B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Review of BPA Voltage Control Review of BPA Voltage Control ConferenceConference

Steve Enyeart, Dmitry Kosterev, Terry Oliver,

Eric Heredia, Bart McManus and Steve Hitchens

Bonneville Power Administration

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

BPA Voltage Control ConferenceBPA Voltage Control Conference Held on August 23, 2011 in Portland OR

About 100 participants from Western, Eastern Interconnection and ERCOT

Presentations and Next Steps are posted at

www.bpa.gov/ti

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 3

Presentation SubjectPresentation Subject The presentation focus is on control needs for “stability” time

frame – i.e. making sure that the power system survives and performs during the power system disturbances

Specifically, the presentation will focus on:

– Voltage stability and controls

– Monitoring and modeling

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 4

Many of BPA Paths are Stability-LimitedMany of BPA Paths are Stability-Limited

California – Oregon Intertie

Pacific HVDC Intertie

Montana - Northwest

Cross-Cascades North

Cross-Cascades South

Wind

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 5

August 10 1996 OutageAugust 10 1996 Outage

425

450

475

500

525

550

575

15:41:00.00 15:42:00.00 15:43:00.00 15:44:00.00 15:45:00.00 15:46:00.00 15:47:00.00 15:48:00.00 15:49:00.00 15:50:00.00 15:51:00.00

Vo

ltag

e [k

V]

Time

Slatt Voltage

Keeler-Allston outage

McNary tripping begins

COI sepration

Oscillations

“Those who forget the past are condemned to repeat it”

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 6

August 10 1996 OutageAugust 10 1996 Outage Outage

– A combination of voltage collapse and unstable oscillations resulted in COI separation and PDCI shutdown

– 28,000 MW of load lost, 7.5 million customers

Lessons learned:

– There are interactions among several transmission paths in Pacific Northwest, on-line generation and load levels

– Dynamic reactive reserves in lower Columbia are critical to support the high power transfers and this is where much of the wind is located

– Operating in POWER FACTOR instead of VOLTAGE control is highly detrimental

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 7

Impact of August 10 1996 OutageImpact of August 10 1996 Outage Interties were de-rated:

– COI and PDCI was de-rated by 33% Significant investments were made by BPA to improve voltage

stability in the Pacific Northwest– Synchronous condenser capabilities at two Lower Columbia plants– Shunt capacitor additions– AC Reactive RAS and Fast AC Reactive Insertion– PDCI control modification

Operating procedures– COI / PDCI Operating Nomogram– Dynamic reserve monitoring for generators in Lower Columbia– Equivalent unit monitoring for voltage swing support

Bakeoven series capacitors are added in 2011 to increase COI utilization

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 8

Voltage ControlVoltage Control Voltage control goes beyond maintaining plant

voltages during power ramps

– This is given

Voltage control helps the system to survive “macro” disturbances

– the system ability to survive disturbance sets the System Operating Limits

– at times wind will be the predominant generation line, and we depend on them for voltage control

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 9

Early Wind Generation (pre-2006)Early Wind Generation (pre-2006)

Voltage control was not required because amount of wind generation was small

Wind turbine technology was induction generators (type 1 or type 2)

Unfortunately, several events of voltage instability were observed with these generators

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 10

Voltage Control ProblemVoltage Control Problem

Pow

erVo

ltage

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 11

Wind Ramp EventWind Ramp EventThis is actual data, not simulationsAll lines in serviceWind ramping up eventWind power plants are in power factor control mode

Power goes upVoltage goes downRecipe for disaster

0 50 100 150 200 250 300234

236

238

240

242

244

Time (min)

Vol

tage

(kV

)

Voltages

0 50 100 150 200 250 3000

100

200

300

400

500Wiind Total Power

Time (min)

Pow

er (

MW

)

0 50 100 150 200 250 300-20

-10

0

10

20Wiind Total Reactive Power

Time (min)

Rea

ctiv

e (M

VA

R)

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 12

Wind Ramp EventWind Ramp EventPV-Curve for same event

0 50 100 150 200 250 300 350 400 450233

234

235

236

237

238

239

240

241

242

243PV Plot - Voltage versus Wind Total

Power (MW)

Vol

tage

(kV

)

The edge

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 13

2006 Study…2006 Study… How integrating 3,000 MW of wind generation will

impact the system “macro” stability Conclusion 1: The Interties will be de-rated if the wind power

plants do not provide dynamic voltage control– Voltage stability

– Damping of inter-area oscillations

– Transient stability

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 14

……2006 Study2006 Study Conclusion 2:

To integrate reliably 3,000 MW of wind

– Wind power plants have adequate reactive capabilities

– Wind power plants are operated in voltage control mode

New wind generation technologies are capable of voltage control, and can offer performance better than synchronous machines. But they must be set and remain in proper control mode.

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 15

Voltage Control RequirementsVoltage Control Requirements BPA requires wind power plants to provide voltage

control and reactive power to support the grid – just like other large generation projects

How much reactive is needed: – Dynamic reactive sized to provide continuous +/- 0.95 power

factor at 34.5-kV bus – Switched shunts to compensate for reactive power losses

between WTGs and Point Of Interconnection, high switching duty

How to control reactive resources:– Operate in continuous (no dead-band !) voltage control mode,

control POI voltage to BPA schedule with reactive droop– Shunt switching to maximize the availability of dynamic

capabilities at the plant

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 16

Observed ChallengesObserved Challenges A plant is commissioned in high side voltage control

Plant experiences large MVAR swings for small voltage fluctuations

Power plant operator turns plant into MVAR control

BPA is working with large wind power plant operators on correcting their voltage control responses

How to test the voltage control response ?

– Trust but verify

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 17

Performance MonitoringPerformance Monitoring BPA installs Phasor Measurement Units (PMUs) at plant POIs

– PMUs provide time-synchronized measurements 60 times per second versus 2-second SCADA – often compared to MRI vs. X-ray technology

WECC is investing more than $108 M in the PMU technology

Applications include:

– Performance monitoring, including voltage response verification

– Model Validation

– Detection of operational issues (e.g. voltage flicker)

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 18

2011 Study2011 Study How integrating 8,000 MW of wind generation

will impact the system “macro” stability

Studies done up to now has not shown “macro” stability issues with 8,000 MW of wind in Pacific Northwest - as long as wind power plants follow BPA voltage control requirements and study models are correct

… Will reality agree with the studies ?

B O N N E V I L L E P O W E R A D M I N I S T R A T I O N

Slide 19

BPA Voltage Control ConferenceBPA Voltage Control Conference Wind generation technologies (type 3 and 4) are

certainly capable of delivering the required voltage control performance

BPA is working with the wind power plant operators on settings wind plant controls

BPA is following up on the Next Steps

www.bpa.gov/ti