Third Report of the National Academy of Engineering/National Research Council Committee on New Orleans Regional Hurricane Protection Projects (2006)

Thank you for your comprehensive second letter report titled “Second Report of the National Academy of Engineering / National Research Council Committee on New Orleans Regional Hurricane Protection Projects” providing an evaluation of the information presented in the Interagency Performance Evaluation Task Force (IPET) second report dated 10 March 2006. Your review is critical in synthesizing the facts presented in the IPET reports. We are pleased to read in your overall assessment that IPET has made progress on several fronts and believe you will find many of your review comment have been integrated into the draft IPET Final Report released 1 June 06. The purpose of this letter is to provide information on how your review comments positively influenced the draft IPET Final report.

The use of a geographic information system (GIS) within IPET has progressively increased since IPET first report. The GIS continues to be beneficial in unifying regional spatial analysis and promoting system wide analysis, visualization, and communication. The GIS was used to organize and facilitate system wide evaluation of designed, pre and post Katrina levee crest elevations, and surge and wave conditions. GIS is also the core capability for the IPET system – wide risk assessment, and has been especially important in the integration of the many information sources needed for the reliability, consequence and inundation analyses. To more effectively implement the GIS, a GIS team composed of members from IPET, Task force Guardian, Task Force Hope, and the New Orleans district was established on 7 February 06. This GIS team had weekly meetings to promote system wide analysis, visualization and communication. We recognize that there are many more avenues for exploiting GIS and to use this capability in a more global sense for tying all IPET information and products together. Time has prevented exploitation of some of these options, however, the GIS is being transitioned over to the New Orleans District for continued development and use in the project life cycle process.

As you note, IPET has made excellent progress in modeling the large-scale hydrodynamics of Hurricane Katrina. Wave-surge model interaction was considered in the regional hydrodynamic modeling that is documented in the draft IPET Final Report. ADCIRC was run without wave forcing to compute the regional storm surge field as a function of time. All near shore wave models (the four STWAVE model domains) were then run using the time/space varying surge field to change the water depth field considered in the wave modeling. Time and space varying radiations stresses were computed from these wave model simulations. ADCIRC was re-run with the time-and space-varying wave radiation stresses. The contribution of wave radiation stresses to the regional storm surge field is presented and discussed in the draft IPET Final Report, Volume IV, “The Storm” main text and supporting appendices.

The vertical datum information was used to modify the topography and bathymetry information used in the final ADCIRC and STWAVE modeling. Levee and floodwall crest elevation data, relative to the NAVD88 2004.65 datum, was incorporated into the ADCIRC model as best we could in the time available (some of these data became available very late in the study process). The work done to incorporate datum adjustments is documented in an appendix to The Storm, Volume IV. Some additional work to incorporate levee/wall crest elevations has been done subsequent to the draft IPET final modeling, and these data will be factored into current ADCIRC model of the region, which will be applied in the S.E. Louisiana Coastal Protection Restoration (LaCPR) study.

A number of sensitivity tests were done to examine the sensitivity of model results to input data uncertainty, model coefficients and model formulation for certain processes, and tide/wave contributions to storm waves and water levels. Those results are documented in The Storm (Volume IV) main text and supporting appendices. A number of statistical comparisons (model results compared to measurements) were done to examine computed wind, wave, and water level accuracy, in response to an earlier request by the NRC reviewers. Those results are documented in the main text of Volume IV and appendices.

The draft IPET Final Report integrates two of the three detailed hydrodynamic study methods by coupling the physical model and the numerical models in the sense that the results from each were compared in the entrance regions of the canal. Engineering analysis were used to set the water levels within the canals from the water levels at the boundaries in Lake Pontchartrain, so these results were also coupled into all of the final detailed hydrodynamic wave runs.

Although more sensitivity tests can always be of value, a fairly wide range of tests with ADCIRC, and the other detailed hydrodynamic models was used for water level analysis. We did not have the chance to run many sensitivity tests on the wave predictions within the canals or along the major flood protection levees in IPET Report 2. For this reason, we tried to avoid interpreting the wave information in a completely deterministic manner within that report. We agree completely with the comment that it will be important to run such studies and that information will be provided in the IPET Final Report.

The geotechnical investigation has progressed significantly since IPET Report 2. The draft IPET Final Report contains additional information on regional geology including varying rates of subsidence across the region. This detailed regional geology provides significant information that was used to determine the soil profiles along the hurricane protection system. The geotechnical investigation is closely integrated into the risk and reliability analysis to provide a system wide assessment of vulnerabilities. A system wide assessment of levee vulnerability will be provided in the risk and reliability analysis. The decision to place closure gates at the outfall canals significantly reduces the vulnerability of these canals. Findings and lessons learned from the IPET performance evaluation were continuously shared with Task Force Guardian benefiting the reconstitution of the hurricane protection system. The key vulnerabilities of existing structures within the system are identified in the draft IPET Final Report with the lessons learned.

Alternate breaching mechanisms have been addressed in the draft IPET Final Report to include the potential for instability at other locations. The risk analysis is addressing the statistical and probabilistic procedures for charactering the vulnerabilities of the soil conditions throughout the system. The most reliable IPET field testing results came from the CPT tests for the clays. Results from the DSS and field vane test are being added to the IPET report. The National Science Foundation sponsored results are also being reviewed to assess applicability to the IPET report. To date, the information received from the University of California at Berkeley Team has been consistent with the IPET geotechnical data.

The risk studies are intended to provide a relative measure of the hurricane risks and vulnerabilities that exist in New Orleans (NO) due to the performance of the Hurricane Protection System (HPS). This will provide USACE decision makers with additional information to use in determining future investments in the HPS. As with any risk analysis, the values determined in the analysis are not used to make final engineering decisions on the configurations, locations or design of projects. Additional detailed engineering studies are required to finalize any project design.

The risk and reliability analysis is challenging. The IPET Risk and Reliability Team (R&R) is comprised of leaders in the field of risk analysis and all methods used have been thoroughly reviewed by the team members and the American Society of Civil engineers External Review Panel. Aleatory and Epistemic uncertainties are considered in the analyses and the results will reflect these uncertainties. The risk analysis will not be used as the basis for reoccupying the region. This will be a state and local government decision that is beyond the control of IPET and USACE.

The hurricane rates used in the risk analysis are based on the probability of hurricanes striking the NO area. The probability of a single hurricane affecting NO is approximately 23% for any given year therefore the sum of the hurricanes rates total 23%. The joint probability method used does not assign rates such as 1/100, 1/50 or 1/400 to any given storm in the 1800 storm sample. The elevation-frequency relationship is determined based on the entire 1800 sample of possible storms. Public and government officials should be educated concerning the meaning of the recurrence numbers. The IPET Risk team’s risk communication plan has been developed with this in mind.

In an effort to further discuss the joint probability method, an attachment is provided with some additional information on the IPET methodology. This will be expanded upon in the IPET final report.

The IPET R&R team will provide some additional information in the final report concerning the history of hurricanes in the region. The characteristics of the historical hurricanes are included in the parameter sets used in the suite of ADCIRC runs and several historic storms, for which accurate parameters are available, and are being used in the calibration process. The standard project hurricane has not been considered as a specific run since it is not an historic hurricane but was determined using an alternative method of analysis.

The IPET risk analysis considers more than 180 reaches and more than 330 individual gates, transitions and features. Fragility curves have been developed to determine the probabilities of failure as a function of water elevation for each reach and feature. The hurricane protection system has been modeled as a series system as described in your letter report section titled “Risks to the Hurricane Protection System”.

The difficulties described in your committee’s report pertaining to establishing the probabilities of failure of individual components of the protection system and the need for validation and verification of the results is understood. The Risk team is currently in the process of conducting validation studies that will be described in detail in the IPET final report. As previously stated, the risk model does not specifically consider the SPH as used for the original design of the HPS.

With regard to the deterministic approach proposed in your letter report, IPET sees this approach as being similar to the EST methodology used to develop the SPH. It is not IPET’s opinion that this approach “will provide a better understanding of surge and wave response potentials for a representative set of hurricanes. This approach is likely to be more valuable for assessing the near-term state of the hurricane protection system than a highly questionable probabilistic analysis.” The proposed approach would be primarily based on a short period of record of hurricane events and would not consider the full range of hurricanes considered to be possible in the NO area. In addition, Cat 5 hurricanes are defined primarily based on wind speed and not in the context of the six parameters used to define a hurricane so there is no clear definition of what constitutes a Cat 5 hurricane. The approach would also not consider the performance of the many features that make up the HPS so the potential for breaching or component failure would not be included.

For these reasons, the Risk team selected the joint probability approach to define possible hurricanes and a probabilistic reliability method to evaluate component performance. The dialogue currently underway between the NRC Committee and IPET concerning hurricane modeling and parametrics is a highly valuable forum to explore these issues and to ensure that the hazard definition for the Risk Analysis is appropriate for the purpose of the effort.

As encouraged in your report, it is the intent of IPET and USACE to use their considerable knowledge to help inform the public and elected officials about risks associated with hurricanes and flooding. The Risk team has prepared a Risk Communication plan to assist USACE leadership in informing state and local officials of the results of the risk analyses. It is hoped that this information will assist them in their emergency preparedness and future planning for the NO region.

We are committed to understanding the behavior of the New Orleans hurricane protection system in response to Hurricane Katrina and using that knowledge to reconstitute a more resilient hurricane protection system. Thank you

for your continued assistance in helping achieve this important and challenging objective. We look forward to your future overall assessment and recommendations on the draft Final IPET Report.