Safety Cases and Safety Reports

Meaning, Motivation and Management

Richard Maguire

• Publisher: CRC Press
• Year: 2017
• Print ISBN: 9781138075320
• eBook ISBN: 9781317059882
• Edition: 1
• Topic: Technology & Engineering
• Subtopic: Transportation Industry
• Index: Technology & Engineering

The safety case and its associated reports are quickly becoming not only a mechanism for achieving safety goals, but also a valuable decision-support asset, and a vital industrial liability management tool. Recent developments in industry have led to safety cases being frequently required as contractual deliverables as part of large and complex commercial programmes. A safety case consists of a rational argument and detailed evidence to justify and demonstrate that a system or product is tolerably safe in its use, and that it has a management programme to ensure that this remains so. The safety case report is the snap-shot presentation of the arguments and evidence demonstrating the contemporary safety performance of the system and the programme that is in place. This book, written from personal experience and reference, provides a concentrated source document for assessing and constructing safety cases and safety case reports - from understanding their purposes, through their development and on to their presentation.

Chapter One

Accidents and Safety

Introduction

At whatever stage in your life you are starting to read this book, you will have been aware of disasters in the world. Ever since William Huskisson MP became the first person to be killed on UK Railways in September 1830 on the opening day of the Liverpool to Manchester line, the record of industrial accidents and disasters has been added to with frightening regularity. Even in recent history when disasters have become global media events the list keeps on growing. Table 1.1 contains a list of relatively recent events that may be classed as disasters – certainly by those effected.

Probably everyone reading this now will be recalling memories of these or some dreadful accident that occurred to them, someone they knew, at some place they knew or something else that became a national tragedy, to the extent that it was lead story for days and actually has anniversary memorials. I can think of far too many of these.

However, with each occurrence of harm, injury or loss that takes place, engineers grow more informed about what happens in the world that they build. Design and operating improvements are mandated, codes of better practice are developed and protection and information schemes are put in place. The goal of all these approaches is to not only ensure that similar events do not happen again, but that as time progresses, the world becomes collectively more safe. Each replacement product, system or process should be safer than the one it replaces; each brand new product, system or process should be compared with existing items to benchmark and improve on its safety performance.

Of course it is far better not to have to wait for an accident to occur in order to prevent any similar future ones happening. Humanity is thinking very hard about how accidents initiate, develop and propagate into disasters, such that they can be prevented before they have opportunity to cause harm, injury or loss. Many industries and countries have authorities and inspector organisations that research and police hazardous areas of work and judge safety performance. Evidence is often called for in demonstration of safety performance and this has many beneficial features from identifying areas for improvement to actually providing defence evidence in legal cases.

Table 1.1 Examples of Recent Accidents and Disasters

Industry	Description	Date	Cause(s)	Impact
Rail	Kings Cross	1987	Fire / smoke	31 fatalities
	Arizona	1997	Bridge failure	116 injuries
	Paddington	1999	Training / signal design	31 fatalities, £2m fine
Chemical	Flixborough	1974	Explosion	28 fatalities
	Bhopal	1984	Toxic gas	2500+ fatalities
	Piper Alpha	1988	Fire	167 fatalities
Nuclear	TMI	1979	Component failure	Political disaster
	Chernobyl	1986	Radiation	31 fatalities
	Tokaimura	1999	Radiation / human error	2 fatalities
Defence	Dhahran	1991	Missile software	28 fatalities
	Chinook ZD576	1994	Human error / Software (?)	29 fatalities
	Osprey Marana	2000	Craft stability / human error	19 fatalities
Construction	Milford Haven	1970	Design flaw	Policy change
	Daegu subway	1995	Gas explosion	101 fatalities
	Toledo Ohio	2004	Anchor procedures	4 fatalities, $280k fine
Aircraft	Kegworth	1989	Component failure /Human error	47 fatalities
	Florida	1996	Oxidiser in hold	110 fatalities
	Concorde	2000	Foreign object	113 fatalities, commercial closure
Space	NASA 51-L	1986	Component failure	7 fatalities
	Arianne 5	1996	Software	Mission loss
	NASA Mars probe	1999	Software	Mission loss
Tourism	Hyatt hotel	1981	Design change flaw	114 fatalities
	Herald of Free Enterprise	1987	Procedural failure	193 fatalities
	Indiana train ride	1996	Component neglect	1 fatality, commercial closure

This compilation of evidence has several names across the many industries and nations of the world, but its focus is always concerned with understanding the safety status of a system with the familiar goal of avoiding future accidents. Some of the titles given (not an exhaustive list) to these processes and documents are as follows:

1. Contemporary Safety Status Report

2. Safety Case & Safety Case Report

3. Annual Safety Report

4. Control of Major Accident Hazards Report

5. Occupational Safety and Health Plan

6. Health and Safety Plan (HASP)

7. Health Hazard Assessment Report

8. System Safety Approach Documentation

9. Safety Assessment Report (SAR).

This book will make reference to many of these, but will inevitably concentrate on just a few as vehicles for discussing the issues relevant to all safety regimes.

The Safety Case

The precise meaning of the term ‘safety case’ rather depends on your particular relationship with the safety case and the particular purpose the safety case is intended to satisfy. It is likely that each person approaching the phrase ‘safety case’ will have some preconceived idea about what they are getting involved with. For a safety virgin, this idea is unlikely to be well developed – that is to be expected and is perfectly acceptable. For a seasoned guru or safety ‘black-belt’ the meaning of ‘safety case’ will be quite familiar. However, it is of value to review the definitions contemporary with this text so that the readers become familiar with them in general and in the context of the book.

Before approaching the more technical and specialist areas for detailed definitions, it is worth a cursory look through a language dictionary. Mine, published by the Longman Group twenty years ago [Longman 1986] doesn’t contain ‘safety case’ as an entry, I would not expect it to. However, it does contain both ‘case’ and ‘safety’. The combination offers a powerful starting point for a very useful definition.

Case: n b(1) the evidence supporting a conclusion; b(2) an argument, especially one that is convincing.

Safety: n 1 the condition of being safe from causing or suffering hurt, injury or loss.

This combination of ‘convincing argument and evidence supporting a condition of being safe from hurt, injury or loss’ is certainly not trivial. With the addition of a few specific terms for individual areas, this combination from pretty standard dictionary definitions may be seen to be the root of many more complicated and technical descriptions of the subject. Well done Longman.

The most recent available technical definition from a UK military standard [MoD 2004] cites the safety case as being;

Safety Case: A structured argument, supported by a body of evidence that provides a compelling, comprehensible and valid case that a system is safe for a given application in a given operating environment.

The comparison of the dictionary and military standard statements, with over a twenty year gap, highlights an unexpected (to this author at least) but welcome similarity.

The principle aim of a safety case is to derive and present an argument that the system in question will be acceptably safe in a given context. The concept of a safety case is not industry specific, the system could be from any industry. It just needs to be an entity with boundaries, for example a physical system – an engine, a factory, a weapon or a washing machine; it could be procedural for example an oil production facility, a transport network or an assembly line; or it can even be related to some specific event, for example a sports game, a prototype test flight or the demolition of a building. The safety case should contain all necessary information to enable the safety status of the entity to...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Dedication
• Table of Contents
• List of Figures
• List of Tables
• Preface
• Acknowledgements
• 1. Accidents and Safety
• 2. The Language of Safety
• 3. The Safety Management System
• 4. The Purpose of a Safety Case
• 5. The Requirement for a Safety Case
• 6. Setting a Safety Boundary
• 7. Measuring Safety Performance
• 8. Safety Targets
• 9. So Far as is Reasonably Practicable
• 10. Individual, Group and Population Risk
• 11. The Safety Team
• 12. Costs in Safety
• 13. Techniques and Tools for Safety Cases
• 14. The Hazard Log
• 15. Human Factors in Safety Cases
• 16. Software Factors in Safety Cases
• 17. Management Factors in Safety Cases
• 18. Independent Safety Review
• 19. Presentation of the Safety Case
• 20. Maintenance of the Safety Case
• Epilogue
• Index