distributed generation & hosting capacity · nserc industrial research chair in bibliography j....

Pow

erS

yste

mS

imu

lati

on

NSERC Industrial Research Chair in

Distributed Generation &Hosting Capacity

Need for stochastic analysis

Tomas Yebra [email protected]

April 24, 2012

University of ManitobaPower System group

[email protected]

http://umanitoba.ca/faculties/engineering/departments/ece/index.html

Pow

erS

yste

mS

imu

lati

on


1 Introduction

2 Hosting Capacity & Power

3 Hosting Capacity & Voltage

4 Hosting Capacity & Protection

5 Conclusions

Pow

erS

yste

mS

imu

lati

on


Table of contents

1 Introduction




5 Conclusions

Pow

erS

yste

mS

imu

lati

on


Goals for Delivering Electric Energy

Power System

PRIMARY GOALS(Objectives)

• Voltage Quality

• Reliability

• TariffsHow is electricity

deliver?

SECONDARY GOALS(Engineering)

• Overload

• Stability

• Operational security

• Current quality

• . . .

Consumers TransmissionSystemOperators

DistributionSystemOperators

Regulator

CommandsEquipment

Deliverselectric energy

PERFORMANCE INDICATORS(INDICES)

Pow

erS

yste

mS

imu

lati

on


Performance IndicatorsIndices

Voltage

1 Overvoltage

2 Undervoltage

3 Voltage distortion

4 Imbalance

5 Flicker

Power

1 Overloded equipment

2 Losses

Current

1 Ampacity in cables

2 Harmonic distortion

Frequency

1 Overfrequency

2 Underfrequency

Reliability

1 Number of Outages

2 Number of voltage sags

3 Number of failures in equipment

Pow

erS

yste

mS

imu

lati

on


Hosting capacity

IRREVERSIBLE DAMAGE

UNACCEPTABLE DETERIORATION

ACCEPTABLE DETERIORATION

NORMAL OPERATION

IMPROVEMENT

Limit of Improvement

Amount of generation (AG)

Per

form

ance

ind

ex(P

Idx)

Definition

The amount of distributed generation for which the performance becomesunacceptable.

Pow

erS

yste

mS

imu

lati

on


Hosting capacity

Definition

The amount of distributedgeneration for which the

performance becomesunacceptable or does not

improve.IRREVERSIBLE DAMAGE

UNACCEPTABLE DETERIORATION

ACCEPTABLE DETERIORATION

NORMAL OPERATION

IMPROVEMENT

Limit of Improvement

Amount of generation (AG)

Per

form

ance

ind

ex(P

Idx)

(PIdx)

(AG)HG

(PIdx)

(AG)HG1 HG2

(PIdx)

(AG)

Pow

erS

yste

mS

imu

lati

on


Disturbance levels

Dis

turb

ance

Lev

el

Per

form

ance

Ind

ex

Immunity

Immunity level

Reference level

Planning level

Emission level

Emission

Load System Operator

SafetyNetwork

SafetyClient

ProbabilityEquip. Failure

Do notCross

Increase

Do notIncrease

ExpensiveEquipment

ExpensiveEquipment

• Reference level: The probability of equipment failure is low. International standards.

• If the Reference level is moved there are no incentives to improve the network.

• ⇑ Planning Level −→ ⇑ Hosting Capacity −→ ⇑ Emissions −→ ⇑ Risk System Operator

Pow

erS

yste

mS

imu

lati

on


Table of contents

1 Introduction




5 Conclusions

Pow

erS

yste

mS

imu

lati

on


OverloadingFeeder

No Distributed generationPconsumed = PLB + PLC

Pdeliverd = PAB + PBC

No overload conditions

Pmaxdeliverd < Pmax

LB + PmaxLC ⇐ Consumers

Pmaxdeliverd < Pmax

AB + PmaxBC ⇐ Ampacity

ExampleA B CPmax

AB = 9MW PmaxBC = 3MW

PmaxLB = 4MW

PminLB = 1MW

PmaxLC = 2MW

PminLC = 1MW

Hosting capacityPdeliverd = Pconsumed − Pgenerated

Pmax1delivered = Pmax

consumed − Pmingenerated ⇐ Same case that no generation

Pmax2delivered = Pmax

generated − Pminconsumed ⇐ Change direction of power flow

Pmaxgenerated = Pmax2

delivered + Pminconsumed

No overload conditions

Pmaxgenerated < Pmax

consumed + Pminconsumed ⇐ 1stHosting capacity (HC1)

Pmaxgenerated < Pmax

feeder + Pminconsumed ⇐ 2stHosting capacity (HC2)

ExampleA PAB B

PLB

≤ 7MW

PGB

CPBC

PLC

4MW

PGC

PHC1AB = (4 + 2) + (1 + 1) = 8MW

PHC1BC = (2) + (1) = 3MW

PHC2AB = (9) + (1 + 1) = 11MW

PHC2BC = (3) + (1) = 4MW

Pow

erS

yste

mS

imu

lati

on


OverloadingImplications

Managing the risk

• Limits are deterministic.

• Minimum consumption is only a few hours per year.

• Maximum production occurs over a fraction of time.

• Probability of reaching the worst case scenerio (maximumproduction, minimum load) is very low.

• Stochastic approach• Measurement: Accurate data of consumer patterns.• Consumer: Acknowledge possible interruptions.• System Operator: Penalties for interruptions.

Pow

erS

yste

mS

imu

lati

on


Table of contents

1 Introduction




5 Conclusions

Pow

erS

yste

mS

imu

lati

on


Voltage regulation

Parameters• Cross-section (A)

• Length, location (l)

• Reactance (α)

• Power factor (k)

• Nominal voltage (U)

Definition

R = ρl

A

α =X

R

k =Q

P

ZL = R + Xj = R(1 + αj)

I =1

V(P + Qj) =

P

V(1 + kj)

GRID

U

ZL = R + Xj

V

∼P,Q

Equations

∆V = |U − V | = |ZLI |

∆V = |R(1 + αj)P

V(1 + kj)|

∆V =PR

V|1− αk + (α + k)j |

∆V

V=

PR

V 2

√(1− αk)2 + (α + k)2

P =δVV 2

R

1√(1− αk)2 + (α + k)2

Pow

erS

yste

mS

imu

lati

on


Voltage regulationCross-section & Length

0 2 4 6 8 10 P Hosting Capacity (MW)

0.9

1.0

1.1

1.2

1.3

1.4

1.5

Volt

age (

p.u

.)

50 mm2

100 mm2

200 mm2

50 100 150 200 250 300 350 400 450 500 S(mm2 )

0

5

10

15

20

25

P H

ost

ing C

apaci

ty (

MW

)


0.9

1.0

1.1

1.2

1.3

1.4

1.5

Volt

age (

p.u

.)

500 m1 km2 km

0 2 4 6 8 10 Position (km)

0

2

4

6

8

10

P H

ost

ing C

apaci

ty (

MW

)

Pow

erS

yste

mS

imu

lati

on


Voltage regulationNominal Voltage & Reactive power

0 5 10 15 20 P Hosting Capacity (MW)

0.8

0.9

1.0

1.1

1.2

1.3

Volt

age (

p.u

.)

1.9 kV3.3 kV5.7 kV

0 2 4 6 8 10 12 Nominal voltage (kV)

0

10

20

30

40

50

60

P H

ost

ing C

apaci

ty (

MW

)

0 5 10 15 20 P Hosting Capacity (MW)

0.8

0.9

1.0

1.1

1.2

1.3

Volt

age (

p.u

.)

XR

=0.2Q

P=0.15

XR

=0.85Q

P=0.4

XR

=1.5Q

P=0.667

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6XR

feeder or QP generator

2.5

3.0

3.5

4.0

4.5

5.0

5.5

6.0

6.5

P H

ost

ing C

apaci

ty (

MW

)

XR

Q

P

Pow

erS

yste

mS

imu

lati

on


Voltage regulationimplications

Factors

• HC ∝ cross-section.

• HC ∝ 1/location.

• HC ∝ voltage level.

• HC ∝ XR

and Q.

Stochastic approach

• The deterministic method:• Overvoltage: Lowest consumption and

maximum production.• Probability of suffering overvoltage or

undervoltage is zero.

• Worst-case scenerio is highly unlikely.

• Voltage regulators decrease HC, andincrease uncertainty.

• Risk analysis. Immunity level is high:Probability of failure is low.

Pow

erS

yste

mS

imu

lati

on


Table of contents

1 Introduction




5 Conclusions

Pow

erS

yste

mS

imu

lati

on


Hosting Capacity & ProtectionMargin of coordination

GRID

Iup

BRK

ItripIdown

∼

Coordination

No generationImindown > Itrip Iup = 0

GenerationImindown > Itrip > Imax

up


0

200

400

600

800

1000

I up, I dow

(A

)

Iup

Idow

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 P Hosting Capacity (MW)

0

200

400

600

800

1000

I margin

(A

)

Pow

erS

yste

mS

imu

lati

on


Hosting Capacity & Protection

0.0 0.5 1.0 1.5 2.0 P Hosting Capacity (MW)

0

200

400

600

800

1000

I margin

(A

)

5 km8 km

2 3 4 5 6 7 8 9 10Length of the feeder (Km)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

P H

ost

ing C

apaci

ty (

MW

)

0.0 0.5 1.0 1.5 2.0 2.5 3.0 P Hosting Capacity (MW)

0

200

400

600

800

1000

I margin

(A

)

1 km4 km

0 1 2 3 4 5Position generation (Km)

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

5.5

P H

ost

ing C

apaci

ty (

MW

)

Pow

erS

yste

mS

imu

lati

on


Hosting Capacity & ProtectionImplications

Summary

• ⇑ Generation =⇒ Uncoordination.

• ⇑ length feeder =⇒ ⇓ HC.

• Generation far from substation =⇒ ⇑ HC .

• Current protection is not a solution when the penetration is high.

Stochastic approach

• Different type of generation =⇒ Different contributions of faultycurrents.

• Generation is not present in the system at all times.

• ⇑ # trips 6=⇒ ⇑ # failures loads.

• Controlled island operation ⇑ reliability.

Pow

erS

yste

mS

imu

lati

on


Table of contents

1 Introduction




5 Conclusions

Pow

erS

yste

mS

imu

lati

on


conclusions

Hosting capacity calculation

• The Hosting capacity concept can be applied to any otherperformance index that can be measured in the network suchas harmonics (THD, TDD,. . . ), imbalance, reliability (voltagesags, outages. . . ). . .

• A deterministic approach to calculate the hosting capacityunderestimates the potential of the current distributionsystem to integrate distributed generation in most ofperformance indecies.

Pow

erS

yste

mS

imu

lati

on


Bibliography

• J. Deuse, S. Grenard, and M.H.J. Bollen.”EU-DEEPintegrated project - technical implications of the“hosting-capacity” of the system for DER”. InternationalJournal of Distributed Energy Resources, 4(1) : 17-34, 2008.

• Bollen, M.H. and Hassan, F. “Integration of DistributedGeneration in the Power System”. Wiley-IEEE Press. 2011.

• Etherden, N. and Bollen, M.H.J.. “ Increasing the hostingcapacity of distribution networks by curtailment of renewableenergy resources”, PowerTech, 2011 IEEE Trondheim, 2011.

Pow

erS

yste

mS

imu

lati

on


The end

Thank you for your attention.

Are there any Questions?

distributed generation & hosting capacity · nserc industrial research chair in bibliography j....

Documents