A Case Study of the British Petroleum (BP) Oil Spill in the Gulf of Mexico
British Pteroleum’s Deepwater Horizon disaster was the worst oil spill in American history. Such was the magnitude of this disaster that its repercussions were felt throughout the global oil industry. In the aftermath of the spill, investigations were conducted into the incident to identify the causes that led to the spill, the reasons behind the poor response to the disaster, as well as to formulate ways in which safety standards could be improved so as to prevent similar disasters from taking place in the future. This study reviewed extensive literature pertaining to the prevalent Health, Safety and the Environment standards in the offshore oil drilling sector. It was found that although there is a substantial knowledge base on the subject that could act as guidelines for HSE practices, there is no standard in place that may be uniformly followed by offshore oil drilling across the world. Several weaknesses were identified. In order to address these challenges, the author has recommended the use of an evaluation criteria to analyse the effectiveness of safety systems with a focus on accident prevention and disaster response.
The Deepwater Horizon oil spill has raised several questions pertaining to the HSE standards being following by oil companies around the world. This research will assist in the better understanding of the current HSE standards in place for offshore drilling units such as the Deepwater Horizon, their weaknesses as well as possibilities of improvement that could prevent future oil spills of this magnitude. The research will aim to cover key stakeholders namely: a) oil companies, b) governments, c) regulatory authorities, and d) local communities, in order to gain a more holistic understanding of current HSE processes, identification of gaps and recommendations for a more systemized and organized approach towards disaster prevention and response in case of future oil spills.
The Deepwater Horizon, BP’s mobile offshore drilling rig, was the location of one of the biggest maritime and environmental disasters in the history of America. Located in New Mexico at the Macondo exploratory oil well, a major explosion rocked the rig when a blow-out occurred out of the oil well. The immediate impact was the unfortunate death of 11 platform employees and the injury of 17 others. However, the long term impacts of the disaster were catastrophic. As Deepwater Horizon burned, it took rescue workers 36 hours to curb the fire. The rig eventually sank along with portions of the safety systems that were in place to prevent blow-outs as well as the upper pipe of the well . With the exposed oil well constantly spilling oil into the ocean, by the time BP capped it in July 2010, 4 million barrels or 206 million gallons of oil had been spewed. Not only was the environmental impact of the spill devastating, but it also shook the local as well as global economy considering the ceaseless depletion of natural fuel and skyrocketing petrol prices .
The degree as well as wide scope of impact of the Deepwater Horizon oil spill was unprecedented in America. It necessitated responses in terms of containment and recovery, clean up of impacted beaches, skimming, booming and burning of oil floating on the ocean surface as well as recovery of wildlife impacted. Considering the massive quantity of oil released, affected areas other than the Gulf of Mexico itself included Alabama, Florida, Louisiana, Mississippi and Texas. Several thousands of workers from various organizations falling under multiple jurisdictions strived to respond in an organized and effective manner . Experts in the field were shocked that a catastrophe of this degree could occur considering the advancements in health and safety standards since the 1989 Exxon Valdez oil spill which was the largest American oil spill before Deepwater Horizon . The fact that the oil rig was operated by BP, a company that was considered to be a benchmark in safety standards until that time, was even more surprising. During the three months that it took to seal the well, there was widespread public debate on the causes of the blow-out, as well as the decisions made pertaining to safety, risk, stakeholder involvement in time and money being spent, and management and operational issues. Questions arose whether lapses in judgement and caution were unique to BP or were common within the oil and gas sector. Several investigations and task forces were organized to identify the causes of the catastrophe and bring to justice those who were responsible to justice . Figure 1 reveals some of the most immediate responses to by major countries and oil companies soon after the disaster .While a majority of nations clearly stated their concern for current safety standards being followed at offshore drilling rigs, the also revealed a determination to verify and upgrade processes being followed by oil companies drilling within their borders. Similarly, although Exxon-Mobil and Chevron reiterated that their existing safety standards were sufficient to prevent a disaster of this level, Total, Conoco-Phillips and Shell were more forthcoming about the need to re-evaluate the safety processes of their organizations as well as the industry at large.
Figure 1: International and industry response to Deepwater Horizon Disaster ).
One of the most discussed issues after Deepwater Horizon was the Health, Safety and the Environment or HSE standards being followed at the rig. The effects of the disaster were far reaching. While UK increased the number of mandatory rig inspection by double, Norway placed its proposed deepwater drilling projects until the results of the Deepwater Horizon investigations were revealed . Considering the wide geographic area impacted by massive spills such as Deepwater Horizon, it is crucial to understand the prevailing health and safety culture and standards in the oil and gas industry. The Deepwater Horizon serves as an ideal case study to analyse the causes of major accidental oil spills, the response to such a disaster, as well as the impact on health, safety and the environment.
- Background of the Study
Health, Safety and the Environment or HSE is the focal point of this thesis. The focus of this study lies on accidental oil spills, hence, it excludes any such incident that may have been caused by intentional sabotage or malicious intent. Although operational or intentional spills also occur, they are relatively rare in nature. Hence, the primary interest of the thesis will remain on accidental disasters. Instead of addressing the overall Oil and Gas industry, the thesis will address only the Oil sector as oil tends to have a greater environmental impact as compared to natural gas. The role of government agencies and HSE authorities will be specifically be studied. BP’s Deepwater Horizon disaster forms that basis of the research conducted and how a catastrophe of this calibre may be prevented in the future.
It can be safely said that spills of this level are commonly caused by accidents. It is also more feasible to identify the probable as well as exact causes of such disasters and hence provide tangible preventive measures thereof. Further, as the Deepwater Horizon was an oil rig, it is natural that the focus of this study will be the oil industry. On the same lines, oil spills caused by shipping accidents have been excluded from this study. Although accidents of this nature are common causes of oil spills, the scope of such research would be wide enough to solicit a study of its own. In order to prevent the dilution of the scope of the current study and its focus on the Deepwater Horizon, the research will address only drilling platform accidents and not shipping accidents.
Regulations and standards play a major role in the implementation of HSE practices and hence it is important to understand their functioning and efficiency. As these regulations are commonly created and enforced by governmental agencies, such authorities will be given emphasis in this research. In addition, corporate policies and processes pertaining to health and safety standards are mostly crafted to comply with government regulation. Corporate documentation relevant to this issue is generally confidential in nature and is hence difficult to access. However, considering that regulation and its implementation does not solely rely on government policies, the influence of the civil society as well as the oil industry will be covered to the maximum degree possible.
- Research Aim and Objectives
This thesis aims to build awareness about the current HSE standards in practice on a global plain on offshore oil drilling platforms, identifying strengths and weaknesses through the study of the Deepwater Horizon disaster. For the weaknesses identified, recommendations for the prevention of oil spill disasters will be made To achieve this aim the research objectives were:
1.3 Chapter Summary
This chapter introduced readers to the foundation of the thesis, the BP Deepwater oil spill disaster. The background and importance of the research into HSE practices as well as disaster prevention and emergency response procedures being followed within the industry, the effectiveness thereof, identification of weaknesses within the disaster prevention and response system as well as the scope for improvement were discussed. The objectives and aims of the research were also communicated.
2. Literature Review
Offshore oil drilling is a complicated process that requires management of high disaster risk levels. While accidents are common on drilling platforms, it is on a rare occasion that they develop into disasters that solicit wide scale response and control. The economic, environmental and social impact of oil spill disasters is enormous, which is why it is crucial to prevent such disasters. In the scenario of disasters taking place, there needs to be an effective, well planned and executive response in order to minimize the damage caused by the disaster. There are several theories on the concept of disasters, their prevention and emergency response planning that will be discussed at length in this chapter. Theories on environmental justice and its economic feasibility has also been discussed as this sheds light on some of the barriers to sufficient disaster prevention and emergency response planning on the part of oil corporations.
2.1The Oil Drilling Process and Related Risks
Oil exploration, drilling and extraction are complex processes that involve a tremendous degree of risk. In order to understand why oil drilling is a high risk operation, it is importance to know the oil exploration process itself. Oil exploration consists of six key mining processes, namely: a) Seismic Exploration, b) Field Development, c) Establishing or Drilling Installations, d) Oil Drilling, e) Maintenance, and f) Field Decommissioning or Abandonment. Varied degrees of expertise are required for effective running of each of these processes. Although all active parties in each process will not be discussed in this study, it is importance to know the characteristics of contractors, operators and licensees as this information is highly relevant to the Deepwater Horizon case. A license to carry out oil drilling in a geographical area can be granted to a single or multiple licensees. In order to mitigate financial risks, oil companies often share a single license. The actual oil drilling is carried out by the operator. An oil company can be a licensee and an operator at the same time, although every company using a shared license may not be involved in the drilling process. Finally, contractors may be hired by operators to perform the drilling in their stead. Contractors play a major role in the Deepwater Horizon case study. Due to the complicated ownership of oil drilling platforms, the accountability of implementing and maintaining effective HSE and disaster prevention systems may be diluted. This factor has been studied in the context of the Deepwater Horizon during the course of the research.
Water depth plays a significant part in determining the risks involved in drilling oil. The temperatures and pressure prevalent at the oil well being drilled are also important components of the risk factor. In the case of Deepwater Horizon, oil was drilled at a depth of over 1500 meters. The Macondo oil well where the platform was located had a history of high temperatures and pressure. The great water depth and extreme pressure and temperature made Macondo more unstable when compared to other wells and hence more vulnerable to oil leaks. In the Deepwater Horizon disaster, Macondo’s extremes of temperature and pressure proved to be a more critical factor when compared to the drilling depth .
- Disaster and Safety Theory
Accidents occur commonly but few convert into disasters. In most cases, the immediate steps taken after an accident takes place can prevent it from developing into a disaster. However, in order to study the processes that apply to accidents on drilling platforms that may lead to oil spills, it is important that the concept of ‘disaster’ itself be understood.
Saleh et al classify the prevention of accidents as ‘theories of accident causation and system safety’ . After studying a large number of accident cases, it became evident that accidents may be caused by technical, organizational and human errors. It has also been found that, when compared to the individual actions of these factors, it is the interaction between them that play a greater role in accidents that develop into disasters. The interaction between these factors has been classified as ‘socio-technical systems’, whereas the accident itself is categorized as system accidents or organizational accidents. Although there is a sizable body of literature on system safety and causes of accidents, it has not been consolidated .
Various approaches to safety and prevention against accidents have been developed over time. These were initially non-holistic in nature, addressing individual factors in a non-holistic manner. These were later integrated into a comparatively comprehensive holistic system and control theory. Further practices were added to this theory until where it stands today. Based on the latest approaches to safety, standard criteria for what should be considered an efficient approach towards accident response and control. These criteria can then be applied to mobile drilling platforms such as Deepwater Horizon.
- Non-holistic Theories
In the year 1978, Barry Turner developed what is considered to be the pioneer theory on how and why disasters take place. The Man Made Disasters Model or MMD stated that accidents having human involvement bore great similarity to those accidents arising from flaws in the organizational and human interactions in the socio-technical system. According to Turner, disasters were sociological events that differed in extremity with the culturally accepted norms pertaining to risks governing governments, organizations and the society itself, and the situation in reality. The theory puts forth that disasters develop over a prolonged period of time during which unchecked erroneous processes amalgamate and culminate as a failure in the system. This developmental phase is termed as ‘disaster incubation period’. The vulnerability towards disaster results from unintentional and complicated system interactions, their gradual magnification as dictated by the structure in which the system is organized, and the societal processes that unknowingly veil the failure, thereby making organizational intelligence incompetent in comprehending the unfolding events . Pidgeon and O’Leary studied the MMD and deduced that the answer to the question of safety lies at the junction where institutional design and culture meet. According to them, ‘safety culture’ in an organization may be strengthened by: a) the commitment of the senior management towards safety, b) common and sincere concern towards dangers and ownership of the effects that they may have on people, c) policies and norms pertaining to hazards that are realistic and adaptable, and d) continuous review of safety practices that relies on monitoring, evaluation and feedback processes. However, Rijpma noted that such a process in the face of third party scrutiny would lead to the growth of malpractices such as deceitful reporting, secrecy and altering statistical facts in case of failure . For example, Halliburton Energy Services Inc., one of the contractors operating BP’s Macondo Well, intentionally destroyed evidence on poor quality cementing work done on the well after the Deepwater Horizon disaster took place in order to avoid prosecution for gross negligence leading to the disaster .
Building on Turner’s theory, Charles Perrow presented the Normal Accident Theory or NAT in1984. After observing that some socio-technical systems were more vulnerable to disasters when compared to others, Perrow put forth that this different results from ‘tight coupling’ and ‘interactive complexity’ . Interactive complexity signifies the number of potential interactions that may occur between various components of a system, while tight coupling signifies the speed with which these interactions take place . Socio-technical systems that have a high interactive complexity as well as tight coupling are most prone to accidents and will eventually face such a circumstance. However, Perrow failed to provide any concrete solution as to how such accidents may be prevented .
Saleh et al. noted that certain organizations, despite having accident prone systems, display an outstanding safety record . By observing such organizations, Rijpma identified system features that are necessary to bolster safety. These features are :
- Redundancy – the ability of one system component to automatically replace another that has failed
- Decentralization of authority in the decision making process – reducing response time by empowering lower level employees to make decisions based on a predefined premise
- Wide conceptual framework – the system relies on multiple theories for prevention and proper response in case of an accident
- Learning and development – the setting up of a process of monitoring and evaluating systems, making changes through accident simulations and using learning’s from these to upgrade existing systems as well as training of staff.
While the above mentioned characteristics are common among high reliability organizations or HROs, Rijpma cautions that they should not be considered hallmarks of an effective safety system . For instance, redundancy can be negatively impacted by the similar nature of back-up systems. Failure of a running component could result in the failure of the back-up too, leading to a domino effect. Furthermore, as back-up systems replace failed ones, the causes of failure may remain unnoticed until an accident actually takes place. In such a scenario, it will become difficult for the organization to identify the precise cause of failure as it will need to evaluate multiple systems that have been in play over a prolonged duration of time. This would, in turn, increase the response time as well as hinder effective decision making, thereby increasing the probability of an accident converting into a disaster.
In order to overcome barriers that prevent effective organizational learning, Cooke and Rohleder formed an incident learning system . This system is based on the identification of incidents, responding to them, developing comprehensive reports of the incident, evaluating the report to determine the causes of the incident, identifying and implementing corrective actions, and documenting the learning outcomes. Based on this model, Cooke and Rohleder identified 10 recommendations that enable organizations to break down the barriers to learning. Of these recommendations, the following five can be considered most applicable to the scope of the present study:
- Developing a culture of transparency and open discussion regarding safety, giving people a fair opportunity to voice their opinions and taking corrective measures as needed
- Instead of identifying a singular cause of the incident, the organization must adopt a systems approach towards safety and learning
- The learning should be converted into best practices that should then be shared effectively across the organization as safety benchmarks
- Reporting events and corrective measures should remain the focus of attention at all times and the ‘blame game’ should be eradicated entirely
- Tools to record and analyse feedback should be in place for enhanced understanding of the effectiveness of managerial commitment and decisions pertaining to safety and their handling of incidents
Cooke and Rohleder insist that these recommendations would prove to be effective only when the required amount of resources is assigned towards learning. Also, if the legal and political framework under which the organization works encourages the placing of blame and resulting prosecution, the focus of the efforts may be drawn away from safety and learning .
As mentioned earlier, Saleh et al. had noted that the aforementioned theories of safety do not address the technical facets of a safety system . It is to address this gap that the Probability Risk Analysis or PRA, and the Defence-in-Depth frameworks were developed. The PRA relies on a combination of analytical tools that aim to evaluate the circumstances surrounding an accident as well as the risks arising from complex systems. It defined the designing of technical systems and can be utilized it examine the succession of incidents that may result in a disaster.
In simple terms, it posits three vital questions: a) what kind of accidents may take place, b) what is the likelihood of these accidents happening, and c) what would be the result of such accidents? By answering these basic questions, an organization can simulate accidental scenarios and the possible preventive measures that may be applied. The PRA attempts to provide an in-depth source of information about the technical and statistical aspects of a system to support proposed identification of risks. However, Saleh et al. also note that the PRA fails to address the possibilities arising from software malfunctions and human errors. Further, it provides limited scope in terms of the realistic prediction of possibilities as well as the tentative nature of the results .
In response to these shortcomings of the PRA, the Defence-in-Depth framework put forth another question – what measures can be taken to prevent possible accidental situations? It seeks to create a multi-pronged approach towards the development of safety and defence barriers to address accidental situations. The purpose of these barriers is to provide varied solutions in preventing, addressing and containing failures in the system for individual accidental situations. While the introduction of these barriers may enhance reliability of an organization, it leads to the creation of greater complexity in systems which can pose new challenges.
- Holistic Theories of Systems Safety
The Systems-Theoretic Accident Model and Processes or STAMP model was developed after studying the salient features of the aforementioned non-holistic theories and models . The NAT and HRO theories failed to provide practicable and reliable suggestions towards enhancing safety systems for preventing accidents . The STAMP model adopts a systems based on a holistic approach towards accidents that enables it to converge organizational and social characteristics, software malfunction and system accidents arising from erroneous interactions between system components, human error, and adaptation of the safety system. The STAMP model does not focus on the causes of accidents, but the comprehensive reasoning behind the accident scenario .
Leveson’s STAMP model holistically lists factors that could result in a disaster. Not only is the list comprehensive in nature but can also be adapted to suit different socio-technical systems. It can also be utilized to evaluate a system in terms of its effective inclusion of features that can prevent possible failures. The STAMP list forms the foundation of the analysis criteria that will be applied against accidental oil spills, specifically the Deepwater Horizon, in later sections of this thesis.
The STAMP model is based on two core concepts. The ‘emergent property’ concept posits that safety cannot be attributed to a single component of a socio-technical system. As such, the level of safety of a system can only be established by adopting a wider spectrum approach that would ideally include the individual constituents of a system, their interactions as well as the environment that they function in. Hence, safety can be said to ‘emerge’ from the overall functionality of the system . The second concept is that of ‘safety constraints’. This concept posits that the behaviour of every component of a system is governed by unique constraints that facilitate safety. Accidents are considered to be an issue of control, while the emergence of safety is driven by these control laws . The safeties constraints are built with the purpose of averting component interactions that are deemed unsafe. It is mandatory for these constraints to be met even when the system is under the process of change or reaction by itself or a similar change occurring in its external environment.
Marais et al. state that, based on the two concepts of STAMP, the following conditions are required to facilitate the understanding and prevention of accidents: a) the identification of safety constraints, b) the designing of a system that strictly adheres to the constraints at all times, and c) formulating a metric that could predict the changes that may take place in processes as well as when such change may be expected to take place . Figure 2 shows how the aforementioned concepts may be applied to ensure safety in a socio-technical system. The arrows in the diagram show the adoptive loops of feedback that create different levels of control. While commands for control move downwards, information of the state of the system or feedback flows upwards. Through this set up, the upper levels provide constraints for lower level behaviours .
Figure 2: Socio-Technical Model of Control (Leveson 2004)
Despite having a well-planned safety system in place, there is always the likelihood of unpredictable and unforeseen circumstances arising that may cause accidents that cannot be controlled. The preventive measures taken thus far cease to hold any meaning and response becomes the pivotal concern. Emergency planning entails the formation of response strategies and resources that can assist in curtailing the dire consequences of an accident. While there is no standard guideline than can be followed by organizations and governments alike for governing the response to disasters, extensive research on the subject had led to the development of several concepts that may aide in the formation of such a standard .
According to Perry and Lindell “emergency preparedness refers to the readiness of a political jurisdiction to react constructively to threats from the environment in a way that minimises the negative consequences of impact for the health and safety of individuals and the integrity and functioning of physical structures and systems” . They presented 10 guidelines for emergency planning which are:
- The process planning should rely on the most reliable and updated knowledge base available pertaining to relevant disasters and the related responses
- The managers who are responsible for emergency responses need to be empowered to focus on the appropriateness of decisions and actions rather than the speed of their execution
- The plan should permit a level of flexibility whereby actions may be altered when circumstances present varied requirements for response
- Coordination between organizations needs to be achieved in terms of tasks to be performed, building of efficient communication systems as well as the allocation of resources
- Rather than following plans based on singular hazards, multiple hazard plans should be devised to enhance efficiency
- Prospective plans should be comprehensively tested by conducting drills in order to identify potential flaws and rectify them before a plan is finalized
- Constant and regular reviews of emergency plans should be mandatory in order to account for changes in the internal and external systems environment
- Opposing and conflicting factors in terms of worst case scenarios, allotment of resources and determining of roles should be taken into account
- The difference between emergency planning and emergency management should be thoroughly understood, as the best of plans can fail without well managed implementation
Alexander built on the recommendations of Perry and Lindell by emphasizing the role that stakeholders, government and legislation would play in an effective emergency plan . He noted that an effective emergency plan would need to be evaluated with these factors taken into consideration, aspects that were left unaddressed by Perry and Lindell. Alexander’s Emergency Planning Standard for evaluation posits several questions that are placed under five categories. There are:
- Legislative and Organizational Context
- Are policies for disaster mitigation sufficient and applied?
- Does the plan strictly adhere to legislative instruments and their provisions?
- Have the jurisdictional and legal roles of the participants of the plan been fully defined?
- Clarity of Objectives
- Have the general objectives and scope of the plan been clearly defined?
- Have the circumstances in which the plan is to be auctioned been clearly specified?
- Hazard, Vulnerability and Risk Analysis
- Has the plan taken into consideration the historical data of hazards at the location of the site?
- Has the plan been formed after conducting a thorough hazard probability analysis?
- Have vulnerability and risk analysis been sufficiently conducted?
- Have potential disaster possibilities and circumstances for the location been identified and constructed, and have they been found to be adequate?
- Has the availability of resources to be allocated for emergencies been audited?
- Has the hierarchy of systems and centres of command been clearly defined?
- Have procedures, protocols and equipment pertaining to communication been clearly specified?
- Have pre-disaster systems such as warning and evacuation been defined?
- Have search and rescue processes been thoroughly organized?
- Are there sufficient measures in place to ensure public order in case of a disaster?
- Have the provisions for public and media information been established, verified and approved?
- Have services such as medical aide, mortuary assistance and logistics for the injured been organized?
- Have inter-organizational pacts for mutual aide been included in the plan?
- Have basic services and efforts in terms of recovering infrastructure been established?
- Has the involvement of private entities such as airports and hospitals been taken into account?
- Have arrangements been made for the testing of the plan?
- Have arrangements been made for the effective communication and distribution of the plan?
- Have arrangements been made for the plan to be regularly reviewed and updated?