wind power intermittency - powerworld.com · 2 june 14-15, 2006 2006 power world client conference:...

1

Scott R. Dahman, [email protected]://www.powerworld.com

2001 South First StreetChampaign, Illinois 61820+1 (217) 384.6330

Wind Power Intermittency

2006 PowerWorld Client ConferenceJune 14-15

Chattanooga, Tennessee

2006 Power World Client Conference: Wind Power Intermittency 2June 14-15, 2006

Overview

• Extension of work with Davis Power Consultants and the California Energy Commission

• Prior Analysis– Examine magnitude and frequency of

contingency overloads as a simple measure of system reliability

– Compare base cases to those with different penetrations of distributed renewable resources

2


Overview

• Intermittency Analysis– Integrate the contingency overloads over time

to study the impact of wind intermittency on system reliability

– Use Simulator’s Time Step Simulation (TSS) to model time-varying wind production

– TSS enables collection and analysis of a large amount of data with a modest investment in setup time


Intermittency Challenges

• Wind plants are generally developed… where the wind blows!

• If EHV transmission is in the vicinity, it is usually by coincidence

• Wind production is unpredictable• Transmission network must support a multitude

of wind production patterns across varying:– load conditions – production patterns of thermal and hydro resources

3


Wind Turbine String


Typical Wind “Transmission”

4


Intermittency Challenges

• Wind production is usually highest during off-peak hours– Fast load following resources are off-line– Moderated in California, where even simple

cycle units may serve as base load resources• Voltage regulation

– Var-producing generators are often distant– LTCs and caps must switch frequently


Summary of Prior Analysis

• Identify links between electricity needs in the future and available renewable resources.

• Optimize development and deployment of renewable resources based on their benefits to:– Electricity system– Environment– Local economies

• Develop a research tool that integrates spatial resource characteristics and planning analysis.

5


Objectives

• Investigate the extent to which renewable distributed electricity generation can help address transmission constraints

• Determine performance characteristics for generation, transmission and renewable technology

• Identify locations within system where sufficient renewable generation can effectively address transmission problems


Objectives

• Determine the impact of large-scale distributed projects on grid security.

• We need to:– Identify weak transmission elements and

define metrics that assess system security.– Find locations where new generation would

enhance the security of the grid.– Combine maps of beneficial locations with

maps of energy resources.

6


Normal Operation Example

100 MW

50 MW

280 MW 187 MW

110 MW 40 Mvar

80 MW 30 Mvar

130 MW 40 Mvar

40 MW 20 Mvar

1.0

1.01 pu

1.04 pu1.04 pu

1.04 pu

0.9930 pu1.05 pu

A

MVA

A

MVA

A

MVA

A

MVA

A

MVAA

MVA

A

MVA

A

MVA

67 MW

67 MW

33 MW 32 MW

57 MW 58 MW

21 MW

21 MW

66 MW 65 MW

11 MW

11 MW

23 MW

42 MW

43 MW 28 MW 29 MW

23 MW

23 MW

1

200 MW 0 Mvar

200 MW 0 Mvar

A

MVA

29 MW 28 MW

OneThree

Fo

Two

Five

Six Seven

23 MW

87%

A

MVA

82%

A

MVA

System does not have normal operation thermal violations


Contingency Example

100 MW

50 MW

280 MW 188 MW

110 MW 40 Mvar

80 MW 30 Mvar

130 MW 40 Mvar

40 MW 20 Mvar

1.00 pu

1.01 pu

1.04 pu1.04 pu

1.04 pu

0.9675 pu1.05 pu

A

MVA

A

MVA

A

MVA

A

MVAA

MVA

A

MVA 45 MW

45 MW

55 MW 53 MW

0 MW 0 MW

58 MW

56 MW

52 MW 51 MW

26 MW

25 MW

43 MW

36 MW

37 MW 24 MW 25 MW

30 MW

30 MW

150 MW

200 MW 0 Mvar

200 MW 0 Mvar

A

MVA

25 MW 24 MW

OneThree

Four

Two

Five

Six Seven

44 MW

83%A

MV A

83%A

MV A

95%A

MV A

156%A

MVA

Suppose there is a fault and this line is disconnected

Planning Solutions:New line to bus 3

OR New generation

at bus 3

Then this line getsoverloaded

(is a weak element)This is a serious problem for the

system

7


Contingency Analysis

• Security is determined by the ability of the system to withstand equipment failure.

• Define weak elements as those that become overloaded during contingency conditions (congestion).

• Perform N-1 contingency analysis simulation.• Apply ranking to help prioritize transmission

planning: Aggregate MW Contingency Overload (AMWCO)


Results Organized by Lines, then Contingencies

Sum each value-100 to find the Aggregate Percentage ContingencyOverload (APCO)

Then multiplyby limit to getthe Aggregate MW ContingencyOverload (AMWCO)

8


100 M W

50 M W

280 M W 187 M W

110 M W 40 M v ar

80 M W 30 M v ar

130 M W 40 M v ar

40 M W 20 M v ar

1 .00 p u

1.01 pu

1.04 pu1.04 pu

1.04 pu

0.9930 p u1 .05 pu

A

MV A

A

MVA

A

MVA

A

MV A

A

MV AA

MV A

A

MVA

A

MV A

67 M W

67 M W

33 M W 32 M W

57 M W 58 M W

21 M W

21 M W

66 M W 65 M W

11 M W

11 M W

23 M W

42 M W

43 M W 28 M W 29 M W

23 M W

23 M W

150 M W

200 M W 0 M v ar

200 M W 0 M v ar

A

MVA

29 M W 28 M W

O neTh re e

Fo ur

Tw o

Fiv e

S ix Se v e n

23 M W

8 7 %A

M V A

8 2 %A

M V A

28211470

AMWCO

Weak Element Visualization


Weak Element Visualization

9


Intermittency Analysis:Time-Step Simulation

• 2100 MW wind capacity across five regions• Historic hourly wind production capacity factors are

known for each region • Apply known capacity factors to existing and planned

capacity to create simulation inputs• Match production patterns to characteristic power flow

cases– Apply summer peak pattern to summer peak case, winter peak

pattern to winter peak case, etc.– Hold load constant within each case

• Follow the wind production with natural gas resources


Intermittency Analysis:Time-Step Simulation

• Run contingency analysis at each step• Store results for each line and transformer at each

hourly time step– AMWCO for contingencies– % of limit used for each base case (non-contingent

operation)• Extend AMWCO metric over time: Aggregate

MWh Contingency Overload• Examine variability of results• Identify potentially vulnerable transmission paths

10


Wind Generation in California

San GorgonioPass: 500 MW

Altamont Pass: 600 MW

Solano County: 230 MW Pacheco Pass:

13 MW

Tehachapi: 800 MW (additional 4000 MW planned)Note: capacities shown are from 2005

power flow case used in simulation


Input Data:Hourly Gen MW for each Wind Generator

336 time points representing summer peak hours

Values based on historical regional production capacity factor, applied to simulated regional capacity

11


Input Data:Follow Wind with Natural Gas

AGC set to YES for Natural Gas units, NO for others


Contingency Analysis

• Set Each Hourly Field (Run CTGs) to YES• Time Saving Measures

– Filter the N-1 contingency list • Include only contingencies that cause overloads in

the base case (before applying any hourly scenario)

• Reduced number from 5000+ to 195– Use linearized lossless DC calculation

12


Results Grid

Contingency analysis shown in process


System AMWCO Duration

-

5,000

10,000

15,000

20,000

25,000

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

% Hours AMWCO Exceeded

AM

WC

O

Median Hour AMWCO = 16,494

A few hours experience unusually high level of overloads; rest of curve is fairly flat

13


Hourly Line Loadings

Line Timepoint Custom Results Variables

24402 (ANTELOPE) TO 24466 (TAP 81) CKT 1 Used % of Limit 25642 (TAP819) TO 25625 (TAP809) CKT 1 Used % of Limit35202 (USWP-WKR) TO 33776 (SOUTH BY) CKT 2 Used % of Limit

DateTime8/28/20048/21/20048/14/20048/7/20047/31/20047/24/20047/17/20047/10/2004

Valu

es

110

105

100

95

90

85

80

75

70

65

60

55

50

45

40

35

30

25

20

15

10

These three lines show very different loading patterns:• Red: high average

loading, low variability

• Green: low avg loading, high variability

• Blue: low avg loading, low variability overall, but heavily impacted during certain conditions


Line Loading Duration Curve

Transmission Line Loading Duration

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

110%

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

% Hours Loading Equals or Exceeds L%

Line

Loa

ding

L (%

of M

VA L

imit)

24402 (ANTELOPE) TO 24466 (TAP 81) CKT 1

25642 (TAP819) TO 25625 (TAP809) CKT 1

35202 (USWP-WKR) TO 33776 (SOUTH BY) CKT 2

Shown are the same three lines from prior slide

14


Aggregate MWh Contingency Overload (AMWhCO)

Red indicates lines most severely overloaded over time


Overload Variability

• Shows AMWhCO relative to a reference (median hour)

• Many overloaded lines experience consistent levels at nearly every hour – wind intermittency is not a big factor

• Red indicates lines most adversely affected by intermittency

• Green indicates lines favorably affected by intermittency

15


Time-Step Simulation

• TIP: It is helpful to use an external database for managing TSS inputs and results– Perform calculations to determine inputs– Paste inputs into TSS grid– Query and filter based on statistical criteria– Create time-integrated graphs– Paste time-integrated results into Simulator’s

custom float fields for contouring


Database:Manage Inputs and Results

Query can be transferred to Excel, then pasted directly into Simulator TSS grid

16


Conclusions

• As integration of wind power on the grid increases, intermittency issues will become more significant

• Simulator’s TSS offers a robust and efficient environment for studying time-scale phenomenon such as wind intermittency– Easy to set up, especially for a modest number of

input variables– Many results can be stored and accessed within the

Simulator TSB file without having to catalog multiple power flow cases


Conclusions

• An external database may be very helpful for preparing inputs, screening results, and integrating results over time

• Some plots are available in Simulator for viewing inputs and results

• Complex plots may be readily produced from a good external database

wind power intermittency - powerworld.com · 2 june 14-15, 2006 2006 power world client conference:...

Documents