1 Open University

Integrating Renewables Conference

24 January 2006

Wind power on the grid…What happens when the wind stops

blowing?

David Milborrow

david.milborrow@btinternet.com

2

Scope

•MUST examine electricity networks Problems with intermittent generation sources,

e.g. nuclear, gas, cross-channel link, etc

•Behaviour of wind plant

•Assimilating wind into networks Issues and costs Storage Capacity credit

•Denmark and Germany: Lessons to learn?

3

Electricity systems

4

Why integrated systems?

•Smoothing of Demands and Generation sources

•Peak/average ratios House: 15 UK: 1.5

•Lower plant margins needed - House: at least 2*peak Large electricity system: ~1.2* peak

•All leads to LEAST COSTS

5 Firm power is a concept ONLY

1. UK-France link 5 trips, Jan-Jun 2005,

"cause unknown"° 24 Jan, 03:37, 13:51,

9 May, 06:56, etc Outage times cover

maintenance+faults Source: UCTE

2. Typical utility (ERCOT) forced outages: ~2%, + ~7% planned outages

Nov 04Dec

Jan 05Feb

MarApr

MayJune

0

2,000

4,000

6,000

8,000

10,000

12,000

14,000

Outage time, minutes

6 Reserves in a power system

Pumped storage is used as reserveAll can cope with demand increase or decreaseVoltage reductions may be used in emergencies

Type Amount (UK) % of max demand

Inertia Small

Frequency response

~ 900 MW 1.5

Spinning reserve ~1000 MW 1.6

Standing reserve ~800 MW 1.4

Demand-side management

Varies

7

Wind characteristics

8 Smoothing makes a difference

• Wind turbines smooth wind variations

• Wind farms smooth them more

• Wind farms over the country smooth them even more!

• We now have data from Denmark that illustrates this

0 10 20 30 400

500

1,000

1,500

2,000

2,500

Time in hours

Wind power output

1 farm(kW)

West DK(MW)

9 Smoothing of power swings

-100 -50 0 50 100

0.01

0.1

1

10

Change, % rated capacity

Time, %

1 farm

WesternDenmark

Time interval: 1 hour

10 Running electricity systems

•Managing electricity systems is all about managing risks

•All estimates of uncertainty come with a range of probabilities, and

•Uncertainty margins do not add arithmetically – a “sum of squares” law applies

•So the extra impacts of wind are small

11

Costing the effects of wind

•Scheduling error with wind enables extra reserve capacity needs to be estimated

•Establish cost of extra reserve, based on Reduced efficiency of part-loaded plant Cost of plant, or, Market rates

•System operators do not care what the reserve is – as long as it can increase or reduce output when asked

12

Extra back-up capacity

0 10 20 30 40 500

2

4

6

8

10

Wind capacity/peak demand, %

Back-up capacity/wind capacity, %

Ireland(Doherty)

NRELPersistence

Perfect

US (BPA)

AuthorUpper

Lower

13

Extra costs for backup

0 2 4 6 8 10 120

1

2

3

4

5

Wind Energy penetration, %

Cost of extra balancing, $/MWh

NGC

Ilex

PacifiCorp

BPA/Max

/Min

EPRI/Xcel

GRE

MN

14

Capacity credits

The “Firm power” issue

+ + =

?

15 Capacity credits depend on:-

•Amount of wind on system

•Wind speeds

•Wind turbine types

•Winds at time of peak demand

•Utility operating procedures

When “normalised” for differences in wind speed, good agreement between most estimates for northern Europe

16

Capacity credits - UK

0 5 10 15 20 25 300.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Wind energy contribution, %

Capacity credit/annual capacity factor

NGC

SCAR

CEGB

17 5 days in the life of west Denmark (January 2001)

0 20 40 60 80 100 120500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

Hours

Demand, MW

Demand

Demand-wind

• Even with 2400 MW of wind, demand variations still predominate•Wind reduces net demand at peak times

18

Storage

19

Storage

•"Renewables need storage" ? Rather misleading!

•Only the intermittent sources

•"Storage can transform the economics of the intermittent renewables" ?

•Only if they are very low cost!

•Most studies conclude that economics must be studied separately; may be useful to system, or as reserve

20

Storage - problems

• With dedicated storage: How do you size the store? Can you be sure it will not "overflow" Or run out during calms

• Very difficult to get best value from a store UNLESS USED FOR BENEFIT OF SYSTEM

• Even then Value>cost? is the acid test US DoE and UK SYSTEM cost targets ~ £500/kW more if paid for ancillary services

• Economics of isolated systems are site-specific, so dedicated storage may be worthwhile

21

Total extra cost impacts

Offshore/onshore split: equal; offshore cost 50% higher35% capacity factorsCarbon dioxide: €/24/tonne (Oct 2005)

0.4 0.5 0.6 0.7 0.8 0.9 1700

800

900

1,000

1,100

1,200

1,300

1,400

Gas price, €/therm

Onshore wind cost, €/kW

10% windExtra cost to consumers

Lower cost to consumers

Source: Windpower Monthly, January 2006; method, Power UK 109

22 Lessons from Denmark and Germany

•West Denmark (20% wind) “Irrelevant”, say critics…… but links with Sweden, Germany and Norway

are finite, so “effective” penetration about 10%

and S.O. could manage with 100% wind –without the links, or storage

•Germany (~6% wind)but wind speeds lower than UK; so higher balancing costs, low capacity credit…………

23 Wind integration – conclusions

•Thermal power sources (and consumers) will determine bulk of reserve costs for many years to come

•Capacity credits? – Yes, roughly=average power; declines as wind % increases

•Problem areas? May be preferable, once wind input exceeds

~10%, to curtail wind output on a few occasions

•…..but wind will NEVER impose “jolts” on the system comparable with loss of a circuit of cross-channel link, or a 1320 MW nuke

24

Thank you!

The End