Terrorism and the Chemical Infrastructure: Protecting People and Reducing Vulnerabilities (2006)

Since the attacks of September 11, 2001, federal, state, local, and tribal governments, aided by the private sector, have undertaken an unprecedented review of the nation’s infrastructure to determine potential targets for future terrorist attacks. At the national level, the Department of Homeland Security (DHS) has divided the nation’s infrastructure into 17 categories of critical infrastructures and key resources, one of which is the chemical industry and hazardous materials sector.

The chemical sector is a key part of the national economy. Although its products represent only 2 percent of the U.S. gross domestic product, they underpin most other manufactured goods. Direct products of the chemical industry include plastics, fibers, and drugs, and many more products such as paper, fabrics, cosmetics, and electronics are dependent on the products of the chemical industry.

The chemical sector includes firms that manufacture huge volumes of chemicals intended for many uses—such as major refineries processing thousands of tons of petrochemical feedstocks daily—as well as firms that produce small quantities of materials with highly specific uses, such as small pharmaceutical companies producing products in gram or kilogram quantities after many days of processing and purifying. Some of these chemical products are toxic, flammable, or explosive. The facilities in which chemicals are produced are similarly varied—from refineries covering square miles of land with many high-volume chemicals on-site, to startup companies occupying thousands of square feet in light industrial parks. Products can

be transported to their final place of use by truck, rail, pipeline, marine vessel or other means in both large and small quantities.

This study was requested by DHS to assist the department in characterizing and mitigating the vulnerabilities faced by the nation from its chemical infrastructure (see Appendix B for full statement of task). The study has examined classes of chemicals and chemical processes that are critical to the nation’s security, economy, and health; identified vulnerabilities and points of weakness in the supply chain for these chemicals and chemical processes; assessed the likely impact of a significant disruption in the supply chain; identified actions to help prevent disruption in the supply chain and actions to mitigate loss and injury should such disruption occur; identified incentives and disincentives to preventive and mitigating actions; and recommended areas of scientific, engineering, and economic research and development that might advance the nation’s capability to protect against such losses and minimize their impact.

This report addresses the most significant general types of vulnerabilities associated with the chemical infrastructure, not site-specific vulnerabilities. Other government and private sector efforts are developing vulnerability and risk assessments that account for site-specific factors such as the amount of chemical on a site and the size of the potentially affected population near a site. This study is intended to supplement those efforts.

This report adopts the definition of catastrophic incident outlined in DHS’s National Response Plan—one that “results in large numbers of casualties and/or displaced persons, possibly in the tens of thousands.” Similarly, an economic impact on the order of tens to hundreds of billions of dollars would be considered catastrophic. A catastrophic event is one whose consequences are so extensive that they overwhelm the ability of emergency responders, local and federal government officials, and/or the general public to adequately and/or fully respond in a timely fashion.

Toxic, flammable, and explosive materials present the greatest risk of catastrophic incident.

Casualties within the chemical sector are most readily caused by exploiting the toxic, explosive, or flammable properties of chemicals. By far the largest number of casualties would be anticipated from situations involving toxic inhalation hazards, that is, large-scale release of toxic chemicals in a gaseous form. Damage to infrastructure and subsequent economic loss is caused more readily by the flammable and explosive properties of chemicals.

All scenarios envisioned by the committee with potentially catastrophic consequences were some variation on or combination of one or more of three basic scenarios: (1) release from high-volume storage (either fixed site or in transit); (2) shortage of key chemical or chemical product; and (3) misuse of a small quantity of chemical (tampering or theft). Terrorists could conceivably cause a single terrorist incident or multiple terrorist incidents, geographically co-located or dispersed, simultaneously or over a period of days or weeks. Consequences would be greater with a greater number of events carried out; the difficulty and complexity of such an attack and risk of interception would similarly be greater.

In the absence of specific threat information, it will be most appropriate to invest in mitigation and preparedness for general classes of vulnerabilities.

In the case of either casualties or economic loss, catastrophic levels of consequences are expected only where large quantities of chemicals with toxic, explosive, and hazardous properties are involved. However, social response may amplify the effects of an incident involving even a small quantity of chemicals to the point at which its economic effects, not its casualties, become catastrophic, or at least of national concern.

By analogy with past accidents involving the chemical industry, it is possible that a single terrorist incident involving the chemical infrastructure could result in catastrophic loss of life or injuries.

This report discusses scenarios based on historical chemical incidents that serve as existence proofs (but not necessarily upper bounds) for possible consequences of a disruption to the chemical supply chain. By using this approach it is easy to determine that a single chemical event could

cause catastrophic casualties. For example, approximately 4,000 people died in the immediate aftermath of the methyl isocyanate gas leak from the Union Carbide India Limited Bhopal plant in December 1984. Injuries have been estimated to range from 200,000 to 500,000 and contributed to an accumulation of 15,000 to 20,000 disaster-related deaths in subsequent years. In this country, an explosion involving 2,300 tons of ammonium nitrate in a Liberty ship at a loading dock in Texas City, Texas, on April 16, 1947, cascaded into widespread destruction of nearby petroleum refineries, chemical production facilities, and another fertilizer Liberty ship, ultimately claiming nearly 600 lives and causing approximately 3,500 injuries—America’s worst chemical catastrophe.

The economic effects of a single terrorist incident involving the chemical infrastructure could be significant, but multiple terrorist events would be required to achieve nationally catastrophic economic consequences.

The chemical industry is quite diverse, with redundancies that mitigate the effects of loss of production due to major shutdowns. Where stockpiles do not exist, market forces quickly compensate for loss of production by increased production at another facility of the same or a different company, or by temporary substitution in industrial processes of another chemical with similar properties.

Although a single incident might not result in a nationally catastrophic economic loss, such an incident could result in changes to business and manufacturing processes across the industry, either voluntarily or through regulation. The costs associated with such changes could be significant to individual companies and to local economies, but would not have a major impact on the national economy.

Public response is significant in determining the consequences of attack on the chemical infrastructure.

Public response to any act of terrorism in this country involving the chemical infrastructure will undoubtedly be significant and could invoke both positive and negative consequences. While the impact of a terrorist incident in itself may be linear (that is, the loss of life and injury will be directly related to the size of a chemical release within a given category of chemicals), there may be significant nonlinear social consequences of the incident. These consequences could significantly affect sectors of the

economy, such as the negative impact on the travel and airline industries after September 11th. It may also impact public morale, affect the level of trust and confidence in the government’s ability to protect its citizens, and exacerbate feelings of vulnerability leading to social (sociological) and psychological effects. Conversely, a well-informed general population that is adequately prepared for such events could decrease negative consequences and unnecessary casualties.

Public authorities will need an understanding of social amplification and attenuation if they wish to successfully manage the aftermath of a chemical attack. Research can support the development of specific guidelines for limiting, and even mitigating, consequences by stimulating a positive public response and preventing negative social amplification.

Accurate information analysis and communication before, during, and after an event between parties charged with responding and with the public may be the best short-term means to mitigate the possible consequences of an event.

Accurate information analysis and communication consists of several components:

• Acquiring reliable data—in an emergency, this includes reducing data errors and ambiguity to the greatest extent possible;

• Converting the data into integrated information and conclusions;

• Deciding on and communicating appropriate actions; and

• Communicating promptly to the public in an accurate, comprehensible, and believable fashion.

Effective emergency response depends on the rapid analysis of information received in a crisis to determine its relevance and accuracy. This information must also be communicated rapidly and accurately to all necessary parties involved in decision-making, and then integrated into the

decision making process to determine an effective response. This is especially true in a chemical attack because event-specific conditions such as the type of chemical, quantity of material, and release location will be critical to determining the appropriate course of action. Multiple attacks on the chemical infrastructure may not immediately be recognized as such. Prompt recognition that an incident is an actual case of terrorism and may be part of a series of attacks is critical to taking actions that may limit the consequences of such attacks. Recognizing that an attack is part of a larger coordinated effort may be hampered if incidents are widely dispersed, involve different types of attacks, or otherwise present challenges to recognizing a larger pattern, particularly if communication between affected parties is significantly impeded.

The perception of disasters among members of the public sometimes escalates as a consequence of a breakdown in the communication process. Conversely, a well-informed public can often take action to minimize the effects of a disaster. Information must reach the end users in a comprehensible and useful form; it must be perceived by them as relevant to their situation; and they must have the capacity and the necessary resources to use this information to better prepare for, respond to, and recover from a hazard or disaster situation. Research to determine the most critical information to be communicated between responders and to the public, and the means to gather and disseminate that information, can result in a rapid improvement in emergency response capabilities. Such research would be universally applicable to all chemical emergencies—independent of the type of incident or chemical involved.

Near-term benefits can be obtained from research efforts directed toward enhancing emergency preparedness, emergency response, and disaster recovery. This offers an immediate means to mitigate the effects of a terrorist attack on the chemical infrastructure.

Through study of past events, social science research has derived significant understanding of the components required to prepare a community or a populace for hazardous events and to effectively respond to and

recover from those events. Efforts to increase this knowledge and to expand its use in practice can rapidly enhance our capacity to mitigate the effects of a chemical event. This could, in turn, make the chemical infrastructure a less attractive target for terrorists. It has the further benefit that such efforts are “dual use”—of near or equal value in the case of a chemical accident.

The most desirable solution to preventing chemical releases is to reduce or eliminate the hazard where possible, not to control it. This can be achieved by modifying processes where possible to minimize the amount of hazardous material used, lower the temperatures and pressures required, replace a hazardous substance with a less hazardous substitute, or minimize the complexity of a chemical process.

Many of the advances required for development of practical alternatives to today’s chemicals and chemical processes are fundamental and pre-competitive. The economic incentives for industrial funding are frequently absent, which leads to the need for either a government investment in research or government-provided financial incentives for industrial investments. Inherently safer chemistry, such as process intensification, “just-in-time” chemical manufacturing, and the use of smaller-scale processes, offers the potential for improved safety at chemical facilities. While applications show promise and have found use within the chemical industry, these applications at present are still quite limited in scope.

The chemical sector is complex, with many links and interdependencies between operators. Economic research demonstrates that in an interdependent system, firms may have a disincentive to invest in security if all other operators in the system do not do likewise. Economic analysis of this sector and of the incentives and disincentives that firms have to take protective measures could help determine how some combination of regula-

tion and private sector measures, such as insurance with third-party inspections, can be utilized to maximize firms’ willingness to invest in appropriate security measures.

A container holding significant quantities of a hazardous chemical provides an obvious terrorist target. Although efforts to strengthen existing containers against intentional rupture are ongoing, there may be opportunities to fundamentally change the means by which hazardous chemicals are stored. For example, methods to store chemicals in adsorbents are currently available but are generally limited to small quantities. Research could seek to enable the use of adsorbents at the cylinder scale and to use such storage methods for larger volumes involved in truck or rail shipments or on-site storage. Other possibilities for fundamental change in storage include low pressure storage (which would reduce the release rate given an unintended rupture) or underground storage technologies (which would reduce the storage tank profile presented to terrorists).

The near-term objective of enhancing emergency response effectiveness can be furthered through efforts to develop reliable detection techniques that can be distributed widely, are easy to use, and would give accurate results quickly and clearly. These can aid in “early warning” of chemical

releases before they become catastrophic and would aid in decision making and response, prevention of catastrophic release, or more timely and effective emergency response. In research and development for chemical sensors, DHS should focus on furthering technologies that are relatively inexpensive to deploy and easy to use.

Using inventory controls as a means to quickly identify theft of hazardous chemicals may provide a fundamental means to prevent a terrorist attack. This capability may prove difficult if not impossible to mimic with sensor technology. Investments aimed at improving compliance with such procedures would be appropriate.

Current disaster impact models have been generalized from natural disasters and accidents and may have features that do not apply in the case of a deliberate attack. Further research is needed to confirm that the models’ assumptions and relationships are valid in these situations. Previous disaster research supports an all-hazards approach, in contrast to the focus on specific hazards that has emerged in recent approaches to homeland security. Furthermore, there is speculation, but little research, on whether human responses to intentional terrorist events differ significantly from responses to natural disasters or accidents. Such incongruities between current disaster models and current security concerns need to be identified and examined to determine what, if any, changes are required to our current understanding of mitigation planning, response, and recovery.

The presence and use of toxic chemicals create vulnerabilities and could result in catastrophic casualties. Effective predictive models of casualties could greatly reduce these vulnerabilities and improve emergency planning and response. Different levels of accuracy and precision are needed for different levels of emergency planning and emergency response. The accuracy and precision of situational models and consequence analysis currently in use must be better understood. Although the physics of a hazardous materials release can be described using models, effects on populations are not yet well characterized. Limitations in understanding the toxic effects of many substances and in understanding the dose-response relationship of hazardous chemicals over time, especially for vulnerable populations such as children, the elderly, and the poor limit current capacity to model casualties. Further efforts will be needed to understand the dispersion and toxicity of chemical mixtures. Furthermore, the reliance of early security risk assessments on the outputs of emergency planning efforts such as the Risk Management Plans submitted to Environmental Protection Agency has led to misimpressions of the potential consequences of individual events. While such data may have been useful for initial screening, they have also led to significant confusion and alarm among various decision makers and the public. Better and more appropriate data should be used, and clear explanations of the change should be provided to different stakeholders.

While the consequences of a terrorist attack on the chemical infrastructure are of significance to the population affected, there is no reason to deviate from the principles and approach of good risk assessment and management decision making when prioritizing investments to mitigate these consequences. Each assessment should consider a realistic scenario and its vulnerabilities, likelihood of occurrence, and consequences if it were to occur. The scenario should be processed through a series of tests to assess if it can be significantly disruptive or catastrophic. These tests should consider loss of life, economic impact, and the ability of state and local governments to respond to the event, and should also consider the impact of

social amplification. This should be followed by an analysis to assess the trade-off between expected benefit and cost of the proposed solution.

The findings and recommendations in this report emphasize the importance of the development of new technology and of investment in current technology and also highlight the need to combine this technology with effective communication strategies, reliable and effective mitigation techniques, and preparedness and response strategies. This combination is necessary to minimize the possibility of a terrorist attack and its effects or consequences.

When confronting the potential for terrorist attack, it is essential to constantly reassess both the progress being made and the possibility of unintended consequences when implementing a “solution.” The threat from terrorism is not static, and it is not unreasonable to assume that terrorist tactics will evolve with emerging technologies designed to defeat their threat. Some strategies to address terrorism reduce the chance of a successful attack, some reduce the consequences of such an incident, and some relocate the vulnerability—that is, these strategies may reduce the chance of direct casualties, but still leave financial and cascading impacts. All of these factors must be taken into consideration when assessing vulnerabilities of the chemical infrastructure.