Hydropower Variability in the Western U'S': Consequences and Opportunities

1
Hydropower Variability in the Western U.S.
Consequences and Opportunities

• Nathalie Voisin, Alan Hamlet, Phil Graham, Dennis
  P. Lettenmaier
• UW-UBC Fall Hydrology Workshop
• University of Washington
• October 1, 2004

2
Background

• Climate
• Increasingly predictable up to 6 months (or more)
  in advance
• West coast U.S. climate more predictable than
  other regions, due to strong ocean influence
• California and the Pacific Northwest are out of
  phase for some climate events such as El Nino
  Southern Oscillation (ENSO)
• Energy Demand
• California has regular peaks in winter and summer
  while energy consumption in the Pacific Northwest
  (PNW) has a strong winter peak
• Question How can climate predictions be used to
  manage West Coast energy transfers more
  efficiently?

3
Outline

• 1/ Data and Models
• Meteorological data
• Hydrological model
• Reservoir models
• 2/ Observed covariability
• Streamflow and Climate
• Hydropower and Climate
• Energy demand and Climate
• Hydropower and Energy Demand
• 3/ Opportunity more efficient inter-regional
  energy transfers?
• Currently climate information is not used in
  planning West Coast energy transfers
• Some ideas for an energy transfer model that
  exploits climate information
• 4/ Conclusions

4
1/ The Data

• 1/ Data and Models
• Meteorological data
• Hydrological model
• Reservoir models
• 2/ Observed covariability
• 3/ Opportunity more efficient inter-regional
  energy transfers?
• 4/ Conclusions

5
Meteorological Data

• Station Data sources National Climatic Data
  Center (NCDC)
• Extended time series from 1916 to 2003
• Forcing data sets gridded to the 1/8 degree
• Adjustment of forcing data sets for orographic
  effects based on PRISM (Parameter-elevation
  Regressions on Independent Slopes Model )
  approach (Daly and colleagues at Oregon State
  University)
• Adjustment to reflect long-term trends that are
  present in the carefully quality controlled
  Hydroclimatic Network (HCN) and a similar network
  for the Canadian portion of the Pacific Northwest
  (PNW) region (Hamlet and Lettenmaier 2004)

6
Hydrologic Model VIC (1/2)
1/ Water Balance
2/ Runoff Routing
7
Hydrological Model VIC (2/3)
Simulated Flow Red Observed Black
8
Hydrological Model VIC (2/3)
Simulated Flow Red Observed Black
9
Reservoir Models CVMod and ColSim

• Represent physical properties of the reservoir
  systems and their operation
• Assume fixed level of development
• Monthly time step

Monthly Natural Streamflow
Flood Control, Energy Demand
Water Demand
CALIFORNIA CVMod (Van Rheenen et al 2004)
PACIFIC NORTHWEST ColSim (Hamlet and Lettenmaier
1999)
Hydropower
10
2/ Observed Covariability

• 1/ Data and Models
• 2/ Observed Covariability
• Streamflow and Climate
• Hydropower and Climate
• Energy demand and Climate
• Hydropower and Energy Demand
• 3/ Opportunity more efficient inter-regional
  energy transfers?
• 4/ Conclusions

11
Streamflow Covariability
North CA peak in winter South CA peak in
spring ENSO 17 annual flow difference PDO 2
12
Streamflow Covariability
PNW peak in early summer ENSO/PDO 12-16
annual flow difference
13
Hydropower Covariability
PNW peak in J CA peak in M
14
Energy Demand Covariability

• 2 types of demand
• Peak hour demand
• Daily total Demand
• Demands are out of phase in CA and in the PNW!!

15
Energy Demand Covariability

• How predictable is the energy demand?
• Regression of observed energy load with
  temperatures

Monthly average of daily total demand
Warming/Cooling degree days S (T-18.7)day
Daily Peak Hour Demand Tmax
R20.68
R20.60
16
Timing

• Interannual variability winter and summer
• Energy demand is out of phase in CA and in the
  PNW
• PNW energy production and energy demand are out
  of phase
• PNW hydropower and CA peak energy demand are in
  phase
• Interannual variability ENSO events
• ENSO warm Higher temperatures and less
  precipitation in the PNW
• ENSO cold Higher energy demand in the PNW in
  winter and higher summer hydropower production

17
3/ Energy Transfers

• 1/ Data and Models
• 2/ Observed Covariability
• 3/ Opportunity more efficient inter-regional
  energy transfers?
• Currently climate information is not used in
  planning West Coast energy transfers
• Some ideas for an energy transfer model that
  exploits climate information
• 4/ Conclusions

18
The Pacific NW-SW Intertie

• 8000 MW capacity
• Reliable transmission
• Southward transfer during peak hour
• Northward transfer overnight, if needed
• Notes
• The energy transfer follows the energy demand
• Transfers are decided on an hourly basis during
  the day
• Currently climate information is not used in
  planning West Coast energy transfers

19
More efficient energy transfers?

• Based on a decision making process following the
  demand, a relation exists between climate and a
  10 year intertie time series
• BUT complications appears when using the above
  climate-intertie

Temperature
Precipitation
Climate
(timing)
Energy Demand
Hydropower
?
Energy Transfers
20
Energy transfer model (in progress)

• Monthly time step, daily sub time step ( peak
  hour complication)
• Principles
• Assumes perfect forecast ( monthly hydropower
  production known)
• Transmission line capacity limits the energy
  transfers

Temperature
Precipitation
Climate Forecast
(timing)
Energy Demand
Hydropower
Derived daily and peak hour
Disaggregation to daily based on temperature
Energy Transfers
Energy Transfer Model
21
Conclusions

• Observed Covariability
• Streamflow and Climate (precipitation,
  temperature)
• Hydropower and Climate (precipitation and
  temperature)
• Energy Demand and Temperature
• Consequences Energy supply and demand are out
  of phase within the same Region ( California or
  PNW)
• Opportunities Temperature is (relatively) highly
  predictable. How can long-range (out to a year)
  forecasts of air temperature anomalies be used to
  better manage energy transfers between the two
  regions?
• Future work
• Evaluate the potential for increased transfers
  using statistical methods, combined with a simple
  model for incorporating (uncertain) forecasts of
  energy demand and supply for lead times up to one
  year
• Evaluate the worth of (energy production and
  demand) forecasts via an economic analysis based
  on the price difference between hydropower and
  conventional resources

22
Additional slides for eventual questions
23
Meteorological Data NCDC
HCN/HCCD Monthly Data
Topographic Correction for Precipitation
Correction to Remove Temporal Inhomogeneities
Preprocessing Regridding Lapse Temperatures
Temperature Precipitation
Coop Daily Data
PRISM Monthly Precipitation Maps
Extended time series from 1916 to 2003
24
Energy Demand Model (1/2)

• Derived peak hour energy demand time series in
  the Pacific Northwest skill in wintertime

25
Energy Demand Model (2/2)

• Derived peak hour energy demand time series in
  California skill in summer

26
Overall Covariability
27
Energy transfer model (in progress)
Scenario 1 total daily energy ( hydropower
Conventional Resources) meet PNW total daily and
peak hour energy demands.

• Daily time step
• Results aggregated to monthly time step
• Principles
• Assumes perfect forecast ( monthly hydropower
  production known)
• Transmission line capacity limits the energy
  transfers

Hydropower Conventional Resources over peak
hour period
Meet PNW Peak Hour Demand ?
How much energy needed to meet remaining daily
energy demand?
Compute Potential Transfer during Peak Hour
Enough time/capacity to send energy back
eventually?