Wind and Solar Integration into the Bulk Power System

uwig.org

Wind and Solar Integration into the Bulk Power System

Wind and Solar Integration

into the Bulk Power System

Solar and Wind Interconnection

Workshop

Michael Milligan and Debbie Lew

Feb. 22, 2012

NREL is a national laboratory of the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, operated by the Alliance for Sustainable Energy, LLC.


Frequently Asked Questions

2


Source Article in IEEE PES Magazine

•Michael Milligan, NREL

•Kevin Porter, Exeter

Associates

•Edgar DeMeo,

Renewable Energy

Consulting Services

•Paul Denholm, NREL

•Hannele Holttinen,

VTT Technical Research

Center, Finland and

chair of IEA Task 25:

Large-Scale Integration

•Brendan Kirby,

Consultant, NREL

•Nicholas Miller, GE

Energy

•Andrew Mills,

Lawrence Berkeley

Laboratory

•Mark O’Malley,

University College,

Dublin, Ireland

•Matthew Scheurger,

Energy Systems

Consulting

•Lennart Soder, Royal

Institute of Technology,

Stockholm,Sweden

3


4

Where do the Answers Come From?

• Extensive scientific and engineering

analysis

• International Energy Agency Task 25

Report: Design and operation of

power systems with large amounts of

wind power State of the art report.

o http://www.vtt.fi/inf/pdf/tiedotteet/2009

/T2493.pdf

• Utility Wind Integration Group

o http://www.uwig.org

• NREL Transmission and Grid

Integration Group

o http://www.nrel.gov/publications


5

Questions addressed

1) Can grid operators deal with the

continually changing output of wind

generation?

2) Does wind have capacity credit?

3) Does the wind stop blowing everywhere

at the same time?

4) How well can you predict the wind?

5) Is it expensive to integrate wind?


6

Questions addressed

6) Doesn’t wind power need new transmission,

and won’t that make wind expensive?

7) Does wind/solar power need back-up

generation? Isn’t more fossil fuel burned with

wind/solar than without, due to back-up

requirements?

8) Does wind/solar need storage?

9) Isn’t all the existing flexibility already used up?

10) Is wind power is as good as coal or nuclear

even though the capacity factor of wind power

is so much less?

11) Is there a limit to how much wind/solar can be

integrated?


7

1) Can grid operators deal with the

continually changing output of wind

generation?


Geographic diversity reduces sharp ramps

Output Normalized to Mean

8

15 Turbines Stdev = 1.21, Stdev/Mean = .184

200 Turbines Stdev = 14.89, Stdev/Mean = .126

215 Turbines Stdev = 15.63, Stdev/Mean = .125

1.6

1.4

1.2

1.0

0.8

0.6

1.6

1.4

1.2

1.0

0.8

0.6

1.6

1.4

1.2

1.0

0.8

0.6

0

5

10

15

Seconds

20

25

30x10 3

Source: NREL Wind Plant Data

(Approximately 8 hours)


MW

We can manage high instantaneous penetrations

9

West Denmark January 10-16, 2005

4000

Wind

3500

Load

3000

2500

2000

1500

1000

500

0

1 25 49 73 97 121 145

hour

Source: Soder, Royal Institute

of Technology, Sweden. Steve

Beuning, Xcel, 2011


2) Does wind have capacity credit?

10


Wind typically provides modest capacity credit

EWITS, 2010

IEEE Task Force Paper, IEEE Transactions on Power Systems, Vol. 26, No. 2, May 2011

NERC Integrating Variable Generation Task Force Report 1.2

11


12

3) Does the wind stop blowing

everywhere at the same time?


13

Geographic diversity reduces sharp ramps

Source: ERCOT, WindLogics

Ela, E.; Kirby, B. (2008). ERCOT Event on

February 26, 2008: Lessons Learned. 13 pp.;

NREL Report No. TP-500-43373.

http://www.nrel.gov/docs/fy08osti/43373.pdf

www.osei.noaa.gov


4) How well can we predict the wind?

14


15

Wind forecasts are improving

• Easier to predict

wind for short time

steps: ~5% hour

ahead

• More difficult day

ahead: ~8-15%

• Relative forecast

errors are reduced

for large geographic

footprints

Source: John Zack, AWS, UWIG 2011


5) Is it expensive to integrate wind?

16


It is very difficult to calculate integration costs, but if you

try:

Year

Study

Wind

Capacity

Penetration

Regulation

Integration Cost ($/MWh)

Load

Following

Unit

Commit.

Gas

Supply TOTAL

2003 Xcel-UWIG 3.5% 0 0.41 1.44 - 1.85

2003 We Energies 29% 1.02 0.15 1.75 - 2.92

2004 Xcel-MNDOC 15% 0.23 - 4.37 - 4.60

2005 PacifiCorp-2004 11% 0 1.48 3.16 - 4.64

2006 Calif. (multi-year)* 4% 0.45 trace trace - 0.45

2006 Xcel-PSCo 15% 0.20 - 3.32 1.45 4.97

2006 MN-MISO** 31% - - - - 4.41

2007 Puget Sound Energy 12% - - - - 6.94

2007 Arizona Pub. Service 15% 0.37 2.65 1.06 - 4.08

2007 Avista Utilities 30% 1.43 4.40 3.00 - 8.84

2007 Idaho Power 20% - - - - 7.92

2007 PacifiCorp-2007 18% - 1.10 4.00 - 5.10

2008 Xcel-PSCo*** 20% - - - - 8.56

2009 Bonneville (BPA) + 36% 0.22 1.14 - - 5.70

2010 EWITS ++ 48% - - 1.61 - 4.54

2010 Nebraska +++ 63% - - - - 1.75

Source: * Regulation 2009 costs Wind represent Technologies 3-year Market average. Report, DOE/GO-102010-3107, Aug 2010; Milligan, M.; et al,

(2011).

** Highest

Cost-Causation

over 3-year evaluation

and Integration

period.

Cost Analysis for Variable Generation, NREL TP-5500-51860

*** This integration cost reflects a $10/MMBtu natural gas price scenario. This cost is much higher than the

17


18

6) Doesn’t wind need new transmission

and won’t that make wind expensive?


Transmission provides a lot of system benefits and is

not a large portion of consumers’ bills

19

• Transmission is needed for most new

generation sources

• Joint Coordinated System Plan found

benefit/cost ratio of 1.7/1 for

transmission that would support a 20%

wind energy penetration. Transmission

was 2% of the wholesale energy cost.

• Consumers often will benefit by lower

energy costs

• Transmission build-out can reduce the

need for new generation

• A new resource adequacy model at NREL

shows that significant installed

generation in the West would not be

needed if there were a robust

transmission system and coordinated

operations


20

7) Does wind/solar need back-up

generation?

Isn’t more fossil fuel burned with

wind/solar than without,

due to back-up requirements?


System Load (MW)

21

1-to-1 Backup is not needed

• Individual generators are not

backed up: but reserves are

provided on a system basis

Wind and solar need dynamic

reserve, not constant

Wind displaces generation,

freeing some generation to

provide up-reserves

• Generators that change

dispatch as a result of wind

may have reduced efficiency,

but total fuel burn and

emissions will decrease

seconds to minutes

Regulation

Source: Western Wind and Solar Integration Study, 2010, www.nrel.gov/wwsis

Time (hour of day)

0 4 8 12 16 20 24

tens of minutes to hours

Load

Following

Days

Unit

Commitment

day

Scheduling


8) Do wind and solar need storage?

22


Storage is only one of the mitigation options for wind/solar

• Storage is always

useful, but may not

be economic

• Simulations of

system operations

up to 35%

penetration find no

need for storage

• Experience with

more than 40,000

MW of wind in the

US shows no need

for storage

Source: DOE

23


24

9) Isn’t all the existing flexibility

already used up?


Excess load following capability often exists

1

Kirby & Milligan, 2005 Methodology for Examining Control Area Ramping Capabilities with Implications for Wind

http://www.nrel.gov/docs/fy05osti/38153.pdf

2

Kirby & Milligan, 2008 Facilitating Wind Development: The Importance of Electric Industry Structure.

http://www.nrel.gov/docs/fy08osti/43251.pdf

3

Milligan & Kirby 2007, Impact of Balancing Areas Size, Obligation Sharing, and Ramping Capability on Wind Integration .

http://www.nrel.gov/docs/fy07osti/41809.pdf

25


26

Planning for the future

• Additional sources of flexibility may be needed

at high penetration rates

• Newer types of generation: CTs, reciprocating engines

• As inflexible generation retires, replace with more

flexible generation

• Institutional flexibility

• Fast, large energy markets

• Sub-hourly scheduling protocols with neighboring

balancing areas

• Demand response

• Plug-in hybrid electric vehicles in the future


Schedule and dispatch speed matters

Average Total Regulation (MW)

27

10000

9000

Average Total Regulation for 6 Dispatch/Lead

Schedules by Aggegation (Dispatch interval -

Forecast lead time)

Faster Faster Faster

8000

7000

6000

5000

4000

3000

2000

10-10

30-10

30-30

60-10

30-40

60-40

1000

0

Footprint Regional BAU

Large Medium Small

Milligan, Kirby, King, Beuning (2011), The Impact of Alternative Dispatch Intervals on Operating

Reserve Requirements for Variable Generation. Presented at 10th International Workshop on Large-

Scale Integration of Wind (and Solar) Power into Power Systems, Aarhus, Denmark. October


28

10) Is wind as good as coal/nuclear

even though

wind’s capacity factor is much smaller?


2008 $/MWh

29

Wind and solar are principally energy resources

90

80

70

Wind project sample includes

projects built from 1998-2008

60

50

40

30

20

10

0

Nationwide Wholesale Power Price Range (for a flat block of power)

Cumulative Capacity-Weighted Average Wind Power Price (+/- 1 standard deviation)

2003 2004 2005 2006 2007 2008

49 projects 62 projects 80 projects 98 projects 117 projects 145 projects

2,268 MW 3,069 MW 4,083 MW 5,165 MW 7,654 MW 9,873 MW

Average Cumulative Wind and Wholesale Power Prices Over Time. Source: Wiser, Ryan and Mark

Bolinger. Annual Report on U.S. Wind Energy Markets: 2008. U.S. Department of Energy,

http://www1.eere.energy.gov/windandhydro/pdfs/46026.pdf.


Units with low capacity factors play an important role

in system operations

30

Midwest ISO Plant Capacity Factor by Fuel Type

(June 2005–May 2006)


31

11) Is there a limit to how much

wind/solar can be integrated?


Energy penetration [%]

Physical limits have not been found

25

Wind Energy Penetration in 2010

20

15

10

5

0

Denmark Portugal Spain Ireland Germany

(2009)

Source: Holttinen, 2011 and EIA 2011

US

(2009)

• Studies to date in the U.S. have not identified a physical

limit, up to 35% energy penetration

• However, changes in standard operational and planning

techniques may need to change

32


Western Wind and Solar Integration Study

33


Can we integrate 35%

wind and solar

in the West?

Goal - To assess the operating

impacts and economics of wind and

solar on the WestConnect grid.

• How do local resources compare to remote, higher quality resources via long

distance transmission?

• Can balancing area cooperation help manage variability?

• Do we need more reserves?

• Do we need more storage?

• How does geographic diversity help?

• What is the value of forecasting?


35

Scenario penetration levels

Name WestConnect Rest of WECC

Wind Solar Wind Solar

10% 10% 1% 10% 1%

20% 20% 3% 10% 1%

20/20% 20% 3% 20% 3%

High

renewables

30% 30% 5% 20% 3%

Penetration levels are by energy (MWh), not capacity (MW).

Solar is 70% concentrating solar power with thermal storage and 30% rooftop

photovoltaics.


Geographic

Scenarios

(high renewables case)

•In-Area - each state meets target from

sources within that state.

•Mega Project - concentrated projects in

best resource areas.

•Local Priority - Balance of best resource

and In-Area sites.


37

What do we model?

• Model the year 2017 three times

• Use historical load and weather patterns from

2004, 2005, 2006 (need correlated

load/wind/sun data!)

• Statistical analysis of variability. Focus on

extreme events.

Power simulations of all of WECC on hourly

basis and down to 1 minute for extreme events.

• Need high resolution (in time and space) wind

and solar data


How does the system operate

with 35% wind/solar?

38

38


How does the system operate in the high

renewables case?

39

Mid-July

Load

The operator formerly

managed to load but now

has to manage the net

load.

Net Load

= Load - Wind - Solar


How does the system operate in the high

renewables case?

Mid-July

Mid-April

Mid-April shows the challenges of

operating the grid with 35% wind and solar.

This was the worst week of the 3 years studied.

40


41

Operations during mid-April

No Wind/Solar

High renewables case


42

What are the benefits

of 35% wind and solar?


High renewables case saves 40% in fuel and

emissions costs

43

40%, or $80

per MWh of

wind/solar

Note: Assumes $30/ton CO2 tax.


44

High renewables case reduces CO2 by 25-45%

At a $9.50/MBTU gas

price, wind/solar

displace gas, resulting

in modest emissions

reductions.

15% reduction

25%

5%


45

High renewables case reduces CO2 by 25-45%

At a $9.50/MBTU gas

price, wind/solar

displace gas, resulting

in modest emissions

reductions.

30% reduction

50%

45%

At a $3.50/MBTU gas

price, coal is

displaced, resulting in

high emissions

reductions.


High renewables case reduces CO2 by 25-45%

At a $9.50/MBTU gas

price, wind/solar

displace gas, resulting

in modest emissions

reductions.

30% reduction

50%

45%

At a $3.50/MBTU gas

price, coal is

displaced, resulting in

high emissions

reductions.

That’s like taking 35 million cars off the road!

46


What do you need to make 35% wind and

solar work?

47

47


You need Balancing Area Cooperation

48


You need subhourly scheduling

Regulation duty with hourly

scheduling

Decreased regulation duty when

5 minute redispatch allowed

49


50

You need subhourly scheduling

Hourly scheduling puts more stress on the fast regulating reserves than the wind and solar

variability does.


51

You need forecasting

Forecasting saves up to 14% in annual operating costs. If

forecasts were perfect, an additional 1-2% could be saved.


You need demand response

(or more expensive measures like more reserves)

52

Wind Forecast Error Distribution


You need demand response (or more expensive measures)

• Using existing rules for contingency reserves, there were

occasional reserve shortfalls

• Increase spinning reserves for every hour of the year.

• Add storage like pumped hydro storage.

• We only have shortfalls for 89 hours of the year (1%), so these

options can be expensive.

• Demand response (paying

loads to turn off) was found

to be effective and was less

expensive.

53


A new 100 MW PHS plant does not recover costs

This assumes storage has perfect knowledge of spot prices.

We have only evaluated hourly price arbitrage value of

storage. We have not evaluated use of storage to provide

reserves.

54


55

It is operationally feasible for WestConnect to

accommodate 30% wind and 5% solar if:

• Substantially increase balancing area cooperation

• Increase use of subhourly scheduling

• Increase utilization of transmission.

• Enable coordinated commitment and dispatch over wider regions.

• Use forecasts in operations.

• Increase flexibility of dispatchable generation.

• Commit additional operating

reserves as appropriate.

• Implement/expand demand

response programs.

• Require wind to provide

down reserves.

Source: DOE


56

Next Steps

Phase 2 of WWSIS:

• Capture true wear and tear cost of cycling

and ramping coal/gas plants and include

those costs in operations optimization

• Capture emissions impacts from cycling and

ramping of coal/gas plants and refine avoided

emissions

• Compare impact of wind vs solar

• Better representation of existing operational

practices


57

Wear and Tear Cost Data

• Data results from analysis of APTECH on 400 coal and

gas plants:

• Costs of hot, warm and cold starts

• Costs of ramping down to minimum output

• Forced outrage rate impacts

• Long term heat rate degradation

• Costs for 7 types of plants

• Coal – small subcritical

• Coal – large subcritical

• Coal – super critical

• Gas – combined cycle

• Gas – large frame combustion turbine

• Gas – aeroderivative combustion turbine

• Gas – steam

• Plus ‘best in class’ units


58

APTECH analysis completed

Cold Start Cost Lower Bounds-Includes Outliers

(Maintenance and capital cost per MW capacity)

400

300

200

100

0

1: Coal - Small Sub Critical

2: Coal - Large Sub Critical

3: Coal - Super Critical

4: Gas - CC [GT+HRSG+ST]

5: Gas - Large Frame CT

6: Gas - Aero Derivative CT

7: Gas - Steam

Source Steve Lefton, Intertek-APTECH.


Increased Costs

($/MWh renewable)

Ceiling on wear and tear costs

• Apply APTECH upper

and lower bounds of

wear and tear costs to

WWSIS Phase 1 results

for ceiling on wear and

tear impacts

• Results show that

cycling costs are up to

$2 per MWH of

renewables produced,

or a reduction in value

of the renewables of up

to 2.4%

2.00

1.80

1.60

1.40

1.20

1.00

0.80

0.60

0.40

0.20

0.00

Increased Cycling Costs

I10R I20R I2020R I30R

Scenario

Lower Bound

Upper Bound

59


EPA Continuous Emission Monitors (CEMs)

• Use measured emissions from each fossil fuel

plant in the U.S.

• Determine for each plant:

• Heat rate (and CO 2 emissions) as a function of

generation

• Emissions (NO X , SO 2 ) as a function of generation

• Eliminate units with obviously clustered data

60


61

Preliminary Results - Emissions Analysis

Applying these new emissions data to WWSIS phase

1 results only reduces emissions benefits of

renewables by 2.3%. Most of that reduction is due to

start-ups not part-load


62

Thank you!

Debbie Lew

+1 (303) 384-7037

debra.lew@nrel.gov

http://www.nrel.gov/wwsis


Appendix

• How much wind energy is installed in the U.S.

today?

• What really happened in ERCOT?

• Key outcomes: large-scale wind integration

studies in the U.S.

• High-penetration (instantaneous) in actual

systems

• Proposed Energy Imbalance Market (EIM) in

the West

63


MW

64

Net Load = load – wind – solar

16x10 3

Steeper ramps

Lower turn-down

14

12

10

8

6

4

Load

Net Load

Wind

2

0

1200

1220

1240

1260

1280

Hours

1300

1320

1340


65

How much wind energy is installed in the

U.S. today?


Wind in the US Today

66


From LBL 2010 Wind Technologies Market Report (2011 report to be released)

67


What really happened in ERCOT?

68


69

What happened in TX, Feb 2008?

• Event was

forecasted but

ERCOT’s

forecasting system

had not yet been

integrated into

system operations.

• Source: ERCOT


70

What happened in TX, Feb 2008?

• A frequency drop resulted in a call upon reserves including

loads that had voluntarily signed up for a Load-Acting-As-

Resource program in which they are paid to curtail.

• There were 3 major contributors:

o

o

o

Wind generation dropped from 2000 to 360 MW in 3.5 hours

A conventional unit with 370 MW capacity tripped offline

The load forecast was wrong

• The wind ramp (500 MW/hour) is not the same as a

reliability event with instantaneous loss of gen or

transmission.

o

o

Conventional contingencies require dedicated, expensive spinning

reserve and immediate response.

Multi-hour wind ramps allow for use of less expensive resources., such

as a 15-minute non-spinning reserve product


71

Lessons Learned From Texas Event

• Load forecast failed to

predict the large ramp

up in demand

• The accurate wind

energy forecast was not

used in scheduling (this

has been rectified)

• LAaR was very effective

in economically

reducing demand

demand response

Wind event was a ramp

event, not a

contingency event

(similar to 2007 event)

Ela, E.; Kirby, B. (2008). ERCOT Event on

February 26, 2008: Lessons Learned. 13 pp.;

NREL Report No. TP-500-43373.

http://www.nrel.gov/docs/fy08osti/43373.pdf


72

Large-Scale Wind Integration Studies

• Sponsored by US DOE, managed by NREL

• Eastern Wind Integration and Transmission

Study, released Jan 20, 2010.

www.nrel.gov/ewits

• Western Wind and Solar Integration Study,

released in Mar 2010. www.nrel.gov/wwsis

• These studies show that up to 30% (and 5% solar

in the west) can be integrated reliability and

economically if operational practices can

provide additional flexibility thru institutional

changes


MW

Wind Penetration

Wind as a Percent of Load

• Public Service of

Colorado (PSCo)

reached record

levels of wind

penetration in

October 2011

o Hourly Record: 55.6%

October 9, 2011 HE4

o Daily Record: 37.0%

October 8, 2011

7000

6000

5000

4000

3000

2000

1000

0

1 25 49 73 97 121 145

60.0%

HE4 = 55.6%

50.0%

40.0%

30.0%

20.0%

10.0%

0.0%

1 25 49 73 97 121 145

Public Service of Colorado - October 4-10, 2011

73


MW

74

Record Wind Penetration in Ireland

4000

3500

3000

2500

2000

1500

1000

500

0

Penetration over 50% between 2330 and 0600

Sept 6 & 7 th 2011

Time


http://erc.ucd.ie/RealTimeData/chartsirish.html Nov 12/13, 2011

75


EIM

76


Overview: Sumo-wrestler Theory of Integration

77

• Two main characteristics of markets help with

efficient integration of variable generation (VG,

consisting of solar and wind energy) in the bulk

power system

o Big

o Fast

• Proposed Energy Imbalance Market (EIM) in the

West has potential to significantly reduce the

challenge of large-scale solar and wind

integration: it is big and fast

o TGIG involvement: Reserve assessment and

production simulation (Plexos runs Spring 2012)


Reduction in Ramp Demand

(MW/hr)

78

The Big (Net load = load – wind – solar )

Unnecessary ramping of generation can be

minimized with a larger market area

3000

1000

-1000

Net

Load

-3000

1 25 49 73 97 121 145 169

Hours

• This means:

o Overall variability (per unit) of net load declines

o Operating reserve (required generation flexibility)

requirements are reduced, saving money


79

Value of storage depends on the system

Value of storage with current resource mix.

Value of storage with new flexible resource mix.

Source: Andrew Mills, LBNL, UWIG Fall Technical Workshop, 2008


The Fast…

Average Total Regulation (MW)

80

10000

9000

Average Total Regulation for 6 Dispatch/Lead

Schedules by Aggegation (Dispatch interval -

Forecast lead time)

Faster Faster Faster

8000

7000

6000

5000

4000

3000

2000

10-10

30-10

30-30

60-10

30-40

60-40

1000

0

Footprint Regional BAU

Large Medium Small

Milligan, Kirby, King, Beuning (2011), The Impact of Alternative Dispatch Intervals on Operating

Reserve Requirements for Variable Generation. Presented at 10th International Workshop on Large-

Scale Integration of Wind (and Solar) Power into Power Systems, Aarhus, Denmark. October


…And The Proposed Solution in the West

• Energy Imbalance

Market (EIM)

• Key operational

impacts:

o Big: Large electrical

footprint (depending

on participation in

the EIM)

o Fast: Short dispatch

period (5 minutes)

81


Appendix

82


83

NREL/TGIG Analyses

• Partners: Xcel Energy, WIEB, WECC, Plexos

• Flexibility reserves analysis for multiple footprints,

scheduling periods

• Input to WECC’s consultant E3 modeling

• Assistance to individual BAs for BA analysis

• Multiple webinars with various stakeholders

• Support to WECC TAS MWG to incorporate flexibility

reserves into TEPPC cases

• Plexos modeling for Western PUC meeting to be held in

Spring 2012

• Multiple conference papers, reports. Full footprint EIM

cuts flexibility reserve deployment by approximately 50%

compared to business as usual (peer reviewed)

• Initial work funded primarily by Wind/Solar programs; OE

now funding significant effort


84

References

• NREL Technical report: “Market Characteristics for

Efficient Integration of Variable Generation in the

Western Interconnection,”, by M. Milligan and B.

Kirby available at

http://www.nrel.gov/docs/fy10osti/48192.pdf

o (Report was adopted by the Variable Generation

Subcommittee, Market Workgroup of the Western

Electricity Coordinating Council)

• King, J.; Kirby, B.; Milligan, M.; S. Beuning (2011).

Flexibility Reserve Reductions from an Energy

Imbalance Market with High Levels of Wind

Energy in the Western Interconnection. 100 pp.;

NREL Report No. TP-5500-52330