Updating EBMUD’s Water Supply Planning Model to Meet Future Challenges
CWEMF 2016
GARY PALHEGYI, P.E., D,WRE BEN BRAY, PH.D., P.E. Why a New Planning Model?
• Previous model built in FORTRAN – Many years of edits, changes and manipulation – Difficult to make code changes and integrate new components – Loosing experts that can manage the model
• Update to modern software (RiverWare) – Improved system and temporal resolution – Improved capabilities and flexibility – Improved transparency
2 Overview of System
Lower Mokelumne River
Woodbridge ID Lodi
Walnut Creek Pump Station
3 RiverWare Schematic
PG&E Regulated Inflows
Amador County CCWD Calaveras County Intertie
Mokelumne Hill G Walnut Creek PP Jackson Valley WD Local 3 2 1 hydrology Pardee Reservoir FRWP WC FP
Local USL hydrology Camanche Reservoir
G Lafayette P Chabot FP Moraga Channel Seepage, Riparian & Senior PP Release Evaporation G Appropriators USL Briones WTP Briones Lafayette North San Joaquin PP G P County WD Orinda Control Center Woodbridge ID, Woodbridge Dam San Pablo G City of Lodi
Release Orinda Sobrante FP WTP ><((((‘> Customer Demands ><((((‘> ><((((‘> Flows to Bay Delta 4 Operational Priorities
• Meet obligations for water right holders • Comply with environmental requirements – Flows and temperature • Meet USACE flood reserve requirements • Meet EBMUD customer demands
5 Water Right Holders
• PG&E (non-consumptive use) – Unimpaired Runoff Regulated Undiminished Flow
• Upcountry – Amador County Water Agency – Calaveras County WD, Calaveras Public Utility District – Jackson Valley Irrigation District • East Bay Municipal Utility District • Lower Mokelumne River – North San Joaquin Water Conservation District – Woodbridge Irrigation District, City of Lodi – Riparian and Senior Appropriators
6 Environmental Flow Requirements
Below Camanche Below Woodbridge Below Camanche AN/N BN D CD AN/N BN D CD 푅 = 푚푎푥 Below Woodbridge + diversions and losses Month [cfs] [cfs] Oct 1-15 325 250 220 100 100 100 80 15 Oct 16-31 325 250 220 130 100 100 80 75 Nov 325 250 220 130 100 100 80 75 Dec 325 250 220 130 100 100 80 75 Jan 325 250 220 130 100 100 80 75 Feb 325 250 220 130 100 100 80 75 Mar 325 250 220 130 100 100 80 75 Apr 325 250 220 130 150 150 150 75 May 325 250 220 100 300 200 150 15 Jun 325 250 100 100 300 200 20 15 Jul 100 100 100 100 25 20 20 15 Aug 100 100 100 100 25 20 20 15 Sep 100 100 100 100 25 20 20 15
Woodbridge Golf Year Type AN Above Normal N Normal BN Below Normal D Dry CD Critically Dry
7 Environmental Flow Temperature Requirements
Goal: Maintain cold water temperatures in the Lower Mokelumne River • Maintain Camanche hypolimnetic volume 28 TAF thru October • Colder than 16.4 degrees Celsius • Supported by Pardee Reservoir
Observed
Modeled
28 TAF
8 Flood Reserve Requirements
Sept 15th May 31st Jul 31st
Rain Reserve
Snow Melt 130 TAF Reserve
200 TAF Nov 5th Mar 15th Jun 30th
9 Fixed Demand Model
Level of Development Discretized into 5 treatment plants Annual Average Demand For example: Walnut Creek Filter Plant
Modeled Drought Contingency Planning
• Drought Planning Sequence • Customer Demand Rationing • Freeport Regional Water Project – CVP Drought Contingency Supply • Need for Water – Transfers
11 Drought Planning Sequence
• 1976, 1977 and modified 1978
560 TAF
• Which drought LOD = 2040
is more severe?
(TAF)
12 Rationing Guidelines
Drought Projected End-of-September Rationing Goal Stage Total System Storage (TAF)
Normal 500 TAF or more None
Moderate 500 – 425 TAF 0 to 10% - Voluntary
Significant 425 – 390 TAF 10% to 15% - Voluntary
Severe 390 – 325 TAF 15% to 20% - Mandatory
Critical Less than 325 TAF 20% - Mandatory
13 Freeport Regional Water Project
• Eligible when EOS-TSS < 500 TAF
• Contract limits (1-year, 3-year)
Drought Year First Year July Start at 85.4 cfs First Year July Start at 110 cfs
1st 41 TAF 53 TAF 2nd 62 TAF 80 TAF 3rd 62 TAF (165 TAF limit) 32.4 TAF (165 TAF limit) 4th Repeat pattern Repeat pattern etc.
14 Need for Water – Example Demand Reduction
2040 LOD Demand Deficiency
Drought Planning Sequence
15 Pardee Reservoir
Modeled Observed RMSE = 22 TAF
16 Camanche Reservoir
Modeled Observed RMSE = 32 TAF
17 Mokelumne Aqueduct Draft from Pardee Reservoir
2010 LOD, 216 MGD
Comparison to OBSERVED daily data 18 Briones Reservoir
San Pablo Reservoir
USL Reservoir
19 Questions?
[email protected] [email protected] 510-287-2068 510-287-0206 Forecasting Future Storage
Carryover Storage CVP drought supply EOS TSS March 1st
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Rationing EOS TSS April 1st
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
JSA Year Type (Oct thru Mar) P+C Storage Oct 1st
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
21 Comparison of Flows at Golf
5000 Difference in Volume = -0.050% 4500 RMSE = 41 cfs 4000
3500 y = 1.006x R² = 0.996 3000
2500
2000
Predicted Flows (cfs) Flows Predicted 1500
1000
500
0 0 1000 2000 3000 4000 5000 Observed Flows (cfs) 22 Year Type Designation
• Environmental Flow Requirements (JSA) – AN, BN, Dry, Critical – April thru September (runoff from Oct thru Sept) – October thru March • Oct 1st: forecast Nov 5th Pardee+Camanche storage
• Riparian & Senior Appropriators Diversions – Runoff volume (runoff from Oct thru June) – RSA year type (< or ≥ 250 TAF)
• Woodbridge Irrigation District/Lodi Diversions – Pardee Inflows (Oct thru Sept) – WID year type (< or ≥ 375 TAF)
23 TRA Calibration – Runoff (base flows and storm flows)
Comparison of Computed Local Runoff to Observed San Pablo Creek Flow 0.60 Measured Flows
Computed by Rational Formula with Baseflows added 0.50
0.40
High flows exceeded gage capacity, no data recorded, gaged damaged
0.30
storm flows
Discharge (AF/acre /month) (AF/acre Discharge 0.20 base flows
0.10
0.00 Apr-83 Oct-83 Apr-84 Oct-84 Apr-85 Oct-85 Apr-86 Oct-86 Apr-87 Oct-87 Apr-88 Oct-88 Apr-89 Oct-89 Apr-90 Oct-90 24 Pardee+Camanche Storage - OBSERVED
RMSE RiverWare = 50 TAF
RMSE 2000 to 2012 = 23 TAF RMSE using EBMUDSIM = 75 TAF
25 Total System Storage - OBSERVED
RMSE RiverWare = 58 TAF RMSE 2000 to 2012 = 25 TAF RMSE using EBMUDSIM = 80 TAF
26 TRA Calibration
Local hydrologic parameters Measured Inflows
Precipitation Watershed runoff (storms, base flows) Measured Measured Evaporation Storage Draft Seepage/leaks
Measured Outflows
27 RiverWare Workspace
28 True Natural Flow & Year Type
29 Calibration Procedure
USGS @ Mokelumne Hill G
Local hydrologic parameters
Pardee Reservoir JVID Aqueduct Draft Precipitation (Seepage = 16-18 cfs) Computed Pardee Outflows Test for Storm runoff G Goodness-of-Fit Observed Pardee Outflows Evaporation Seepage/leaks Camanche Reservoir (Seepage = 8-12 cfs)
Computed Camanche Outflows Test for G Goodness-of-Fit USGS @ Camanche Dam
30