupture.
Figure 5.74 Layers of defence against a possible industrial accident.
Figure 5.75 A comparison between ETA and LOPA's methodology.
Chapter 06
Figure 6.1 Emergency management is a crucial part of the overall safety
management system.
Figure 6.2 Flowchart for implementation and follow up.
Figure 6.3 Recommendations flowchart.
Figure 6.4 Workflow for recommendations and their monitoring.
Figure 6.5 Fault Tree Analysis, current configuration (ANTE).
Figure 6.6 Fault Tree Analysis, a better configuration (A configuration).
Figure 6.7 Fault Tree Analysis, the best configuration (POST configuration).
Figure 6.8 Frequency estimation of the scenario “Oxygen sent to blow down,
during start up of reactor of GAS1”.
Figure 6.9 Risk based cost optimization.
Figure 6.10 Proactive and reactive system safety enhancement.
Figure 6.11 Relationship among incidents, near misses and nonincidents.
Chapter 07
Figure 7.1.1 Area involved in the accident. Right, unwinding section of the line, left,
the front wall impinged by flames.
Figure 7.1.2 The flattener and the area involved in the accident. Details of the area
struck by the jet fire, view from the front wall.
Figure 7.1.3 Details of the hydraulic pipe that provoked the flash fire.

Figure 7.1.4 Map of the area struck by the jet fire and by the consequent fire. The


dots represent the presumed position of the workers at the moment the jet
originated.
Figure 7.1.5 Footprint of the jet fire on the front wall.
Figure 7.1.6 Timescale of the accident. F.1 is the time interval in which the ignition
occurred. F.2 is the time interval in which it is probable that the workers noticed
the fire. The group 5 and group 6 events are defined as in Table 7.1.2.
Figure 7.1.7 The domain used in the FDS fire simulations.
Figure 7.1.8 Simulated area, elevation [1].
Figure 7.1.9 Jet fire simulation results: flames at 1 s from pipe collapse.
Figure 7.1.10 Jet fire simulation results: flames at 2 s from pipe collapse.
Figure 7.1.11 Jet fire simulation results: flames at 3 s from pipe collapse.
Figure 7.1.12 Jet fire simulation results: temperature at 1 s from pipe collapse.
Figure 7.1.13 Jet fire simulation results: temperature at 2 s from pipe collapse.
Figure 7.1.14 Jet fire simulation results: temperature at 3 s from pipe collapse.
Figure 7.1.15 Scheme of the hydraulic circuits with two position (a) and three
position (b) solenoid valves.
Figure 7.1.16 Event tree of the accident. The grey boxes indicate a lack of safety
devices.
Figure 7.1.17 Damages on the forklift.
Figure 7.1.18 Frames from the 3D video, reconstructing the incident dynamics.
Figure 7.2.1 Longitudinal section of the ship, with fire compartments.
Figure 7.2.2 Left: open fire damper of the garage ventilation. Right: local command
at deck 4 for closing the fire dampers.
Figure 7.2.3 Closed intercept valve between the emergency pump and the drencher
collector.
Figure 7.2.4 The valves opened in the valve house are those activating the drencher
at deck 3 (instead of deck 4).

Figure 7.2.5 Left. The drencher plan located in the drencher room. Right. Details of
the instruction on the plan.
Figure 7.2.6 Recognition and collection of evidence about the power supply on
board.
Figure 7.2.7 Localised bending of transversal beams and V shaped traces of smoke
on the bulkhead. The majority of the fire load is attributable to the olive oil tanks.
Figure 7.2.8 Lateral openings on deck 4.


Figure 7.2.9 CFD simulations: single truck combustion and 3D pictures of the first
instants of fire at deck no. 4, with smoke emission and flames from the openings
on the starboard side of the ferryboat.
Figure 7.2.10 CFD simulation describing the heat transfer by radiation through the
metal plate between decks no. 3 and no. 4. Conditions of the plastic boxes inside a
truck on deck no. 3.
Figure 7.2.11 General RCA logic tree.
Figure 7.2.12 Detailed RCA logic tree.
Figure 7.2.13 Part of the timeline of the incident.
Figure 7.2.14 Photos taken inside the ferryboat from Villa to Messina, 2016.
Figure 7.2.15 Collection form used during the discharge operations.
Figure 7.2.16 The reconstructed cargo plan at deck no. 3 and no. 4.
Figure 7.2.17 An example of a vehicle identity record.
Figure 7.2.18 Functional diagram of Rutter VDR 100G2 and corresponding IMO
requirements.
Figure 7.2.19 “Propulsion” screen example from system VDR Playback Version
4.5.4.
Figure 7.2.20 Connections schematic between DPU and the partially
undocumented Data Discrete acquisition Units.
Figure 7.2.21 Extract from MSC/Circ. 1024.
Figure 7.2.22 Example 1 of RAW data from FRM with bogus characters.

Figure 7.2.23 Example 2 of RAW data from FRM with bogus characters.
Figure 7.3.1 Causal factors diagram (part 1/4).
Figure 7.3.2 Causal factors diagram (part 2/4).
Figure 7.3.3 Causal factors diagram (part 3/4).
Figure 7.3.4 Causal factors diagram (part 4/4).
Figure 7.4.1 Damages of the piping uphill the road. Gash caused by BLEVE.
Figure 7.4.2 Some damaged pipes downwards the road. There is also the pipe of the
fire system.
Figure 7.4.3 Transversal section of the subway before the incident. Taken from [2].
Figure 7.4.4 Photos of the extinguishment operation. Used by permission.
Figure 7.4.5 An helicopter view of the area. Used by permission.
Figure 7.4.6 Graphical visualization of the found shortcomings.


Figure 7.4.7 Graphical visualization of the defined fire strategy.
Figure 7.4.8 Transversal section of the subway after the incident.
Figure 7.5.1 Area involved in the accident.
Figure 7.5.2 The bottom crawl space, with a discrete part of the sawdust bulk
collapsed, generating a dust cloud ignited probably from a pool of burning sawdust
inside the silo. The water is spayed by fire service after the flash fire event.
Figure 7.5.3 The sequence of the underestimated and unespected hight speed
discharge event, generating the saw dust cloud, with the flash fire ignited in the
last image.
Figure 7.5.4 The smouldering combustion in the saw dust discharged by the silo, in
the occurrence of the event.
Figure 7.5.5 Footprint of the flash fire on the front wall of the shed in front of the
discharge hole.
Figure 7.5.6 The development of the flash fire could be deducted by the burned
trees. The parked bobcat resulted in being ignited.
Figure 7.5.7 The silo with the baghouse filter at its top. See the vents.

Figure 7.5.8 Elements of a Flash Fire and the Explosion Pentagon.
Figure 7.6.1 The van after the accident.
Figure 7.6.2 Gas cylinders removed as exhibits.
Figure 7.6.3 Valve P.R. TA W brev. DN 1/4”.
Figure 7.6.4 Copper pipe and fittings found on the ground behind the van.
Figure 7.6.6 Cylinder A with details of the Fire Brigade labelling, top photo, and of
the Expert, photo below.
Figure 7.6.7 Cylinder B with details of the Fire Brigade labelling, top photo, and of
the Expert, photo below.
Figure 7.6.8 Cylinder C with details of the Fire Brigade labelling, top photo, and of
the Expert, photo below.
Figure 7.6.9 Cylinder D, in particular the base (in the background cylinder A), the
ogive and the coating with labelling of the Expert.
Figure 7.6.5 LPG system diagram indicating the 3 points of possible catastrophic
rupture hypothesised during simulations.
Figure 7.6.10 Series of frames from “Guastalla tragedia al mercato.avi”.
Figure 7.6.11 Still image from “video0054.mp4”.
Figure 7.6.12 Still image from “Untitled.avi”.


Figure 7.7.1 Ruptured steel box.
Figure 7.7.2 Process unit tridimensional layout.
Figure 7.7.3 Process unit involved in the incident tridimensional layout from the
3D laser scanning of the area and the identification of the piping containing
hazardous substances.
Figure 7.7.4 Model setup of 2D simulations in rotational symmetry: supported box
(left) and free box (right).
Figure 7.7.5 Velocity profiles of the top plate for different pressure temperature
combinations (case “box supported”).
Figure 7.7.6 Velocity profiles of the top plate for different pressure temperature

combinations (case “unsupported box”).
Figure 7.7.7 Numerical model for launch velocity investigation (steam pressure –
plate velocity relation has been studied with a specific hydrocode named SPEED©
by Numerics GmbH).
Figure 7.7.8 Launch velocity of the top plate versus box internal pressure.
Figure 7.7.9 Numerical model for stress investigation.
Figure 7.7.10 Plastic deformation of intact box at different internal pressures: 35
bar (left), 50 bar (middle) and 65 bar (right).
Figure 7.7.11 Main stresses in the weld determined from the numerical
simulations.
Figure 7.7.12 Box deformation: simulation 35 bar (top), 50 bar (middle) and real
box measurements (bottom).
Figure 7.7.13 Total box deformation versus internal pressure.
Figure 7.7.14 Autodyn 3D© model of the box and plate.
Figure 7.7.15 Numerical model with partly connected top plate, representing the
delay condition observed during the box rupture.
Figure 7.7.16 Results of simulations with delayed failure of the welds (in the
pictures 1,5 ms delay and 2,0 ms delay).
Figure 7.7.17 Evaluation of top plate velocity from simulation with 2.0 ms failure
delay.
Figure 7.7.18 Impact Conditions (tip, edge, face).
Figure 7.7.19 FE Model showing symmetry along the shotline.
Figure 7.7.20 Validation activity.
Figure 7.7.21 Maximum plastic strain. Top picture 99 m/s impact into the pipe.


Bottom picture – 143 m/s impact into the pipe. Plastic Strain level held constant in
both simulations.
Figure 7.7.22 Crack / perforation criteria in the FE method.
Figure 7.7.23 Damage evaluation using cut planes at 2mm increments: 6 mm long

hole as per the damage criteria described in paragraph 7.7.4.4.
Figure 7.7.24 Plastic strains & deformation: 8_40_BA1002.
Figure 7.7.25 Modifications of the FI BLAST© code.
Figure 7.7.26 FI BLAST© tool: impacting trajectories and pipe damage indicated in
grey as shown inside the calculation code to the user.
Figure 7.7.27 Damage Function for Pipe 8 40 BA Edge Impact. Damage = 1
indicates a hole in the pipe. Damage = 0 indicates possible plastic deformation but
no holes and no cracks. Cracks begin to form, but they do not create a hole.
Figure 7.7.28 Indentation function (crack depth due to loss of material from the
impact) for Pipe 8 40 BA in the impact location. Black diamonds indicate
simulation results. Linear interpolation is used between know points.
Figure 7.7.29 Incident Effects Results.
Figure 7.7.30 Comparison of consequences: Top Events from Safety Report Vs. new
HYPs from fragment study. Flammable top events comparison.
Figure 7.8.1 Block Flow Diagram of the light fuel treatment section, before the
incident
Figure 7.8.2 Block Flow Diagram of the heavy fuel treatment section, before the
incident
Figure 7.8.3 Photos of the incident
Figure 7.8.4 Steel structure damaged
Figure 7.8.5 Block Flow Diagram of the light fuel treatment section, after the
incident
Figure 7.8.6 Block Flow Diagram of the heavy fuel treatment section, after the
incident
Figure 7.8.7 Plan view before the incident
Figure 7.8.8 Plan view after the incident
Figure 7.8.9 Unit 1700. Arrangement of equipment before the incident
Figure 7.8.10 Unit 1700. Arrangement of equipment after the incident
Figure 7.8.11 Forensic engineering highlights about evidence collection, tagging,
and movement

Figure 7.8.12 Simulations carried out to validate the accidental hypothesis about
the fire dynamics Radiation at 5 (top) and 10 meters (bottom) by pool fire, in
different weather condition (2F and 5D).
Figure 7.9.1 Oil Pipeline near Genoa, affected by the rupture. It is evident the crater
formed in the soil due to leaked oil pressure.
Figure 7.9.2 Oil pipeline formed by two pipes with different diameter: 16” pipeline
was affected by the rupture. Images show the pipeline after the excavation to
sample the broken segment.
Figure 7.9.3 Detail of the segment affected by fracture and fluid (oil and water,
alternate) direction when the accident occurred.
Figure 7.9.4 The segment affected by fracture after sampling and details of external
corrosion, related to the age of the pipe.
Figure 7.9.5 Pipeline portions destined to mechanical tests and chemical analysis.
Figure 7.9.6 Pipe segment in which the fracture along the longitudinal line “h 6”
and the letter “A” identifying one of the two edges of the pipe (the other one is
called “B”) are shown. Along the length of the fracture, different positions named
from A1 to A33 are marked.
Figure 7.9.7 Thickness measured with ultrascan along four longitudinal lines on
the pipe.
Figure 7.9.8 Crack face thickness measured by ultrascan. Similar data were
obtained with a mechanical comparator.
Figure 7.9.9 Outer diameters (in light grey, in mm) and corresponding thickness
(in white, in mm).
Figure 7.9.10 FEM Model – Global view.
Figure 7.9.11 Deformed Mesh – Global view.
Figure 7.9.12 Von Mises stresses and deformed mesh – Global view.
Figure 7.9.13 Principal stress σ1 (circumferential) along generator “h 6”.
Figure 7.9.14 On the left: Principal stress σ2 (longitudinal) along line “h 6” – On

the right: Principal stress σ3 (radial) along line “h 6”. It is noted that maximal
values are on the edge, at the external supports (so they are fictitious), here not
visible.
Figure 7.9.15 Von Mises stresses calculated along the longitudinal line “h 6”.
Figure 7.10.1 Photo of the burned roof and the installed PV system.
Figure 7.10.2 Curve of the maximum fire spread rate values v on roof surface
(surface composed of modules of area equal to 1 m2 placed continuously one to
another one). Cases with bottom surface temperature Te equal to 200 °C and


300 °C. The case with more heating (300 °C) is clearly with a bigger rate.
Figure 7.10.3 The PV thin film.
Figure 7.10.4 The burned layers of the roof.
Chapter 09
Figure 9.1 Virtual recognition of some signs due to the heat.
Figure 9.2 Record on the timeline of the performed actions during the geometric
survey.


Principles of Forensic Engineering Applied
to Industrial Accidents
Luca Fiorentini
TECSA S.r.l.
Italy
LucaMarmo
Politecnico di Torino
Italy


Copyright

This edition first published 2019
© 2019 John Wiley & Sons Ltd
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any
form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by law. Advice
on how to obtain permission to reuse material from this title is available at />The right of Luca Fiorentini and Luca Marmo to be identified as the authors of this work has been asserted in accordance
with law.
Registered Offices
John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA
John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
Editorial Office
The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK
For details of our global editorial offices, customer services, and more information about Wiley products visit us at
www.wiley.com.
Wiley also publishes its books in a variety of electronic formats and by print on demand. Some content that appears in
standard print versions of this book may not be available in other formats.
Limit of Liability/Disclaimer of Warranty
While the publisher and authors have used their best efforts in preparing this work, they make no representations or
warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all
warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No
warranty may be created or extended by sales representatives, written sales materials or promotional statements for this
work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of
further information does not mean that the publisher and authors endorse the information or services the organization,
website, or product may provide or recommendations it may make. This work is sold with the understanding that the
publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be
suitable for your situation. Y ou should consult with a specialist where appropriate. Further, readers should be aware that
websites listed in this work may have changed or disappeared between when this work was written and when it is read.
Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but
not limited to special, incidental, consequential, or other damages.
Library of Congress Cataloging in Publication Data
Names: Fiorentini, Luca, 1976 author. | Marmo, Luca, 1967 author.

Title: Principles of forensic engineering applied to industrial accidents /
 Luca Fiorentini, Prof. Luca Fiorentini, TECSA S.r.l., IT, Luca Marmo,
 Prof. Luca Marmo, Politecnico di Torino, IT.
Description: First edition. | Hoboken, NJ, USA : Wiley, 2019. | Includes
 bibliographical references and index. |
Identifiers: LCCN 2018034915 (print) | LCCN 2018037469 (ebook) | ISBN
 9781118962787 (Adobe PDF) | ISBN 9781118962794 (ePub) | ISBN 9781118962817
 (hardcover)
Subjects: LCSH: Forensic engineering. | Industrial accidents. | Accident
 investigation–Case studies.
Classification: LCC TA219 (ebook) | LCC TA219 .F57 2018 (print) | DDC
 363.11/65–dc23


LC record available at />Cover Design: Wiley
Cover Image: © Phonix_a/GettyImages


Dedication
To my wonderful family: to my beloved wife Sonia and to my incredible children
Riccardo, Lodovico and Ettore.
To all those who, thanks to this book, will take their first steps in the world of forensic
engineering or increase their interest in this fascinating discipline.
Luca Fiorentini
To Baba, Beat, Bibi, Chicco.
To all those guys that believe in science, evidences and knowledge.
Luca Marmo