Difficulties Integrating Wind Generation Into Urban Energy Load

1
Difficulties Integrating Wind Generation Into
Urban Energy Load

• Russell Bigley
• Shane Motley
• Keith Parks

2
Currently in 2009 Xcel Energy is the 1
utility provider of wind in the nation

• 2,876 MWs
• of Wind Generation on Xcel Energy system

3
Utility Overview

• Primary goal
• Keep the lights on
• Secondary goals
• Run at peak efficiency
• Prepare for plant maintenance and other outage
  issues such as transmission

4
Utility Overview-Load

• Understanding Power Usage (load)
• Power Load Forecasts
• Highly dependent on weather conditions
• Temperatures
• Cloud Cover
• Precipitation

5
Utility Overview-Load

• Load Forecast Error
• Error comes from 2 sources
• Model Error
• Weather Forecast Error
• Load forecast Error (MAE) is typically less than
  3-averaged over the 24 hour period (even day
  ahead)

6
Generation Forecasting

• Optimizing Power Plant Output for forecasted
  LoadTypically this involves scheduling
• Coal Power Plants
• Gas Power Plants
• Hydro/Geothermal Facilities
• Wind Plants--highly variable output

7
Generation Assets

• Many physical differences in power producing
  assets
• Main concern Assets that can be dispatched and
  assets that cannot be dispatched
• Wind Generation is non-dispatchable
• wind generation can be curtailed
• Wind Generation is forecasted and scheduled
• Thus there is risk associated with the
  generation

8
Scheduling Wind Generation?

• Many Issues with wind generation
• 1) Generation is dependent on wind
• Generation is typically not static
• Requires an excellent wind forecast
• Even a great wind forecast doesnt result in an
  accurate generation forecast
• Accurate Power Curves for wind turbines
• A better understanding of generation output on a
  large farm scale basis
• Many estimates for total farm output are
  overestimated (Danish Wind Industry)

9
Wind Generation Forecast Error

• Wind Generation forecast Error average around 20
  for the 24 hour day ahead period
• Persistence is a good forecast in real time, but
  misses the ramps
• How can the forecast be sooo bad!!!

10
Why is generation so variable the forecast
performance poor.

1. Wind speeds are variable
2. Terrain differences
3. Elevation and hub height difference
4. Turbine availability/turbine types
5. Turbine induced wake effects
6. Turbulent eddies induced by terrain
7. Wind speed variations with height
8. Turbine blades build up debris and affect the
   aerodynamics
9. Weather model resolution
10. Data Data Data
11. Communication with wind farm operators.and
    theres more!!!!!

11
Peetz/Logan Wind Farm
Wind farm over 40 miles across and over 200
turbines
12
(No Transcript)
13
Turbines size HUGE!! These are 2.3MW Seimens
turbines located near Adair, IA.
14
Generation Forecasting

• Wind fields tend to be variable and output is
  even more variable
• Small changes in wind speed tend to make large
  differences in power generation
• Air Density differences also affect the power
  output (i.e. Summer vs. Winter)
• Power Curves are not well documented and are
  performed at sea level and at standard
  temperatures

15
Pa 1/2 ? µ A v3         (2)where µ
efficiency of the windmill (in general less than
0.4, or 40)
16
Wind Forecasting

• Wind direction can make a huge impact on power
  generation as turbine placement enhances turbine
  wake effects
• Wake effects can propagate up to 10 times the
  blade diameter of the turbine (Danish Wind
  Industry Assocation)

Blade Lengths are 35 meters (114 ft) long
Wake can propagate up to 700 meters (2296 ft)
The Diameter is then over 70 meters (230 feet
17
A rare, aerial photo of an offshore windfarm in
Denmark clearly shows how turbulence generated by
large turbine rotors continues to build with each
successive row of turbines.
18
(No Transcript)
19
Weather Impacts

• High Winds
• Turbines cut-out at a predetermined wind speed
  to prevent damage to the turbine (blades,
  generator, etc.)
• Cold Temps
• Turbines cut-out at predetermined temperatures
  to prevent damage
• Precipitation
• Rain and snow reduce power output
• Freezing Rain may damage blades and throw ice
• Decreases power output

20
Other impacts

• Debris buildup on blades
• Dirt and insect buildup reduce the aerodynamics
  around the blade

21
Communication

• Information from the wind plant operators is
  critical in this whole process
• Downtime due to different causes
• Maintenance
• Weather
• Weather
• Weather

22
Key Issues and Solutions

• Wind and generation data
• Attempting to acquire all wind speed, wind
  direction, and generation data by turbine
• 1000s of pieces of data to stream to a database
• Modeling
• Acquired the assistance of NCAR and NREL
  (National Central for Atmospheric Research and
  the National Renewable Energy Lab)
• Use latest modeling technology and bias
  corrections to achieve better results for
  real-time and day-ahead wind and generation
  forecasts

23
Without improvements in Communication with wind
plant operatorsData at the Turbine Level
Modeling we head down a dangerous path if we
plan on integrating even more wind on our systems.

• youtube video turbine failure

24
(No Transcript)