Neuse River Basin – Hydrologic Model Update

1. Neuse River Basin –Hydrologic Model Update Brian J. McCrodden Casey Caldwell Steven Nebiker March 25, 2009

2. Agenda Project status Model overview Inflow development Calibration Preliminary model results • Eno River system • Falls Lake • Durham Remaining steps

3. Two-year effort, initially projected to be complete by Feb. 2010 Project now 90% complete in terms of deliverables Remaining project components Finalize calibration and operating rules (if deemed necessary) Install model on DWR server to provide stakeholder access Provide documentation Conduct training Project Status

4. Evaluation of: Alternative operating protocols Combined effects of water supply plans Interbasin transfer permit applications Development of individual water supply plans Platform for risk-based drought plans, starting with Falls Lake Not intended for real-time flood control Intended Uses of Model Model Overview

5. Geographic Scope of Model Model Overview

6. Upper Basin Model Overview

7. Middle Basin Model Overview Added all Crabtree Creek impoundments to improve calibrationwith downstream flow gage

8. Lower Basin Model Overview Inflow from node 700 upstream

9. Inflow Development Inflow Development Develop daily inflow record from 1930 to present • Captures extreme drought event of 2007 Adjust gage flows for upstream impairments • Lake operations • Water withdrawals (municipal, industrial, agricultural) • Wastewater returns Unimpaired inflows required if system is to be operated differently than in the past Inflow data set ensures monthly unimpaired gage flow is preserved • Assumes error is embedded in the impairment data

10. Agricultural Demand Inflow Development Historic time series developed by county Basin delineated into sub-basins based on inflow reaches Agricultural demand node placed in model for each reach of interest

11. Neuse Gage Timeline Inflow Development

12. Long Term Gages Inflow Development Gage is significantly impaired Flat River at Bahama Gage is moderately impaired Eno River at Hillsborough Gage has little to no impairment Neuse River nr Clayton Little River nr Princeton Middle Creek nr Clayton Contentnea Creek nr Hookerton Neuse River nr Goldsboro Neuse River at Kinston

13. Long Term Gages Inflow Development Gage is significantly impaired Flat River at Bahama (115, 0.05) Gage is moderately impaired Eno River at Hillsborough (44, 1.6) Gage has little to no impairment Water Year 2007 statistics: (Avg. cfs, 7 day min avg. cfs) Neuse River nr Clayton (994, 223) Little River nr Princeton (176, 0.01) Middle Creek nr Clayton (90, 6.6) Contentnea Creek nr Hookerton (735, 35) Neuse River nr Goldsboro (2196, 226) Neuse River at Kinston (2710, 303)

14. Model Nodes with Inflows ( Reservoirs and Gages) Inflow Development 010 Upstream Pond 050 WFER 060 Lake Orange 200 Little River Reservoir 250 Lake Holt 140 Lake Michie 740 Little River Reservoir (Raleigh Proposed) 080 Corp. Lake 270 Lake Rogers 100 Lake Ben Johnston 230 Beaverdam Lake 110 Hillsborough Eno Gage 115 Durham Eno Gage 290 Wake Forest Lake 300 Falls Lake Needed for Eno Capacity Use requirements 750 Little River nr Princeton Gage Needed for Teer Quarry withdrawal 500 Buckhorn Reservoir 400 Lake Crabtree 630 Clayton Gage 445 Lake Johnson 450 Lake Raleigh 560 Contentnea at Hookerton Gage 780 Goldsboro Gage 420 Lake Wheeler 440 Lake Benson 800 Kinston Gage 900 – No gage here but inflows needed for the Weyerhauser intake. Note downstream gages are tidally influenced 480 Middle Creek gage

15. Overview of Basin-wide Impairments Inflow Development

16. Usable Storage Comparison Inflow Development * * Falls Lake volume includes Beaverdam sub-impoundment

17. Usable Storage Comparison – Excluding Falls Lake Inflow Development

18. Usable Storage Comparison Inflow Development

19. Basin Demand and WW Return Overview Inflow Development 2004 Data * ** * Includes Raleigh WW return through Neuse River and Little Creek WWTPs ** Johnston Co. demand includes Clayton; WW return includes Clayton WWTP; Smithfield WW returned through Johnston Co. WWTP

20. Basin WW Return Comparison Inflow Development 2004 Data Not associated with water supply withdrawals * * Falls Lake volume includes Beaverdam sub-impoundment

21. Agricultural Demand by County Inflow Development 2004 Data

22. Unimpaired Gage Flow Examples Inflow Development Hillsborough gage (October 2007) • Gage flow = 4.5 cfs • Total impairment upstream (mainly decrease in storage in Lake Orange and West Fork Eno Reservoir) = -2.5 cfs • Unimpaired gage flow = 2.0 cfs Clayton gage (April 2008) • Gage flow = 1080 cfs • Total impairment upstream (mainly increase in Falls Lake storage) = 703 cfs • Unimpaired gage flow = 1783 cfs

23. Unimpaired Gain Inflow Development Since unimpaired flow at gage is forced to match historic flow, gain between two gages also matches historic

24. Spreadsheet Showing Gage Unimpairment Inflow Development

25. Daily Inflow Estimation Inflow Development Since impairment data are often only available monthly, daily variation preserved by using locally unimpaired gage Goal: to develop a representative set of daily inflows while preserving monthly unimpaired gage flow as “ground truth”

26. Inflow Record Inflow Development Extends from January 1930 to April 2008 Fill in missing gage records based on correlations with other gages Model equipped with inflow update provision • Since impairment data are time-consuming and costly to collect, data collection anticipated only every 5 years • In meantime, to generate real-time forecasts, use provisional inflow development approach to update inflows through present day • User only needs to input gage data and major impairment data (like Falls change in storage)

27. Calibration Calibration Required to test accuracy of inflow estimates Possible for reservoirs where historic operating data and/or gage flow data are available Focus on Eno River, Durham, and Falls More calibration results shown in appendix

28. Eno River Calibration West Fork Eno Reservoir • Lake level and release data available since 2001 (but not always daily) • Calibration run: use estimated inflows, match release, and compare computed and historic storage • For inflows, use drainage-area adjustment of unimpaired Hillsborough gage • Unimpaired monthly, disaggregated to daily with Flat River

29. West Fork Eno Reservoir Calibration Dead storage

30. Eno River Calibration Lake Orange • Lake level data available since 2001 • Release data not available • Calibration run: use estimated inflows, incorporate operating policies based on Capacity Use Area Rules, and compare computed and historic storage • Results shown at previous TRC meeting showed poor agreement • Very sensitive to release policy • Influenced to some degree by inflows from upstream agricultural ponds

31. Lake Orange (Without Adjustments) Calibration Dead storage

32. Adjustments Calibration Added agricultural pond upstream • Uses cumulative storage of all ponds upstream based on Ken Terlep’s estimates Release policy • Most of Hillsborough demand is during 8 hour periods, 5 days a week • During these periods, releases from Orange must be 2.4 times higher than the weekly average to meet instantaneous instream flow requirements • Factor = [( 5 “peak days” * 3) + (2 ”normal days” * 1 ) ] / 7 total days = 2.4

33. Lake Orange (With Adjustments) Calibration Dead storage

34. Eno River Calibration Hillsborough gage flow • Simulate upstream lake operations based on Capacity Use Area Rules (see Appendix) • Compare computed and historic gage flows

35. Hillsborough Gage Calibration

36. Durham Calibration Operating data available since 2000 Data on flows between reservoirs and treatment plants was often challenging to compile Calibration run: use estimated inflows, match release, and compare computed and historic storage • For inflows, use daily drainage-area adjusted Flat River and Little River gage flows

37. Lake Michie Calibration Dead storage

38. Little River Calibration Dead storage

39. Lake Michie and Little River Calibration Dead storage

40. Falls/Beaverdam Calibration Each project is modeled explicitly so it can be evaluated independently Operating data for Falls available since project inception (1981); data for Beaverdam available since 2000 Calibration run: use estimated inflows, match release, and compare computed and historic elevation and storage • For inflows, back-calculate from change in storage and release

41. Water Supply and Water Quality Pool Accounting Calibration Elev.251.5 (top of conservation pool) Falls Elev.249 Beaverdam 101,705 af WQ/WS 4617 af WQ/WS + 2663 af WQ (58%) 58,659 af WQ (58%) 1954 af WS (42%) 43,046 af WS (42%) Elev.236.5 (bottom of conservation pool) Used Corps’ spreadsheet (developed by Terry Brown) as guidance Corps’ calculation of Falls storage accounts includes Beaverdam accounts when elevation >= 249 feet OASIS model breaks out storage accounts by project Total storage = 106,322 acre feet (af) • WQ = 61,322 af • WS = 45,000 af

42. Beaverdam Calibration

43. Beaverdam WQ Storage Calibration

44. Beaverdam WS Storage Calibration

45. Falls Lake Calibration

46. Falls Lake WQ Storage Calibration Note: Differences due to minor adjustments to Corps’ WQ accounting

47. Falls Lake WS Storage Calibration Note: Differences due to minor adjustments to Corps’ WS accounting

48. Falls Lake (for full period of operation) Calibration Note: Beaverdam data not available pre-2000, so estimated inflows to Beaverdam (which affect Falls elevation) are slightly different than those from the Corps for that period

49. Falls/Beaverdam Calibration Calibration run, but not matching historic releases • Use estimated inflows (back-calculated) • Incorporate nominal Corps operating protocols • Rule curve • Minimum Falls release (100 cfs from April to October, 60 cfs otherwise) • Minimum flow at Clayton (254 cfs from April to October, 184 cfs otherwise) • Drought management (specifying releases from Beaverdam) • Flood control rules controlling lake levels and downstream flows • Use historic Raleigh withdrawals • Compare computed and historic elevation, storage, and downstream flows

50. Falls Lake Calibration Note: Simulation run uses current Falls rule curve adopted in 2000, resulting in some disagreement before then