ML19010A274

Final Safety Evaluation by the Office of Nuclear Reactor Regulation Topical Report- TVA Overall Basin Probable Maximum Precipitation and Local Intense Precipitation, Calculation, CDQ0000002016000041 Tennessee Valley Authority

ML19010A274
Person / Time
Site:	Browns Ferry, Watts Bar, Sequoyah
Issue date:	03/18/2019
From:	Andrew Hon Plant Licensing Branch II
To:	James Shea Tennessee Valley Authority
Hon A
References
Download: ML19010A274 (22)

Text

FINAL SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION TOPICAL REPORT TVA OVERALL BASIN PROBABLE MAXIMUM PRECIPITATION AND LOCAL INTENSE PRECIPITATION, CALCULATION, CDQ0000002016000041 TENNESSEE VALLEY AUTHORITY Enclosure 2

Contents

3.0 TECHNICAL EVALUATION

...................................................................................... 3.1 Storm Selection ......................................................................................................... 3.2 Transposition Limits ................................................................................................... 3.3 Observed Storm Precipitation Data ........................................................................... 3.4 Storm Representative Dew Point Selection and Moisture Maximization Using In-Place Maximization Factor .................................................................................. 3.5 Dew Point Climatology and Moisture Transposition Adjustment ............................. 3.6 Terrain Adjustment using Orographic Transposition Factor .................................... 3.7 Final Probable Maximum Precipitation Results ....................................................... 4.0 LIMITATIONS AND CONDITIONS ..........................................................................

1.0 INTRODUCTION AND BACKGROUND

By letter dated September 20, 2016 (Reference 1), as supplemented (Reference 2) on April 19, 2018 and (Reference 3) on June 22, 2018, Tennessee Valley Authority (TVA, the applicant) submitted a Topical Report (TR), TVA Overall Basin Probable Maximum Precipitation and Local Intense Precipitation Analysis, Calculation CDQ0000002016000041 for review. The purpose of the TR is to provide Probable Maximum Precipitation (PMP) analyses for the operating nuclear plants in the Tennessee Basin. Upon the U.S. Nuclear Regulatory Commission (NRC) staffs approval, any license holder, or applicant for a license, wishing to utilize this methodology can submit a license amendment in accordance with the appropriate regulatory requirements for site specific applications which demonstrates that methodology is applicable to their site. During the review NRC staff issued Requests for Additional Information (RAIs), as discussed throughout this report. TVA responded (Reference 2) and updated the entire TR to Revision 1 (Reference 3). The focus of this evaluation and the processes and results described in this safety evaluation (SE) are related to Revision 1 of the TVA TR.

PMP is defined by the World Meteorological Organization (WMO) as the greatest depth of precipitation for a given duration meteorologically possible for a design watershed or a given storm area at a particular time of year (Reference 4). Operationally, when sufficient historical extreme rainfall observations are available, PMP is estimated based on a commonly used method combining storm moisture maximization, transposition (i.e., relocating patterns of storm precipitation to other areas), and envelopment (i.e., identifying maximum storm precipitation values) (Reference 3). For critical infrastructure such as dams classified as high hazard and nuclear power plants, PMP has been used as input to simulate the probable maximum flood (PMF) as a conservative design criterion.

Historically, PMP across the U.S. has been estimated primarily through work products issued by the National Weather Service (NWS) and predecessor agencies. These estimates were based on data collected over a number of decades and for a variety of extreme rainfall events.

Fundamentally, PMP estimates can be classified as either theoretical or operational. Although the formal definition of PMP assumes a theoretical upper limit for precipitation, in practice a theoretical PMP cannot be directly computed or verified. Instead, most conventional PMP estimates follow an operational approach in which historical data and professional judgment may result in PMP estimates lower than the theoretical upper limit (Reference 6).

The basic approach and detailed methods used in developing operational PMP estimates have been described in numerous Hydrometeorological Reports (HMRs) published by the NWS.1 For example, HMR 51 (Reference 5) provides generalized all-season PMP estimates for the U.S.

east of the 105th meridian for drainage areas from 10 to 20,000 square miles (mi2) and for durations of 6-72 hours. The NWS HMRs identify two types of PMP estimates: generalized PMPs and individual drainage PMPs. The PMP estimates provided in most HMRs (e.g.,

HMR 51) are termed generalized estimates. In these HMRs, isolines of PMP are given on a map, allowing determination of basin-average PMP for any drainage basin. Typically, simplifying assumptions regarding the influence of topography and orographic processes were used in lieu of a detailed analysis. Other HMRs and studies produced by NWS (e.g., HMR 41

[Reference 7], HMR 46 [Reference 8], and HMR 56 [Reference 9]) provide PMP estimates for individual drainage basins that are specifically adjusted for the area and physical influences of the drainage basin under consideration. The reasons for analyzing individual drainage basins 1

www.nws.noaa.gov/oh/hdsc/studies/pmp.html

include (1) generalized PMP studies were not available, (2) the watershed was larger in size than those covered by available generalized PMP studies, or (3) detailed studies indicated orographic effects would yield PMP estimates significantly different from those based on available generalized PMP charts (e.g., watersheds in the Appalachians).

The TVA drainage basin covers an area of approximately 41,900 mi2 along the Tennessee River system and includes land in seven states. Figure 1 shows the TVA project domain used for the PMP development in the TR.

Figure 1. TVA project domain used for PMP development. (Source: TVA, 2018)

The purpose of this staff evaluation is to provide the details related to the review of the methodology and resulting precipitation values at the three operating nuclear sites located in the basin. There are TVAs three operating nuclear power plants (BFN, SQN, and WBN) in the TVA basin that provide an average of 7,800 megawatt of electricity. WBN is located upstream in the drainage basin, at Tennessee River mile 528.0 near Spring City, TN (Reference 10). SQN is located roughly 43.5 miles downstream of WBN, at Tennessee River Mile 484.5 near Soddy-Daisy, TN (Reference 11). BFN is located roughly 190.5 miles downstream of SQN, at Tennessee River Mile 294.0 near Athens, AL (Reference 12).

Figure 2 shows the location of each nuclear power plant and their respective upstream watersheds. Since all three plants are located along the same river, the contributory watershed area increases for the more-downstream locations. The WBN watershed covers 17,293 mi2, the SQN watershed covers 20,653 mi2, and the BFN watershed covers 27,213 mi2.

a) b)

c)

Figure 2. TVA nuclear watersheds.

The TVA drainage basin covers a topographically complex region of the southeastern U.S. The western portion, which contains relatively minor topographical relief, includes parts of the Gulf Plains and Central Plains; elevation in the western TVA drainage basin ranges from roughly 350 to 1,900 feet mean sea level (MSL). The central portion, which contains moderate topographic relief, includes parts of the Central Plains, Cumberland Plateau, and Great Valley; elevation in the central TVA drainage basin ranges from roughly 550 to 4,200 feet MSL. The eastern portion, which contains high topographic relief, includes parts of the Appalachian and Blue Ridge Mountains; elevation in the eastern TVA drainage basin ranges from roughly 850 to 6,700 feet MSL. WBN and SQN are located in the Great Valley, between the Cumberland Plateau and the Appalachian Mountains. BFN is located in the Gulf Plains.

2.0 REGULATORY EVALUATION

TVA submitted a TR that will support development of calculations to support future License Amendment Requests for revision of its operating nuclear plant design basis river and local flooding analysis, in accordance with Title 10 of the Code of Federal Regulations (10 CFR)

Section 50.90, Application for amendment of license, construction permit, or early site permit.

It is stated, in part, in Title 10 of the Code of Federal Regulations (10 CFR) Part 50, Appendix A, General Design Criterion (GDC) 2, that structures, systems, and components important to safety shall be designed to withstand the effects of natural phenomena such as floods without loss of capability to perform their safety functions. The design bases for these structures, systems, and components shall reflect appropriate consideration of the most severe of the natural phenomena that have been historically reported for the site and surrounding area, with sufficient margin for the limited accuracy, quantity, and period of time in which the historical data have been accumulated.

NUREG-0800, Standard Review Plan for the Review of Safety Analysis Reports for Nuclear Power Plants: LWR Edition, Section 2.4.3, Probable Maximum Flood (PMF) On Streams and Rivers, states that to meet the requirements of GDC 2 with regards to design bases for flooding in streams and rivers, the PMP on the drainage area that contributes to runoff on the stream network adjacent to the plant site should be determined. Similarly, NUREG-0800 Section 2.4.2, Floods, states that estimates of potential local flooding on the site and drainage design should be based on estimates of local intense precipitation (LIP) or local PMP.

Regulatory Guide 1.59, Design Basis Floods for Nuclear Power Plants, Revision 2, describes the design basis floods that nuclear power plants should be designed to withstand in accordance with GDC 2.

Since TVAs analysis in the TR deviated from the NRC guidance provided in NUREG-0800 of using the currently applicable National Oceanic and Atmospheric Administration (NOAA) NWS HMRs, a detailed NRC staff review was required. The NRC staffs review is described in this SE. There is currently no existing guidance to the staff or to licensees on how to produce or review a site specific probable maximum precipitation (SSPMP) study. Therefore, the staff reviewed the TR on the basis of meteorological conservatism and reasonableness. NRC staff subject matter experts held discussions with TVA during the course of the TR review regarding technical areas for which the NRC staff needed further clarification.

3.0 TECHNICAL EVALUATION

TVA computed PMP values across the TVA Basin (see Figure 1) using a uniform grid with a resolution of 0.025 x 0.025 degrees. A total of 17,938 grid cells were analyzed, with each grid covering an area of approximately 2.5 mi2.

From Section 1.0 of the TR (Reference 3):

This study provides Probable Maximum Precipitation (PMP) values for any drainage basin within the overall Tennessee Valley Authority (TVA) domain. The PMP values are valid for specific seasons based on storm type (see Section 4.5, Table 11), which is the time of the year when 100% of the PMP rainfall could occur. The PMP values are used in the computation of the Probable Maximum Flood (PMF). PMP values provided in this study are provided in place of PMP values in the four Hydrometeorological Reports (HMRs) for locations within the jurisdiction of TVA.

Since the latest HMR for the Tennessee River watershed was developed in the mid-1980s, much of the knowledge, data, and tools available now offers advantages over the NWSs HMR products. TVA identifies several issues with the HMRs, including the following:

• The limited number of analyzed storm events

• Lack of inclusion of storms that have occurred since the 1980s

• Inadequate processes used to address orographic effects

• Inconsistent data and procedures used among the HMRs

• Outdated procedures used to derive PMP (Section 1.1 of Reference 3)

As stated in Section 1.2 of the TR (Reference 3), the TVA PMP study aims to deliver reliable and reproducible PMP estimates for areas ranging from 1/3 square mile through the project domain and for durations ranging from 1 to 120 hours0.00139 days <br />0.0333 hours <br />1.984127e-4 weeks <br />4.566e-5 months <br />. In general, the TVA PMP study follows many procedures used in HMR development. Several updated procedures were also applied with consideration of meteorological and terrain interactions within the TVA drainage basin.

Consistent with the HMRs and other PMP approaches, TVA followed a storm-based approach, by which PMP is estimated based on historical storm observations. The major features of TVAs PMP estimation procedures are described in the remaining sections of this SE as follows:

• Storm selection

• Observed storm precipitation data

• Storm representative dew point selection and moisture maximization

• Dew point climatology and moisture adjustment

• Terrain adjustment using an orographic transposition factor (OTF)

Each section includes a description of TVAs submittal and the NRC staffs review findings.

NRC staffs review of the TR is site-specific and is limited to the application of aspects of the methodology and calculations of final SSPMP values that could potentially result in consequential flooding at TVAs three operating nuclear power plant sites. The NRC staffs review focuses only on assessing the adequacy of final SSPMP values given by the methodology with site-specific modifications as applied explicitly to the operating nuclear power plants with the TVA basin. The NRC staff did not assess the generic reasonableness of applying the methodologies for other sites and applications outside the watersheds directly affecting the TVA plants. Potential applications of the TR methodology for SSPMP estimates at other nuclear facility sites or watersheds, which are outside the scope of this review, would require further evaluation by NRC staff on a site-specific and case-by-case basis.

Appendix H of the TR (Reference 3) describes the use of a PMP Evaluation Tool to derive PMP values for watershed application. This tool is described as a Python scripting language-based tool designed to be run within the ArcGIS environment. The review of this PMP tool is outside the scope of this review of the PMP TR. Any licensee or applicant that seeks to make use of the PMP Evaluation Tool will be reviewed as part of the NRC staffs site-specific license amendment or application review.

The following SE sections describes the following topics:

• Storm Selection (Section 3.1)

• Transposition Limits (Section 3.2)

• Observed Storm Precipitation Data (Section 3.3)

• Storm Representative Dew Point Selection and Moisture Maximization (Section 3.4)

• Dew Point Climatology and Moisture Transposition Adjustment (Section 3.5)

• Terrain Adjustment using OTF (Section 3.6)

• Final PMP Results (Section 3.7)

• Limitations and Conditions (Section 4.0)

Each section includes a summary of TVAs submittal and the NRC staffs review findings.

3.1 Storm Selection Storm selection involves the process of identifying and selecting historical storm events that are appropriate for inclusion in PMP development. This process is the essential first step in a storm-based PMP approach. The use of a particular storm for computing PMP across a region should consider the storms transposition limits, or the geographic extent to which the storm can reasonably be applied for PMP estimation and whether it could ultimately influence the PMP values for the location being analyzed.

3.1.1 Topical Report Summary To assess alternative PMP occurrences that may result from the variable weather patterns present in the TVA Basin, TVA evaluated three different storm types: general, local, and tropical storms. TVA associated general storms with heavy rainfall occurring over large areas and for long durations; they typically are associated with stationary or slow-moving fronts and are strongest and most active in the fall, winter, and spring (TVAs season of occurrence is September 15 through May 15). TVA associated local storms with extremely heavy rainfall occurring over small areas and for short durations; they are typically associated with severe weather systems known as mesoscale convective complexes and individual thunderstorms energized with Gulf of Mexico moisture transported into the storms by low-level jets. This storm type is most active in the warm season (TVAs season of occurrence is April 15 through October 31). TVA identified that tropical storms occurring in the TVA Basin can produce heavy rainfall from remnant tropical systems and are most active in summer and fall (TVAs season of occurrence is June 1 through October 31) (Sections 3.3 and 4.5 of Reference 3).

To develop storm-based PMP estimates, TVA first assembled a set of extreme storms identified through a review of historical storm archives and data. TVA leveraged information from previous and ongoing PMP studies, NOAA HMRs, US Army Corps of Engineers (USACE) storm studies, precipitation data from the NOAA National Centers for Environmental Information,2 TVA rain and flood documents, and various other sources (Section 3.4.2 of Reference 3).

Information about the rainfall data and data analysis is provided in Section 3.3 of this SE.

TVA identified historical extreme storms first by identifying a storm search domain (i.e., a region in which observed storm events could similarly have occurred in the TVA Basin). Through its initial search domain, TVA identified storms that could be transpositionable (i.e., reasonably relocated) to the TVA Basin and for which further analysis was needed. TVA identified the region in this study to include areas from the Canadian border to the Gulf of Mexico and from approximately 100 degrees west longitude to the Appalachians. TVA did not consider direct tropical storm landfalls as transpositionable to the TVA Basin.

TVAs identification of historically extreme storms resulted in a long-list of storms, which it subjected to additional review and refined through a series of investigations and discussions.

TVAs common reasons for excluding a particular long-list storm from the analysis were that the storm was smaller than a larger nearby storm and that the storm was not transpositionable to the TVA Basin. In the end, TVA included 31 general storms, 19 local storms, and eight tropical storms on the short-storm list used for PMP development (Section 5.1 of Reference 3).

2 Formerly (before 2015) the National Climatic Data Center.

3.1.2 NRC Staff Evaluation The NRC staff finds the general storm selection approach to be reasonable. The staff reviewed the procedures used by TVA and found that they closely follow the HMR approach; however, the staff acknowledges that TVA developed a gridded PMP product that required high-resolution data sets and greater computational capabilities that were unavailable when the HMRs were developed. Using geographic information systems (GIS) and other resources, the transposition zones could be more easily defined based on high-resolution elevation and climatology datasets. Although these updated tools provide better accuracy and better definition of transposition zones, professional judgment is still required to assign transposition limits to each storm, as a more objective approach has not yet been developed.

Using data collected from the USACE Black Book (Reference 14), NRC staff analyzed the long-list storms to assess how close the USACE-reported observed precipitation characteristics were to the final TVA PMP values. The staff used a similar process to evaluate all USACE Black Book storms. The staff found that several storms (four long-list storms and three USACE Black Book storms) were identified that were near the PMP but were excluded from the final PMP short-storm list without clear justification. As a result, the staff issued RAI #2 (Reference 13) to request further explanation for excluding those storms. TVA provided the requested information related to each of the excluded storms and the reason for the exclusions.

The staff found TVAs justification for excluding all seven storms in question to be reasonable.

3.2 Transposition Limits As documented in NOAA HMR 51 (Reference 3), storm transposition is defined as relocating isohyetal patterns of storm precipitation within a region that is homogenous relative to terrain and meteorological features important to the particular storm rainfall under concern.

3.2.1 Topical Report Summary Given the variable weather patterns and topography across the TVA Basin, TVA subjected all storms to a transposition process. TVA defined transposition limits broadly across four transposition zones shown in Figure 3, with TVAs professional judgment being used to develop custom transposition limits for specific storms. TVA stated that these four zones are predominantly separated based on topography and climatology. TVA made decisions as to where to move specific storms based on the storm type, seasonality, isohyetal patterns, and moisture source.

Additional details on TVAs storm transpositioning process can be found in Section 4.3 of the TR (Reference 3).

Figure 3. TVA drainage basin transposition zones.3 3.2.2 NRC Staff Evaluation Based on the NRC staffs review of information provided in response to RAI #1 (Reference 13),

the staff found that the majority of storms included transposition limits that conform to the TVA Zone boundaries; however, four storms contained custom transposition limits that did not conform to the TVA Zone boundaries. Consequently, the staff issued RAI #10 to seek justification for imposing custom transposition limits. In response to RAI #10 (Reference 13),

TVA provided justification for imposing these custom transposition limits. Although staff found the justification for imposing latitudinal transposition limits for two storms to be reasonable, the staff needed additional information to justify the other two storms. With respect to TVAs decision to exclude two tropical storms from the protected region of the watershed based on the Tropical Storm Remnant 0.24 L-Cv contour developed by Schaefer et al. (Reference 4), TVA clarified that the L-Cv contour was used to inform transposition limits, but that the limits were ultimately informed by differences in the meteorological and topographical environments between the original storm center locations and the sheltered region of the TVA Basin. The references to the Tropical Storm Remnant 0.24 L-Cv contour were removed, as discussed in the response to RAI #10 (Reference 13). These clarifications were reflected in the final revised TR version. Therefore, the staff found TVAs justification for imposing custom transposition limits to be reasonable.

TVAs PMP calculation based on general, local, and tropical storm typing was considered reasonable by NRC staff due to the varying storms types commonly occurring in the TVA basin.

The staff also found TVAs selection and application of transposition limits reasonable.

Therefore, the staff found TVAs PMP calculation process and its selection of storms reasonable.

3 Zone naming based on Section 4.3 of the topical report (TVA, 2018)

3.3 Observed Storm Precipitation Data The observed storm precipitation data quantify the temporal and spatial distribution of rainfall for historical storms. Such data are typically quantified in the form of Depth-Area-Duration (DAD) data and presented as DAD curves or tables. A DAD value is expressed as a depth of precipitation occurring over a given area for a specified duration.

Following a storm-based approach, TVA further processed these observed data to estimate the maximum rainfall that could have occurred for that storm under an assumption of increased moisture availability and storm transposition.

3.3.1 Topical Report Summary TVA analyzed all short-list storms included for PMP development using a suite of computer scripts called the Storm Precipitation Analysis System (SPAS). SPAS provided spatial and temporal characteristics for each storm using gridded storm analysis techniques. SPAS was originally developed in 2002 and provides DAD values for storm analyses. It uses precipitation gauge records and radar rainfall estimates with a base map interpolation approach to generate spatial and temporal rainfall information at high resolution. Using data from the location of maximum storm precipitation, TVA developed mass curves to show temporal rainfall distribution and characterize storms on a temporal basis. A more detailed description of the SPAS data sources and procedures used by TVA is available in Appendix G of the TVA TR (Reference 3).

In some cases, a single storm event may produce rainfall across wide areas and produce significant rainfall events that are separated in time or space. Using professional judgment, TVA geographically separated such storms into multiple storm centers represented by separate DAD tables. In these cases, the SPAS number (i.e., the unique SPAS-analyzed storm identification number) remained the same in TVAs analysis, but TVA categorized the separate storm centers into different zones and analyzed them as individual events.

3.3.2 NRC Staff Evaluation ENERCON provides safety-related products and services to the nuclear power industries. The ENERCON quality assurance program is comprised of the Quality Assurance (QA) Manual and the associated implementing procedures. The quality assurance program meets the following requirements:

• Title 10 Part 50 Appendix B and Title 10 Part 21 of the Code of Federal Regulations: QA requirements of the U.S. Nuclear Regulatory Commission for safety applications

• ANSI [American National Standards Institute] N45.2-1977 and NQA-1-2008 /

NQA-1a-2009: QA program standards endorsed by the US Nuclear Regulatory Commission as acceptable methods for meeting the requirements of 10 CFR 50 Appendix B.

ENERCON applied their quality assurance program to SPAS for the purpose of dedicating the software for use in a nuclear licensing action.

NRC staff completed an inspection of the SPAS software from November 14-18, 2016, in order to assess ENERCONs compliance with provisions in 10 CFR Part 50, Appendix B, Quality Assurance Criteria for Nuclear Power Plants and Fuel Reprocessing Plants. The initial NRC inspection report (Reference 16) provided a notice of nonconformance, finding that ENERCON failed to verify the processing of radar data input into SPAS. A subsequent response to the

NRC inspection report (Reference 17) alleviated these concerns for using SPAS results in an NRC licensing action. This inspection specifically evaluated ENERCONs implementation of quality activities associated with the commercial grade dedication of the SPAS 9.5 and 10.0 software.

NRC staff did not review the raw or quality-controlled data used for the TR storms and instead reviewed the final DAD results. For historical storms previously evaluated in the Black Book (Reference 14), staff compared the SPAS DAD against the USACE DAD and found TVAs DAD data to be reasonable.

NRC staff reviewed the separation of single storm events into multiple storm centers and found TVAs treatment to be acceptable. In addition, the transposition limits applied to each storm center were found to be reasonable. Overall, staff found TVAs storm precipitation data to be reasonable.

3.4 Storm Representative Dew Point Selection and Moisture Maximization Using In-Place Maximization Factor Storm representative dew point data are often used as a surrogate to estimate the theoretical atmospheric moisture supply (i.e., precipitable water) available during historical storm events.

Although many factors contribute to a precipitation event, moisture availability is a critical component. It was used in the HMRs for maximizing an observed storm event by increasing the observed dew point temperature to a theorized maximum value based on dew point temperature climatology. TVA used a similar storm maximization process. This section describes the process TVA used to analyze storm representative dew point temperature and perform moisture maximization.

3.4.1 Topical Report Summary For each short-list storm, TVA estimated the storm representative dew point using either surface dew point or sea surface temperature (SST) data. To identify the storms moisture source, TVA used the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT)4 model for storms occurring in 1948 or later. TVA stated that HYSPLIT computes moisture travel paths using three-dimensional wind speed data from archived data sets; TVA ran HYSPLIT using reanalysis data5 (available since 1948). For storms occurring before 1948, TVA selected moisture sources based on meteorological information available (e.g., daily weather maps) or historical storm assessments (e.g., NWS documents). Based on the identified moisture source region, TVA selected a set of suitable stations for determining the storm representative dew point. Since heavy precipitation is assumed to correspond to periods of high moisture availability, TVA selected data corresponding to high dew point or SST. In general, TVA selected these data during periods before significant rainfall occurred and at locations upwind of the storm center and outside the rainfall region.

For a storm with a land-based moisture source, TVA computed a storm representative dew point based on weather station observations of dew point temperatures. While the HMRs computed a 12-hour persisting dew point temperature, TVA computed a maximum average dew point. For each storm evaluated, TVA computed the maximum 6-, 12-, or 24-hour average dew point, with local storms generally using 6- or 12-hour values and general/tropical storms using 4

http://ready.arl.noaa.gov/HYSPLIT.php, accessed 6/27/2018 5

https://www.esrl.noaa.gov/psd/data/gridded/data.ncep.reanalysis.html, accessed 6/27/2018

12- or 24-hour values. Typically, TVA averaged data from multiple stations to provide a more reliable spatial representation of dew point.

For a storm with a sea-based moisture source, TVA computed a storm representative dew point based on average daily SST observations because dew point observations are typically not available over water. These SST observations are typically taken aboard moving vessels.

For two historical storms, TVA stated that sufficient data were not available to compute storm representative dew point temperatures. In these cases, TVA converted historical 12-hour persisting dew point temperatures from the USACE Black Book (Reference 14) to maximum average dew point temperatures using a heuristic estimate developed during previous SSPMP analyses. TVA made heuristic adjustments to one general storm and one local storm whereby the 12-hour persisting dew point was converted to a maximum average dew point using a + 2 °F and + 7 °F adjustment, respectively.

For all storms, TVA rounded the storm representative dew point to the nearest 0.5°F. For additional information on TVAs storm representative dew point determination process, see TR Section 4.1.2 (Reference 3).

Following traditional PMP calculation procedures, TVA performed moisture maximization, which is defined by WMO (Reference 4) as the process of adjusting observed precipitation amounts upward based on the hypothesis of increased moisture inflow to the storm. TVA calculated the storm maximization using the In-Place Maximization Factor (IPMF). TVA limited IPMF values to a maximum of 1.5, which was largely consistent with the HMRs; a minimum IPMF of 1.00 is possible. Since the IPMF is a factor applied to the storm itself, the value remains constant regardless of where the storm is transpositioned.

To maximize the storm, TVA used dew point climatology for storms with land-based moisture sources, whereas SST climatology was used for storms with sea-based moisture sources.

The IPMF is calculated according to Eq. (1):

PWMax,Srep,SE IPMF = (1)

PWStorm,Srep,SE where PWMax,Srep,SE = the precipitable water calculated using the 100-year recurrence interval dew point (or +2 SST) at the storm representative dew point location from the storm elevation to the top of the atmosphere.

PWStorm,Srep,SE = the precipitable water calculated using the storm representative dew point (or SST) at the storm representative dew point location from the storm elevation to the top of the atmosphere.

TVA estimated precipitable water depths based on the relationship between dew point and precipitable water provided in the HMR 55A precipitable water tables (Reference 18). TVA rounded dew points (or SSTs) to the nearest 0.5°F. TVA used the storm elevation for both the numerator and the denominator in the IPMF calculation and it used the elevation at the storm center location rounded to the nearest 100 feet (or nearest 500 feet for elevations above 5,000 feet MSL).

For additional information on TVAs moisture maximization process, see TR Section 5.1.1 (Reference 3).

3.4.2 NRC Staff Evaluation NRC staff found TVAs use of maximum average dew points (instead of 12-hour persisting dew points) to be a reasonable departure from the HMRs. Also, the staff found TVAs use of HYSPLIT in determining moisture inflow trajectories and identifying moisture source locations to be reasonable.

NRC staff issued RAI #2 (Reference 13) to ask TVA to provide the observed hourly dew point data sheets for all short-list storms. This information helped staff review and understand the raw data used to develop storm representative dew points for all storms. RAI #13 (Reference 13) was issued asking TVA to clarify the duration analyzed for the Warner Park, TN, storm representative dew point; TVA verified that a 12-hour duration was used.

As a part of its assessment, the staff reviewed the rainfall mass curves, HYSPLIT trajectories, and storm representative dew point information that TVA provided in response to RAI #1 and RAI #2 (Reference 13). Staff independently evaluated these features to assess the reasonableness of TVAs application and found them to be acceptable.

NRC staffs review of this information revealed that, for some storms, TVAs original storm representative dew point selection used dew point data that were observed at locations far upwind of the storm center and during time frames in which significant rainfall had already occurred. The staff found that conducting the analysis in this way could inadequately represent the storm characteristics and (in these cases) result in PMP underestimation (a higher storm representative dew point will result in a lower IPMF value), since the relatively higher moisture observed could not have induced the observed rainfall. The staff performed sensitivity analysis, which revealed moderate increases in PMP when alternative, more physically justified storm representative dew points were used for these storms. Consequently, RAI #11 (Reference 13) was issued asking TVA to provide justification and correction of the selected dew point values.

In response to RAI #11, TVA corrected the storm representative dew point values for four controlling storms to adopt NRCs values and recomputed the general and tropical PMPs (see Section 4.1.2 of Reference 3). Staff found TVAs remaining storm representative dew point values to be reasonable. After confirming the reasonableness of TVAs dew point climatology values (see SE Section 3.5 for more information), staff found TVAs IPMF values to be reasonable.

3.5 Dew Point Climatology and Moisture Transposition Adjustment Using dew point climatology data for storm adjustment offers a method to increase a storms observed DAD from an observed measure of storm moisture availability (i.e., storm representative dew point) to a reasonable upper limit of moisture availability (e.g., 100-year recurrence interval dew point climatology). Since historical events often occur when moisture availability is below maximum levels, a conventional HMR assumption is that more rainfall could occur if more moisture is available. By using dew point climatology data, the increased rainfall production can be quantified through storm moisture maximization.

The HMRs used a moisture transposition adjustment process to estimate the difference in maximum moisture available to a storm when it is moved from the in-place moisture source location to a transpositioned moisture source location. A similar storm maximization process was used by TVA.

3.5.1 Topical Report Summary Dew Point Climatology To estimate a maximized level of moisture that could have been available to a storm and to compare moisture availability across wide areas in transpositioning storms, TVA developed climatologies of dew point temperature and SST. TVAs process of computing dew point climatology involved several steps, which included calculating annual maximum dew points for each calendar month, performing frequency analysis, and conducting spatial interpolation and manual smoothing.

Historically, the HMRs used monthly maximum 12-hour persisting dew point temperatures to maximize storms. In this study, TVA produced monthly dew point climatologies for the 100-year recurrence interval of the annual maximum series data for individual stations. TVA developed these climatologies for storms with land-based moisture sources for 6-, 12-, and 24-hour durations using the NOAA TD3505 Integrated Surface Data set. For storms with sea-based moisture sources, TVA used monthly SST climatologies for two standard deviations above the mean SST. For both the land-based dew point climatologies and SST climatologies, TVA produced maps using manual smoothing of point-based climatologies.

TVAs climatological maximum dew point and SST maps are provided in Appendix C of the TR.

For additional information on TVAs dew point climatology development and application processes, see TR Sections 4.1 and 4.2 (Reference 3).

Moisture Transposition Adjustment When moving a storm from its originally observed location to a new location using storm transposition, TVA accounted for changes in moisture availability by using the moisture transposition factor (MTF). The MTF represents the ratio between the maximized moisture available at the storm transposition dew point location and the moisture available at the storm representative dew point location. TVA based its precipitable water estimates on the 100-year recurrence interval dew point (or +2 SST) rounded to the nearest 0.5°F. MTF values above and below 1.00 are possible, and TVA placed no limits on the calculation.

TVA calculated the MTF according to Eq. (2):

PWMax,ST,SE MTF = (2)

PWMax,SRep,SE where PWMax,ST,SE = the precipitable water calculated using the 100-year recurrence interval dew point (or +2 SST) at the storm transposition dew point location from the storm elevation to the top of the atmosphere.

PWMax,SRep,SE = the precipitable water calculated using the 100-year recurrence interval dew point (or +2 SST) at the storm representative dew point location from the storm elevation to the top of the atmosphere.

TVA rounded the dew points (or SSTs) to the nearest 0.5°F. TVA used the storm elevation for both the numerator and the denominator in the MTF calculation and rounded the elevation at the storm center location to the nearest 100 feet (or nearest 500 feet for elevations above 5,000 feet MSL).

For a few storms, TVA modified the MTF application. Since precipitation frequency data were not available for three general storms from Texas, TVA did not calculate an OTF (see SE Section 3.6). Instead, TVA modified the MTF values to include an elevation-based vertical adjustment to account for the elevation-based differences in moisture availability between the storm center location and the target grid locations. In addition, the MTF for the Holt, MO, local storm was set to 1.00 because TVA concluded moisture availability would not increase in moving from west to east.

For additional information on TVAs moisture transposition adjustment process, see TR Section 5.1 (Reference 3).

3.5.2 NRC Staff Evaluation NRC staff found TVAs use of 100-year recurrence interval dew point climatologies based on station annual maximum series data to be a reasonable departure from the HMRs.

The NRC staff issued RAI #4 (Reference 13) asking TVA to provide a copy of the digital dew point climatology GIS data layers used for PMP development. Staff reviewed the dew point climatology data provided by TVA in response to RAI #1 and RAI #4 and independently evaluated these features to assess the reasonableness of TVAs application. The staffs primary climatology-related concern with TVAs original TR was the use of a data set (NOAA TDL [Techniques Development Laboratory]) that was not quality controlled for land-based dew point temperatures although a higher quality-controlled dataset was available (NOAA TD3505).

Based on the NRC staffs independent analysis results that suggested higher PMP values would result from using NOAA TD3505 data, the staff issued RAI #12 (Reference 13) requesting that updated dew point climatologies be developed using the NOAA TD3505 data set and applied to the general, local, and tropical PMPs. In response to RAI #12, TVA updated the TR and its dew point climatology using the NOAA TD3505 data set with data available through 2017 (see Section 4.1.1 of Reference 3). The staff compared TVAs initial submittal using NOAA TDL data to TVAs final dew point climatologies and found that TVAs updated dew point climatology was more reliable and yielded higher PMP values.

3.6 Terrain Adjustment using Orographic Transposition Factor The use of terrain adjustment also is necessary when a historical storm is transpositioned from its original location to a point of interest for PMP development. Several different approaches have previously been used in hydrometeorological studies to quantify a terrain transposition adjustment, with a common goal of accounting for how differences in topography or orographic effects may influence PMP.

3.6.1 Topical Report Summary TVA used an OTF to account for terrain adjustments. Use of this method represents a major departure from the HMRs process-based methodology and philosophy.

The calculations described in SE Sections 3.4 and 3.5 represent processes similar to those used in the HMRs in which an observed precipitation event is (1) moisture maximized using the IPMF and (2) geographically transpositioned (on a latitude-longitude plane) using the MTF.

Although it captures spatial variation in moisture, the MTF may not adequately capture the effects of terrainhence the need for a terrain adjustment. Some NOAA HMRs account for terrain effects on PMP adjustment using the Storm Separation Method, whereas previous PMP studies use the barrier adjustment factor (BAF). For this study, TVA used the OTF.

TVA computed the OTF by assessing the relationship between precipitation frequency depths at the target and source locations. TVA collected precipitation frequency data from Volumes 2, 8, and 9 of NOAA Atlas 14 (Reference 19; Reference 20, Reference 21, respectively). For General and Tropical storms (except the three Texas storms), TVA calculated the OTF using a linear regression developed based on 24-hour precipitation frequency depths for various recurrence intervals (10 to 1,000 years). For the three Texas storms, since NOAA Atlas 14 precipitation frequency depths were not available, TVA uniformly set the OTF to 1.00.

TVA fitted the NOAA Atlas 14 values to a linear regression line, shown in Eq. (3), to estimate m and b:

PAtlas14,Site = m*PAtlas14,SC +b (3) where PAtlas14,Site = target grid point rainfall frequency depth (in.) across various selected durations and return periods from Atlas 14.

PAtlas14,SC = storm center grid point rainfall frequency depth (in.) across various selected durations and return periods from Atlas 14.

m = slope.

b = intercept (in.).

TVA then used the linear relationship determined through the target-source regression fit in Eq. (3) to determine the orographically adjusted rainfall for all grid points based on the SPAS-analyzed in-place rainfall using Eq. (4):

PTVA,Site = m*PTVA,SC +b (4) where PTVA,Site = orographically adjusted rainfall (in.) at the targeted grid point.

PTVA,SC = SPAS-analyzed in-place rainfall (in.).

m = slope from Eq. (4).

b = intercept (in.) from Eq. (4).

Rearranging Eq. (4) yields the OTF, as shown in Eq. (5):

PTVA,Site b OTF = =m+ (5)

PTVA,SC PTVA, SC Since NOAA Atlas 14 values (and hence OTF values) show discrepancies across some volume boundaries, TVA adjusted the raw general and tropical storm OTF calculations throughout the southwest portion of the TVA Basin to uniformly set all OTF values to a single value based on what was considered a representative location.

For the calculation of LIP at the three nuclear power plant sites, TVA followed a different OTF calculation approach, which has been used in other recent PMP studies. Rather than performing linear regression, TVA calculated the OTF using the ratio between the target and source 6-hour, 100-year recurrence interval precipitation frequency depth. In addition, since PMP values in parts of the TVA Basin would otherwise exceed world record rainfall values, TVA

set the OTF to 1.00 for the Simpson, KY, storm and added a vertical elevation component to the MTF calculation.

From Section 5.1.1.5 of the TR (Reference 3):

For the development of the LIP values at SQN and WBN, the OTF adjustment was not used for the Simpson, KY July 1939 storm (the OTF was set to 1.00). Instead, an adjustment using the HMR standard vertical evaluation (approximately 0.8% per 100 feet difference) was applied as part of the MTF process to account for differences in elevation between the source storm location and the nuclear site. In addition, the 6-hour 100-year precipitation frequency climatology was used in the OTF calculations for all storms used in the LIP analysis. This replaced the use of the linear fit method used in the OTF calculations for all other grids within the TVA domain. This was done based on discussion with the NRC to better align with current OTF calculation processes utilized since the completion of this study. The result was a transposition factor of 1.06 between the Simpson, KY storm center and the SQN and WBN sites. This approach was a more conservative application to the Simpson, KY July 1939 storm transposition than the normalized OTF approach applied to the PMP calculations. This adjustment had no effect on the BFN location because the Simpson, KY July 1939 storm is not transpositionable to that location.

In addition, the Smethport, PA July 1942 storm was not transpositionable to the SQN, WBN, or BFN locations and therefore no other adjustments were applied to that storm for LIP considerations.

The OTF calculation using the 100-year recurrence interval ratio is shown in Eq. (6):

PAtlas14,100-y,Site OTF = (6)

PAtlas14,100-y,SC For additional information on TVAs terrain adjustment process, see TR (TVA, 2018),

Sections 3.5, 5.1, and 5.5.4.

3.6.2 NRC Staff Evaluation Despite the NRC staffs concerns regarding the lack of a physical basis for the OTF and its high sensitivity to the internal variability of Atlas 14, the staff found TVAs final SSPMP results to be adequate for application to the three operating TVA nuclear power plant sites. The staff made this determination following its detailed review of TVAs OTF calculations, modifications included in the revised TR, and review of sensitivity analyses.

The NRC staff issued RAI #6 (Reference 13), which asked TVA to provide justification for applying sizable reductions in OTF in transpositioning some example storms across orographically similar zones. The examples cited in the RAI were the Warner, OK, and Fall River, KS, storms for which OTF values in TVA Zone 1 were approximately 0.80 and 0.75, respectively; these represented large decreases, although the storm centers shared similar orographic and climatologic characteristics with Zone 1. In addition, OTF values for the Smethport, PA, storm were much lower than expected, although the OTF values had been manually adjusted using normalization. In response to RAI #6, TVA provided meteorological reasoning for why OTF values varied for the examples cited. To better understand the sensitivity of OTF, the staff conducted independent evaluation using the conventional BAF-based approach. Overall, the staff considers the OTF to be a highly uncertain and

sensitive adjustment for PMP calculation purposes, particularly in relatively non-orographic regions (e.g., western parts of the TVA drainage basin). However, given that the OTF-based SSPMP values were similar to the estimated BAF-based SSPMP values for the three nuclear power plant watersheds, the staff found TVAs use of OTF for the three nuclear power plants to be adequate for the purposes of the TVA TR.

The NRC staff issued RAI #7 (Reference 13), which asked TVA to provide justification for including significant OTF reductions for two local storms. In the original TR (Reference 1), TVA included normalization adjustments to the Smethport, PA, and Simpson, KY, OTF values whereby all OTF values were decreased proportionally, so that the maximum OTF for any grid in the TVA Basin was 1.00. This effectively decreased all OTF values for both storms by over 50%. In response to RAI #7, TVA revised the TR to address this concern by updating the LIP calculation for the three nuclear power plant sites to include Simpson, KY, OTF values of 1.00 and elevation-adjusted MTF values. Since the Smethport, PA, storm was not transpositioned to any of the nuclear power plant sites, TVA did not change its OTF calculation. Staff found TVAs final LIP calculation at the three nuclear power plant sites to be reasonable and these changes were included in an update to the TR.

The NRC staff issued RAI #8 (Reference 13), which asked TVA to justify the use of NOAA Atlas 14 for calculating the OTF and for using the best fit linear trend method in lieu of the 100-year recurrence interval ratio method for calculating the OTF. In response to RAI #8, TVA provided additional details related to NOAA Atlas 14 and its use in computing OTF values.

Although staff remains concerned with some technical details of using NOAA Atlas 14 values for PMP application, the NOAA Atlas 14 data represent the best-available public data and are reasonable for use. TVA also implemented changes in how the OTF was calculated for LIP at the three nuclear power plants, choosing to use the 100-year recurrence interval ratio method instead of the best fit linear trend method. Consequently, staff found TVAs calculation of OTF using NOAA Atlas 14 to be reasonable for the purposes of the TVA TR.

The NRC staff issued RAI #9 (Reference 13), which asked TVA to explain some OTF calculation inconsistencies identified during the review of TVA files. In response to RAI #9, TVA confirmed that the inconsistencies identified were not problematic and were the results of manual adjustments to the OTF calculation in portions of the TVA Basin to correct two features.

One manual adjustment ensured that OTF values remained consistent along the border between NOAA Atlas 14 Volume 2 and Volume 9. The other adjustment ensured that OTF values in the complex terrain near the Georgia-North Carolina border resembled the underlying topography and that they were reasonable given disparities between the two volumes. Staff found TVAs explanations reasonable.

Altogether, the NRC staff found TVAs final OTF-based SSPMP values to be adequate for the general and tropical PMP and for the LIP of the three nuclear power plants.

3.7 Final Probable Maximum Precipitation Results Following the PMP development and calculation approaches described above, TVA produced final, gridded PMP depths for general, local, and tropical PMP. The results are provided in Appendix A of the TR (Reference 3) and include a 72-hour, 17,306-mi2 WBN basin-averaged General and Tropical PMP of 12.57 inches and 10.50 inches, respectively. Compared with the HMR 41 June 72-hour, 7,980-mi2 PMP of 17.05 inches, these values are 26.3% and 38.4%

lower, respectively. This (and additional) summary information is provided in Section 5.8 of the TR.

A separate LIP calculation specific to the three nuclear power plants was also conducted by TVA. Compared with the local PMP, the site-specific LIP calculations included a unique treatment of the Simpson, KY, storm as described in Section 3.6 of this SE and Section 5.1.1.5 of the TR (TVA, 2018). While the local PMP provides localized, small-scale PMP estimates across the TVA Basin, nearly all data provided is outside the scope of the NRCs review, as stated in Section 3.0; instead, the LIP depths provide site-specific calculations relevant for on-site flooding at each of TVAs three operating nuclear power plants which may result from LIP over the powerblock. The final 1-hour, 1-mi2 LIP depths were 11.60, 13.81, and 13.81 inches at BFN, WBN, and SQN, respectively. Additional information is provided in Section 5.4.4 of the TR.

Section 5.6.1 of the TR (Reference 3) describes a PMP evaluation tool which is used to summarize SSPMP values for specified watersheds in the TVA drainage basin. The functionality of this PMP evaluation tool is outside of the scope of the NRCs review for TVAs TR. It is anticipated that TVA will use the tool for future hydrological modeling work submitted to NRC and the tool would be reviewed by the NRC staff at that time.

4.0 LIMITATIONS AND CONDITIONS The scope and NRC staff review of the TR is limited to the evaluation of PMP depths calculated in the TR (Reference 3) that are relevant to potential flooding at the operating nuclear plants in the TVA basin. Any nuclear power plant wishing to utilize the methodology discussed in this SE must submit a license amendment under 10 CFR 50.90 for site-specific review of any amendment or application. As part of the amendment or application to utilize this methodology, any licensee or applicant needs to fully explain and validate the approach and methods used in developing the PMP values. Any PMP values not relevant to the operating TVA reactors are not subject to NRC approval under this review. As stated in Section 6.0 of the TR (Reference 3):

[The TR] provides PMP values which can be used to support a license amendment request to perform an assessment of river flooding effects and local intense precipitation effects at the operating nuclear plant sites in the Tennessee Basin. NRC review of this report is applicable only to these effects and their impact on these three sites.

TVA did not include examples or illustrations to address how the gridded-precipitation values derived in the TR will be applied during future surface hydrology evaluations.

Therefore, the staff did not review the aforementioned PMP evaluation tool and its use in the hydrologic analysis of the TVA nuclear sites. An NRC staff review of TVAs PMP evaluation tool may be performed during the subsequent LAR hydrologic reviews.

5.0 CONCLUSION

Following detailed review and discussion with TVA and a subsequent update to the PMP calculation, NRC staff found TR TVA-NPG-AWA16 (Reference 3) to provide reasonable estimates of PMP for operating nuclear power plants in the TVA basin.

The storm-specific DAD data, transposition limits, IPMF values, MTF values, and OTF values used as input to the PMP evaluation tool were reviewed and deemed adequate for use in TVAs subsequent hydrological modeling work related to PMP. The PMP evaluation tool was not assessed within the scope of the NRC staffs review and would require subsequent review and approval. The NRC staff concludes that the TR PMP estimates, when correctly applied, will contribute to the assurance that the three TVA operating plants will maintain compliance with

requirements in 10 CFR 50 Appendix A, GDC 2, governing that structures, systems, and components important to safety shall be designed to withstand the effects of natural phenomena such as floods without loss of capability to perform their safety functions. Therefore, the NRC staff finds the TR PMP estimates acceptable, with the above noted limitations and conditions, for compliance with the NRC regulations.

REFERENCES

1. Shea, J. W., Tennessee Valley Authority, letter NRC, TVA Overall Basin Probable Maximum Precipitation and Local Intense Precipitation Analysis. Topical Report No.

TVA-NPG-AWA16. Calculation CDQ0000002016000041, [Rev. No. 000], September 20, 2016. ADAMS Accession No. ML16264A454.

2. Shea, J. W., Tennessee Valley Authority, letter NRC, Response to NRC Request for Additional Information Related to Topical Report TVA-NPG-AWA16, TVA Overall Basin Probable Maximum Precipitation and Local Intense Precipitation Analysis, Calculation CDQ0000002016000041, April 19, 2018. ADAMS Accession No. ML18117A225.

3. Shea, J. W., Tennessee Valley Authority, letter NRC, TVA Overall Basin Probable Maximum Precipitation and Local Intense Precipitation Analysis. Topical Report No.

TVA-NPG-AWA16. Calculation CDQ0000002016000041, Rev. No. 001, June 22, 2018.

ADAMS Accession No. ML18192A510.

4. WMO (World Meteorological Organization). 2009. Manual for Estimation of Probable Maximum Precipitation.

5. Schreiner, L. C., and J. T. Riedel, 1978. Probable Maximum Precipitation Estimates, United States East of the 105th Meridian (HMR No. 51). Washington, DC: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

6. Micovic, Z., M. G. Schaefer, and G. H. Taylor. 2015. Uncertainty Analysis for Probable Maximum Precipitation Estimates. Journal of Hydrology. 521:360-373. 56

7. Schwarz, F. K. 1965. Probable Maximum and TVA Precipitation over the Tennessee River Basin above Chattanooga (HMR No. 41). Washington, DC: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

8. Schwarz, F.K. and J.T. Riedel, 1970. Probable Maximum Precipitation, Mekong River Basin (HMR No. 46). Washington, DC: US Department of Commerce, National Environmental Science Services Administration.

9. Zurndorfer, E. A., Schwarz, F. K. Hansen, E. M. Fenn, D. D. and Miller, J. F. 1986.

Probable Maximum and TVA Precipitation Estimates with Areal Distribution for Tennessee River Drainages Less Than 3,000 Mi2 in Area (HMR No. 56). Silver Spring, MD: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

10. TVA (Tennessee Valley Authority). 2012. Watts Bar Nuclear Plant (WBN), Unit 2Final Safety Analysis Report (FSAR), Amendment 109, Section 2, Site Characteristics, ADAMS Accession No. ML12244A077.

11. TVA (Tennessee Valley Authority). 2016. Sequoyah Nuclear Plant, Units 1 and 2 Updated Final Safety Analysis Report, Amendment 26 (Redacted Version), Section 2, Site Characteristics, ADAMS Accession No. ML16172A186.

12. TVA (Tennessee Valley Authority). 2017. Browns Ferry Nuclear Plant Updated Final Safety Analysis Report, Amendment 27, Section 1.6, Plant Description, ADAMS Accession No. ML18018A787.

13. Electronic Mail from NRC to TVA, Request for Additional Information Related to TVA Fleet Topical Report TVA-NPGAWA16 (EPIC: L-2016-TOP-0011), February 23, 2018 ADAMS Accession No. ML18057A637.

14. USACE (US Army Corps of Engineers). 1973. Storm Rainfall in the United States 1945-1973. Washington, DC.

15. Schaefer, M., B. Barker, S. Carney, J. Day, W. Gibson, D. Martin, T. Parzybok, and G. Taylor. 2015. Regional Precipitation-Frequency Analyses for Mid-Latitude Cyclones, Mesoscale Storms with Embedded Convection, Local Storms and Tropical Storm Remnant Storm Types in the Tennessee Valley Watershed. Tennessee Valley Authority.

16. Nuclear Regulatory Commission Vendor Inspection of Enercon Services, Inc. Report No.: 99901474/2016-201, December 22. 2016. ADAMS Accession No. ML16351A422.

17. N. Eggemeyer, Enercon Services, Reply to a Notice of Nonconformance Enercon Services, Inc. Docket No. 99901474 Vendor Inspection of Enercon Services, Inc. Report No. 99901474/2016-201, January 18, 2017. ADAMS Accession No. ML17019A391.

18. Hansen, E. M., D. D. Fenn, L. C. Schreiner, R. W. Stodt, and J. F. Miller. 1988. Probable Maximum Precipitation EstimatesUnited States between the Continental Divide and the 103rd Meridian (HMR No. 55A). Silver Spring, MD: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

19. Bonnin, G. M., D. Martin, B. Lin, T. Parzybok, M. Yekta, and D. Riley. 2004.

Precipitation-Frequency Atlas of the United States. NOAA Atlas 14, Vol. 2. Silver Spring, MD: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

20. Perica, S., D. Martin, S. Pavlovic, I. Roy, M. St. Laurent, C. Trypaluk, D. Unruh, M. Yekta, and G. Bonnin. 2013. Precipitation-Frequency Atlas of the United States.

NOAA Atlas 14, Vol. 8. Silver Spring, MD: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

21. Perica, S., D. Martin, S. Pavlovic, I. Roy, M. St. Laurent, C. Trypaluk, D. Unruh, M. Yekta, and G. Bonnin. 2013. Precipitation-Frequency Atlas of the United States.

NOAA Atlas 14, Vol. 9. Silver Spring, MD: US Department of Commerce, National Oceanic and Atmospheric Administration, National Weather Service.

Principal Contributor: K. Quinlan Date: March 18, 2019