Search

International audience; La période d'icubation (Turner, 1978) est un concept reconnu qui indique que des précurseurs d'accidents peuvent être détectés et traités à temps pendant une période ou des signes de dégradation de la sécurité sont observables. Cette publication discute d'enseignements tirés d'une évaluation du management de la sûreté où une période d'incubation a été détectée.

International audience; Cette communication se propose d'analyser des marqueurs de perte de mémoire relatifs aux accidents de Saint-Laurent-des-Eaux de 1969 et 1980. Elle met en exergue des « facteurs d'oubli » engendrés par les transitions (technologiques, organisationnelles et sociétales) survenues sur la période considérée.

This Deliverable is the result of a joint effort by experts, working in the fields of risks management, accident analysis, learning from experience and safety management. They come from 10 countries mainly from Europe and also from USA and Australia. Their expertise covers several industrial sectors. They attempted to provide useful information, both from a theoretical and a practical point of views, about "Foresight in Safety". Safety is still an ongoing issue for which a number of subjects remain under debate (e.g. is goal of safety to ensure that 'as few things as possible go wrong' or to ensure that ‘as many things as possible go right’?). Anyway, we can assume that safety is to act in a way for both the process continues to be run right and that errors and failures to not lead to a major accident. Even if "foresight in safety" is the implicit underlying goal of every practitioner in safety, the outlines of its domain remain blurred and the relevant topics associated with it have never been clearly defined. A humble ambition of this Deliverable is to display some aspects of "foresight in safety" according to the current state of practices and scientific knowledge., JRC.G.10-Knowledge for Nuclear Security and Safety

National audience; La présente réflexion intègre le fait que la prévention des accidents ne se joue pas seulement dans les interactions ‘ici et maintenant’ mais qu’il s’agit d’un phénomène distribué dans l’espace et le temps. En particulier, malgré les défis de la complexité et du caractère éminemment dynamique de l’évolution des systèmes, la dimension rétrospective, notamment historique, devrait être une dimension centrale de l’analyse des risques organisationnels. De plus, l’histoire des accidents industriels peut éclairer certains signes de dégradation de la sécurité d’une organisation. Sur la base de ces considérations, une étude empirique de la pratique d’analystes d’un institut d’expertise et de recherche, l’IRSN, portant sur l’évaluation du management des risques mis en œuvre dans les centrales nucléaires, permet d’illustrer des modalités de mobilisation de cette connaissance pour diagnostiquer et anticiper des risques organisationnels.

International audience; In this paper, after defining what the ‘knowledge of accidents’ is, our contribution addresses how ‘knowledge of accidents’ can be used in practice to interpret ‘weak signals’ especially weaknesses in risk management. This research is supported by a case study: during an assessment of safety and radiation protec-tion management of nuclear power plant maintenance outages, conducted by IRSN, (a French public institute, i.e. the technical support organization that supports the nuclear safety authority (ASN)), IRSN experts have used the knowledge of accidents to conduct an organizational diagnosis.

International audience; Our research explores why and how a very special kind of knowledge – i.e. the recurring lessons learned from past industrial accidents- could be better capitalized and used in the riskwork of public sector experts when assessing industrial risk management practices. After defining the concept of ‘knowledge of accidents’, we provide empirical insights from its use by experts for in-depth organizational diagnosis within a regulatory safety assessment performed by IRSN (the technical support organization of the French nuclear safety authority) on practices of risk (safety and radiation protection) management of nuclear power plants.

International audience; Summary The techniques that use algorithms and statistics in association with linguistic approaches that cover "big data" and automated natural language processing are being implemented. These recent techniques could bring confirmed or potential benefits to the current issues and remaining challenges of learning from experience. They can also lead to unwanted and perverse effects that we should characterize. Beyond technical aspects, their use should enable as well to better address human and organizational factors in learning from experience.; Les techniques utilisant largement les algorithmes et les statistiques en association avec des approches linguistiques qui recouvrent le « Big Data » et le traitement automatique du langage sont en cours de déploiement. Ces techniques récentes sont de nature à fournir des apports avérés ou potentiels aux problématiques actuelles et aux défis persistants du retour d'expérience. Elles peuvent également être sources d'effets non souhaités voire pervers qu'il convient de caractériser. Au-delà des aspects techniques, leur usage devrait permettre également de mieux prendre en compte les facteurs organisationnels et humains dans le retour d'expérience.

International audience; Summary The techniques that use algorithms and statistics in association with linguistic approaches that cover "big data" and automated natural language processing are being implemented. These recent techniques could bring confirmed or potential benefits to the current issues and remaining challenges of learning from experience. They can also lead to unwanted and perverse effects that we should characterize. Beyond technical aspects, their use should enable as well to better address human and organizational factors in learning from experience.; Les techniques utilisant largement les algorithmes et les statistiques en association avec des approches linguistiques qui recouvrent le « Big Data » et le traitement automatique du langage sont en cours de déploiement. Ces techniques récentes sont de nature à fournir des apports avérés ou potentiels aux problématiques actuelles et aux défis persistants du retour d'expérience. Elles peuvent également être sources d'effets non souhaités voire pervers qu'il convient de caractériser. Au-delà des aspects techniques, leur usage devrait permettre également de mieux prendre en compte les facteurs organisationnels et humains dans le retour d'expérience.

Risques à faibles probabilités ayant des conséquences de grande ampleur, "événements extrêmes et rares", "cygnes noirs", "risques résiduels", "risques technologiques et accidents majeurs", catastrophes naturelles ; industrielles ; sanitaires, "désastres",... Depuis quelques décennies, les discussions et les échanges entre les scientifiques, les experts et les décideurs sur ces sujets n'ont cessé de s'intensifier. Cet ouvrage collectif, coordonné par les quatre animateurs du groupe de travail et d'échanges commun à IMdR et l'AFPCN "Incertitudes et Décisions", a été l'occasion pour des communautés de pratiques différentes (risques naturels, industriels, environnementaux et sanitaires) et des communautés de chercheurs en sciences humaines et sociales, en sciences de la vie et en sciences de l'ingénieur de dépasser les clivages entre le monde de la "recherche" et le monde de "l'action" sur la prévention, la mitigation et la gestion des risques et des crises liées aux catastrophes naturelles et sanitaires, et aux accidents industriels majeurs. Quels sont les pratiques, les insuffisances, les problèmes et les attentes des parties prenantes et des décideurs en la matière ? Quels sont les résultats des recherches des différentes disciplines scientifiques ? Comment accompagner le changement de pratique notamment en France ? Quelles suggestions faut-il proposer pour améliorer la pratique ainsi que la recherche en la matière ? Pour examiner ces questions, les auteurs prennent appui sur de nombreux dossiers des industries de la pétrochimie au nucléaire, des risques accidentels aux risques chroniques, des risques dits naturels et environnementaux aux risques dits technologiques et sanitaires.

The European Safety Reliability and Data Association (ESReDA) has since 1993 set up a series of Project Groups dealing with the different angles of ‘accident investigation’ and ‘learning from events’. With the 25th Anniversary of ESReDA now in 2016, the core of this group is still active, and has just initiated a new phase with its latest Project Group on ‘Foresight on Safety’. With the objective of improving the quality of accident investigation and the efficiency of learning from experience process and ultimately raising safety performance, the successive groups tasked themselves at two levels: the first one, at a societal, institutional and legal level, on the public accident investigation and societal learning; the second one, at a methodologicaland organisational level, on the conduct of accident investigation, the enablers and barriers to learning. This article summarises the Project Groups' achievements (reports, books, papers and ESReDA seminars) on the various aspects of of accident investigations and dynamic learning from events. This article presents a synthesis of the approach and main results, the lessons learned, some dilemmas and conflicts, future challenges, recommendations and suggestions for action. Although varying in composition over time, the main participants remained involved in the development of the issue: participants from the European and member state authorities, industries, research centres and universities and professional practitioners represent a unique, voluntary cooperation across sectors, actors and disciplines that has lasted for almost 23 years now. At last, with the rise of ‘big data’ it is valuable to recall the interest of single case investigation and address the complementarity of the two approaches to learning.

This paper aims at sharing some lessons learned by practitioners and researchers in terms of risk regulation when facing uncertainties and ambiguities inherent and induced by the occurrence and the management of major risks, either technological, natural or health diseases. On one hand, we have noticed that there are significant differences when analyzing the several frameworks in these domains, either from risk regulations (risk averse approach) and from business point of view (risk taking approach). On the other hand, we have observed how scientific segmentation of problems can increase blindness or some blind spots which can weaken major risks prevention policies. The paper is structured as following. - We will first discuss some key aspects of public risk regulation focusing on how an analytical framework to assess the “quality” of decision process can contribute to a more sustainable major risks’ governance. - We will then describe the different types of risk regulation situations based on real case studies analysis (ex. Air pollution). - We will finally present what we means by “quality of decision process” and conclude by suggesting some key challenges for risk and policy analysts.

Analysant ou réanalysant un grand nombre d'événements graves, incidents et accidents, les auteurs dressent le " portrait type " d'un accident industriel. Loin de l'erreur humaine comme facteur explicatif ultime, ils mettent en lumière les défaillances profondes de l'organisation qui conduisent, après une période d'incubation, à la catastrophe. Ils établissent ainsi le " portrait robot " de l'organisation dysfonctionnelle. Les auteurs ne s'en tiennent pas là : à partir de " cas d'école " d'accidents, ils jettent les bases d'une méthode nouvelle et efficace d'analyse des organisations à risques, et dessinent en creux les voies de renforcement de la sécurité industrielle.

This paper attempts to share some lessons learned by practitioners and researchers studying challenges of decision-making in the domain of major risks, either technological, natural or health diseases. Varying regulatory frameworks, different decision-making levels and scientific disciplines lead to a diversity of approaches which complicates the design of generic principles for decision-aiding for the prevention of low probability-high consequence events. In order to delineate the potential for improvement, several biases affecting prevention are addressed (in risk modelling and risk perception, especially of low probabilities). The proposed approach focuses on the analysis and quality assurance of the conditions of validity and legitimacy of decisions, implying several stakeholders and suggests the need for a better use of case studies.; Cet article est l’occasion pour nous de partager certains enseignements acquis par les praticiens et les chercheurs étudiant les défis de la prise de décision dans le domaine des risques majeurs, qu’ils soient technologiques, naturels ou sanitaires. Des différences dans les cadres réglementaires, dans les niveaux et les objets de décisions, dans les cadres disciplinaires fondent une diversité d’approches ce qui complique la conception d’un dispositif générique d’aide à la décision pour la prévention de ces événements rares et catastrophiques. Afin de situer des axes de progrès, de multiples biais affectant celles-ci sont évoqués (dans la modélisation du risque, dans la perception du risque et des faibles probabilités). L’approche proposée vise à fournir des éléments de réponses en vue de l’analyse et de la fiabilisation des conditions de validité et de légitimité des décisions, impliquant de multiples parties prenantes et plaide pour un usage renforcé des études de cas.

Analysant ou réanalysant un grand nombre d’événements graves, incidents et accidents, les auteurs dressent le « portrait type » d’un accident industriel. Loin de l’erreur humaine comme facteur explicatif ultime, ils mettent en lumière les défaillances profondes de l’organisation qui conduisent, après une période d’incubation, à la catastrophe. Ils établissent ainsi le « portrait robot » de l’organisation dysfonctionnelle. Les auteurs ne s’en tiennent pas là : à partir de « cas d’école » d’accidents, ils jettent les bases d’une méthode nouvelle et efficace d’analyse des organisations à risques, et dessinent en creux les voies de renforcement de la sécurité industrielle.

Public expertise in safety, security and environment (SSE) is a process that is increasingly submitted to control and transparency. As decision-making, the exercise of expertise involves subjectivity and judgment. An oversight, a monitoring and an aiding approach is therefore required for its conduct and its governance. This paper proposes a novel way of embedding ethical aspects and participative decision-making elements into the process of risk assessment and risks management. Based on their experience feedback, the authors first propose some early contributions to study the validity and the legitimacy of expertise in SSE. In the second part of the paper, the authors give an overview on how public expertise is organized in France and how the problems and the conclusions are framed in SSE. Finally, the authors propose a generic integrated framework for public expertise that constitutes the “responsible exercise of expertise”. This framework allows framing a valid and a legitimate expertise process and its conclusions.

In 2007, the French government organized a series of meetings dealing with environment and sustainable development problematics under the title " Grenelle de l'Environnement ". The conclusions of these meetings were introduced in a new law published in August 2009 (Law N° 2009-967) known as 'Grenelle 1'. Article 1 requires that the State designs a so-called Sustainable Development National Strategy (SDNS) structured around 9 targets. Target 1 consists in fighting against climate change. The SDNS is used as a basis for the involvement of public and private agencies in this perspective. The struggle against climate change and implementation of sustainable development principles within the agency framework raises a number of practical questions: how are adaptive methods and tools to be developed to help the agency reach a new balance given the environmental, economic and social risks induced by these broad targets? Are there any standards of reference and if not can an innovative approach be developed to support decision-making in situations of uncertainty? How can a management dashboard be developed for a proportional allocation of resources adapted to the various stakeholders and decision levels within the agency? How does one compare actions that can have different impacts in different sub-systems and with different time frameworks ? Following a brief historical review of the origin of the challenges of sustainable development and climate change in the introduction part of the paper, we discuss the methodological approach selected and the investigator's intervention program in the first part of the paper. In the second part, we address practical issues for an agency facing the challenges of sustainable development and climate change. To this end we relate the experience of designing a sustainable development framework within a public agency using an organizational approach and a multiple criteria decision aid methodology. Finally, we discuss the decision-makers' choices and the lessons learned by implementing an innovative approach which we set up to face these new challenges.; Les organisations font face à de nouveaux défis soulevés par la mise en application du principe de développement durable. Cet article offre un retour d'expérience et une réflexion sur la mise en place d'une stratégie et d'un plan d'actions développement durable, au sein d'un institut public d'expertise dans le domaine de l'environnement industriel et des risques, et révèle le besoin d'un accompagnement au changement et sa nécessité pour parvenir à ce changement.

Some of the key points from iNTeg-Risk ERRA C4 will be described, with an emphasis on what points can be generalised to emerging risk in general. These will cover two main facets: The first line of discussion covers how many times does an atypical event need to occur before it is no longer considered atypical and how does one then convince an established industry to change it's view ? A succession of vapour cloud explosions are used to illustrate this point. The second part will look at a range of atypical events and examine common threads among the contributing factors that often are seen with atypical events. Managing these factors would be a starting point for managing generic unknown or emerging risks.

National audience; Organizations are facing new challenges raised by the implementation of the sustainable development principle. This paper points out the need to support organizational change in response to these Environmental Demands. New methods are proposed and discussed through a real case study of implementing a strategical plan for sustainable development within a public institute of expertise in the field of industrial environment and risks.; Les Organisations font face à de nouveaux défis soulevés par la mise en application du principe de développement durable. Cet article offre une réflexion sur le besoin en accompagnement au changement des Organisations qui relèvent ces défis. Nous avons proposé une méthodologie basée sur une approche de type recherche-intervention. Cette méthodologie prend appuie sur de l'analyse organisationnelle et de l'aide multicritère à la décision. Pour finir, nous relaterons l'histoire d'un cas de mise en place d'un plan stratégique et d'un plan d'actions développement durable au sein d'un institut public d'expertise dans le domaine de l'environnement industriel et des risques.

National audience; Organizations are facing new challenges raised by the implementation of the sustainable development principle. This paper points out the need to support organizational change in response to these Environmental Demands. New methods are proposed and discussed through a real case study of implementing a strategical plan for sustainable development within a public institute of expertise in the field of industrial environment and risks.; Les Organisations font face à de nouveaux défis soulevés par la mise en application du principe de développement durable. Cet article offre une réflexion sur le besoin en accompagnement au changement des Organisations qui relèvent ces défis. Nous avons proposé une méthodologie basée sur une approche de type recherche-intervention. Cette méthodologie prend appuie sur de l'analyse organisationnelle et de l'aide multicritère à la décision. Pour finir, nous relaterons l'histoire d'un cas de mise en place d'un plan stratégique et d'un plan d'actions développement durable au sein d'un institut public d'expertise dans le domaine de l'environnement industriel et des risques.

Several of the major incidents that have occurred within Europe in recent years have shown unexpected incident scenarios not considered by their site Seveso II Safety Case (e.g. Toulouse, Buncefield,...). For instance at Buncefield, the scenario of an overflowing petroleum storage tank generated a large flammable cloud and gave rise to a large explosion that would not have been kept by standard assessment techniques. A standard safety assessment of the Buncefield site (before 11 December 2005) would have predicted that the worst credible scenario would have been a large pool fire. We know this missed the actual accident that occurred. This is an example of how improving hazard identification in safety reports has become progressively more difficult to achieve, despite the considerable investment in response to the relevant European Directives and their implementation in member states. One of the reasons is that available hazard identification (HAZID) methodologies, such as Hazard and Operability studies (HAZOP), fault trees, 'what if' scenarios etc, take no notice of less likely or unknown events. Unidentified scenarios thus constitute a latent risk, whose management is nowadays extremely complex and open-ended. For this reason one task of the European research project "INteg-Risk" addresses the general issue of improving Hazard Identification, also including specific lessons learned from the Buncefield accident and implications for industry. The purpose of this paper is thus analyzing and describing the progress of work made until now to accomplish the task and showing the preliminary results obtained. An in-depth analysis of some of these unusual incidents was performed, in order to outline general features of process plant in which they occurred, their causes, consequences and lessons learned. At first a historical incident analysis was carried out to find similarities among the atypical incidents analyzed and other cases in the past. On the basis of the findings, well-known HAZID methodologies were studied and applied, with the purpose to assist future HAZID processes to identify atypical scenarios. In order to determine all the causes and consequences of a loss of containment, without preconceptions on occurrence probability, bow-tie diagrams were thus obtained following the methodology for the identification of major accident hazards (MIMAH), which was developed as part of the ARAMIS project. This allowed the methodology to be reviewed and recommendations developed, which should help to consider warnings of possible atypical scenarios

International audience

International audience

Proper hazard identification in safety reports has become progressively more difficult to achieve. Several major incidents in Europe in recent years, such as Buncefield and Toulouse, were not even considered by their site Seveso II Safety Case. One of the reasons is that available hazard identification (HAZID) methodologies take no notice of least likely or unknown events. Non identified scenarios thus constitute a latent risk, whose management is extremely complex and open-ended. Hence this work aims to investigate the issue of atypical scenarios and explain how they could have been identified. An in-depth historical analysis of some of these incidents has been performed, in order to outline general features of plants in which they occurred, their causes, consequences and lessons learned. This analysis has followed a precise common scheme, which allowed a systematic approach to the problem. Based on the findings, failures connected to risk management and risk appraisal have been identified. Checklist questions and recommendations will be compiled, making efforts to improve the general safety of industry. Attempts will be made to review well-known HAZID methodologies, and thus the process of risk appraisal, with the purpose of improving the HAZID processes to identify atypical scenarios.

These guidelines have been prepared for investigators, investigation managers, people who order investigations, responsible persons who will have to learn from the event, victims and researchers. These guidelines provide a minimum, current and recognised cross-sectorial best practices oversight to conduct investigations related to industrial, technological and organisational events. These guidelines give practical and theoretical advices related to the different stages of such investigations. In-depth analysis of accidents, incidents and crises clearly showed that any event is generated by direct or immediate causes (technical failure and/or ¿ "human error"). Nevertheless their occurrence and/or their development is considered to be induced, facilitated or accelerated by underlying organisational conditions (complex factors) found in socio-technical system and organisational networks. It implies to deal with different natures of causalities: mechanistic met in technical installations and more complex met in human and social systems. Addressing those causalities requires various competencies (with disciplines from exact sciences, to engineering and social sciences) to investigate and to learn from an accident. Accident investigation can be performed for various reasons and objectives. They depend on the stakeholders (many types among companies, authorities, or public parties) and their perspectives. Several investigations could often be managed simultaneously (e.g. judicial, technical,¿) and it may lead to interest and operational conflicts such as the access to the accident scene and witnesses, the collection of the facts, the preservation of evidence, the findings, the communication of findings,¿ Corporate, political, cultural and societal requirements will shape the context in which the investigation is conducted. This should be cleared and stated when defining the mandate of the investigation. Despite all this variety of contexts, investigation obeys to general principles (protocols, coordination, competence, data and evidence, reporting, follow-up of lessons learned and communication) and to process phases (defining terms of reference, appointing team, collecting data, hypotheses generation, analysis, findings and recommendations). Connection between phases is not a linear process but rather an iterative one. Any investigator gets skills and know-how, so accident investigation is influenced by a priori knowledge or initial models. Specific methodologies have been developed to facilitate some investigation tasks. They use different logical constructions, different underlying models, and address different levels of phenomena with various perspectives (what happened, why it happened and what is recommended to prevent its repetition). Organisations should prepare their protocols and train their investigators before the event occurs in order to be ready to undertake the investigation in an effective way. These guidelines aim to support those processes. To communicate during and after the investigation process is an important issue. Aim is to provide stakeholders with findings through a report or other supports in order to initiate and facilitate the learning process. Many barriers to learn lessons can be faced by organisations. Turning findings into recommendations (corrective measures) is a specific task that requires specific knowledge of the organisational network and the socio-technical system behaviour. The corrective measures can be challenged by reality and a specific follow-up should be put in place. Finally, some future challenges that would have to be addressed by the investigation and safety management communities conclude the present guidelines., JRC.F.3-Energy security

Current operating feedback Systems (le. systematic and organised analysis of events related either to process malfunction or to industrial safety) in industry show some limits as, in numerous industrial companies, the numbers of events do not significantly decrease anymore and as similar events seem to recur. Despite their recognition as one of the main cornerstones for safety, one can wonder if lessons from both incidents and major accidents are or were satisfactory learned and implemented. These difficulties lead some companies to particularly poorly react to signs that herald an accident. In this paper, we will address several cases in varions high-risk industries that highlighted some failures in their operating feedback Systems. After ïdentifying some of their organisational weaknesses, we will argue about what we assume is foundation of this flaw: the lack of "culture of accidents ".