New best available and safest technologies (BAST) will come from many sources, as described above. As new technologies are identified that could be deployed as BAST, the Bureau of Safety and Environmental Enforcement (BSEE) will bear the responsibility for evaluating their potential to increase safety offshore. To do so will require testing, modeling, and analysis that characterize the efficacy of the new technologies and their impact on offshore systems. BSEE will need access to people with the right experience and skills so that the agency understands not only the technologies being evaluated but also how they are incorporated into the complex systems used in offshore exploration and production.
Candidate technologies will need to be evaluated in many ways. As standalone technologies, the performance of the mechanical and material components themselves is often easiest to characterize fully. This characterization will often come from the original source of the technology, but in the event that it is incomplete or missing, BSEE should be prepared to perform the necessary tests or to utilize external laboratories and technical resources to have this work done.
Just as important, BSEE should take a system-level view of any technology and its impact on safety that considers not only the individual technology but also the overall complexity of the integrated drilling or production system and the interactions of individual components, subsystems, and systems, including human factors. Such evaluations must recognize the complexity and implications of the limited understanding of the geologic environment in which the engineered and human systems are embedded and operating (Recommendation 3-1).
The performance and failure modes of individual components can be straightforward and easy to assess. The behavior of complex systems, however, can be much harder to predict, and unintended consequences of altering one component of a system can lead to failures elsewhere. In addition, the interfaces of individuals, work crews, and organizations with technology are key determinants of system-level safety. BSEE has the responsibility for evaluating technologies given this complexity and for recommending as BAST those that materially improve offshore safety and whose incremental costs can be justified given the effect on safety, health, and the environment.
A concept such as “technology readiness level” (TRL) can be useful in assessing the maturity of technologies and could be of value to BSEE in the development and assessment of BAST.1 TRL is typically assigned on a numeric scale. For example, a TRL of 1 may designate basic research, while a TRL of 9 may indicate technologies that have been tested within operating systems and are fully operational. In the middle range, a TRL of 5 through 7 can indicate demonstration projects of varying complexity and maturity. BSEE should consider using a metric such as TRL, with levels established on the basis of explicit criteria, in categorizing BAST and communicating with industry on technology maturation (Recommendation 3-2).
Reliability and Risk Analysis
Whenever a technology component is introduced into oil and gas operations, the question of how it will affect systemwide reliability and safety arises. For proposed new technologies, trustworthy answers may be difficult or impossible to obtain before the decision is made whether to deploy the technology. Although many methods of engineering reliability and risk analysis have been developed to help anticipate and reduce risks of failures in technological systems, none completely overcomes the complexity and uncertainty inherent in managing risks of new technologies in oil and gas exploration and production.
Electronic, hydraulic, mechanical, software, hardware, and human components and subsystems interact at multiple points in the normal control and safe operation of a drilling rig, production platform, or other offshore facility. They may also interact in different, unexpected ways during an accident (e.g., due to common-mode failures), so it is necessary to consider not only whether new technologies introduced into one subsystem will perform as planned within that subsystem but also how they might interact in unintended ways with other subsystems, especially during emergencies. Perrow’s theory of “normal accidents”2 suggests that the complexity and tight coupling of interactions among these sub-
systems can be expected to cause accidents that cannot easily be foreseen or prevented.
Methods of risk analysis can help to anticipate and prevent at least some accident scenarios, and full advantage of this limited help should be taken. The following are examples:
• Risk matrices, priority lists, rankings, ratings, and scores are often used to document expert opinions and perceptions about the frequencies and severities of types of accidents and accident precursors. Although their validity may be difficult to establish in the absence of data, such qualitative and semiquantitative methods can at least help those who use them to remain mindful of the risks that have been identified.
• Fault tree analysis can help to reason systematically about how undesirable end states (e.g., failure of an electronic control module) might occur. It reasons backward from supposing that such a “top event” happens to identify combinations of events and conditions that could cause the event. This can help to identify potential failure paths and suggest countermeasures to prevent them in systems with well-understood components and possible causes of failure.
• Event tree analysis helps to reason forward systematically from the assumed occurrence of one or more initiating events (e.g., failure of a component, fire in a control room) to their possible consequences, again identifying paths that lead to catastrophic outcomes.
• Bayesian networks, influence diagrams, and probabilistic expert systems provide a flexible set of software and computational tools for identifying possible (and, data permitting, most likely) failure paths in complex systems. They are able to integrate expert judgments, statistical analyses, and probability models (e.g., for component reliabilities and failures).
• Stochastic simulation models of systems operations and rare failure events can help to quantify the time until (or between) different kinds of failures, including cascades of events leading to catastrophic outcomes—if useful input data are available on conditional failure rates and dependencies. For new technologies, such data are usually not available, and even methods such as accelerated life testing cannot easily furnish dependable surrogate data for large, complex systems.
• Design of experiments and testing protocols: Statistical and operations research methods have been developed to optimize sequential and adaptive testing protocols for reliability systems (e.g., by testing first the components of a series system with the greatest failure probability per unit of testing cost, to minimize the expected cost of determining whether the series system will perform when needed).
• Statistical risk models and data analysis can be applied to accident precursors and near misses to help make best use of experience as it accumulates and provide early warnings of potential failure modes.
All of these methods can offer some insight into system safety associated with particular technologies, but the limitations inherent in each cannot be overstated. The offshore operating environment, particularly during the exploration phase, cannot be fully or accurately characterized or modeled, and all of the methods described above are limited by significant uncertainties in the characterization of the subsurface environment. Other analytic approaches emphasize the importance of human and organizational factors and safety culture in complex sociotechnical systems (e.g., Qureshi 2007). However, the major insights from these approaches are most applicable to assessment of ongoing operations rather than to incorporation of new technologies.
An incomplete understanding of potential interactions of new components or subsystems with the larger systems and operating environments into which they are integrated will also limit the effectiveness of all of these methods of risk analysis. Substituting a new, possibly safer, technology for an established one raises the possibility that partly known old risks are simply exchanged for less well-known new ones. Important concerns have arisen in other areas of risk analysis—for example, green chemistry (where some believe that regulatory programs intended to prevent the use of old chemicals suspected of possibly harming health have led to the “regrettable substitution” of new chemicals that harm health),3 pharmaceutical safety [where some observers have expressed concerns that bans on the use of animal antibiotics, intended to reduce the spread of resistant organisms, have instead led to more animal and human illnesses and to increases in therapeutic antibiotic use for both animals and people (Hayes and Jensen 2003)], and complex engineering systems (e.g., Chernobyl, where testing of shutdown power to the main circulating pumps contributed to loss of control of the reactor).4
The practical lesson from much of applied risk analysis is that models and methods such as those just mentioned can help to reduce some risks, especially risks that can be identified by systematic consideration of possible event sequences and behaviors of well-understood systems, but they cannot eliminate all of the major uncertainties that surround introduction of new technologies into complex, tightly coupled systems. In characterizing the offshore environment and technologies deployed in that environment, important uncertainties will inevitably remain.
BSEE is required to assess the economic impacts of alternative options for BAST. Quantitative economic analyses of safety technologies for offshore drill-
3http://www.nytimes.com/2013/03/31/us/osha-emphasizes-safety-health-risks-fester.html?pagewanted=all&_r=0. Accessed September 25, 2013.
4http://www.world-nuclear.org/info/Safety-and-Security/Safety-of-Plants/Chernobyl-Accident/#.UfdWcJXn-Uk. Accessed September 25, 2013.
ing are considerably more challenging than are similar efforts in other domains, such as transportation, where large numbers of accidents afford well-established statistical data. The scarcity of data with regard to low-probability, high-impact offshore accidents makes it exceptionally difficult to quantify risks and therefore ascribe dollar values to safety technologies.
The economic test established in the statute is consistent with these analytic realities. The statute establishes a cost test for evaluating a technology option, mandating “the use of the best available and safest technologies which the Secretary determines to be economically feasible, wherever failure of equipment would have a significant effect on safety, health, or the environment.…”
In carrying out the economic feasibility analysis—the test that is required under this provision—the committee notes that BSEE will need to consider three types of costs:
• Capital or initial acquisition costs of the technology,
• Operating and maintenance costs associated with the technology, and
• Potential impacts on the reliability and efficiency of the drilling and production systems.
The assessment of the capital and operating costs of candidate technologies will be relatively straightforward. However, many of these technologies will be fairly new, and therefore the uncertainties in these cost estimates will be greater than uncertainties in estimates for technologies that have longer track records. The data limitation with respect to newer technologies is even more of an impediment in assessing potential impacts on reliability and efficiency. Estimates of the costs of a disruption or degradation of operations will inevitably be coupled with considerable use of “best engineering judgment” and qualitative assessments concerning the likelihood of these effects occurring.
Although the impacts on reliability could be treated as an operating cost, separating them is useful because of their economic importance. The committee heard from several industry representatives that adverse impacts on reliability and resultant shutdowns in operations can impose large costs.5 This factor therefore weighs heavily in industry considerations with regard to the required introduction of new technologies and practices into offshore drilling and production operations that may not be fit for purpose. BSEE will need to decide how much weight to give this factor in its own economic feasibility deliberations.
In those instances where the Secretary uses her or his discretion to proceed to the second step described in the statute to determine whether the “incremental benefits are clearly insufficient to justify the incremental costs,”6 BSEE will need to develop an assessment of the incremental benefits of a proposed technology and compare them with the incremental costs. Benefits could include
5Presentations to the committee on May 30, 2013, at its meeting in Houston, Texas.
6Section 21(b), Public Law 95-372, as amended on September 18, 1978.
• Reduced risk of accidents,
• Mitigation of impacts if an accident were to occur, and
• Any ancillary benefits associated with the new technology (such as a reduction in unplanned outages).
The committee notes the extreme difficulty of constructing quantitative estimates of the reduced risk of offshore accidents or likely reductions in their severity as the result of the installation of a safety technology. For example, in a benefit–cost analysis, a key benefit of a better safety technology would be the product of two factors: the cost of an accident if it were to occur and the reduction in the probability of it occurring if the new technology is adopted. A simplistic example is a new technology that would reduce the risk of a $1 billion accident by 10 percent. The benefit would be valued at $100 million. Quantifying the two factors would be exceedingly difficult. First, the costs of a hypothetical accident would be difficult to predict beyond qualitative statements that the type of accident prevented would probably result in small, medium, or large costs. Attaching a number to the second factor—how likely is it that an accident would occur without the new technology and how much less likely would it become if the technology is adopted—is equally problematic if not more so. The previous section on probabilistic risk assessment provides insights into how difficult these types of determinations will be. In view of the challenges associated with technology assessments and economic analyses and of the role played by expert judgments, BSEE should seek access to the requisite expertise, including a multidisciplinary group of individuals with economic, engineering, and scientific skills; access to experts with unique technical skills; and the ability to request independent reviews (Recommendation 3-3).
The committee also notes that, since the economic analyses will tend to be qualitative, the acquisition by BSEE of senior staff with the skills for both understanding these complexities and uncertainties and effectively communicating them to senior U.S. Department of the Interior (DOI) officials will be important.
It would be unfortunate if the regulatory and permitting processes become bogged down by the unrealistic expectation that economic analyses will create a “bright line” for decision makers with regard to what constitutes BAST. Additional data and analyses will provide valuable insights and help BSEE evaluate the roles of qualitative and quantitative risk assessment approaches in narrowing the range of risk possibilities and making BAST determinations. Recognition of the need to apply sound judgment will be needed.
Personnel, Skills, and Experience
Building new and necessary competencies within BSEE will be an enormous challenge. Historically, the regulator has relied heavily on operating companies to provide the bulk of technical work. As BSEE enhances its role in the evaluation of BAST, it will need access to personnel with experience not com-
parable with that of personnel in its predecessor agencies. Potential solutions will be discussed in Chapter 4. For the evaluation of BAST, the key will be to have access to staff with knowledge of and experience with the specific systems and technologies being developed as well as a working knowledge of how the technology is used and the environment in which it will be deployed. BSEE will need to know what working on an offshore facility or onshore in a real-time operations center is like so that it can bring the necessary judgment into the BAST evaluation process.
In addition, BSEE will need to know where available skills and expertise are in industry, academia, and government. Within the federal government, there are significant resources in the national laboratories that can be utilized through cooperative agreements, but BSEE needs to know who these people are to access them when needed. As BSEE builds closer technical relationships with industry and academia, it can also become aware of centers of expertise that can be called on. This and other personnel issues will be discussed in more detail in Chapter 4.
In November 2010, Secretary of the Interior Ken Salazar proposed the formation of an “Ocean Energy Safety Institute” (OESI) in partial response to the Deepwater Horizon oil spill to assist BSEE by facilitating R&D, training, and implementation of operational improvements in offshore drilling safety and environmental protection. BSEE has initiated a solicitation for the operation and maintenance of an OESI that is intended to be a source of technical support to BSEE for BAST implementation. The committee considers this to be a suitable vehicle for identifying, evaluating, and maturing new technologies that would materially improve safety in offshore operations. If properly organized, staffed, and supported, OESI could be a key source of advice to BSEE for BAST development and evaluation and could help in solving problems associated with a government agency competing with industry for top talent and expertise.
Other government agencies and departments have effectively used different models to manage new and developing technologies, such as federally funded research and development centers and university-affiliated research centers. These models will be addressed in some detail in Chapter 4.
Through several decades of offshore exploration and development, technology has been critical in enabling industry to move into progressively more complex and challenging environments. Basic and applied research has resulted in key advances in exploration, production, and safety technologies incorporated into offshore operations today.
In carrying out its BAST responsibilities, BSEE should consider the acquisition and maintenance of an in-depth understanding of existing industry and government capabilities for development, evaluation, and testing of technologies to be a priority (Recommendation 3-4). Resources in in-
dustry, government, and academia as well as joint and international facilities need to be assessed. An accounting of these capabilities will include what exists, who has access to it, what organizations and people have the knowledge and skills to carry out testing and development activities, and where the gaps are. This knowledge will allow BSEE to set priorities between basic and applied research as well as steer funding toward work on BAST that can have the greatest impact.
Roles and Processes
Investment in technology maturation, whether for operational or safety systems, has in the past come predominantly from industry. It has been carried out in operating and service company research centers, in joint industry projects (e.g., DeepStar, Gas Research Institute), and through company-sponsored research at universities. Over the past two decades, industry research has shifted more to technology development and deployment targeting specific assets or asset classes (deep water, tight formations) and away from basic research, leaving a gap that can be filled by government-sponsored research.
In considering the development and maturation of BAST, multiple paths and processes are likely. Technologies and research necessary for the development, maturation, and approval of BAST will vary with the technologies involved. Different approaches (subject to the availability of funds) could be established for different categories of technologies, such as the following:
• High-priority critical technologies [e.g., blowout preventer (BOP)7 and wellhead instrumentation];
• Long-term technology development goals;
• Out-of-the-box ideas that lend themselves to a model similar to the Defense Advanced Research Projects Agency;
• Big-picture questions that might be addressed with an X-Prize model; and
• Small, short-term seed funding of novel ideas from many quarters [e.g., U.S. Department of Defense Small Business Innovation Research (SBIR) model].
Whatever the categories, they will require different levels of funding and management support, will vary between basic and applied research, and will operate over different time frames. BSEE will need a knowledgeable advisory group to allocate its own limited resources effectively and manage the flow and maturation of ideas coming from these different processes.
7See Chapter 3 of NAE and NRC 2012 for recommended improvements to BOP systems.
The amount of money available for government-sponsored research is small compared with industry R&D spending. These smaller budget levels lend themselves to basic research, where the costs are modest compared with maturing engineered systems for deployment, and there can be great leverage in funding early and basic technology development. Historically, this has been the principal area of government-funded research investment in oil and gas as well as in other areas of science and technology; the typically much larger investments required to mature technologies can be made by companies with a commercial interest in their success and deployment.
There are significant potential roles here for OESI in advising BSEE: to steer federal spending in safety-related research toward early-stage technology development and to make that research known to industry so that BAST can be matured and deployed. Applied research and development is largely conducted by private interests, to whom an assessment of risks associated with their investments is of critical importance.
Available Resources and Incentives
In funding research for new technology offshore, the federal government invests on behalf of the public, and industry invests on behalf of its shareholders. While federal funding comes from many sources [e.g., the Department of Energy (DOE), DOI], it has been limited compared with that of industry, and the committee believes that this may well continue to be so.
In the case of BAST, the Secretary will make the determination that a specific technology meets the requirements for BAST and mandate its use. However, a stronger business case for the adoption of new safety technologies could result in a greater industry focus on technology for BAST and shorten their development and deployment times. Although industry may develop a business case for potential technologies, including those considered for BAST, BSEE should consider using legislative or regulatory incentives to speed the deployment of new safety technologies (Recommendation 3-5). Broad and focused incentives collectively would afford BSEE influence on technology development paths across “broadly applicable,” “category-specific,” and “well-specific” areas. Examples of incentives, some of which have been applied in the past, are the following:
• Favorable tax treatment for investments in research in safety technologies (e.g., a research tax credit model);
• Placement of permits at the front of the queue for wells and facilities that incorporate new safety-enhancing technologies or that are used to develop or demonstrate new technologies, or the incorporation of “best value” concepts in leasing and permitting (this would require close cooperation between BSEE and operators during the well planning and deployment phases);
• X-Prize-type incentives for members of academia and other independent entities to develop deepwater candidate technology (e.g., the Wendy Schmidt Oil Cleanup X Challenge);8
• Modest royalty relief for projects that incorporate new technologies (as has been done in the past to motivate industry to develop marginal deepwater fields);9
• Awards within DOE, DOI, and other federal agencies for valuable federal, state, and local government employee contributions to outer continental shelf BAST;
• A BAST prize cosponsored by DOI and industry through the Offshore Technology Conference or a similar visible venue; and
• The establishment of an SBIR program or a small business technology transfer program, which could help broaden participation to smaller industry participants.10
Favorable treatment might be extended to low-risk operating environments where new safety technologies could be deployed first with minimum risk to gain valuable operating data and experience.
Organization and Facilities
As discussed in Chapter 2, there are several examples of organizations that address exploration and production technology development and maturation. The oil and gas industry has successfully used joint industry projects, public– private partnerships, and academic consortia to address many technology challenges. In the past when new technology development was critical for the success of a major project, operating companies have formed alliances with equipment manufacturers to develop and test the necessary technology and move it quickly through maturation to deployment.
In considering technology resources, test facilities merit special attention. Because of the size and complexity of many BAST systems, such as existing and future BOPs, test facilities tend to be large and costly. Hence, the efficient and effective use of existing capabilities is important to industry in minimizing the costs of proving that BAST systems are ready to be deployed for uses offshore.
BSEE should consider creating and maintaining a compendium of worldwide test facilities for determining where best to test introductions into the BAST family (Recommendation 3-6). Such a compendium can be created effectively through international cooperation and agreement on how these
930 CFR 203 Subpart A—Relief or Reduction in Royalty Rates.
10http://www.acq.osd.mil/osbp/sbir/about/index.shtml. Accessed September 25, 2013.
facilities can be used. It would need to include such items as capabilities, potential effectiveness, location, and availability and would need periodic updating. The effort could begin with creation of the U.S. portion of the compendium for use by the U.S. industry, and BSEE could take the lead in promoting the implied international cooperation. The compendium process should proactively seek and discuss industry plans for using test facilities more effectively for BAST introduction. BSEE should strive to gather information from any relevant existing compilations.
The compendium of available facilities should be accompanied by a review that identifies the staff with the expertise to use them. As BSEE builds closer technical relationships with industry and academia, it can become aware of centers of expertise that can be called on. OESI could have the role of compiling a characterization of offshore skills and competencies and maintaining it on an ongoing basis.
BSEE should consider testing for new BAST capabilities by using combinations of scale models and full-size prototypes, systems, subsystems, and modeling and simulation. Such testing should be done in static and dynamic environments (Recommendation 3-7). Testing on a given BAST will likely involve a combination of verification (i.e., whether it meets its design intent and specification) and validation (i.e., whether it satisfies the needs and intent of the customer, including necessary margins). In addition, specific reliability testing will likely be required to show that a given BAST meets its reliability objectives.
Facilities for the integrated and full-scale wet testing of offshore technologies will have special challenges given the scale of offshore systems. Individual companies may have difficulty in justifying the expense, so such large test facilities might be operated by industry consortia or might be U.S. government facilities, if funding permits.
The essence of the above is that several types of tests are required to ensure that a potential BAST is certified for deployment.
Simulation of actual application environments will be difficult. Therefore, modeling and simulation will be important adjuncts to mechanical and electrical testing. However, modeling and simulation are useful only when the model’s predictive capability has been verified by test and operational data. In complex offshore systems, data adequate for full evaluation of models and simulations are unlikely to exist. Therefore, judgment will be required in assessing the maturity of BAST, which will again necessitate access to personnel with the right skills and experience.
In the specific area of offshore safety, the post-Macondo creation of the Marine Well Containment Company11 and the Helix Well Containment Group12 may offer models that can be followed in other areas where significant invest-
ment is needed that will have a broad impact on industry’s license to operate.13 In both cases, industry has been successful in focusing significant resources on a critical and complex problem and in building industrywide solutions and capability without government assistance.
With the above background, it is the committee’s view that BSEE needs a trusted agent to assist BSEE in evaluating test plans and assessing the effectiveness and reliability of new systems before recommending certification that any new BAST is ready for operational use. This function could logically be accomplished under the expanded role being suggested for OESI, which will be discussed in Chapter 4.
13NTL No. 2010-N10. Statement of Compliance with Applicable Regulations and Evaluation of Information Demonstrating Adequate Spill Response and Well Containment Resources. http://www.bsee.gov/Regulations-and-Guidance/Notices-to-Lessees-and-Operators.aspx. Accessed September 25, 2013.