Hazard Mitigation

Appeal Brief

Appeal Letter

Appeal Analysis

Background

In May 2010, severe storms impacted the City of Nashville.  Overland floodwaters from the Cumberland and the Stones rivers submerged the grounds at the K.R. Harrington water treatment facility.  The K.R. Harrington facility was shut down on May 2, 2010, when floodwaters were projected to rise above the on-site electrical substation and emergency power generators.  It remained out of service until June 1, 2010, when it was brought back online in a limited capacity. FEMA initially prepared Project Worksheet (PW) 5504 to fund repair of damaged items at the facility.[1]  In addition to the repair work, Nashville-Davidson County (Applicant) submitted a Hazard Mitigation Proposal (HMP) in July 2011 in the amount of $493,372.44.[2]  The Applicant proposed to elevate critical components of the pumping station to two feet above the 500-year flood elevation.  FEMA prepared Version 1 of the Project Worksheet (PW) for $456,172.44[3] in March 2012 to add the HMP.  FEMA did not obligate funding for the HMP due to the proposal “not being cost effective to the project.”[4]

First Appeal

The Applicant appealed FEMA’s determination in a letter dated May 10, 2012, arguing that FEMA made an error in determining that the proposed mitigation project was not cost effective.  In support of its argument, the Applicant cited both the benefit cost ratio (BCR) of 15.86[5] that it had determined for this HMP and FEMA Recovery Policy RP9526.1, Hazard Mitigation Funding Under Section 406 (Stafford Act)[6] which lists benefit cost analysis (BCA) as an acceptable methodology for determining cost effectiveness.

The Tennessee Emergency Management Agency (Grantee) transmitted the Applicant’s appeal to FEMA Region IV in a letter dated August 13, 2012.  The Grantee concurred with the Applicant and asserted that the documentation provided on first appeal supported approval of the Applicant’s HMP.

The FEMA Region IV Regional Administrator (RA) denied the first appeal in a letter dated April 11, 2014.  The RA found that the Applicant’s BCA was not performed in accordance with RP9526.1.  Specifically, the RA cited three reasons for the denial.  First, that the Applicant incorrectly included an estimated future loss of function in its BCA as a cost avoidance benefit under RP9526.1.  Second, that FEMA was not able to recalculate the BCA because the Applicant did not provide adequate documentation to show how it calculated loss of function for the declared event.  Third, that the Applicant did not provide sufficient documentation to explain how it calculated the return interval, which is typically determined based on a Flood Insurance Study and the Applicant did not provide such a study.[7]

Second Appeal

The Applicant submitted a second appeal to the Grantee in a letter dated June 26, 2014, and submitted supplemental material in an e-mail and attached letter, both dated June 27, 2014.  The Grantee transmitted this appeal to FEMA in a letter dated July 2, 2014.  The Applicant asserts it has submitted the information FEMA required on first appeal to approve the HMP.  Specifically, the Applicant responds to FEMA’s three reasons for denying the first appeal as follows:

1. Future loss of function: The Applicant provides a newly developed BCA on second appeal that does not include historic loss of function as a benefit, as well as the supporting documentation for this revised BCA and resulting cost effective BCR. With this, the Applicant asserts that the project does not need loss of function to show cost effectiveness.The Applicant also separately argues that loss of function should be applicable for 406 mitigation projects, presenting the methodology that it used to obtain future loss of function.

2. Recurrence interval documentation: The Applicant asserts that this information was included in the original submittal, but provides again the Flood Insurance Study as well as a regression analysis and other relevant documentation supporting how it determined recurrence interval in the revised BCA.

3. Loss of function documentation: The Applicant asserts that sufficient documentation to determine loss of function was included in the original submittal, but submits this and additional supplemental information, including master flow data for the K.R. Harrington and Omohundro water treatment facilities from 2000 through 2011.[8]These facilities together serve Nashville-Davidson County.

Discussion

Under the Public Assistance (PA) Program, “work to restore eligible facilities on the basis of the design of such facilities as they existed immediately prior to the disaster”[9] is eligible for PA funding.  In approving PA funding, the “Regional Administrator may require cost effective hazard mitigation measures not required by applicable standards.”[10]  Hazard mitigation means “any cost effective measure which will reduce the potential for damage to a facility from a disaster event.”[11]  Pursuant to RP9526.1, mitigation measures must be determined to be cost effective.[12]  The policy lists three means for determining cost effectiveness, including a BCA.[13]  It states that the BCA will be based on a comparison of the total project cost to the total cost of the following projected benefits: (1) damage to the facility and its damaged contents, (2) emergency protective measures required as a result of that damage, (3) temporary facilities required due to the damage, (4) loss of function, (5) casualty (loss of life and injury), and (6) cost avoidance (damages avoided in the future due to mitigation measures).  It also includes a reference to FEMA’s Benefit Cost Analysis software, noting that the software provides an appropriate methodology for conducting a BCA.[14]  According to FEMA’s BCA Reference Guide, the “BCA is the method by which the future benefits of a mitigation project are determined and compared to its cost.  The end result is a BCR, which is derived from a project’s total net benefits divided by its total project cost…A project is considered to be cost effective when the BCR is 1.0 or greater.”[15] 

BCA Reanalysis

There was confusion between FEMA and the Applicant about how to apply loss of function in the BCA.  The Applicant cites a BCR of 15.86 in its first appeal, and a BCR of 30.48 in its second appeal, both of which are above 1.0 and consequently would be viewed as cost effective.  However, the Applicant uses a non-standard methodology to account for loss of function in its BCA, which FEMA was unable to verify on first appeal because of the nature of the methodology and lack of documentation provided by the Applicant.  Specifically, FEMA in its first appeal decision found that the Applicant inappropriately attributed loss of function as a cost avoidance benefit and the applicant, on second appeal, interpreted this to mean that loss of function could not be used at all.  The Applicant has provided new information to FEMA on second appeal to substantiate the values it used in the BCA and to describe its methodologies.  There are two underlying issues with the Applicant’s BCA methodology: (1) how it calculated loss of function and (2) how it assessed the after-mitigation recurrence interval and damages.[16]  The following subsections explain how FEMA resolved each of them in preparing a revised BCA.  In addition, Table 1 presents a summary of all BCA inputs and any differences between the values used by the Applicant and those used by FEMA in revising the BCA.

Loss of Function

The Applicant calculated a lump sum value for loss of function benefits instead of allowing FEMA’s BCA software to calculate loss of function benefits based on standard inputs.  The Applicant used this approach because the HMP submitted in Version 1 of PW 5504 is only one portion of a large-scale mitigation effort.  In total, the mitigation proposals aim to mitigate 27 components across two water treatment facilities, the K.R. Harrington and Omohundro water treatment facilities, that serve Nashville-Davidson County.  The HMP on appeal includes mitigation for five of these components.  The Applicant asserts that 19 percent (5/27 = 19%) of total loss of function benefits should be attributed to this mitigation project.[17]  

The Applicant submitted new documentation on second appeal to substantiate its method and assumptions for determining loss of function.  FEMA found that the Applicant’s method of attributing 19 percent of the loss of function benefits to this mitigation project is appropriate because the five mitigation components at the K.R. Harrington facility account for only a fraction of the mitigation project across the entire system.  However, instead of using this method to compute a lump sum value for loss of function benefits as the Applicant had, FEMA applied the 19 percent assumption on reanalysis to modify the standard value of water treatment service that is embedded in FEMA’s BCA software.  This allowed the software to calculate loss of function benefits.[18] 

After-Mitigation Recurrence Interval and Damages

The documentation provided on second appeal demonstrates that the Applicant used an after-mitigation recurrence interval of 1083 years.  The Applicant also assumed that after mitigation damages would be 25 percent of before-mitigation damages, but did not provide documentation on second appeal to substantiate this assumption.  Two changes to the Applicant’s BCA were made on reanalysis, and are described here. 

In keeping with standard BCA practice, FEMA used a conservative after-mitigation recurrence interval of 501 years, which is the Applicant’s before-mitigation design level of protection for the mitigation project plus 1 year, instead of the Applicant’s proposed 1083 years.  The proposed mitigation project brings the facility components to an elevation of 431 feet, which represents the 500 year event (429 feet) plus two feet of freeboard.[19]  The Applicant was required to build to at least the 500 year elevation per FEMA’s environmental review[20] and consistent with applicable codes and standards.[21]  The addition of freeboard is a best practice that will reduce the likelihood of flooding further.  However, while the Applicant has used an acceptable methodology to determine the 1083-year recurrence interval and it is possible that the Applicant’s proposed mitigation protects up to the 1083-year flood event, a more conservative recurrence interval of 501 years is used to ensure that benefits are not incorrectly overstated.  Using this conservative estimate, the project remains cost beneficial.

As the Applicant did not provide documentation to justify its assumption that after-mitigation damages would be 25 percent of before-mitigation damages, FEMA used 100 percent in its reanalysis, representing the worst case scenario (e.g. the mitigation reduced damages by zero percent).[22]  An assumption of after-mitigation damages being 100 percent of before-mitigation damages means that if a flood were to exceed the design level of protection of the proposed mitigation, the damages would be the same as those experienced before the mitigation project was completed.  In contrast with the Applicant’s unsupported assumption that these damages would only be 25 percent of those experienced before mitigation, the assumption of 100 percent is more conservative.

Summary of BCA Reanalysis Inputs

The values used in FEMA’s BCA reanalysis along with the values used by the Applicant on second appeal are described in Table 1 below.  The table is organized by the fields in the BCA software and explains where the reanalysis concurred with the Applicant’s assumptions and where changes were made.  The BCA reanalysis yielded a BCR of 5.62, demonstrating that the Applicant’s HMP is cost effective.

Table 1. Comparison of Applicant and Reanalysis Values in BCA  

In addition to the BCA reanalysis described above, FEMA performed a supplemental BCA sensitivity analysis to account for the fact that the Omohundro facility provided redundant service to a portion of customers who lost service as a result of the K.R. Harrington facility failure.  The two facilities together serve customers in Nashville-Davidson County, and the K.R. Harrington facility is the facility where the proposed HMP would be implemented.  Evaluating flow data from both facilities, it can be argued that the total lost function as a result of the K.R. Harrington facility being offline was only approximately 20 percent of the service provided by the two facilities combined.[27]  This is in comparison to the 60 percent loss in overall service that would have been expected if the Omohundro facility had not provided any redundant service.  Reanalyzing the BCA with the K.R. Harrington facility only providing service to 20 percent of the customers in Nashville-Davidson County resulted in a BCR of 1.87, which is still cost effective.

Additional Review

Pursuant to RP9526.1, in addition to determining that a HMP is cost effective, FEMA will also evaluate proposed HMPs for technical feasibility and compliance with statutory, regulatory and executive order requirements, as well as “ensure that the proposed HMPs do not cause a negative impact to the facility’s operation, surrounding areas, or susceptibility to damage from another hazard.”[28]  This appeal analysis did not consider the additional reviews required pursuant to relevant FEMA policy and referenced here.

Conclusion

The Applicant has sufficiently demonstrated that the HMP is cost effective.  However, pursuant to RP9526.1, FEMA must review the HMP for technical feasibility and compliance with statutory, regulatory and executive order requirements, as well as ensure that the proposed HMP does not cause a negative impact to the facility’s operation, surrounding areas, or susceptibility to damage from another hazard.  In accordance with 44 C.F.R. § 206.206, the Applicant maintains the right to appeal any new eligibility determinations that occur during completion of the additional reviews of this HMP.