Xcel Energy’s Wind Energy Program

Wind Energy Prediction – Research & Development WorkshopMay 11-12, 2010

About Xcel EnergyNorthern States Power Company- Minnesota

Northern States Power Company- Minnesota

Northern States Power Company- Minnesota

Northern States Power Company- Minnesota

Public ServiceCompany

of Colorado

Public ServiceCompany

of Colorado

Public ServiceCompany

of Colorado

Public ServiceCompany

of Colorado

Southwestern Public ServiceSouthwestern Public ServiceSouthwestern Public ServiceSouthwestern Public Service

Northern States Power Company- Wisconsin

Northern States Power Company- WisconsinNorthern States Power Company- Wisconsin

Northern States Power Company- Wisconsin

5th Largest Combination Electric and Gas Utility (based on customers)

5th Largest Combination Electric and Gas Utility (based on customers)

Xcel Energy Wind Farms

PSCo – 1234MW

SPS – 680MW (874MW in BA)

NSP-MN – 1285MW

Xcel Energy – 3393MW

Xcel Energy manages output from 3.4GW (2623 turbines) of wind energy across the three operating companies (NSP-MN, PSCo, SPS)

Obligations and Resources Today

Net Load – Load less variable output generation.

Traditional Utility Paradigm

Mid-Merit

Baseload

Peakers Renewables

PSCo - April 26, 2010 to May 2, 2010

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

1 25 49 73 97 121 145

MW

Loads and Resources Today

Loads and Resources are forecast every working day for operational planning purposes. This is called day-ahead commitment.

PSCo - April 26, 2010 to May 2, 2010 Actual – Thick Line; Forecast – Thin Line

0

500

1000

1500

2000

25003000

3500

4000

4500

5000

1 25 49 73 97 121 145

MW

Forecasts of loads, variable energy generation, and unit availability are inputs to the operational planning process. Typically 18 to 42 hours ahead, but as much as five days.

Wind

Loads

Loads and Resources Today

Uncertainty is driven by the wind portfolio.

PSCo - April 26, 2010 to May 2, 2010

Mean Absolute Error

Load = 86MW

Wind = 245MW

Net Load = 240MW

-600

-400

-200

0

200

400

600

800

1000

1 25 49 73 97 121 145

MW

Load ErrorWind Error

Net Load Error

Loads and Resources Today and Tomorrow

0

5001000

1500

20002500

3000

3500

40004500

5000

1 25 49 73 97 121 145

MW

PSCo - April 26, 2010 to May 2, 2010 -1000

-500

0

500

1000

1500

1 25 49 73 97 121 145

MW

Load ErrorWind Error

Net Load Error

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

1 25 49 73 97 121 145

MW

-1000

-500

0

500

1000

1500

1 25 49 73 97 121 145

MW

Load ErrorWind Error

Net Load Error

2010 2013

Wind as a Percent of LoadPSCO Wind as a Percentage of Obligation Load (1/1/2007 thru 4/30/2010)

4/13/2010 HE3 (38.5%)

Oct 1, 2009 HE2 (36.9%)Mar 23, 2009 HE3 (35.4%)

0%

5%

10%

15%

20%

25%

30%

35%

40%

1/1/

2007

2/1/

2007

3/1/

2007

4/1/

2007

5/1/

2007

6/1/

2007

7/1/

2007

8/1/

2007

9/1/

2007

10/1

/200

7

11/1

/200

7

12/1

/200

7

1/1/

2008

2/1/

2008

3/1/

2008

4/1/

2008

5/1/

2008

6/1/

2008

7/1/

2008

8/1/

2008

9/1/

2008

10/1

/200

8

11/1

/200

8

12/1

/200

8

1/1/

2009

2/1/

2009

3/1/

2009

4/1/

2009

5/1/

2009

6/1/

2009

7/1/

2009

8/1/

2009

9/1/

2009

10/1

/200

9

11/1

/200

9

12/1

/200

9

1/1/

2010

2/1/

2010

3/1/

2010

4/1/

2010

5/1/

2010

Hourly Monthly Annual

In real-time, more wind means fewer dispatchable resources are on-line to manage the variability. Those resources that remain on-line are typically cheap and inflexible baseload facilities

Who’s in Charge of Uncertainty? Balancing Authorities have final responsibility regarding

managing uncertainty Market structure determines the forecast timeframe of most

significance for market participants Uncertainty costs ultimately flow to customers in the form of

increased fuel costs, O&M, starts, and market volatility. Xcel Energy Operating Company & Balancing Authority

NSP – MISO w/ day-ahead and real-time markets SPS – SPS w/ SPP real-time market PSCo – PSCo w/ WECC hour-ahead market gates

Wind Prediction System

Goal: Create a new state-of-the-art wind forecasting system covering the entire Xcel Energy service territoryMinimize Dispatch Costs Increase Reliability

This is primarily a function of reducing uncertainty associated with short and long-term wind energy production

NSP Benefits

$-

$1

$2

$3

$4

$5

$6

$7

$8

$9

$10

Janu

ary

Febr

uary

Mar

ch

April

May

June

July

Augu

st

Sept

embe

r

Oct

ober

Nov

embe

r

Dec

embe

r

$Mill

ions

2009 2010

Opportunity Cost is a measure of real-time market exposure. Reducing forecast error reduces exposure to real-time price deviations.

∑∑ −−=∆j

RTij

i

DAijijij LMPLMPDARTR )(*)(

Opportunity Cost

PSCo BenefitsFebruary - Dispatch Cost due to Wind Uncertainty

y = 256957xR2 = 0.3274

$(50,000)

$-

$50,000

$100,000

$150,000

$200,000

0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 30.0% 35.0% 40.0%

Daily Normalized Mean Absolute Error

Dai

ly C

ost

0.0%

5.0%

10.0%

15.0%

20.0%

25.0%

Janu

ary

Febr

uary

Mar

ch

Apr

il

May

June

July

Aug

ust

Sep

tem

ber

Oct

ober

Nov

embe

r

Dec

embe

r

200820092010

In PSCo, savings are due to more optimal dispatch of resources. This saves fuel through more efficient loading of units, leaner commitment, and avoided wind curtailments.

The rate of savings is a function of natural gas price…

$5.50/mmbtu = $800k/%MAE

Currently reducing error by 4% over 2009

~$3.2M annualized

Mean Absolute Error30-Day Rolling nMAE (May 11, 2010)

0

2

4

6

8

10

12

14

16

18

20

1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70

Forecast Hours

PSCoNSPSPS