Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook (2020)

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.

68 Introduction to Levels of Analysis Historically, many DOTs have used CBA only for large or complicated projects; however, with increased emphasis on asset management planning, transportation practitioners increasingly recognize that CBA would be useful in design and planning. They also realize that CBA could be applied to a betterment decision on an FHWA ER project, but CBA is generally regarded as cumbersome and costly, and transportation practitioners are reluctant to undertake the analysis for decisions regarding smaller capital projects. Yet, changes in weather and climate patterns are influencing assets designed for smaller events. More frequent significant rain events and increases in sea levels are resulting in increased incidents of nuisance flooding, causing trans- portation assets to be inundated more frequently. Transportation practitioners need a short and simple method to evaluate if adaptation strategies will be considered, and if so, what level of investment might be cost-effective. Chapters 7 and 8 describe an approach for conducting an initial screening to evaluate if adap- tation strategies might be cost-effective, and if implemented, what level of damages might be expected with and without adaptation. The focus of climate change impacts considered in this methodology is on the increased design flood discharges for transportation facilities, including culverts, bridges, and stormwater control facilities. The general assumption is that the rela- tionship between discharges and their frequency (i.e., return interval) for current conditions is based on a historically stationary hazard, while future values may be subject to non-stationarity brought about by climate change. The discussion that follows focuses on riverine flooding, but it is independent of flow rates or water surface elevations, so can be equally applicable to riverine or coastal analysis. The design criteria commonly used for transportation facilities, including drainage work and flood control projects, is based on annual exceedance probability and its reciprocal, the return period. Table 13 summarizes the design criteria most transportation agencies use. The value of annual exceedance probability or return period for each design level represents an acceptable level of risk at that level. Although design discharges and flood levels may increase under climate change scenarios in comparison with current conditions, facilities do not necessarily need to mitigate against events with larger return periods. It simply means the same return period (same failure prob- ability) will feature higher discharges and flood elevations. While the potential damages associated with a specific flood discharge will not change in the future, the overall hazard level will increase if the same discharge will occur more frequently (i.e., will have a smaller return period) (Figure 20). Ensuring the system will accommodate the increased discharges will make it resilient against the impacts of climate change, but attempts to adapt to climate change impacts need to be balanced with approaches that make economic sense. In some C H A P T E R 7 Study Level 1 Climate Resilience Cost-Benefit Analysis

Study Level 1 Climate Resilience Cost-Benefit Analysis 69 cases, designing for the absolute worst case scenario might not be cost-effective, as discussed in Box 3. Flooding is one of the most frequent and costliest natural hazards to damage transportation assets and systems, as well as one of the natural hazards likely to be most affected by climate change; average annual flood losses in the five most at-risk cities in the United States are expected to be approximately $8 billion by 2050 (Hallegatte et al., 2013). Consequently, this CBA methodology focuses on flooding. The CBA analysis levels follow the approach taken in HEC-17, “Highways in the River Environment: Floodplains, Extreme Events, Risk, and Resilience,” which provides technical guidance and methods for assessing the vulnerability of transportation facilities to extreme events and climate change in riverine environments. The focus in HEC-17 is quan- tifying exposure to extreme flood events, considering climate change and other sources of 50 100 500 33 50 62 100 260 500 1,000 3,000 5,000 7,000 9,000 11,000 13,000 15,000 17,000 1,00000101 Q p (c fs ) Current Future Figure 20. Climate change could result in a given level of an event occurring more frequently in the future. Table 13. Design criteria example for annual exceedance probability (AASHTO, 2014). Roadway Classification Annual Exceedance Probability (percent) Return Period (years) Interstate, Freeways (Urban/Rural) 2% 50 Principal Arterial 05%2 Minor Arterial System, ADT>3,000 VPD 2% 50 Minor Arterial System, ADT<3,000 VPD 52%4 Collector System with ADT>3,000 VPD 4% 25 Collector System with ADT<3,000 VPD 01%01 Local Road System 20%–10% 5–10 ADT = average daily traffic. VPD = vehicles per day.

70 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook non-stationarity. HEC-17 offers five levels of analysis with increasing complexity and accuracy for estimating the future discharges: 1. Historical Discharges, 2. Historical Discharges + Confidence Limits, 3. Precipitation Projection Trend Test, 4. Projected Discharges using CMIP tool, and 5. Customized Projected Discharges with Climate Scientist. HEC-17 Levels 1 and 2 are simple analyses, resulting in an amplified design discharge. HEC-17 Levels 3, 4, and 5 are more complex analyses, resulting in amplified discharges for each return period that are calculated to account for climate change: • HEC-17 Level 3 is a transition level that involves T-year, 24-hour precipitation projection using various CMIPs, which may result in staying with Level 2 (if the trend is weak) or moving to Level 4 to compute future discharges using future precipitation (if the trend is strong). • HEC-17 Level 4 involves incorporating rainfall projections into rainfall/runoff hydrology. It could be as simple as using future mean annual precipitation in a regional discharge regression equation or as complicated as using projected rain and temperature in a full hydrologic model. • HEC-17 Level 5 is an advanced version of Level 4, which is only appropriate for larger, costlier projects or infrastructure and requires expanded expertise in hydrologic modeling, climate science, and land use planning for custom, site-specific projections. This guidebook focuses on HEC-17 Levels 1, 2, and 3, which are the levels of analysis most likely to be completed for planning or comparison of design alternatives. Approaches for HEC-17 Levels 4 and 5 analyses are probabilistic and robust; they require the generation of peak flow using Monte Carlo simulations, determination of the flood elevation resulting from the gener- ated peak flow, estimation of the flood cost for each event when the elevation overtops the low point of the roadway, and calculation of the flood cost savings for each improvement option. These approaches appear to be reasonable and robust, but DOTs may generally consider them too complex and time-consuming for making funding decisions during the planning and design alternatives analysis phases. In this guidebook, HEC-17 Level refers to the level of analysis defined during this study and described in Chapters 7 and 8. Box 3. MaineDOT Culverts Maine looked at two inland corridors and the crossings involved, considering culvert and bridge sizing for extreme events. It used the ECOS tool to do CBA on two to three structures per corridor, along with depth-damage functions. Surprisingly, the most efficient solution in terms of cost/choice was the 25-year storm sizing. The 100-year storm has been a default upgrade/adaptation in the past. MaineDOT’s study used five different sizing scenarios, starting with a 25-year storm, a 100-year storm, then plus 25 percent, plus 50 percent, and 1.25 bankfull. It found that 25-year storm sizing could handle the amount of water generated in this watershed and that there were risks to a larger structure that might hold back more water. MaineDOT decided it was cheaper to replace the riprap, clean the corridor, and replace the pipe, as the modeled storm would not damage the road structure (MaineDOT, 2015).

Study Level 1 Climate Resilience Cost-Benefit Analysis 71 Process Walk-Through with an Example for Riverine Flooding Select Data Inputs and Data Sources Establish Base Conditions Study Level 1 analysis is basically an approximate test to see if it would be cost-effective to upgrade the hydraulic structure for the future conditions posed by climate change. The central point in this approach is that a given discharge, Q, will cause a given level of damages, D, with or without climate change. However, considering climate change, the given discharge of Q may have a smaller return period in the future than its current value, that is, the same flow, Q, will occur more frequently in the future than it does now, resulting in the level of damages, D, Parameter Value Used in Scenario Data Source(s) Facility of concern Culvert Project file Geographic location of the facility/corridor under consideration Chesterfield, VA Site plan, maps Hazard(s) of concern Flood Hazard analysis Current design criteria—flow rate 9,000 cfs Engineering designs and plans Current design criteria—recurrence interval 50-year event AASHTO design manual, DOT design manual Discount rate(s) to be used in the analysis 7% OMB A-94 Expected useful life of current facility Less than 2 years Capital plan, O&M records Expected useful life of replacement facility 50 years Virginia DOT design guides Anticipated time frame for implementation of adaptation strategies Less than 2 years Capital plan Scenario(s) to be used for analysis Precipitation conditions in 2049 NOAA Atlas 14, SWMM-CAT for warmer, wetter conditions 2045– 2075 Design concepts of adaptation strategies Enlarge culvert, add multiple culverts, use box or arch culvert Engineering department Cost estimate for each adaptation strategy (life-cycle costs, including any long-term adverse impacts from the adaptation strategy) Cost estimates Historical data, recent bids for similar work, cost-estimating software Identification of any non- quantifiable costs associated with the project None DOT analysis Estimates of damages sustained from the hazard of concern Loss estimates Historical data, engineering analyses, O&M records, depth-damage curves Estimates of additional benefits resulting from the project, separated by physical/social/environmental if using multiple discount rates Benefits estimates FEMA benefit-cost analysis tools for drought, ecosystem services, and post-wildfire mitigation Identification of any non- quantifiable benefits associated with the project None DOT analysis

72 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook occurring more often (Figure 21). Therefore, the goal of a Study Level 1 analysis is to identify how to improve the performance of the hydraulic structure or the resilience of the roadway, so that for a given return period the future flow, Q′, under climate change conditions has the same return period as the current flow, Q, so that the level of damages, D, is approximately the same in the future for the higher discharge rate as it is now for the current discharge rate. For damages to remain the same in the future, the improved structure needs to accommodate the addi- tional discharge. The basic premise for this analysis is that even though the relationship between frequency and discharge changes with time, the relationship between frequency and damages remains somewhat constant (e.g., damages sustained from a future 50-year event under changing climate conditions are the same as damages for a 50-year event under current conditions). This approach assesses damages for three event categories: (1) medium probability, low consequence (i.e., base case), (2) low probability, medium consequence, and (3) very low prob- ability, very high consequence, as summarized in Table 14. The following steps summarize the basic inputs and calculations required for this approach. It is intended to be simple enough to calculate by hand, although use of a computer spreadsheet program such as Excel will make calculations easier. This approach uses several variables; a summary of the variables used and their meanings is included in Table G-1 in Appendix G. A blank worksheet to complete a Level 1 analysis by hand is included in Appendix I. A list of sources where needed data might be found is included in Appendix J. In the steps below, the approach is applied to an example to demonstrate its use. The example is based on data for Chesterfield County, Virginia, using rainfall data from NOAA Atlas 14 for the watershed. Future discharge flows were calculated using the EPA’s SWMM-CAT model. However, other approaches can be used to calculate future discharges; some suggested methods Low Consequence Medium Consequence Very High Consequence Very Low Probability X Low Probability X Medium Probability X Table 14. Event categories most likely for application of a Level 1 analysis. Figure 21. Climate change and sea level rise will result in flood events of a given magnitude occurring more frequently in the future (after Vitousek et al., 2017). 7 6.5 5.5 4.5 3.5 6 5 4 3 0 25 50 75 100 125 150 Return Period [years] Fl oo d Le ve l [ m et er s] decrease in return period: sea-level rise Current sea level Future sea level fixed elevation f –1inc

Study Level 1 Climate Resilience Cost-Benefit Analysis 73 are included in HEC-17, Chapter 7, such as Rational Method, Unit Hydrograph, and the Natural Resources Conservation Service’s Peak Graphical Method. 1. Identify the largest return period for which there will be no damages. Typically, this is the design return period. Typical design return periods for transportation hydraulic structures are 10, 25, or 50 years depending on road classification (Table 13). This is Tcnd. For example, if a bridge is designed to safely pass the 50-year discharge, Tcnd would be 50 years. 2. Identify a return period associated with an event that would cause moderate but consider- able structural damage or roadway flooding and traffic interruption. Typically, this would be the next-highest standard return period to Tcnd, defined as Tcmod. For the bridge designed to a 50-year profile, Tcmod might be set to 100 years. 3. Identify a return period for which damages would be practically maximized. Larger or more significant events might cause greater damages, but their probabilities are so small that they do not add much overall risk. This maximum, realistically occurring return period is Tcmax. For example, for the bridge designed to 50 years, Tcmax might be the 500-year flood, which causes bridge structural failure, road embankment erosion, and loss of roadway function for several weeks or months. 4. Estimate total damages associated with Tcmod and Tcmax. Typical damages, Dcmod, at the Tcmod level, could include loss of riprap, short-term road closure, traffic control and road cleanup costs, and so on. Typical damages, Dcmax, at the Tcmax level could include the failure of the hydraulic structure leading to large structural damage and loss of road service and possibly injuries or fatalities. These damages may be estimated based on historical damage records for the same or similar structures or based on expected damages assessed by engineers. The damages are stated in terms of constant dollars by applying the appropriate present value interest factor (Appendix B). For this example, assume Dcmod is equal to $1,630,000 and Dcmax is $3,227,000. 5. Use Equation 10 to calculate the expected annual damages between Tcnd and Tcmod. Annual damages are the damages expected per year over the life of the asset or corridor, or the useful life of the adaptation project. “Expected” annual damages does not mean that these damages will occur every year. Equation 10. Calculating expected annual damages for an event of moderate damage in a Level 1 analysis. 2 1 1 D D D T T acmod cnd cmod cnd cmod = + −   For the example, 0 1,630,000 2 1 50 1 100 $8,150 D D acmod acmod = + −   = 6. Use Equation 11 to calculate the expected annual damages between Tcmod and Tcmax: Equation 11. Calculating expected annual damages for an event of severe damage in a Level 1 analysis. 2 1 1 D D D T T acmax cmod cmax cmod cmax = + −   For the example, 1,630,000 3,227,000 2 1 100 1 500 $19,428 D D acmax acmax = + −   =

74 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook 7. Use Equation 12 to calculate the total annualized damages, which is the sum of Dacmod and Dacmax: Equation 12. Calculating total annualized damages for a Level 1 analysis. D D Dac acmond acmax= + For the example, $8,150 $19,428 $27,578Dac = + = 8. Find the present value coefficient for the remaining project useful life (i.e., the remaining service life during the period of projected climate change) from Appendix B. For this example, the project useful life is 50 years and assumes the OMB A-94 rate of 7 percent. So, for this example: 13.801PVC = 9. Calculate the present value of total expected damages under current conditions using Equation 13: Equation 13. Calculating the present value of total expected damages under current conditions. D D PVCTc ac = For this example, $27,578 13.801 $380,604DTc = = DTc is also equal to the value of a hazard mitigation or resilience project that would eliminate all damages for even the 500-year return interval discharge in the absence of climate change. A hazard mitigation or resilience measure costing more than this would not be cost- effective if discharges (and hence damages) do not increase in the future. 10. Associate discharges with each of the three return periods Tcnd, Tcmod, and Tcmax under current (no climate change) conditions. This step will provide Qcnd, Qcmod, and Qcmax. For this example, assume: Q 9,000 cfs Q 10,505 cfs Q 13,982 cfs cnd cmod cmax = = = Table 15 summarizes this example for current climate conditions: 11. Create a graph by plotting the return periods Tcnd, Tcmod, and Tcmax on a logarithmic scale on the x-axis against the associated discharges on the y-axis. This can be done manually using logarithmic graph paper or on a computer using a spreadsheet program with graphing capabilities. The graph creates a straight-line trend showing return periods and expected discharges. For this example, the graph is shown in Figure 22. 12. Create a second graph by plotting the discharges on the x-axis (with a “normal” as opposed to a logarithmic scale) and the estimated damages associated with each discharge on the y-axis. For the example, the graph is shown in Figure 23. Establish Future Climate Conditions 1. After establishing baseline information for current climate conditions, begin to calculate future flows and associated expected damages for future climate conditions. To do this,

Study Level 1 Climate Resilience Cost-Benefit Analysis 75 start by identifying the climate change scenario or level of risk to be used for analysis (see Chapter 3 in this guidebook and Chapters 4–6 in HEC-17). For this example, a Gumbel distribution was applied to data from EPA’s SWMM-CAT model (U.S. EPA, 2014), which allows users to apply monthly adjustment factors that can represent future changes in climatic conditions. SWMM-CAT uses climate models from CMIP5 and downscaled data from Climate Resilience Evaluation and Awareness Tool (CREAT) 3 for return periods of 5, 10, 15, 30, 50, 100, and 500 years. Analyses can be done for the near term (2020–2049) or far term (2045–2074) for hot/dry conditions, warm/wet conditions, or median conditions. This example assumes warm/wet conditions in the near term. 2. For the selected climate scenario, calculate the estimated future discharges for each return period T′fnd, Tfmod, and Tfmax. This will result in identifying values for Qfnd, Qfmod, and Qfmax (see Table 16). 3. Plot the future discharges under the selected climate change scenario Qfnd, Qfmod, and Qfmax on the same graph as the baseline conditions. For this example, SWMM-CAT discharge outputs for climate-adjusted scenarios under near-term, warm/wet conditions are summa- rized in Figure 24. 4. Extend the linear trend for future climate conditions to the same discharge value for Tfnd (in this example, 9,000 cfs). This provides an estimate of the climate-adjusted return period Current Return Period, Tc Current Damages, Dc Current Annualized Damages, Dac Current Flow, Qc (cfs) Tcnd 50 $0 $0 9,000 Tcmod $8,150 10,505 Tcmax 500 $3,227,000 $19,428 13,982 Total annualized damages 875,72$ 100 $1,630,000 Table 15. Summary of flows for existing conditions for example project. 50, 9,000 100, 10,505 500, 13,982 1,000 3,000 5,000 7,000 9,000 11,000 13,000 15,000 1,00000101 Q (c fs ) T (years) Current Figure 22. Logarithmic graph of return periods and associated flows under existing conditions for example project.

76 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook Figure 23. Peak discharge and associated damages under current conditions for example project. Future Return Period, Tf Future Flow, Qf (cfs) T’fnd 50 9,979 Tfmod 100 11,665 Tfmax 500 15,562 Table 16. Estimated flows under future climate conditions for example project. Figure 24. Estimated return periods and associated flows for current and future climate conditions for example project.

Study Level 1 Climate Resilience Cost-Benefit Analysis 77 for the base flow (in this example, approximately 33 years). Alternatively, the future return period for the selected climate scenario can be calculated using Equation 14 and Equation 15: Equation 14. Calculating the logarithmic value of the climate-adjusted return period for the base flow under future conditions. LogT logT LogT LogT Q Q Q Q fnd fmod fmod fnd fmod cnd fmod fnd ( )= − − − −  and Equation 15. Calculating the value of the climate-adjusted return period for base flow under future conditions. 10Tf LogTf= Using these equations for this example, log 100 100 50 11,665 9,000 11,665 9,979 1.524 10 33.4 years for Q 9,000 cfs1.524 Log T Log Log Log T T fnd fnd fnd ( )( )= − − − − = = = = 5. Set the future damages corresponding to Tfnd to Dfnd = $0, as this value corresponds to the same discharge Qfnd (i.e., Qcnd = Qfnd). Interpolate the damages linearly for each of the revised future discharges using Equation 16, Equation 17, and Equation 18 such that Equation 16. Interpolating damages for future discharges for little damage. D D Q Q Q Q D Dfnd cnd fnd cnd cmod cnd cmod cnd ( ) ( ) ( )′ = + − − − and Equation 17. Interpolating damages for future discharges for moderate damage. D D Q Q Q Q D Dfmod cmod fmod cmod cmax cmod cmax cmod ( ) ( ) ( )= + − − − and Equation 18. Interpolating damages for future discharges for severe damage. D D Q Q Q Dfmax cmax fmax cmax cmax cmax ( )= + − For the example: 0 9,979 9,000 10,505 9,000 1,630,000 0 1,060,312 1,630,000 11,665 10,505 13,982 10,505 3,227,000 1,630,000 2,162,793 3,227,000 15,562 13,982 13,982 3,227,000 3,591,659 D D D fnd fmod fmax    ( ) ( ) ( ) ( ) ( ) ( ) ( ) ′ = + − − − = = + − − − = = + − = Plotting the damages against the peak discharges yields a curve for climate-adjusted flows shown in Figure 25.

78 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook 6. Calculate the annualized damages with climate adjustment using a similar approach to Equation 10, substituting the climate-adjusted values for the current condition values. For the example: D D D D D afnd afmod afmax af Tf $0 $1,060,312 2 1 33.4 1 50 $5,270 $1,060,312 $2,162,793 2 1 50 1 100 $16,116 $2,162,793 $3,591,659 2 1 100 1 500 $23,018 $5,270 $16,116 $23,018 $44,404 $44,404 13.801 $612,820 ( ) ( ) ( ) ′ = + −  = = + −  = = + −  = = + + = = =     7. Table 17 summarizes the climate-adjusted values for the example. 8. Use Equation 19 to compare the additional damages for the base case with and without climate adjustment: Equation 19. Calculating the additional damages for the base case with and without climate adjustment (i.e., value of cost-effective adaptation measures). = $612,820 $380,604 = $232,216 D D D D T Tf Tc T ∆ = − ∆ − Figure 25. Interpolated damages for peak flows under future climate conditions for example project. T Q (cfs) D Da Tfnd 33 9,000 $0 $0 T’fnd 50 9,979 $1,060,312 $5,270 Tfmod 100 11,665 $2,162,793 $16,116 Tfmax 500 15,562 $3,591,659 $23,018 Table 17. Summary of flows and damages for future climate conditions.

Study Level 1 Climate Resilience Cost-Benefit Analysis 79 This value represents the additional present value of the expected damages from climate change during the remaining bridge useful life. A hazard mitigation or resilience measure aimed at maintaining the current frequency-damage structure (design level) while accounting for climate change must cost less than this value to be cost-effective. For this example, such a measure could increase the safe capacity of the hydraulic structure from 9,000 cfs to at least 9,979 cfs and increase the cost over the base case by no more than $232,216. Based on engineering cost estimates (Chapter 4), enlarging the culvert or installing multiple culverts might be cost- effective, while installing a box or arch culvert might not be cost-effective. In addition to performing the analysis using the OMB-recommended 7 percent discount rate for the Level 1 analysis detailed on the previous pages, a sensitivity analysis was performed for the same example using a 3 percent discount rate in accordance with OMB guidance to reflect greater uncertainty associated with future climate risk. Using a 3 percent discount rate for the Level 1 analysis example on the previous pages changes the present value coefficient from 13.801 to 25.730, yielding benefits of $1,142,082 (versus $612,820 for a 7 percent discount rate). Future damages for the base case are calculated as $709,582 (i.e., $27,578  25.730 using Equa- tion 13), which increases the acceptable project cost differential over the base case to $432,500 (i.e., $1,142,082 − $709,582) (Table 18). Under these conditions, the box or arch culvert might also be cost-effective. The sensitivity analysis shows the impacts that uncertainty associated with climate risk can have on acceptable project costs. Practitioners will need to follow current federal guidance on which discount rate to use in analysis when federal funds are used in project funding; however, these individuals will need to determine what discount rate is acceptable and reflects expected risk when funding sources other than federal funds are being used for projects. Case Study As part of FHWA’s climate vulnerabilities pilot studies, the Minnesota DOT (MnDOT, 2014) evaluated the threat of flash flooding to the state’s highway system (https://www.fhwa.dot.gov/ environment/sustainability/resilience/pilots/2013-2015_pilots/minnesota/final_report/index.cfm). Asset types within the system identified as being susceptible to flash flooding included bridges, large culverts, pipes, and roads paralleling streams. MnDOT developed a series of metrics for each asset to evaluate its vulnerability, which allowed MnDOT to score each asset and rank vulnerability according to scores. For the study, two facilities were chosen for further evaluation. Both were large culverts. This case study applies a Study Level 1 analysis to one of the culverts, Culvert 5648, which carries MN-61 over Silver Creek in the Arrowhead region northeast of Two Harbors (Figure 26). Culvert 5648 has two cells, each with a 10-foot span (width) by 10-foot rise (height) and a length of about 90 feet (Figure 27). The culvert was built in 1963 and is at the end of its useful life. It is anticipated that precipitation levels will increase over the life of any new facility installed at this location. 7% Discount Rate 3% Discount Rate Present Value Interest Factor 13.801 25.730 PV of Project Benefits (current climate conditions) $380,604 $709,582 PV of Future Damages (future climate conditions) $612,820 $1,142,082 PV of Acceptable Project Cost (differential) $232,216 $432,500 Table 18. Comparison of Level 1 analysis results using 7 percent and 3 percent discount rates.

80 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook Figure 26. MnDOT evaluated Culvert 5648 for cost-effective adaptations to climate change (MnDOT, 2014). Figure 27. Upstream side of Culvert 5648 (MnDOT, 2014).

Study Level 1 Climate Resilience Cost-Benefit Analysis 81 MnDOT used a software tool called SimCLIM to evaluate future projections for three pre- cipitation scenarios: RCP 4.5, RCP 6.0, and RCP 8.5. All three scenarios considered 24-hour precipitation depths. Storm events with return periods of 2, 5, 10, 25, 50, 100, and 500 years were analyzed. Projections were obtained for three time periods through the year 2100, which coincides with the anticipated end of useful life of the new facility. MnDOT used the U.S. Department of Agriculture Natural Resources Conservation Service’s WinTR-20 program to model peak flows through the culvert for the various storm events analyzed. Hydrologic analyses included assumptions for future land cover based on a build-out of current zoning. A HEC-RAS model was used to evaluate the performance of the culvert under current and future peak flows. MnDOT made assumptions regarding the design of a base case and three potential climate-resilient alternatives: • Base case. Replace the existing culvert in-kind; include upgrades for required freeboard (3 feet) for 50-year flood stage and fish passage per regulatory requirements. Estimated cost: $710,000. • Option 1. Replace the existing culvert with a two-cell culvert having a 16-foot span (width) and 14-foot rise (height). This assumes the culvert will be sunk 2 feet into the stream bed to facilitate fish passage. This option is optimized for the low climate scenario in 2100. Estimated cost: $770,000. • Option 2. Replace the existing culvert with a 52-foot simple span bridge. This approach is optimized to meet the medium climate scenario in 2100. Estimated project cost: $1,130,000. • Option 3. Replace the existing culvert with a 57-foot simple span bridge. This approach is optimized for the high climate scenario in 2100. Estimated project cost: $1,210,000. Depth-headwater elevation curves with and without social costs were developed for each option. The software model COAST was used to evaluate the cost-effectiveness of each option using a 2 percent discount rate. The results indicated that, if social costs are included in the analysis, Option 1 is the most cost-effective approach for all three climate scenarios. If social costs are excluded from the analysis, replacement-in-kind is the most cost-effective approach for the low rainfall scenario, while Option 1 is the most cost-effective for the moderate and high rainfall scenarios. A Study Level 1 analysis was conducted using the data with social costs for the moderate scenario for a project useful life extending to 2100. The projected peak flows are summarized in Table 19. 24-Hour Storm Return Period (years) Existing Discharges (cfs) Medium Scenario Discharges for 2100 (cfs) 2 770 1,230 000,2053,15 10 1,880 2,660 076,3096,252 005,4073,305 100 4,140 5,420 008,7090,6005 Table 19. Summary of discharges for Culvert 5648 for the medium scenario for 2100.

82 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook Depth-damage data and a depth-damage curve were provided for Option 1. Because no data were available for the base case, the Option 1 data were applied to base conditions as well. Table 20 presents the depth-damage data for Option 1. The depth-damage data were correlated with the discharge-elevation curve (Figure 28) and projected peak flows (Table 21) to associate flows with different levels of damages, as shown in Table 22. The data were used to conduct a Study Level 1 analysis. Initial values used are shown in Table 23 and Table 24. The expected annual damages were calculated using Equation 10 through Equation 12. The annualized damages were calculated as D $0 D $570 D $1,712 D $570 $1,712 $2,282 acnd acmod acmax ac = = = = + = Total damages over the project useful life were calculated as D $2,282 39.745 $90,698Tc = = Flood Elevation (ft) Physical Damage and Repair Cost Socioeconomic Costs Property Total Cost Damage (%) Notes Detour Injury Days in Effect Cost 605 $0 0 $0 $0 $0 $0 0% 614 $0 0 $0 $0 $0 $0 0% 615 $30,000 0 $0 $0 $0 $30,000 8% Embank- ment erosion starts 616 $30,000 0 $0 $0 $0 $30,000 8% 617 $40,000 0 $0 $0 $0 $40,000 10% 618 $50,000 0 $0 $0 $0 $50,000 13% 619 $70,000 0 $0 $0 $0 $70,000 18% 620 $80,000 0 $0 $0 $0 $80,000 20% 621 $100,000 0 $0 $0 $0 $100,000 25% 622 $130,000 0 $0 $0 $0 $130,000 33% 623 $160,000 0 $0 $0 $0 $160,000 40% 624 $200,000 0 $0 $0 $0 $200,000 50% 625 $250,000 1 $140,000 $0 $0 $390,000 98% Overtop- ping 626 $320,000 5 $700,000 $80,000 $0 $1,100,000 275% 627 $400,000 15 $2,100,000 $80,000 $0 $2,580,000 645% Table 20. Depth-damage data for Culvert 5648.

628.00 626.00 624.00 622.00 620.00 618.00 616.00 614.00 612.00 Base Case Option 1 Option 2 Option 3 610.00 608.00 606.00 7,000 1,000 0 2,000 3,000 4,000 5,000 6,000 Flow Rate (cfs) H ea dw at er E le va tio n Low Point El. = 625 Figure 28. Depth-flow curves for Culvert 5648 replacement options (MnDOT, 2014). 24-Hour Storm Return Period Existing Dis- charges (cfs) Low Scenario Discharges Medium Scenario Discharges High Scenario Discharges 2040 (cfs) 2070 (cfs) 2100 (cfs) 2040 (cfs) 2070 (cfs) 2100 (cfs) 2040 (cfs) 2070 (cfs) 2100 (cfs) 2-year storm 770 1,070 1,100 1,120 1,090 1,160 1,230 1,180 1,370 1,550 5-year storm 1,350 1,760 1,810 1,830 1,800 1,900 2,000 1,930 2,190 2,460 10-year storm 1,880 2,360 2,420 2,450 2,420 2,540 2,660 2,580 2,920 3,250 25-year storm 2,690 3,260 3,350 3,390 3,340 3,500 3,670 3,550 4,010 4,460 50-year storm 3,370 4,010 4,120 4,170 4,113 4,300 4,500 4,360 4,920 5,480 100-year storm 4,140 4,810 4,940 5,000 4,930 5,170 5,420 5,240 5,940 6,610 500-year storm 6,090 6,870 7,060 7,150 7,040 7,410 7,800 7,520 8,590 9,630 Table 21. Expected flows for various return periods and climate scenarios for Culvert 5648. 24-Hour Storm Return Period Existing Discharges (cfs) Elevation (estimated) (ft) Estimated Damages (Base Case) Medium Scenario Discharges 2100 (cfs) Elevation Estimated Damages (Option 1) 2-year storm 770 608.5 $0 1,230 608.6 $0 5-year storm 1,350 609.5 $0 2,000 609.6 10-year storm 1,880 610.4 $0 2,660 611.4 $0 $0 $0 25-year storm 2,690 613.5 $0 3,670 613.2 50-year storm 3,370 615.1 $30,000 4,500 615 $30,000 100-year storm 4,140 616.8 $38,000 5,420 619.6 500-year storm 6,090 625 $390,000 7,800 627 $2,580,000 $76,000 Table 22. Summary of discharges and expected damages for Culvert 5648 under medium climate scenario conditions.

84 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook This means that the in-kind replacement culvert is expected to sustain damages totaling $90,698 over its useful life of 80 years under current climate conditions. Next the analysis was adjusted to account for climate change. As stated in Table 23, it was assumed that the $0 damage condition would still apply for a design flow of 2,690 cfs, but a new recurrence interval needed to be calculated for this damage-flow combination to incorporate the impacts of climate change. Using Equation 14 and Equation 15, the climate-adjusted recurrence interval was calculated to be 12 years. This same process was used to find the climate-adjusted recurrence intervals for the future flows in the medium scenario to 2100. Table 25 shows the recurrence intervals calculated. The annualized damages were calculated for this data based on Equation 10 and Equation 12. They were found to be D $704 D $423 D $1,502 D $953 D $6,470 D $704 $423 $1,502 $953 $6,470 $10,052 afnd afint1 afmod afint2 afmax af ′ = = = = = = + + + + = Return Period, Tc (years) Current Flow, Qc (cfs) Estimated Damages ($) Tcnd 25 2,690 $0 Tcmod 100 4,140 $38,000 Tcmax 500 6,090 $390,000 Table 23. Initial data used to conduct a Level 1 analysis for Culvert 5648. Project useful life (years) 80 Interest rate (%) 2% Present value coefficient 39.745 Table 24. Additional data used for a Study Level 1 analysis of Culvert 5648. Return Period, Tc (years) Current Flow, Qc (cfs) Estimated Damages ($) Tfnd 12 2,690 $0 T’fnd 25 3,670 $30,000 Tfint1 36 4,140 $38,000 Tfmod 100 5,420 $133,000 Tfint2 157 6,090 $390,000 Tfmax 500 7,800 $2,580,000 Table 25. Interpolated values calculated for Study Level 1 analysis of Culvert 5648.

Study Level 1 Climate Resilience Cost-Benefit Analysis 85 Multiplying $10,052 by the present value coefficient of 39.745 results in total damages of $399,517. So, the expected damages over the life of an in-kind replacement culvert under the medium scenario climate change conditions will be approximately $399,517. The difference in damages to the in-kind replacement of the culvert with and without climate change consider- ations is equal to $308,780. $399,517 $90,697 = $308,820− This means that a climate adaptation project that costs less than the cost of the in-kind replace- ment plus $308,820 would likely be cost-effective. In this case, the cost of the in-kind replacement project is $710,000, so a project costing $1,018,820 would be cost-effective. Reviewing the costs of the options considered by MnDOT, Option 1 is likely to be cost-effective, while Options 2 and 3 are not. These findings are consistent with the analyses completed by MnDOT. Application of Study Level 1 Analysis to Sea Level Rise The same approach used to complete a Study Level 1 analysis for riverine flooding conditions can also be applied to SLR with minor modifications. Instead of using discharges (Qs), flood elevations including wave height (in feet) are associated with recurrence intervals and levels of damage. Sea level rise calculators can be used to estimate future flood elevations. Even though the relationship between frequency and flood elevation changes with time, the relationship between frequency and damage remains somewhat constant (see Figure 21). To summarize the steps in the approach to completing a Study Level 1 analysis for SLR, establish baseline conditions then establish future (sea level rise) conditions. Establish Baseline Conditions 1. Identify the largest return period for which there will be no damages, likely the design return period. Identify the flood elevation associated with this recurrence interval. Set these equal to Tide Elcnd and Tcnd. 2. Identify a return period associated with an event that would cause moderate damages. This will be Tcmod. The corresponding flood elevation will be Tide Elcmod. 3. Identify a return period for which damages would be practically maximized. This maximum, realistically occurring return period is Tcmax. The corresponding flood elevation will be Tide Elcmax. 4. Estimate total damages associated with Tcmod and Tcmax. These will be Dcmod and Dcmax. 5. Use Equation 10 to calculate the expected annual damages between Tcnd and Tcmod. 6. Use Equation 11 to calculate the expected annual damages between Tcmod and Tcmax. 7. Use Equation 12 to calculate the total annualized damages, which is the sum of Dacmod and Dacmax. 8. Find the present value coefficient for the remaining project useful life (i.e., remaining service life during the period of projected climate change) from Appendix B. 9. Use Equation 13 to calculate the present value of total expected damages under current conditions. Establish Future (Sea Level Rise) Conditions 10. Use a sea level rise calculator (or other model) and find the NOAA gauge closest to the loca- tion of interest.

86 Incorporating the Costs and Benefits of Adaptation Measures in Preparation for Extreme Weather Events and Climate Change—Guidebook 11. For the selected gauge and project useful life duration, find the adjusted return period under SLR conditions for the flood elevations used to establish current conditions (i.e., Tide Elcnd, Tide Elcmod, and Tide Elcmax). To get a smoother curve, identify one more point, Tide El′fnd, between Tide Elfnd and Tide Elfmod and the associated flood recurrence interval that includes SLR. 12. Associate levels of damages with the SLR-adjusted return periods. As stated previously, the level of damages associated with a given elevation is likely to remain essentially the same, so Dfmod and Dfmax will remain the same for Tide Elfmod and Tide Elfmax; only the recurrence intervals have changed. Use the SLR calculator to determine the recurrence interval associ- ated with the additional point chosen in Step 11. Interpolate damages associated with this flood elevation and return using a modified version of Equation 16: D D Tide El Tide El Tide El Tide El D Dfnd cnd find cnd fmod fnd fmod fnd ( ) ( ) ( )′ = + ′ − − − 13. Calculate the annualized damages with SLR using a similar approach to Equation 10, substituting the SLR-adjusted values for the current condition values. 14. Use Equation 12 to calculate the total annualized damages for SLR conditions. 15. Find the present value coefficient for the remaining project useful life (i.e., remaining service life during the period of projected climate change) from Appendix B. 16. Use Equation 13 to calculate the present value of total expected damages under SLR conditions. Example Study Level 1 Analysis with Sea Level Rise This example is fictitious and used only for illustration purposes. The City of Galveston wishes to incorporate enhancements into its transit system to reduce damages and service disruptions from future storm events. It also wants to account for SLR in its adaptation planning process. The enhancements will be designed for the current 500-year flood and will have a project useful life of 50 years. Galveston’s transit system currently has an average daily ridership of 5,000 people. The initial project cost is $250,000, and annual O&M costs associated with the project will be $5,000. The improvements will not result in any system-user delays; however, without implementing the project, system users will experience additional one-way trips that are 10 miles longer and take an additional half hour. The delays are estimated to last for 7 days until the system becomes fully operational again. This will affect 200 bus trips per day. Using the FTA recurrence interval adjustment calculator for SLR, recurrence intervals have been found for a project useful life of 50 years and equivalent flood elevations including wave height and SLR based on the NOAA gauge for Galveston Pier 21 (Table 26). Table 27 summarizes the SLR-adjusted recurrence intervals obtained from the RI calculator for recorded flood elevations. Flood Elevation Including Wave Height (ft) Recurrence Interval without SLR (years) Estimated Equivalent Recurrence Interval with SLR (years) 10.12 50 18.35 13.00 100 29.82 17.76 500 145.69 Table 26. Estimated flood recurrence intervals including sea level rise at Galveston Pier 21.

Study Level 1 Climate Resilience Cost-Benefit Analysis 87 The 50-, 100-, and 500-year recurrence interval information for current conditions was used to reflect current pre-resilience conditions. These recurrence intervals were selected because the adaptation/resilience project is intended to protect against the current 500-year event. Next, the sea level rise–adjusted recurrence intervals were used for the same tide elevations to calculate the pre-resilience future (sea level rise) conditions. In addition, one tide elevation in between the current 50- and 100-year events was used for interpolation purposes. The data inputs and results are summarized in Table 28. The results of the analysis suggest that a project costing more than about $173,500 will not be cost-effective. Flood Elevation Including Wave Height (ft) Recurrence Interval without SLR (years) Estimated Equivalent Recurrence Interval with SLR (years) 6.50 2.60 1.79 38.854.7109.8 11.34 90.90 29.32 Table 27. Estimated equivalent recurrence intervals incorporating sea level rise for recorded floods near Galveston Pier 21. Tf (Year) Tide El (ft) Interpolated Damages, Df Base Case Future Annualized Damages, Daf Tide El (ft) Damages (in Current $) Annualized Damages, Dac Tc (Year) 18.35 10.12 $0 $0 Max return period resulting in no damages Tcnd 10.12 $0 $0 50.00 29.32 11.34 $42,361 $432 Next level return period resulting in some damages Tcmod 13.00 $100,000 $500 100.00 29.82 13.00 $100,000 Return period resulting in maximum damages Tcmax 17.76 $1,250,000 $5,400 500.00 145.69 17.76 $1,250,000 Total Annualized Current Damages $5,900 Total Annualized Future Damages $18,475 Project Useful Life PUL 50 Future Damages for Base Case $254,972 Discount Rate (%) i 7 Current Damages for Base Case $81,424 Present Value Coefficient esaCesaBrofsegamaDlanoitiddA108.31CVP $173,548 Present Value of Benefits Benefits $81,424 Max. Acceptable Project Cost $173,548 Current Pre-Resilience Conditions Future (Sea Level Rise) Pre-Resilience $41 $18,003 Table 28. Study Level 1 analysis results for sea level rise adaptation example near Galveston, Texas.