Yadkin_PeeDee_Lumber_Model_Overview_Dec_11_2020

Model Overview –
Yadkin-Pee Dee/Lumber
River Basin Hydrologic ModelDecember 11, 2020
Project Timeline
2
March April May June July August September October November December
Kickoff Training
Meetings
Schematic
Inflow Development
Data Collection
Operating Rules
Basecase Run Development
Documentation
Basecase
Review
Data/Inflow
Review
Project Acknowledgement
3
•Generous financial support by DWR
•Tom Fransen, Pam Behm –DWR
•Generous time commitment from stakeholders, especially TRC members
•Neela Sarwar, Pam Behm –DWR; Jonathan Williams –HDR; Tim Poole –Cube Carolinas; Ed Bruce –Duke Energy;
Tony Young –Corps of Engineers; Aubrey Lofton –Union County; Curtis Weaver –USGS; Chris Goudreau –NC Wildlife Resources
Commission; Brian Fannon –Yadkin Riverkeeper; Jefferson Currie –Lumber Riverkeeper
•Supporting documentation and data from the YPDWMG and its members (utilities and power companies)
Meetings and Purpose
4
•Kickoff Meeting (Mar. 4)
•TRC #1 (Sept. 2): model development overview, plus schematic review
•TRC #2 (Oct. 5): review inflows
•TRC #3 (Nov.12): review basecase run results (including inflows and operating logic)
•Model Overview [today]: provide model overview, applications of model, and results
•Training (mid-Jan): virtual demonstration of OASIS model and scenarios
NC DWR Basin-Wide Modeling Initiative
(Shown to YPDWMG in Nov.2016)
5
CHEOPS model in Catawba; OASIS model elsewhere. First DWR basin model was in the 1970s for the Yadkin Capacity Use Area Study.
Major Modeling and Water Use Assessment in the Yadkin
6
•Relicensing
•Yadkin Project: Alcoa (APGI) early 2000s; new license issued in 2017 to now Cube Hydro Carolinas. OASIS developed by HydroLogics (now Hazen)
•Yadkin-Pee Dee Project: Progress Energy mid 2000s; new license issued in 2015 to now Duke Energy. CHEOPS developed by DTA (now HDR).
•IBT Permitting
•Concord/Kannapolis: OASIS used by NC DWR.
•Union County: CHEOPS used by HDR.
Major Modeling and Water Use Assessment in the Yadkin (cont’d.)
7
From HDR presentation to DMAG in 2019
The Next Model Iteration
8
Reservoir Catchments and
DWR Subbasins
9
Schematic Detail
10
11
Classic OASIS
12
New OASIS
Refinements with YPDL OASIS Model*
13
•Inflow dataset
•Relies on longer inflow record to capture additional droughts
•Uses many more gages in the basin (made possible by extensive data unimpairment), including key gages on the mainstem (Yadkin
College, High Rock, and Rockingham), plus inflows derived from historic operating data for Kerr Scott and High Rock
•Ensures a monthly match with unregulated gage flows
•Automated update to keep inflows current and allow for operations and real-time forecasting
•Ag water use developed explicitly around water use needs of certain crops relative to rainfall and needs of livestock
•Future demands can be updated automatically as crop patterns and livestock counts change
•Tracking the flow of water
•Extensive interconnections (regular, emergency, and/or IBTs) provided
•Automated safe yield routines
•Switch to turn on and off all drought plans, including LIP
•Automated demand adjustment (uniformly applied to all demand nodes)
•WW returns linked to demand nodes get adjusted automatically
* Common to all OASIS models for NC basins
Uses of the Model
14
•Water budgeting (supply and demand) for all significant users, over a long, fully unimpaired
inflow record
•Prior models did not capture this level of detail, including interconnections
•Example: Monroe is one of dozens of systems that can now be evaluated,
capturing the critical droughts of record for each (for Monroe, 1950-51)
Uses of the Model (cont’d.)
15
•Drought plan assessment
•Develop improved operating rules, including probability-based drought triggers
•Drought exercises for the YPDWMG
•Impacts of interbasin transfers
•Planning and operations of facilities (including hydro)
•Ecological flow impacts
•Forecasting of inflows and reservoir storage
•Impacts of reservoir rule curves and storage on downstream flows
•Not a hydraulic model, but can be used for assessing flood control benefits
•Note: routing provided to improve flow estimation to High Rock (based on one day lag of Kerr Scott change in storage) and South Carolina Pee Dee
gage (based on two day lag from Rockingham)
•Generally, routing is not needed because of significant local resolution in inflows due to wide network of gaging stations used in inflow development
Simulation Over Historic Inflow Record
16
Forecast Run –Storage Projection Sample
17
Uses of the Model
of Interest to the YPDWMG
18
•Union County IBT analysis by HDR
Uses Could Include Other Scenarios Considered Before
19
http://cubecarolinas.com/wp-content/uploads/2019/12/2019-12-December-Check.pdf
OASIS Model Accessibility
20
•Available to all stakeholders through accounts to NC DWR server
•Model is a living document, meant to be easily updated
•Provided with automated inflow update
•Changes to system plumbing or operating rules can easily be made
•Adding additional historical data for inflow/operating rule verification
•Model is user-friendly, with easy to define performance measures like elevation, flow, and generation, along with probability tables and plots
and user-defined level of impact (thresholds defining minor or major) like for Union County IBT analysis.
•Model is well documented, including historical detail on hydro operations between the old license, “interim” license, and new licenses
•Tutorial for creating and modifying runs and adjusting input and output
Model Development Process
21
•Develop schematic
•Yadkin Pee Dee: node numbers <= 999
•Lumber: node numbers >= 1000
•Nodes assigned ending number depending on classification (e.g., reservoirs = __0)
•Geographic extent: from headwaters to where rivers join the Pee Dee in South Carolina, with local resolution in North Carolina
•Provide consistency with HDR’s YPDWMG Demand Projections (Tech Memo Update -July 2019) regarding entities, amounts, and sub-basin classifications
•Surface water only (either withdrawals or WW discharges), with facilities in operation or anticipated in the future
•Compile streamflow and precipitation gaging data
•Collect impairment data (withdrawals and WW discharges >= 0.1 mgd for M&I, plus withdrawals for Ag), plus reservoir change in contents,
from databases and information from entities
•Hindcast impairments back to 1930 (start of inflow record), adjusted for facility start/stop dates
•Develop unimpaired inflows on monthly basis
•Match at gages, meaning error is embedded in the impairments
•Disaggregate to daily inflows using mostly reference gages
•Incorporate operating rules
•Develop basecase run (current conditions) –daily timestep, 1930 to Sept. 2019 (with provisional inflow updates to allow for real-time drought forecasting)
Schematic
22
•Inflow nodes: 80 in YPD, 20 in Lumber
•USGS gages = 36
•Reservoir nodes: 30 (all but one in the YPD)
•M&I demand nodes: 40 in YPD, 5 in Lumber
•Agricultural demand nodes: 8 in YPD, 5 in Lumber
•WW return arcs linked to demand nodes: 35 in YPD, 5 in Lumber
•WTP process return arcs linked to demand nodes: 15 in YPD, 1 in Lumber
•WW independent return nodes: 25 in YPD, 20 in Lumber
•Interconnection arcs: 15 regular, 30 emergency, including IBTs
•Future intakes
Compile Gaging Data (for the USGS-designated “Pee Dee River Basin”
23
Gage Map
24
Gage Map
25
Compile Impairments
26
•Water withdrawals*
•Public: LWSP database (1997 through present, with some gaps, on a monthly basis); data collected through 2019.
•Industrial (including power plants): WWATR database (1999 through present, with some gaps, on a monthly basis); data collected through 2018
•Power plants evaluated as “net” withdrawal for consistency with HDR study (= water –wastewater use) since water and WW discharges are in close proximity
•Agriculture: from USDA census data on irrigated crop acreage and livestock counts, + USGS surveys. Key irrigated crops incorporate water use curves in which
irrigation use is dependent on rainfall. Use computed at county level (NC, but also VA and SC) and, in most cases, allocated to each subbasin based on percent coverage.
•Additional data from entities will supersede information from databases
•Wastewater discharges
•Public: NPDES database (early 1990s through present on a monthly basis); data collected through 2018. Some information provided from LWSP databases.
•Industrial: NPDES
•Occasionally, facilities have multiple outfalls which were aggregated to get total discharge
•Additional data from entities will supersede information from databases
•Some entities include NCG (stormwater) permits; stormwater excluded
•Reservoir change in contents and associated net evaporation (using surface area x net evap rate)
•USGS reports provide key information on mainstem reservoir change in contents
•Supplemented with requests of utilities and power companies
* NC Statute in 1991 required WD registration, updated every 5 years, for non-Ag uses > 0.1 mgd (Ag use is > 1 mgd) or transfers from one basin to another.
In 2007, requirement for annual water use reporting.
Criteria For Entities Being Included in the Inflow Unimpairment
27
•All those with historic surface water withdrawals from the basin > 0.1 mgd annual average (Ag not included)
•Seasonality considered when annual average < 0.1 mgd
•Only Lumberton had significant GW withdrawal as well as SW withdrawal. This was accounted for.
•All those with historic surface water WW discharges in the basin > 0.1 mgd annual average
•Also applies to entities that withdraw only GW
•Same note as above on seasonality
Excluded are purchasers that do not have a surface water withdrawal in the basins
•E.g., Yadkin County which purchases water from Jonesville
Note: for schematic inclusion, entities must have used > 0.1 mgd in the last 5 years (or are anticipated to use > 0.1 mgd in the future), or
interconnections like IBTs that have not been used yet or are used only in emergency
•E.g., Charlotte (through Concord-Kannapolis), Union County (from Tillery), Greensboro (through Winston-Salem), and High Point (through
Winston-Salem)
Hindcasting
Sub-Basin Estimates from HDR Study
HDR Base Year = 2017
(Ag based on the highest reported water use
from 5-year USGS reports available from
mid-1980s to 2015)
Breakout from HDR Study
As noted, power discharges are
incorporated in the withdrawal
numbers as a net withdrawal,
so discharges are shown as 0.
We made the same assumption.
Breakout from HDR Study
Aggregation of Impairments
•Current conditions OASIS run (Basecase) uses 2015-2019 averages
•Compare with HDR Baseline conditions which use 2017 data
OASIS Input Data Comparison
Uses 2015-2019 annual averages
for base year, HDR uses 2017 data
for all except Ag, which is based on
largest of 5-year reported USGS
data starting in 1990 (due to wide
variation).
Our Ag based on 2017 Census
data, run with 2015 precip for
comparison to most recent USGS
report.
1.7 (OASIS)0.2 (OASIS)
108 (OASIS)
5.0 (OASIS)
4.3 (OASIS)
0 (OASIS)
9.7 (OASIS)
46 (OASIS)
6.8 (OASIS)
70 (OASIS)
0.6 (OASIS)
0.2 (OASIS)
0.0 (OASIS)
0.4 (OASIS)
49.7 (OASIS)
6.7 (OASIS)Difference is Hendrick Mine by 3 mgd
(ours = 1.8 mgd; HDR = 5 mgd)
[reason is monthly and annual don’t match
in data reports]
Our Ag is 7.0; HDRs is 10.5
Our Ag is 18.6; HDRs is 20.6
Reservoir Storage Summary
34
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
No
r
m
a
l

P
o
o
l

S
t
o
r
a
g
e

(
M
G
}
Ignores flood
control space in
Kerr Scott
(10x conservation
pool size)
Reservoir Storage Summary
35
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
No
r
m
a
l

U
s
a
b
l
e

S
t
o
r
a
g
e

(
M
G
}
Reservoir Impairments
(Using High Rock and Narrows as an Example)
36
Historic net
evaporation
= historic surface
area (converted from
storage-area curve) x
estimated net evap
Reservoir Impairment (using WSACC as an Example)
37
Unimpairment of Gages
38
Fill In Missing Record
Correlation
40
Correlation
41
Annual correlation using
Rockingham unimpaired gage = 0.87
(used mostly from Jun to Sep)
Correlation
42
Annual correlation using
Rockingham unimpaired gage = 0.91
(used mostly from Aug to Nov)
0
500
1000
1500
2000
2500
1/
1
/
2
0
0
2
1/
2
/
2
0
0
2
1/
3
/
2
0
0
2
1/
4
/
2
0
0
2
1/
5
/
2
0
0
2
1/
6
/
2
0
0
2
1/
7
/
2
0
0
2
1/
8
/
2
0
0
2
1/
9
/
2
0
0
2
1/
1
0
/
2
0
0
2
1/
1
1
/
2
0
0
2
1/
1
2
/
2
0
0
2
1/
1
3
/
2
0
0
2
1/
1
4
/
2
0
0
2
1/
1
5
/
2
0
0
2
1/
1
6
/
2
0
0
2
1/
1
7
/
2
0
0
2
1/
1
8
/
2
0
0
2
1/
1
9
/
2
0
0
2
1/
2
0
/
2
0
0
2
1/
2
1
/
2
0
0
2
1/
2
2
/
2
0
0
2
1/
2
3
/
2
0
0
2
1/
2
4
/
2
0
0
2
1/
2
5
/
2
0
0
2
1/
2
6
/
2
0
0
2
1/
2
7
/
2
0
0
2
1/
2
8
/
2
0
0
2
1/
2
9
/
2
0
0
2
1/
3
0
/
2
0
0
2
1/
3
1
/
2
0
0
2
Fl
o
w

(
c
f
s
)
Elkin Gage
unimpaired gage flow reconstituted unimpaired gage flow actual gage flow
Finalize to Daily Timestep
43
Reconstituted = sum of all natural inflows at nodes upstream (690 cfs monthly average)
Unimpaired monthly average = 690 cfs
Actual gage flow monthly average = 714 cfs
Impairment Summary
44
45
Net Impairments (Not Incl. Reservoirs) –High Rock
46
Net Impairments (Not Incl. Reservoirs) –High Rock
47
Reservoir Impairments –High Rock
Other Data for Basecase Run
48
•Physical
•Reservoir storage-area-elevation
•Pumping capacity
•Turbine capacity
•Spill rating curves
SAE for Kerr Scott
Old (from project start)2010 (survey up to 1075 feet)
Operations
50
•Derive from reports, LWSPs, WSRPs, operating licenses, and personal contacts
•Include drought plans and minimum releases
Drought Plans
51
•Almost 20 that are modeled –tied to reservoir storage/elevation; river flow; drought monitor, and/or river stage.
Drought monitor used for LIP when available. % WTP capacity not modeled since that requires distribution system
demand that can vary hourly.
LIP –Yadkin Project Requirements
Verification of Basecase Run
53
•Inflows
•Operating rules
Monroe –Example --At 6 mgd avg.
54
0%
20%
40%
60%
80%
100%
120%
Ju
n
-
0
7
Au
g
-
0
7
Oc
t
-
0
7
De
c
-
0
7
Fe
b
-
0
8
Ap
r
-
0
8
Ju
n
-
0
8
Au
g
-
0
8
Oc
t
-
0
8
De
c
-
0
8
Fe
b
-
0
9
Ap
r
-
0
9
Ju
n
-
0
9
Au
g
-
0
9
Oc
t
-
0
9
De
c
-
0
9
Fe
b
-
1
0
Ap
r
-
1
0
Ju
n
-
1
0
Au
g
-
1
0
Oc
t
-
1
0
De
c
-
1
0
Fe
b
-
1
1
Ap
r
-
1
1
Ju
n
-
1
1
Au
g
-
1
1
Oc
t
-
1
1
De
c
-
1
1
Fe
b
-
1
2
Ap
r
-
1
2
Ju
n
-
1
2
Au
g
-
1
2
Oc
t
-
1
2
De
c
-
1
2
Fe
b
-
1
3
Ap
r
-
1
3
Ju
n
-
1
3
Au
g
-
1
3
Oc
t
-
1
3
De
c
-
1
3
Fe
b
-
1
4
Ap
r
-
1
4
Ju
n
-
1
4
hist
Kerr Scott
55
Note: operation changed in 1993 with Water Control Plan, including new low flow protocol
Kerr Scott
56
Note: shown since 1993 when revised Water Control Plan went into effect
Kerr Scott -Historic
57
Historic release sometimes less than minimum required
(here, normal minimum is 125 cfs only when < 1023 feet)
Historic return to guide curve can be delayed due to hedging on flooding concerns downstream and also holding water
to delay drawdown during drought
Kerr Scott -Simulated
58
Simulated releases the exact required amount
(during flood control based on Wilkesboro gage and
during low flow situations) using perfect foresight of
today’s inflows; plus simulated returns to guide curve faster
High Rock
59
Documents Used to Model Mainstem Operations
60
•HDR Model Logic and Verification Report from 2014
•Pulled some information from the 2002 APGI and 2003 Progress Energy Initial Consultation Documents
•CHEOPS model inputs for 2014 Assessment of Union County IBT
•Relicensing Settlement Agreement for APGI and Comprehensive Settlement Agreement for Progress Energy
in 2007
•License Documents for APGI in 2016 and Duke Energy in 2015
Old License Operation for Yadkin Projects
Alcoa Power operated its Yadkin Project in accordance with a 1968 headwater benefits agreement with the licensee of
the Yadkin –Pee Dee Project. According to the 1968 agreement, Alcoa Power regulates weekly average stream flow
from Falls Reservoir to provide a flow not less than 1,500 cubic feet per second (cfs) during the 10-week period
preceding the recreation season (May 15 through September 15); 1,610 cfs from May ….
Iterations for a New License
Modeled by HDR for Union County IBT work
circa 2014
Basin-Wide Operations
63
•Mostly independent
•Reservoirs upstream will not make releases for users downstream unless minimum flow requirements apply
•Kerr Scott will provide additional release from Winston-Salem’s account during low flow/high demand
•Kerr Scott will limit releases down to Wilkesboro for flood control
•Coordination among entities with multiple reservoirs, intakes, and WW discharges (e.g., WSACC, Anson
County, Moore County in Lumber)
•Coordination through sale and purchase agreements, regular and emergency
•Coordination during drought conditions through Low Inflow Protocol
Hydro Operations
64
•Set up to exploit the permitted operating band per the license agreements
•Model will generate down to the normal minimum elevation (NME) up to turbine capacity
•Limited the operating range based on historic data (since 2017 when both companies were operating with
new licenses)
•Not capturing day-to-day operations that are based on power market prices and demand
•Customized models can be developed as off-shoots to model (e.g., optimal dispatch for Dominion Virginia on the Roanoke River)
YPD Project -Historical Operation
65
Shown is post-2017 after license was renewed so we have representative operations.
Here normal min flow requirement is shown; for Blewett, it would be adjusted if LIP were activated
Maintenance (Ed Bruce believes this was the case for Tillery)
YPD Project -Historical Operation
66
Blewett at 3 feet down
YPD Project -Simulation
YPD Project -Simulation
YPD Project -Simulation
YPD Project -Simulation
YPD Project -Simulation
Yadkin Project -Historical Operation
72
HR at 10 feet down
HR at 4 feet down
Deviation below 4 foot September NME to prepare for Hurricane Florence
Yadkin Project -Historical Operation
73
Yadkin Project -Simulation
Yadkin Project -Simulation
YPD Project -Simulation
More inflow to Falls due to higher releases from
High Rock (during the winter months) when
historically they did not draw down to NME as often.
Yadkin Project -Simulation
Yadkin Project -Simulation
•Implemented in Feb. 2007
•High Rock operations impacting elevations may have changed
between then and when license was issued in 2017
•Monthly determination, influencing Falls and Blewett minimum releases
and water withdrawals (stages >=1) and hydro peaking (stages >= 0)
•Drought monitor based on national product (available
since 2000), potentially refined for regional use
•DMAG to review every 5 years per license conditions
as it relates to drought monitor (national vs. regional),
gaging estimates and long-term averages, and
proportional drawdown of reservoirs
Low Inflow Protocol
79
http://cubecarolinas.com/wp-content/uploads/2019/12/2019-12-December-Check.pdf
LIP (End of Month Assessment) –Matching HR Historic Elevation
0.41 for Trig 1
Both >=1 for Trig 1
High Rock operations post-2007 LIP issuance may have differed from post-2017 license issuance.
Timing of triggers might be offset by a month depending on when calculation is made.
Drought monitor 3-month average may differ from Cube’s calculations (starting Feb 2007); Cube’s used here except later when showing 2000 to 2007.
Actual LIP levels:
Sep and Oct 2007: 2
Jul and Aug 2008: 0
and 1, respect.
Feb 2011: 0
Sep 2011: 0
Nov 2012: 0
Oct and Nov 2014: 0
(due to maintenance)
Oct 2015: 0
Large drawdown in 2014 due to maintenance
Impact of Drought Monitor
LIP -Simulation
LIP –Simulation (With and Without Monitor)
LIP Simulation Back to 2000
0.41 for Trig 1
Both >=1 for Trig 1
LIP –Simulation (With and Without Monitor)
Impact of LIP on Flows
86
Normal = 1000 cfs(rest of year)
Normal = 1500 cfs (May 16 –31)
Normal = 2000 cfs (Feb 1 –May 15)
LIP on is with the drought monitor; No LIP also includes no utility WSRPs on –all set by switch in constants table (drought plans on or off)
Impact of LIP on Flows
87
Normal = 1200 cfs (rest of year)
Normal = 1800 cfs (May 16 –31)
Normal = 2400 cfs (Feb 1 –May 15)
Impact of LIP on Storage
88
Impact of LIP on Storage
89