Chemical accident disasters are unplanned, undesired events involving hazardous substances, causing harm to human health, the environment, and or economic loss or social disruption. While there is a long history of chemical accidents, with events recorded even more than 100 years ago, the study and implementation of technologies and approaches to preventing, preparing and responding to chemical accidents, only gained widespread attention in the last 40 years.
There have been significant advancements in understanding accident phenomena, and in development of technology and management systems to control risks. Nonetheless, beyond a certain level of prevention, meaningful gains in prevention seem to elude our grasp. Indeed, in developed countries, such as European Union (EU) Member States and the USA, & Canada that have by far the most sophisticated understanding and oversight of chemical accident prevention, there are still a high frequency of serious chemical accidents each year, resulting in severe human, environmental and economic consequences.
Moreover, there is an increasing presence of hazardous industries and of volumes of hazardous substances in commercial use in many developing countries where experience with industrial processing hazards and risks is relatively recent and where social and political infrastructures for dealing with the plethora of externalities accompanying industrial production are inadequate.
Most experts do not believe that chemical accidents occur today because our understanding of engineering possibilities runs ahead of our understanding and predictive powers regarding their downsides. Rather, our challenges today stem from a myriad of inputs whose influence on chemical process risks is broadly known and understood, but that go largely unrecognized and unmanaged in organizations and on sites where the risks are present. Hence, it is not our lack of knowledge and understanding of how the technology works, but in many cases a lack of access to such knowledge, and in other cases, a failure to prioritize and use it wisely to prevent serious loss.
Chemical Industries
Chemical accidents will continue to happen in the foreseeable future if chemicals and chemical processing are important for society. In particular, the usage and applications of chemicals is spreading and not decreasing. Moreover, production, transport and storage of dangerous substances are happening in places where these risks were never a problem before. In the meantime, there is evidence from the repetition of accidents from previous generations within industrialized economies that lessons of the past have been forgotten or ignored. This paper outlines the trends that threaten to increase chemical accident risks and proposes some recommendations to address them. Chemical accident prevention, including the Seveso Directive, chemical disaster, process safety, safety culture, economic trends, capacity building, developing countries, multinationals, learning from accidents, know ledge transfer, chemical risk management Introduction In 1921 an explosion of 4,500 tonnes of ammonium nitrate sulfate fertilizer at a BASF site in Oppau, Germany, killed over 500 people and caused considerable damage to the site and surrounding community. At the time, Carl Bosch, BASF’s Nobel-prize winning engineer said,
“The disaster was caused neither by carelessness nor human failure. Unknown natural factors that we are still unable to explain today have made a mockery of all our efforts. The very substance intended to provide food and life to millions of our countrymen and which we have produced and supplied for years has suddenly become a cruel enemy for reasons we are yet unable to fathom. “This statement was no doubt true in 1921 when chemical manufacturing was still a new and growing industry. 100 years later, thanks to the work of generations of dedicated scientists in industry and academia, “unknown natural factors” are rarely an underlying cause or chemical accidents today.
Known Knowns
Accident reports, investigation results, and media reports of recent times give overwhelming evidence that chemical accidents today are mainly caused by failure to apply what is already known, the “known knowns”. Improvements in our understanding of chemical accident risks and chemical accident control technologies and systems have not necessarily led directly to advances in a significant reduction in chemical accident disasters. Indeed, according to a famous study by H.W. Heinrich, 98% of all industrial accidents are preventable. However, technological disasters are by their nature “(hu)manmade” and it can be argued that reduction of chemical disaster risk is particularly affected by the dependence on humans to manage and use the technology appropriately.
Pomfret and Pidgeon argue that disasters arise from an absence of knowledge at some point. They occur because we do not understand enough about those forces, i.e. in industrial processes, which we are trying to harness, and as a result energy is released at the wrong time or at the wrong place. They are also clear that this is not just an engineering issue and that many disasters arise from social or administrative causes or the combination of technical and administrative causes. The science of reducing chemical accident risks is now focused on the underlying causes of human failure to control the risks. Characterising causality in this way adds new dimensions to the study of chemical accident risks.
In attributing causality to control, there is a recognition that further progress in reducing chemical accident risks requires strong involvement of the social sciences. Certainly, there is considerable room for examining new engineering solutions, such as the use of artificial intelligence and adapting existing control technologies to new processes. However, these types of solution are industry and even process specific and do not apply to many sectors in which accidents frequently occur. Indeed, the oil and gas industry are one of the world’s oldest industries and has been the subject of massive technological investment over many decades, and yet globally it is by far the leader in terms of frequency of severe chemical accidents.
Hazardous Industries
The term “hazardous industries” comprises numerous substances, processes and equipment, with considerable variation within each category regarding properties, function, and behavior under different conditions. Petroleum refineries, bulk chemical production (e.g. chlorine and ammonia), manufacture of specialty chemicals (e.g., paints, dyes, plastics and resins), pharmaceuticals are examples of industries that comprise a wide range of processes, each with their own unique risk profile and associated risk management implications. While there is less variety, there is still considerable danger in processes involving hazardous substances in the “non-chemical” industries, such as water and waste treatment, electroplating, and food production.
In addition, distribution activities, including transport by rail, road and pipeline, as well as explorative and extractive activities, on and offshore also are important sources of chemical accident risk. † In societies with mature risk regulation such production and use of hazardous substances is permitted provided that the risks remain at a level deemed acceptable by the local community and society in general. This paper will give evidence that industrialized countries are still far from achieving an acceptable level of chemical accident risk. It will then describe several underlying causes common to all industries and societies that are impeding progress in chemical accident risk reduction.
Chemical Accidents
Chemical accidents are still a relatively frequent occurrence in all industrial countries and raise important questions about the adequacy o f disaster risk reduction efforts. Media monitoring over the last several years shows consistently that at least 25-30 chemical accidents with worker or community impacts are reported each month around the world in industrialized countries. Preliminary results of a study by Wood et al. [53] of accident reports covering all major chemical hazards in fixed facilities and transport over the last 5 years (2012-2016) identifies 29 national and regional chemical accident disasters and 21 chemical accidents with evident high local impact. In total these accidents accounted for 928 deaths, and (where reported) 22,973 injuries. In addition, significant environmental impacts were recorded, with one pipeline disaster reaching USD$ 257 million in restoration costs§. Over 7000 people were evacuated for several months due to a slow leak of natural gas that was finally sealed off in February 2016**. Insurance companies recorded 9 accidents with > USD $100 million in damages, including 2 accidents†† costing >$1 billion. Many other impacts including job loss, environmental impacts, emergency response costs, damage to nearby buildings, market and production losses were sparsely reported, but West Virginia businesses were reported to have lost $USD 61,000,000 in 4 days. “From the perspective of the individual facility manager, catastrophic events are so rare that they may appear to be essentially impossible, and the circumstances and causes of an accident at a distant facility in a different industry sector may seem irrelevant. However, from our nationwide perspective at [U.S.] EPA and OSHA, while chemical accidents are not routine, they are a monthly or even weekly occurrence, and there is much to learn from the story behind each accident.”
Frequency o f severe chemical accidents is at odds with society’s expectations, Society is becoming increasingly risk-averse, and failure is less readily tolerated. There are indications that the frequency of serious chemical accidents is higher than expected in many industrialized countries. In 2015 the number of deaths from major accidents on the ~10,000 EU Seveso sites is already estimated to be at least 15 (see Figure 1). This statistic, if confirmed, means that the frequency of 1 fatality on a major hazard site in the European Union was around 1.5 X 10-3 that is above acceptable limits for individual risk in EU Member States that use quantitative criteria. (e.g., the United Kingdom, the Netherlands) ‡‡.
Executive Order
In 2013 the President of the United States issued an Executive Order to improve chemical facility safety and security following various high profile chemical accidents. In other major industrialized countries, such as China and Brazil, in recent years that chemical accident frequency and severity is approaching or has approached levels which would be generally considered unacceptable in an industrialized economy. Globalization and export o f technology has increased chemical accident risk ou \t side the EU Similar trends are noted in developing countries (See Figure 2) China enacted the Emergency Event Response Law of 2007 because of an important lesson learned from two major chemical accidents in China: the 2003 gas well blow out in Chongqing that caused 243 deaths mainly from hydrogen sulfide inhalation, and the 2005 release of toxic substances into the Songhua River. [55]
New legislation in Brazil covering chemical risks stems from broad-based concerns about problems connected with chemical safety that have grown in intensity and extent in the last two decades. Many developing countries have experienced rapid growth in hazardous operations from growth in particular segments of oil and gas, chemical and petrochemical industries and mining, driven by a combination of factors including increased demand in emerging economies, access to raw materials and the need to lower production costs, facilitated by a decline in trade barriers and government incentives to attract foreign investors.[8]
Chemical Risk Management
Chemical risk management in modern times – the theory is well established, but implementation lags in current times, there is considerable agreement on the fundamental principles of process safety management which, if understood and properly applied, would prevent a large majority of chemical accidents that still occur today. These essential principles are summarised, that are then applied in the context of an ISO 31000:2009 risk management process. From an operational perspective, successful risk management comes from applying layers of protection throughout the process life cycle (design to decommissioning), starting with reduction of the hazard itself, and working outwards to accident prevention, mitigation and response. Above all, it is the organisations and individuals that manage all these elements. For this reason, hazardous sites are expected to have a safety management system in place, a derivative of the well-known “management system” concept, to manage the interface of humans with hazardous processes to minimize process hazard risks.
Chemical accidents nowadays are often derived from the failure of industry, government and society to understand the profound effect that their choices have on risk
Manage Risks
The hazardous industries understand in principle how to manage chemical accident risks. Then why do these industries continue to repeat failures of the past and have accidents and sometimes disasters? A study of accidents of the last few decades and the work of numerous experts in manmade disasters, including chemical accidents, as well as nuclear, space and aviation disasters, suggest that the causes are systemic. Sweeping changes in business philosophy and the explosion of opportunity created by new technology, such as the increasing reliance on computerization of business processes, have brought benefits but also come with their share of risks.
These risks are particularly notable for manmade risks where small changes to complex systems can unwittingly remove barriers to initiation or propagation of a potential hazard event. on the system out their own share of come with also. It is a fact that technological disasters, past and present, not just chemical disasters, have relevant and timeless lessons for risk managers in all industries, many of which have been recently documented by A number of high profile technological disasters occurring after 2000 have challenged some longtime risk management experts to identify the patterns underlying the repeated failures behind the latest round of technological accidents, building on the original work of Perrow[36]
Relationship between the risk management principles, framework and processes (ISO 31000 :2009 Risk management – Principles and guidelines) 10 and Rasmussen on managing risk in complex systems, among others, in the 1970 with new approaches. Hollnagel et al. have introduced the concept of “resilience engineering” for technologically complex industries. They look at risk management from the organizational perspective of the large multinational and government operators that are the owners and operators of these technologies. In resilient systems, individuals and organisations habitually adjust their performance to match the variability of the risk over time, “prior to or following changes and disturbances so that it can continue its functioning after a disruption or a major mishap, and in the presence of continuous stresses.“
Klinke and Renn suggest that “risks must be considered as heterogeneous phenomena that preclude standardised evaluation and handling” in their 2006 paper describing government’s potential role in managing systemic risks. [25] Le Coze proposes that new analytical models for safety assessment to consider the dynamic and systemic aspects of safety. [27] Kletz commented on the pattern of corporate memory loss in UK companies as far back as 1993. [24] More recently, Baybutt’s review of accidents investigated by the U.S. Chemical Safety Board since 1998 concluded “Remarkably, all the reviewed incidents involved some type of deficiency or omission in adhering to established process safety practices. In many cases there were multiple deficiencies and omissions.” [3] Wood et al. also found that where probable causes of the accidents have been ascertained, they are most often associated with predictable circumstances where control measures were insufficient due to poor risk management or, in some cases, a lack of adequate awareness of the risks.
This finding is further substantiated in various lessons learned publications, such as the MAHB Lessons Learned Bulletin, where analyses of recent and older accidents are side-by-side, identifying often remarkably similar findings about what went wrong. [11] The research of Taylor et al. collated and synthesized circumstances and causality associated with twelve significant technological accidents, of which five were chemical accidents, and identified numerous organisational failures associated with leadership, oversight and scrutiny, and communication that were common precursors to the events studied.
Their study identified several factors, including the general decline of safety departments, oversimplification to upper management through aggregation of indicators and other inputs, poor understanding of operational “reality”, lack of processes and systems which ensure that process safety risks are properly assessed, and the influence of commercial interests, as among key forces that shaped the events leading to the accidents. Arstad and Aven [1] point out, “it is dangerous to assume that system boundaries can be limited to the sharp end of the business … wide and open system boundaries recognise the importance of many more risk sources and safety.”
Simplify Risks
They also remark on the tendency to oversimplify risks (“complexity is incompressible”) associated with complex technologies. With petroleum-based industries as a primary candidate, Carnes outlines a High Reliability Governance model based on multiple engagements between government and industry actors, that continually reinforces common performance expectations and a high-level safety culture. [6] A large part of scientific studies of technological disasters focus on big well-resourced organisations. But it is a fact that many serious accidents around the world originate in small and medium enterprises that are operating simple processes (e.g., warehouses, fuel distribution).
While they are not all “disasters”, UNDP’s 2004 report on Reducing Disaster Risks correctly cites that accidents with local impacts are an important part of understanding the scale and dimensions of threats.[50] In addition, there is some evidence that government and society unwittingly, for sometimes very good motivation, accept more risk when it concerns small and medium-size businesses. These companies are often significant challenges for regulators because they lack adequate expertise or even sufficient hazard awareness to manage their risks within acceptable limits.
Typical cases of this type are the small fireworks manufacturers that have been the site of several accidents with multiple fatalities in the last 5 years within the EU. [10][53] Moreover, recent tragedies, such as the disasters of Tianjin, China (2015) [42] and West, Texas (2013) [49] indicate that in these cases, that even though the presence of a significant hazard was known, the government failed at many levels to ensure that adequate prevention, mitigation and preparedness measures were in place. Twelve underlying causes are cited as challenges to controlling chemical accident disaster risk in current times The authors of this paper have identified twelve types of underlying causes based on their own studies of accidents and research of other experts. They are based in part on causal typologies developed by the various experts in manmade risks already cited in this paper. They also reflect the authors’ extensive experience in studying and investigating the causes of chemical accidents, bringing in the small business and governmental dimensions that are sometimes not covered well in research. Causes are not necessarily mutually exclusive since the presence of one underlying cause can make a site susceptible to other dangerous mentalities and conditions.