Investing in Transportation Resilience: A Framework for Informed Choices (2021)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

103 Researchers have sought to develop direct measures of transportation sys- tem resilience, as discussed in the previous chapter. Recovery curves, which are conceptualized in the literature, define and describe resilience in terms of the loss of functionality and the time needed for restoration. Resilience metrics derived from such curves, in theory, would be comparable across transportation modes and systems. However, functionality recovery curves depend on metrics that are specific to the transportation mode or service, and data needs can be extensive. More studies of past hazard events and transportation system disruptions are also needed for these concepts to be refined for practical application. Informed by this study’s reviews of both academic research on resil- ience and practitioners’ efforts to measure resilience, the committee is not optimistic about the prospect of developing a single metric or small set of metrics to support resilience investment choices, at least in the near term. Indeed, the committee concurs with the finding of a 2019 RAND study that “there is no single metric or value that can perfectly reflect all aspects of resilience in all elements of a given system. Instead, decision makers must look at a variety of metrics to assess and understand the impacts of the investments they make … to improve the resilience of the assets in the transportation system.”1 This finding is not surprising. The variations in 1 RAND. 2019. Incorporating Resilience into Transportation Planning and Assessment. Report for National Cooperative Highway Research Program project 08-36, Task 146, p. 29. https://www.rand.org/content/dam/rand/pubs/research_reports/RR3000/RR3038/RAND_ RR3038.pdf. 5 Decision Support Framework

104 INVESTING IN TRANSPORTATION RESILIENCE natural hazards today and over time, geographic settings, and infrastructure characteristics are vast and better measured by a portfolio of metrics. Interest in developing a salient set of metrics is understandable. Invest- ing in resilience can be risky and requires careful consideration because it entails spending and other costs incurred in the present to gain benefits that may or may not be realized in the future. These decisions need to be well reasoned and based on sound analytic principles and processes that help guide prioritization of assets warranting resilience strengthening and inform the choice of specific investments for this purpose. For such decision sup- port analysis, there is indeed a need for a portfolio of metrics. The stakes can be high when investments in transportation resilience are neglected or not made in a deliberate and systematic way. By way of example, the Economics Unit of The Port Authority of New York and New Jersey analyzed the economic costs of hazards to the New York metropoli- tan region.2 The analysis simulated shutdowns of its airports, seaport, tun- nels, and mass transit. Even a 1-day shutdown had significant costs, which increased nonlinearly over 3, 7, and 30 days. Table 5-1 shows the estimates of economic costs for the various time periods and different transportation modes operated by the Port Authority. For the airports, the cost estimates aggregated the costs of trip delay for outgoing and incoming passengers as TABLE 5-1 Economic Costs Associated with Disruptions to New York and New Jersey Port Authority Transportation Facilities (in millions of 2018 dollars)3 Days Shut Down 1 3 7 30 Airports $178.7 $775.5 $1,414.8 $4,048.4 Newark $62.6 $251.9 $460.2 $1,301.3 JFK $75.9 $343.3 $658.0 $1,871.4 LaGuardia $40.2 $180.3 $296.6 $875.7 Seaport Facilities $22.2 $202.7 $535.1 $2,038.6 Trans-Hudson Tunnels $19.1 $56.7 $129.9 $549.2 PATH $2.4 $5.4 $17.4 $80.4 NOTES: Figures may not total due to rounding. Cost estimates should not be added together as estimates were calculated for each mode in isolation. PATH = Port Authority Trans-Hudson. 2 Eshleman, C. 2018. “A Multi-Criteria Decision Index Employing Single-Criterion Features for Evaluation of Transport Infrastructure.” TRB Annual Meeting. 3 The Port Authority of New York and New Jersey, Regional Economic Analysis. 2018. “The Consequences of Facility Shutdowns.”

DECISION SUPPORT FRAMEWORK 105 well as the disruptions to business activity that may result from the cancel- lation of travel altogether. The costs of disruption for seaport facilities were made up largely of the additional inventory costs of extending the supply chains and accommodating delays in bringing goods to market. For mass transit (Port Authority Trans-Hudson) and travel through the tunnels, the evaluation accounted for travel time increases and potential productivity losses as a result of remote work. While the Port Authority’s analysis focused on a select set of measurable economic costs, more complete analyses would also likely have revealed other societal costs such as lost lives, the consequences of injuries, and environmental damages, as well as repair costs for infrastructure and other unmeasured costs resulting from delayed and missed person-trips and freight movements. In this chapter, consideration is given to the structure and elements of a decision support process, or framework, that practitioners like the Port Authority can use to make well-considered investments in the resilience of their transportation infrastructure. Some of the elements, or steps in the pro- cess, are informed by research but are derived largely from existing practice, founded on previous efforts by the federal modal administrations, other federal agencies, state and local transportation agencies, and private industry. Before turning to the framework idea, the next section of the chapter identifies some general principles that the committee believes should under- pin such an effort. The key steps in the framework are then discussed. These steps focus primarily on assessments of resilience benefits. In other words, they are intended to identify and quantify to the extent possible the prospective benefits from making specific investments in resilience that will avoid or lessen the societal costs from natural disasters as they impact transportation systems and their critical functionality. The steps lead to societal benefit measures that can be weighed against measures of the cost of making specific investments, including the resources required to make a resilience improvement as well other relevant considerations such as the opportunity cost of not using those resources for other socially valuable purposes. Thus, while the proposed decision support framework itself will not always produce results that are actionable in the sense that they will provide decision makers with an objective list of resilience improvements that should be made, they can be used to inform such decisions as a key part of a societal benefit-cost analysis (BCA). GENERAL PRINCIPLES OF A DECISION SUPPORT FRAMEWORK The committee believes that a decision support framework should have cer- tain qualities that will ensure that it is generally applicable and sufficiently practical to use. In particular, the framework should be

106 INVESTING IN TRANSPORTATION RESILIENCE • Comprehensive so that it can be applied across modes, locations, time, and hazard types; • Capable of accounting for uncertainties about the future; • Practical to use, requiring data that are reasonable to obtain, and involving analyses that can be readily linked to more informed decision making; • Objective in the sense that quantitative metrics are used where available and reasonable, and qualitative assessments are informed by data or expert judgment and are transparent; • Broadly based by taking into account a locale’s or region’s quality of life and economy in addition to accounting for direct (and often more readily measurable) impacts on infrastructure owners and users; • Attentive to different time dimensions and cognizant of the resil- ience that is needed for immediate response and recovery from disaster as well as the resilience needed over the longer term for disruptions over the life cycle of assets, such as from the effects of climate change; and • Informed by the results of past investments, which can be helpful for understanding where resilience investments have paid off. A MULTI-STEP DECISION SUPPORT FRAMEWORK The steps that make up the framework for measuring resilience benefits—or the societal costs avoided from adding resilience—and costs are logical and straightforward. Figure 5-1 depicts them. They start with the conduct of an inventory of assets, both existing and planned. Next, perhaps integrated with the asset inventory, is evaluation of the criticality (importance or value) of these assets, particularly with respect to their societal functions. This is followed or accompanied by characterizations of the types and like- lihood of hazards that could affect assets in this inventory. In this regard, the framework can be viewed as multi-hazard. Having this information, a transportation agency can then make assessments of the vulnerability to hazards of the most functionally critical assets and characterize the conse- quences should the vulnerability become exposed in a hazard event. One can think of this entire process as a means of estimating or char- acterizing risk, or as tantamount to identifying the prospective benefits and costs of different options to reduce this risk to varying degrees. Decisions about whether to implement these options, with their attendant resilience benefits and the costs associated with their implementation, can then be informed by BCA. More discussion of each step in this process is provided next.

DECISION SUPPORT FRAMEWORK 107 FIGURE 5-1 Components of the proposed decision support framework. Identifying Assets Transportation assets refer to the physical infrastructure, transportation workers, and institutional resources for all relevant transportation modes: road, railroad, maritime, inland waterways, aviation, public transit, bicycle and pedestrian facilities, and pipelines.4 To conduct resilience analysis, agencies need to have up-to-date infor- mation on their assets, including an asset’s location, condition, vulnerability to damage, and history. For transportation agencies, asset management is typically an ongoing process that meets a variety of management goals, involves asset inventory data, and may include vulnerability information. The current regulatory framework for some transportation modes requires maintaining active asset management programs. State Departments of Transportation (DOTs) are required to develop asset management plans 4 USGCRP (U.S. Global Change Research Program). 2021. “U.S. Climate Resilience Tool- kit.” https://toolkit.climate.gov/content/glossary.

108 INVESTING IN TRANSPORTATION RESILIENCE that are then certified by the Federal Highway Administration (FHWA).5 Likewise, the Federal Transit Administration (FTA) requires public transit agencies to develop and implement a Transit Asset Management Plan.6 De- pending on the scale of the envisioned resilience investment, a transporta- tion agency might have a system-level inventory as well as a project-specific inventory, with different levels of detail. For example, Washington State DOT has incorporated resilience analysis at a corridor level and thus has not used detailed inventories of individual assets. Box 5-1 presents some examples of inventory elements useful for resilience analyses at the physical asset and system infrastructure levels. Asset inventories should include information related to an asset’s re- silience. This information identifies whether (and how) an asset is exposed and vulnerable to natural hazards and the asset’s criticality to the opera- tions of the facility. The Port of Long Beach began the development of its Climate Adaptation and Coastal Resiliency Plan with an inventory of criti- cal assets. The inventory included the piers, road and rail transportation, utilities, critical buildings, and the value and type of cargo. Infrastructure outside the port boundaries, such as roads, that are critical to port opera- tions were also included. They then used the asset inventory to analyze which assets were exposed and vulnerable to natural hazards.7 In addition to inventories of physical assets, transportation agencies should also keep an inventory of organizational assets specifically designed for operational resilience, such as procedures, tools, and guidance; continu- ity of operations plans; and staff training resources. Assets should include physical and organizational assets designed to prevent disruption and to speed recovery. Asset inventories need periodic updating to reflect changed assets and asset conditions. Evaluating the Criticality of Assets Criticality can be understood as the importance of an asset to the agency’s mission and to society. Criticality metrics capture this importance from the perspective of business continuity, users, the local or regional economy, health and safety, equity, and other social factors. As described in Chap- ter 3, FHWA encourages agencies to conduct a criticality assessment early in the analysis process to prioritize which assets or parts of the network to 5 FHWA (Federal Highway Administration). 2019. “How TPM and Asset Management Work Together.” https://www.fhwa.dot.gov/tpm/resources/working.cfm. 6 FTA (Federal Transit Administration). 2016. “National Transit Asset Management Sys- tem Final Rule.” https://www.transit.dot.gov/regulations-and-guidance/asset-management/ national-transit-asset-management-system-final-rule. 7 Port of Long Beach. 2016. Port of Long Beach Climate Adaptation and Coastal Resiliency Plan. https://www.slc.ca.gov/wp-content/uploads/2018/10/POLB.pdf.

DECISION SUPPORT FRAMEWORK 109 evaluate for vulnerability. Criticality metrics are typically a composite of several measures, not all of which may be represented in monetary terms. Any process used to score or weight the component parts of criticality metrics should be transparent.8 The Colorado DOT (see Chapter 3) developed criticality metrics for the overall highway system that combined physical inventory metrics with indicators of economic and social value. The Hillsborough County 8 U.S. DOT (U.S. Department of Transportation). 2014. Assessing Criticality in Transpor- tation Adaptation Planning. https://www.fhwa.dot.gov/environment/sustainability/resilience/ tools/criticality_guidance/criticality_guidance.pdf. BOX 5-1 Examples of Asset- and System-Level Inventory Attributes Relevant for Resilience Analysis Asset Level • Asset attributes — Name and number — Location — Description (e.g., design) — Age — Asset class/group — Replacement/renewal value — Design life or expected remaining life — Rehabilitation schedule • Asset condition • Functionality—services provided, volumes carried, traffic mix • Asset history (e.g., prior damages, rehabilitation) • Inspections, maintenance, and rehabilitation resources (including those to increase resilience, such as storm water management improvements, grade improvements, etc.) System Level • Number of inspections and maintenance activities on schedule • Upgrades that have increased resilience (e.g., by raising the facility’s eleva- tion or boosting earthquake resistance) • Capacity (i.e., volumes and loads) • Utilization (i.e., volumes and types of traffic loads carried) • Critical intermodal connections • Redundancy • Interoperability and interdependence with other systems (including connect- ing elements)

110 INVESTING IN TRANSPORTATION RESILIENCE Metropolitan Planning Organization, in its FHWA resilience pilot,9 used its travel demand model to assess criticality based on the regional significance of roads in the county. The analysis calculated an area-based criticality metric made up of the population and employment density of every Traffic Analysis Zone (TAZ). For the Origin-Destination (O-D) criticality measure, the TAZ criticality ratings were used to calculate a criticality score for each O-D pair, which was transformed into the criticality of traffic flows on the road network. Finally, the road network was sorted into three criticality tiers. Box 5-2 presents some examples of the factors to consider when assess- ing criticality, both quantitatively and qualitatively. In the absence of data, stakeholder opinions are often used to score criticality. Asset criticality can be assessed as part of the asset inventory or as a separate step. Criticality metrics can even be imported from other planning processes. As described in Chapter 3, when criticality metrics are combined with metrics for vulnerability or risk, they can also give an indication of overall resilience at the system or agency level. 9 Hillsborough County Metropolitan Planning Organization and Planning Commission. 2014. Hillsborough County MPO: Vulnerability Assessment and Adaptation Pilot Project. https://www.fhwa.dot.gov/environment/sustainability/resilience/pilots/2013-2015_pilots/florida/ final_report/florida.pdf. BOX 5-2 Examples of Factors to Consider for Assessing Criticality • Level of current use (e.g., traffic volume and mix) • Projected future traffic volume • Projected population density • Projected employment density • Projected freight traffic (e.g., volumes, key product types) • Proximity or primary route to major economic and community centers • Part of strategic transportation network (e.g., National Highway Freight Net- work or Strategic Highway Network; hub airports with higher share of con- necting flights) • Intermodal connections • Evacuation route • Link to first response facilities • Transit coverage and ridership • Social and demographic attributes of communities served (e.g., the Centers for Disease Control and Prevention’s Social Vulnerability Index) • Characteristics of redundant routes and modes (e.g., availability, added dis- tance and time, traffic volume and load-bearing capacity)

DECISION SUPPORT FRAMEWORK 111 Characterizing Natural Hazards and Their Likelihood Evaluation and quantification of the character and likelihood of natural hazards with the potential to affect the transportation system under analysis is a key element of the decision support framework. Hazard characteriza- tion is an input to the main resilience investment analysis and typically uses externally provided data. Implementing this step of the framework may in- volve defining a criterion event (e.g., 200-year storm—annual probability of 0.5%) or set of events and requires accounting for changes in environmen- tal conditions due to climate change. As discussed in Chapter 3, criterion events reflect the level or intensity of the hazard chosen as the standard for design and evaluation, relevant to the specific transportation system and assets under evaluation. The criterion event or environmental conditions will differ by location and asset type. The types of natural hazards and their potential to damage infrastructure assets and disrupt travel are covered extensively in Chapter 2. To address uncertainty, a set of criterion events might be defined and used as scenarios in resilience analysis. For example, in some settings it may be appropriate to define and test separate scenarios for riverine flooding, wildfires, and extreme snowfall. Key aspects for characterizing the natural hazard are the type of hazard and its location, scale, intensity, frequency, persistence (such as sea level rise), duration, and the timing of any advance warning. The likelihood or probability of an event has traditionally been determined from the historic frequency of events. As discussed above, uncertainties can be addressed by considering a range of events or scenarios. However, climate change causes the analysis of likelihood based on his- toric data to be inaccurate. The likelihood and character of natural hazards are changing, and forecasts need regular updates using trend analysis with recent data and using scenario modeling, which tests the consequences of a range of future conditions. The uncertainty around the effects of climate change is compounded when using longer analysis horizons, typical of in- frastructure investments with long life cycles. This suggests that if changes in natural hazard risks accelerate, a reexamination of resilience invest- ments may be warranted before the end of asset life is reached. In the face of climate change, regular adaptation is likely to be a safer strategy than “set-it-and-forget-it.” Transportation agencies should obtain and maintain an up-to-date inventory of data describing the specific natural hazards affecting their transportation assets. These agencies depend on other federal and state agencies and private organizations for much of the information, includ- ing trends and forecasts about natural hazards and climate change effects (e.g., the National Oceanic and Atmospheric Administration’s Atlas 14

112 INVESTING IN TRANSPORTATION RESILIENCE precipitation data,10 the Federal Emergency Management Agency’s flood maps,11 FHWA’s Climate Model Intercomparison Project Climate Data Pro- cessing Tool,12 the Colorado Geological Service,13 OpenQuake14). It is es- sential that these significant data be updated and maintained. As discussed in Chapter 2, transportation agencies must augment the external data with local and transportation agency experience. Some of the natural hazard data that transportation organizations should consider in their analysis are identified in Box 5-3. Because many areas of the country are prone to multiple hazards, the possibility of multiple, simultaneous hazards must be addressed in the resil- ience analysis. The analysis should consider the likelihood of several hazard events happening simultaneously or in quick succession and the probability of cascading events, when one event causes or worsens a subsequent event. Evaluating the Vulnerability of Assets Vulnerability refers to the susceptibility of assets and systems to damage and disruption. That is, for a given hazard (e.g., a hurricane) of a given magnitude (e.g., Category 3), how much damage to assets and travel dis- ruption will occur? Vulnerability is influenced by the location, design, materials, and other attributes of the asset and by the characteristics of the natural hazard. While vulnerability assessments for assets focus on the likelihood of failure, damage, or disruption at the specific location of each asset, vulner- ability assessments at the system or network level require a different set of metrics or indicators. Examples of system-level metrics are listed in Box 5-4. Vulnerability assessment should also include assessments of interdependent systems (e.g., an earthquake leading to failure of the power supply needed to run rail transit) and of simultaneous and cascading hazard events. 10 NOAA (National Oceanic and Atmospheric Administration). 2017. “NOAA Atlas 14 Point Precipitation Frequency Estimates.” https://hdsc.nws.noaa.gov/hdsc/pfds/pfds_map_ cont.html. 11 FEMA (Federal Emergency Management Agency). 2018. “FEMA Flood Map Service Center.” https://toolkit.climate.gov/tool/fema-flood-map-service-center. 12 FHWA. n.d. CMIP Climate Data Processing Tool 2.1. https://fhwaapps.fhwa.dot.gov/ cmip. 13 Colorado Geological Survey. n.d. “Colorado Geological Survey: Geoscience for Colorado.” https://coloradogeologicalsurvey.org. 14 OpenQuake. n.d. “The OpenQuake Platform.” https://platform.openquake.org.

DECISION SUPPORT FRAMEWORK 113 Evaluating the Consequences of Hazard Scenarios Consequences measure the economic and social costs resulting from the relevant hazard. Consequences are the values lost or disrupted. The major categories of consequences are the costs to restore functionality and repair or replace the asset, and the value, including criticality, of the functionality that was disrupted because of the hazard. In the context of climate change, BOX 5-3 Examples of Natural Hazards and Climate Change Stressors to Consider (including intensity, duration, geographic extent, and other attributes) Meteorological Hazards • Avalanche • Debris flow • Drought • Fire/wildfire • Flood/flash flood • Hail • Heavy rain • High wind • Ice flow • Lightning • Mudflow • Snow • Storm surge • Tornado • Tropical cyclone • Water table changes Geological Hazards • Earthquake • Land subsidence • Landslide and rockfall • Sinkhole • Tsunami • Volcanic eruption Climate Change–Related Hazards • Precipitation: changes in averages, extremes, and seasons • Temperature: changes in averages, extremes, and seasons • Sea level rise • Interaction of precipitation, temperature, and sea level changes with other meteorological hazards • Freeze-thaw events

114 INVESTING IN TRANSPORTATION RESILIENCE the costs may be ongoing. Hazard-driven morbidity and mortality to those affected by the hazard are part of the consequences. Examples of metrics for consequences are presented in Box 5-5. The costs of lost functionality will generally depend on the transportation agency’s operational resilience— how quickly it can respond and restore service on the infrastructure that is damaged. Other consequences may depend on the agency’s mission. For example, the San Diego International Airport includes the consequences to the wildlife habitat that it maintains. Monetizing consequences is necessary to develop risk-based resilience metrics, but it will not always be feasible. The Colorado DOT monetizes the consequences of damage and disrup- tion by computing the annualized owner costs (e.g., asset replacement and cleanup costs) and user costs (e.g., value of time lost to delays and travel costs of detours); it also includes measures of social vulnerability in its criticality analysis process. BOX 5-4 Examples of System-Level Vulnerability Metrics • Network (route) miles in 100- and 200-year flood zones • Number of critical facilities in 100- and 200-year flood zones • Number of bridges within 100-year floodplain • Coastal railroad route miles less than 2 feet above 2050 projected sea level rise • Areas of inundation due to sea level rise • Percentage of critical equipment affected by high/low temperatures • Number of (or list of) critical components subject to failure due to ambient temperature above X°F • Miles of highways in high wildfire danger areas (wildfires within 5 miles in past 10 years) • Annual percentage of routine facility inspections completed on time • Facility (bridges, highways, airport pavements) condition ratings—number, mileage, or percentages in fair or poor categories • Number of posted bridges (loads limited below standard) on National High- way System

DECISION SUPPORT FRAMEWORK 115 Estimating Risk In the proposed framework, risk is defined conceptually as follows: Risk = Hazard Likelihood × Vulnerability × Consequences where • Risk is the expected value of losses to the economy and society due to the disruption of transportation functionality caused by natural hazards, • Hazard likelihood describes probabilities of relevant natural hazards, • Vulnerability measures asset susceptibility to natural hazards, and • Consequences describe the value of functionality lost because of destruction of assets or service disruptions, including losses to asset owners, asset users, and communities. BOX 5-5 Examples of Metrics for Consequences Owner Consequences • Disruption response costs • Asset replacement costs • Asset repair costs • Cleanup costs • Loss of revenue • Liability for injuries or death • Loss of labor productivity User Consequences • Value of time lost to delay • Cost of added travel for detours and rerouting • Cost of foregone trips Community Consequences • Losses to local and regional economy — Business or tourism sales lost — Workdays lost — Jobs lost • Environment damage • Isolation or loss of access • Other community impacts

116 INVESTING IN TRANSPORTATION RESILIENCE Managing the risks resulting from disruptions due to natural hazards and climate change is a key objective for transportation agencies addressing resilience. This requires having an understanding of the risk associated with an asset or parts of the network due to the relevant hazards. Measuring all of the concepts quantitatively, however, can become difficult or impossible. While for many transportation agencies data avail- ability remains an obstacle to conducting resilience analysis, the complexity of calculating and communicating multi-dimensional relationships is the primary impediment. Because of these complexities, some simplifications might be needed. Some transportation agencies have limited their efforts to evaluating one hazard at a time or to using qualitative scoring to character- ize or rank risks, where that scoring is informed by the best data available. As illustrated in Chapter 3 (see Figure 3-1), this qualitative assessment of risk can then be used to prioritize risks in support of resilience investment decisions. APPLYING THE RESULTS OF THE DECISION FRAMEWORK By identifying risk, or the expected value of losses to the economy and so- ciety due to the disruption of transportation functionality caused by natural hazards, the steps delineated above in essence provide transportation agen- cies with a quantification of resilience benefits. Those benefits, however, can only be realized in part or in full by making the right investments, and it is likely that decision makers will have multiple resilience investment options to consider. Each option will present costs, which must be weighed against the potential for that option to confer resilience benefits. Identifying Options to Increase Resilience and Their Benefits and Costs With an understanding of the risk that natural hazards pose to critical as- sets or portions of the system, an agency can design candidate mitigation actions and identify and assess their benefits and costs. Increasing resilience through investments can be achieved by a number of actions as described in Chapters 3 and 4, and summarized here: • Prevent disruption and destruction of transportation facilities and services by — Building or rebuilding more robust facilities—for example, by designing new facilities with increased resistance to damage by natural hazards or the impacts of climate change, by pro- tecting bridges against scour, by increasing bridge clearances above waterways, or with seismic retrofits;

DECISION SUPPORT FRAMEWORK 117 — Adding redundancy—for example, by adding new routes, im- proving alternative routes, adding or identifying alternative transportation modes, identifying alternative sources of supply of essential resources or services, or acquiring back-up power sources to support critical systems for multiple days (e.g., com- mand and control centers, traffic signals, communication sys- tems, rail crossing barriers, bridge lifts); and — Relocating vulnerable facilities away from areas with high hazard exposure (e.g., rivers, coastal zones, unstable rock formations). • Restore functionality rapidly by — Enhancing response resourcefulness—developing disaster recovery plans and securing adequate resources in advance for rapid restoration of functionality, establishing mutual aid or cooperation agreements, creating secure and redundant communications networks and protocols, and/or setting aside emergency funds specifically dedicated for responding to natu- ral hazard/climate change events; — Improving quick response capabilities, including implementing event prediction and detection, increasing multi-agency disaster response planning and drilling, preplanning detours and modal diversions, establishing decision processes for rapidly invoking detours and diversions, arranging alternative sources for criti- cal supplies (e.g., food, water, medicines, repair materials), and establishing task order contracts for rescue and rebuilding; and — Building or rebuilding infrastructure assets so that they can more quickly recover functionality, including designing bridges and pavements to withstand prolonged immersion in water and installing pumping systems at low-lying airports for quick restoration of operations. Box 5-6 provides an overview of the types of benefits associated with resilience investments. As with the metrics from previous framework ele- ments (e.g., criticality), while quantification is ideal, it might not always be possible. In those cases, agencies should develop judgmental scales based on qualitative assessments. Estimating these benefits for a proposed investment is a complex task. With pre- and post-event data, the estimation will be somewhat easier for addressing the benefits of investments for post-disruption restoration and recovery, especially if a good analysis of the impacts of prior events has been conducted. To evaluate projects intended to reduce future disruptions, it is necessary to construct “with” and “without” scenarios (described in Chapter 4) to estimate the costs of disruptions due to a criterion event and

118 INVESTING IN TRANSPORTATION RESILIENCE those costs that would be avoided because of the investment.15 This requires a detailed understanding of the asset or system being studied, which should come from the asset management plan, as well as a clear specification of the criterion hazard event or events. The difference between “with” and “without” the investment defines the benefit of that investment. While conceptually straightforward, this process presents several chal- lenges. First, estimating the future damage costs requires good informa- tion on the efficacy of the investment. That is why it was suggested that 15 Aerts, J.C.J.H., W.J. Wouter Botzen, K. Emanuel, N. Lin, H. de Moel, and E.O. Michel-Kerjan. 2014. “Evaluating Flood Resilience Strategies for Coastal Megacities.” Science 344:473–475. BOX 5-6 Types of Benefits of Resilience Investments Infrastructure Owner-Operators—Costs Reduced, Avoided • Emergency operations • Recovery and restoration • Reconstruction Users (freight)—Costs Reduced, Avoided • Trip delay costs • Rerouting costs • Canceled trip costs • Inventory costs Users (personal travel)—Costs Reduced, Avoided • Trip delay costs • Rerouting costs • Canceled trip costs • Trip reliability Communities—Costs Reduced, Avoided • Business, tourism sales lost, deferred • Workdays lost, furloughs, jobs lost • Injuries and deaths • Delayed shipment costs (e.g., stockouts, supply chain disruptions) • Canceled shipment costs (e.g., stockouts, supply chain disruptions) • Environmental damage costs • Reductions in damage costs for non-transportation facilities and activities Communities—Positive Changes • Jobs gained in restoration, new construction

DECISION SUPPORT FRAMEWORK 119 focusing on mitigation actions with some proven efficacy is advantageous. Still, design engineers should be able to address changes in structural per- formance under stress brought about by mitigation actions. Addressing the changes in travel costs calls for the application of travel forecasting tools (as in the Hampton Roads Transportation Planning Organization’s use of Volpe’s Resilience and Disaster Recovery Metamodel, described in Chapter 3). Capturing the social and economic benefits is important but requires still different tools from the field of economic impact analysis. For more complex cases, some qualitative analysis driven by local data on social characteristics and vulnerability will be essential to address the social and equity impacts of resilience investments. Integrating these benefits into a single metric also presents a chal- lenge, one that is essentially the same as that faced when making major infrastructure investments. While many of these benefits can be monetized, based on market values, revealed or stated preferences, or other methods, it is likely that some important qualitative benefits will remain and will require judgment. The evaluation time frame, the future period over which benefits are assessed and aggregated, can be defined based on one of several factors, including the expected or design life of the asset, the period for which a reliable forecast can be made (probably shorter than the design life), or a target year determined by local or national policy. A longer time frame may be more appropriate for addressing the benefits of investments to mitigate the effects of climate change, but evaluating investments over a longer time period increases the uncertainty of the estimates. One way to address this is to plan for the long run but periodically reassess system resilience and con- sider if mitigating investments need to be adjusted. Selecting flexible, adapt- able designs will make it easier to adjust system resilience in the future.16 Estimating the costs of options to reduce risk is an essential step in pre- paring for a BCA. The most obvious costs are the costs to the infrastructure owner of modifying the infrastructure to reduce its vulnerability to damage in the event of a hazardous event. These are both capital (initial) costs and ongoing operation and maintenance costs. But the out-of-pocket cost to modify infrastructure is not the only type of cost that should be considered. If robustness of infrastructure is increased by rebuilding, for example, the infrastructure may need to be taken out of service for a period of time while the reconstruction is under way, reducing or eliminating its ability to provide services to users. If redundancy is increased by building new routes, land for that new construction may need to be acquired by eminent domain from property owners, who may consider the compensation for 16 Chan, R., P. Durango-Cohen, and J.L. Schofer. 2016. “Dynamic Learning Process for Selecting Storm Protection Investments.” Transportation Research Record 2599:1–8.

120 INVESTING IN TRANSPORTATION RESILIENCE their property to be inadequate to match their perceived loss. Construct- ing additional highway capacity may increase highway usage, generating increased emissions of greenhouse gases and other pollutants. Relocating vulnerable facilities to less vulnerable locations may have adverse effects on how well those facilities can serve their customers in normal times. Life- cycle costs are also difficult to estimate. Benefit-Cost Analysis Transportation agencies have long used BCA to assess proposed projects; thus, using BCA to analyze resilience improvements adapts a familiar tool to advance resilience. The strengths and the weaknesses of BCA are well known. BCA can incorporate life-cycle—construction, operations, and maintenance—costs for the asset (or operational improvement) and include the life-cycle benefits of resilience to users and society. BCA can also be used to analyze the costs of inaction. The challenge is to capture all of the costs and benefits necessary to give decision makers a comprehensive picture of proposed resilience improve- ments. Some categories of benefits, such as equity considerations (which are not always significant), benefits of protecting endangered species, and benefits of preventing low-probability but high-risk events, cannot always be measured quantitatively. Although some aspects of cost estimation, such as life-cycle costs, are not simple tasks, the major challenge in applying BCA is usually getting a comprehensive analysis of benefits. As described, the benefits to be included in the resilience BCA will primarily come in the form of expected reductions in the costs of disruption, including reductions in adverse social impacts and inequitable distributional effects. The National Cooperative Highway Research Program report Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook provides up-to-date guidance on integrating resilience into BCA and other investment analysis techniques.17 Box 5-7 illustrates some of the evaluation measures that might be applied. Because transportation infrastructure is typically very long-lived, the choice of a discount rate is a critical step in evaluating both the benefits and costs of an infrastructure project. The discount rate converts future benefits and costs to a present value by multiplying the future benefit or cost by 1/ (1 + r)n, where “r” is the discount rate per year and “n” is the number of years between the decision year and the future year in which the benefit 17 NASEM (National Academies of Sciences, Engineering, and Medicine). 2020. Incorpo- rating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook. Washington, DC: The National Academies Press.

DECISION SUPPORT FRAMEWORK 121 or cost occurs. The higher the discount rate, the less those future benefits and costs count in present-value terms. Guidance from the federal Office of Management and Budget has rec- ommended a real discount rate of 7% since 1992. Over the past 20 years, real rates of return on fixed income assets (such as Treasury bonds) have fallen substantially, calling into question the continuing validity of 7% as an appropriate long-term discount rate. Moreover, a basic element of a dis- count rate is the rate of time preference, which reflects the rate at which an individual makes trade-offs between future benefits and present benefits. If most individuals will not be alive at the future time, perhaps 50–100 years in the future, when future benefits and costs are realized, then the rate of time preference becomes an intergenerational trade-off. When the benefits and costs are experienced by different generations, it raises questions as to whether an individual’s rate of time preference is valid as a measure of how a society should trade off future versus present benefits and costs. As a result, many observers have argued that a lower discount rate, perhaps 3%, is appropriate for discounting future benefits and costs that involve intergenerational trade-offs. Some methodological approaches for BCA have recommended the use of declining discount rates over time to capture the issue of intergenerational equity. For instance, the Green Book used in project appraisals in the United Kingdom recommends an initial discount BOX 5-7 Examples of Investment Decision-Making Metrics Derived from BCA • Benefit-cost ratio • Return on investment • Net present value • Costs avoided — Infrastructure damage — Incremental transportation costs—time and money — Economic disruption costs (due to blocked or delayed flows, late or failed deliveries, product spoilage, etc.) — Social disruption costs—social connections, impacts to vulnerable com- munities, health care, education activities delayed or prevented • Equity of distributional effects — Inequities in the distribution of negative impacts across economic and social groups and on vulnerable populations

122 INVESTING IN TRANSPORTATION RESILIENCE rate of 3.5% followed by a declining rate schedule for projects with long- term duration.18 BCA and the Investment Decision Formalizing system resilience concepts and analysis into transportation agency decision making can help decision makers make informed choices to manage the risks of disruptions caused by natural hazards and climate change stressors. The results of BCA can be critical to this process, and the framework proposed in this chapter to measure resilience benefits is conducive to the application of BCA. However, BCA is rarely used as the sole basis for decision making. Typically, there are considerations that are omitted from even a good BCA, such as social impacts, equity con- siderations, and the value to be placed on low-probability but high-risk events. Decision makers in both private and public organizations must make decisions that use judgment to place appropriate weights on these considerations. Nevertheless, a BCA can still be very useful, for example, in distinguishing between options that have different outcomes in terms of measurable costs and benefits but are similar in the more difficult-to- measure considerations. Given the possibility that some impacts of disruptions due to natural hazards will not be assessed in monetary units, either because doing so is too difficult or uncertain or because the deduced monetary values do not reflect the real value of losses to people, augmentation of classic BCA with additional quantitative measures or qualitative descriptions may be neces- sary to reflect the full set of benefits in terms of damage costs avoided and costs of resilience investments. It is possible to conduct BCA not only at the project level but also at the program level. A program-level BCA provides information on what the overall budget of a program should be to achieve certain resilience targets based on BCA principles. No federal tools exist to conduct such analysis for resilience to natural hazards and climate change, but existing models used for condition and performance reporting illustrate the potential for it. For example, FHWA’s Highway Economic Requirements System model uses BCA for program-level assessments, in particular to assess the current and future physical conditions and consider standard options for improving pavements. FTA also uses program-level BCA with its Transit Economic Requirements Model that supports its assessment of future capital invest- ment needs. 18 HM Treasury. 2020. The Green Book: Central Government Guidance on Appraisal and Evaluation. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/ attachment_data/file/938046/The_Green_Book_2020.pdf.

DECISION SUPPORT FRAMEWORK 123 CHAPTER SUMMARY This chapter’s review of both practice and research suggests that more can be done to make the calculus of resilience a more systematic and deliberate part of transportation asset management and investment decision making. The review suggests that resilience should be measured and assessed using a multi-step, multi-hazard analytic framework. The process of assessing the potential benefits of resilience investments includes detailed inventories of assets (or portions of a network) that exist and are planned; assessments of the characteristics and likelihood of natural hazards occurring in the future; and predictions of the vulnerability of the inventoried assets to disruption, damage, and destruction from the hazards. These assessments should be accompanied by determinations of the criticality, or value, of each asset’s functionality and estimations of the consequences of damages to the asset and its lost or degraded functionality. The avoidance of future losses in functionality, as incurred by infrastructure owners and users and the broader community, represents the societal benefits of effective resil- ience investments. Investing in resilience requires spending funds in the present to gain benefits that may or may not be realized in the future. The decision to make a resilience investment must consider its prospective benefits in rela- tion to its life-cycle costs, including financial outlays and other sacrifices. BCA is the analytic tool often used to support such decision making. While translating benefits and costs into monetary values facilitates BCA, resil- ience investments can also be evaluated using quantitative, non-monetary measures and qualitative descriptions to account for the full set of possible outcomes, including equity and distributional impacts. A BCA that yields results showing positive net benefits represents the societal gain from a resil ience investment that takes into account its life-cycle costs and benefits. While BCA is rarely used as the sole basis for making decisions that must take into consideration interests such as equity and distributional impacts, a BCA can nevertheless be an important part of the resilience calculus.