Updating Probable Maximum Precipitation for the Tennessee Valley Authority

Transcription

1 Updating Probable Maximum Precipitation for the Tennessee Valley Authority Bill Kappel, President/Senior Meteorologist Applied Weather Associates, Monument, CO Mike Eiffe, Project Manager Tennessee Valley Authority, Knoxville, TN HydroVision 2014 Annual Conference July 21-25, 2014 Nashville, TN

2 Reasons For Study Correct HMR Errors Update Storm Database State-of-practice data and understanding Potential significant cost savings and maximized utilization of assets

3 Probable Maximum Precipitation Definition: The theoretically greatest depth of precipitation for a given duration that is physically possible over a given storm area at a particular geographic location at a certain time of year (HMR 59, 1999) Types of PMP studies: Generalized (Hydrometeorological Reports) Provides PMP values for a region HMR 51 - East of the 105 th Meridian from Canada to Mexico Regional/Statewide Provide PMP values over regions with varying topography Individual basins are included in the regional/statewide results Site-Specific Provides PMP values for individual drainage basins Considers unique meteorology and topography

4 Background HMR 56-Published in 1988-newest yet still old Based on outdated methods and techniques Better understanding of meteorology Updated storm datasets Improved spatial analysis Methods and techniques updated to quantify topographic effects Major issues with TVA HMRs Highly subjective and not reproducible Several inconsistencies between each report Not clear how storm data used to develop the PMP values Covers a widely varying region Climatologically Topographically

5 HMR Current Coverages

6 Elevations Across TVA, 1,000 Intervals

7 Elevations Across TVA, 1,000 Intervals

8 How Did We Compute PMP? Storm Based Approach-Deterministic Similar to HMR/WMO procedures Reproducible Assumption/subjectivities explicitly stated Consistency with AWA PMP studies Improvements in understanding of meteorology Expanded storm data base Use of computer technologies (HYSPLIT/GIS) Use of NEXRAD weather radar Better understanding of meteorology

9 Not Our First PMP Study

10 Updating PMP-What Did We Do Storm Search Update the storm database Identify the most extreme rainfall events Throughout the basin Surrounding regions Identify Storm Types Local Convective Remnant Tropical General Frontal

11 Updating PMP-Storm Search 1000 s of storms initially captured Grouped by storm type Local Convective, Tropical, Frontal Location Duration Storms used in HMRs included Ensure no potential PMP type storms missed Storms must be transpositionable Similar meteorological & topographical characteristics

12 Short List Storm Locations Used for PMP Development

13 Updating PMP-Storm Analysis Storm Precipitation Analysis System (SPAS) Depth-Area-Duration Table Mass Curves Total Storm and Hourly Isohyetal Patterns Hourly (5-minute rainfall) at 1/3 rd square mile Dynamically adjusted radar and/or basemap for spatial interpolation

14 SPAS Storm Analysis Results

15 How Did We Compute PMP? Each storm maximized Make it as big as physically possible Storm rainfall = dynamics + moisture Can t quantify dynamics, can quantify moisture Assume most efficient storm dynamics Only moisture varies Determine moisture which fed the storm = fuel Ratio: climatological maximum moisture to actual storm moisture = in-place maximization factor

16 Storm Maximization Process Follow same methodology as HMR (e.g. HMR 51 Section 2.3.3) Trying to identify the moisture source which fueled the storm Some location upwind and outside of rain shield Storm rep should best represent timeframe associated with main rainfall event Utilize updated climatological databases HMR 51 used 12-hr persisting climo published in 1968 AWA has updated Dew Points and SSTs climo Dew points 3-, 6-, 12-, 24-hr average at the 100-yr recurrence interval Utilize HYSPLIT model trajectories Not available to HMRs

17 HYSPLIT Trajectory used for Zerbe (Agnes) Storm Rep Analysis

18 How Do We Compute PMP? PMP on a ~2.5mi 2 grid Move maximized storms to each grid Account for differences in moisture and elevation Calculate Orographic Transposition Factor (OTF) Uses Precip Frequency-NOAA Atlas 14 Difference between source (in-place storm) and target (Piru drainage basin) locations OTF-Quantifiable/Reproducible Replaces HMR Storm Separation Method (SSM)/Stippled Region Highly subjective Not reproducible Results in total adjustment factor Apply to the DAD values

19 Example PMP Results 72-hour PMP values Provide basin average, sub-basin averages, or grid values Spatially/Temporally Distributed Using Actual Storm Pattern or Climatology

20 Summary Storm based and reproducible Ability to consider site-specific characteristics Unique topography Needs of TVA operations/hydrology PMP study produces updated/reliable values Developed using the most current methods and data available Higher confidence in results/data Extensive review process Independent board of consultants Nuclear Regulatory Commission/TVA Completed by end of the year

21 QUESTIONS Bill Kappel Cool Science, Happy Clients

22 EXTRA SLIDES