1.1 Introduction
The āCritical Infrastructure Resilience Final Report and Recommendationsā by the National Infrastructure Advisory Council (NIAC 2013) starts its executive summary with the following sentences:[1]
Business and society operate in an increasingly complex world marked by interconnection and interdependence across global networks. This complexity requires that owners and operators of critical infrastructures manage their operational risks in an all-hazards environment across the full spectrum of prevention, protection, response, recovery, and reconstitution activities.
While recognizing that each Critical Infrastructure (CI) sector may operate differently, the report motivates a common definition of Critical Infrastructure Resilience (CIR) for effective governance and policy. The report essentially suggests that CIR is about ādelivering the goodsā despite disruptions. We note that disruptions could be in the form of acute perturbation (e.g., extreme events such as hurricanes, snowstorms, terror attacks) or stresses that build over time (e.g., droughts, social segregation, aging infrastructures). A correspondence in the journal Nature highlighted that resilience has been defined in more than 70 ways, and these definitions are likely to have long-term policy implications to achieve the objective.[2] An operational definition of CIR was proposed by NIAC 2013:
Infrastructure resilience is the ability to reduce the magnitude and/or duration of disruptive events. The effectiveness of a resilient infrastructure or enterprise depends upon its ability to anticipate, absorb, adapt to, and/or rapidly recover from a potentially disruptive event.
To address the similar questions and challenges in critical infrastructures, the risk-based framework has been the tool of choice for engineers and organizations to study threat-impact relationships. For example, in the context of critical infrastructures, the U.S. Department of Homeland Security defines risk as:
The potential for an unwanted outcome resulting from an incident, event, or occurrence, as determined by its likelihood and the associated consequences. Risk is influenced by the nature and magnitude of threat or hazard, the vulnerabilities from the threat and hazard, and the consequences that could result.[3]
Risk from an extreme event results from the interaction of hazards (which includes hazardous and/or extreme events) with the vulnerability and exposure of human and natural systems. Changes in both the nature and magnitude of hazards and socioeconomic processes are drivers of hazards, exposure, vulnerability, and hence risk.
Quantitative estimates of risk assessments are obtained using the Probabilistic Risk Assessment (PRA) Framework to estimate the risk by computing real numbers to determine what is the likelihood of occurrence of hazard and the magnitude of the possible adverse consequences. Consequences are expressed numerically and their likelihoods are expressed as probabilities or frequencies. We will discuss PRA in more detail in subsequent chapters.
PRA has been widely used over the last several decades across sectors ranging from basic sciences and engineering to business and government. Risk management has been used as a tool for preserving ecosystems, securing infrastructures, safeguarding cyberinfrastructures, and making financial decisions. While conceptually generic, risk-based investments and preparedness tend to be threat-centric, situation dependent, and system specific. In contrast, proponents of resilience paradigms have attempted to motivate measures and approaches that are threat-agnostic, adaptable to diverse situations, and ubiquitous across systems. Furthermore, while risk approaches may attempt to reduce system perturbations, the resilience paradigm motivates what has been called graceful degradation or allowing either intentional failure or partial component level collapse to reduce the possibility of permanent or system-wide loss of functionality, albeit with rapid recovery times [4,5]. As the following cases illustrate, the benefits of the resilience paradigm beyond risk management may range from crucial to incremental and occasionally even infeasible to achieve.
Case 1: Threat-centric versus threat-agnostic approach. It may indeed be beneficial to embed resilience across the systems in a way that enhances robustness and increases recovery potential for multiple hazards. However, in many cases, deriving co-benefits across multiple types of threats may not be possible and could even be counterproductive. Let us consider floods, earthquakes, and terror attacks. Investments made in emergency management services such as emergency health care, law enforcement, and communication can result in co-benefits across these hazards. However, co-benefits are not always feasible. For example, consider an organization that is planning to invest to secure its high-value assets such as data server buildings and headquarters from external security threats. Enabling such capacities will entail investments in perimeter fencing and anti-theft systems. While these measures will certainly build capacities against the spectrum of security-related threats, contribution of these investments to enable resilience to natural hazards such as floods and earthquakes may be minimal. There are examples where exclusive focus on co-benefits may be counterproductive or even infeasible. For example, National Electrical Code (NEC) guidelines produced by the National Fire Protection Association of the United States require electric installations (such as power breakers and power backups) in buildings to be placed in well-illuminated, easily accessible areas without any obstructions in the working space. To meet these requirements without compromising aesthetics, it has been a common practice to install these units under stairways, in garages, or in basements. However, during a flooding event, these preferred regions may be most susceptible to flooding. FEMA guidelines specify the following:
Buildings typically respond to an earthquake such that the accelerations at the top of a building can be two or three times stronger than those at the base. Therefore, if flood or storm surge hazards are not present, it is recommended that emergency power equipment and ancillary systems be located at grade since seismic demands will be lower. It is generally not good practice to locate emergency power systems in the basement, since they may be flooded by seismic failure of piping systems.⦠Where both seismic and flood risks exist, critical functions and the equipment needed to support those functions should be located on floor(s) with elevations above Design Flood Elevation and these functions and equipment should also be protected from wind forces, wind-borne debris and seismic effects. All critical equipment and interconnecting piping, wiring, and ducts should be elevated or protected as recommended.[6]
While guidelines of various agencies intend to build resilience to disparate hazards (earthquake and flooding in the preceding example), the safety requirements for the two kinds of hazards are aligned in opposite direction in the sense that where reduction of seismic risk requires the electrical components to be placed at lower elevations, flood risk to these installations may be exacerbated at these elevations. Hence, a threat-agnostic approach to incorporate resilience can neither yield the co-benefits for disparate hazards nor be ubiquitous across systems for certain hazards.
Case 2: āFail-safeā versus āsafe-failā way of planning. One of the key design principles in risk management approaches is preservation of status quo, that is, avoiding the transformative change and minimizing the risk of failure. However, resilience approaches focus on adaptation to changing conditions without permanent loss of functions. Park et al. (2014) distinguish these two approaches by referring to them as āfail safeā and āsafe failā. Let us understand this in the context of levees that are built for flood management. A levee is a man-made structure designed and constructed in accordance with sound engineering practices to contain, control, or divert the flow of water to reduce the risk from flooding for flows up to a certain amount. Historically, reinforcing or temporarily overbuilding levees for increased resistance against rising river stage has been one approach to manage flood risks to ensure that these remain āfail safeā in the event of design flooding. However, as the name design flooding suggests, levees are designed to safeguard the adjoining areas from flows up to a certain amount. However, this approach is inherently limited in its capacity to mitigate flood damage. On the contrary, Park et al. (2014) in their perspective article on resilience provided an example of āsafe failā being a typical aspect of resilience.[4] The Birds PointāNew Madrid Floodway is a flood control component of the Mississippi River and Tributaries Project l...