Accounting for uncertainties and big data

in future power system studies

Accounting for uncertainties and big data

in future power system studiesProf. Jovica V. Milanović

Dipl.Ing., MSc, PhD, DSc, CEng, FIET, FIEEE

Deputy Head of School & Head of Electrical Energy and Power Systems Group

Manchester, M13 9PL, United Kingdom

8th April 2014, National Technical University of Athens, Athens, Greece

School of Electrical & Electronic Engineering

1/40

Overview of the presentation

• Where I am coming from• Future network structure and operation• Modelling and control challenges• Examples of current thinking and potential

solutions• Summary

2/40

Birthplace of the Industrial Revolution

MANCHESTERfrom “Cottonpolis” with first industrial park in the worldand first industrialised city in the world in the 19th century to second city in the UK with growing culture, media, music, sporting and transport connections enjoyed by 2.55 million people in early 21st century.

3/38

The University of Manchester• The University of Manchester created on 1st October 2004 by merging to

leading Manchester universities Victoria University of Manchester (est.1851, generally known as University of Manchester) and University of Manchester Institute of Science and Technology (est. 1824, generally known as UMIST)

38,430 students of which 11,345 (30%) postgraduate students11,080 members of staff of which 5,970 academic & research 25 Nobel Laureates (4 working currently at the university) Income of £827m of which £280m (34%) external research funding

Research intensive university & Largest single site UK university

Currently ranked(SJTU):41st in the world8th in Europe5th in the UK

4/40

School of Electrical and Electronic Engineering• First to teach Electrical Engineering (1905)• One of the largest in the UK

~70 academics & 1100 students~ 646 (60%) undergraduate students

(53% overseas)~ 454 (40%) postgraduate students

(80% overseas)• 6 Research groups• 6 Framework Agreements:

– National Grid, Rolls Royce, – National Instruments, Syngenta,– EDF, Electricity North West

Currently ranked (SJTU - Engineering): 37st - world, 4th - Europe, 3rd - UK2008 UK Research Assessment Exercise : 1st or 2nd out of 38 UK EEE Schools in

all criteria5/40

Electrical Energy and Power Systems Group

1. Prof. Ian Cotton

2. Prof. Peter A. Crossley

3. Prof. Jovica V. Milanović (HoG)

4. Prof. Simon M. Rowland

5. Prof. Vladimir Terzija

6. Prof. Zhongdong Wang

7. Dr. Viktor Levi (SL)

8. Dr. Haiyu Li (SL)

9. Dr. Joseph Mutale (SL)

10. Dr. Qiang Liu (L)

11. Dr. Konstantinos Kopsidas (L)

12. Dr. Pierluigi Mancarella (L)

13. Dr. Luis (Nando) Ochoa (L)

14. Dr. Robin Preece (L) – from 1/7/14

6/40

13 Academics (+1 from July 2014)

74 PhD Students22 Research Associates

5 Academic visitors75 (FT) & 15 (PT) MSc Students £13.3 / €16 / $22 M Active Research Grants

Publications statistics (Jan 2011- Jan 2014):4 Book chapters6 PatentsJournal papers:3.3 / per academic / per year

(1.8 in IEEE Transactions) Conference papers:6.9 /per academic / per year

EPSRC / EU,

5.7 M43%

Industry, 7.6 M 57%

7/40

Electrical Energy and Power Systems Group

Prominent Past Members

Staff Member ContributionColin Adamson Pioneered power system developmentRon Allan, FIEEE Power system reliabilityJose Arrillaga, FIEEE HVDCAlfred Brameller Power system analysis, sparsity techniquesDerek Humpage Power system protectionEd Kuffel HV Techniques and gas breakdown phenomenaJohn Reeve, FIEEE HVDCBrian Stott, FIEEE Fast decoupled power flowMartin Wedepohl Transient phenomena and wave propagationNick Jenkins, FIEEE Application of wind energy and renewablesDaniel Kirschen, FIEEE Power system economics

8/40

EEPS Main Research Areas

Power Systems• Wide area monitoring, protection and

control• Power system dynamics & power quality • Integration of low carbon technologies• Multi-energy networks• T&D networks economics & operation

Future Research Focus

• Risk and Reliability Assessment of Future Energy Networks• Advanced Distribution Network Management • Life management of Transmission & Distribution Asset

Power Plant• Condition monitoring• High voltage insulation• Overhead line design• Power Transformers

9/40

Future network structure and operationFuture network structure and operation

10/40

The question

Considering evolving power system withincreased uncertainties and abundance ofmeasurement data, are the tools currently in usefor system modelling and control adequate, andif not, how should we modify them, or what othertools should we be using?

11/40

Evolving Power Network - Future

• Liberalised market• Increased cross-boarder

bulk power transfers to facilitate effectiveness of market mechanisms

P

P

P

P

P

P

PP

P

P

P

• Increased presence of static and active shunt and series compensation

• Increased deployment of FACTS devices in general

• Increased use of HVDClines of both, LCC and predominantly VSCtechnology (in meshed networks and as asuper grid)

12/40

13/40

Increase of power transfer capability through advanced technologies

Wind power transfer from the north see – DENA study

Evolving Power Network - Future

P

P

P

P

P

P

PP

P

P

P

• Proliferation of non-conventional renewable generation – largely stochastic and intermittent(wind, PV, marine) at alllevels and of various sizes

0 100 200 300 4000

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

Wind Direction (degrees)

Prob

abilit

y

Year 2000

0 5 10 15 20 25 300

1

2

3

4

5

6

7

8x 10-3

Wind speed (m/s)

Pro

babi

lity

Year 2000

• Large on-shore and off-shore wind farms

14/40

Evolving Power Network - Future

P

P

P

P

P

P

PP

P

P

P

T&

D

G

G

HV

Cell

G

MV

Cell

MV

Cell

--

-M

V C

ell

G

G

G

G

P

P

P

P

P

P

P

PP

P

P

P

• Bi-directional energy flow

• Different energy carriers

• Small scale (widely dispersed) technologies in DN

• Active distribution networks• New types of loads within

customer premises (PE, LED)

• Electric vehicles (spatial and temporal uncertainty)

• Integrated “intelligent” PE devices

• Integrated ICT & storage

15/40

16/40

PE in Distribution networks

Source: KEMA Consulting Final Report for The Department of Trade and Industry’s New & Renewable Energy Programme: The Contribution to Distribution Network Fault Levels From the Connection of Distributed Generation May 2004

Distribution

Tran

smis

sion

GenSet Wind Photovoltaic Battery

CommercialIndustrial Residential

DistributionSubstation

SmartController

RegionalDispatch/

Aggregation

CentralGenerating

StationsNuclear, Coal and Gas

FuelCell

Storage

Power Electronics InterfaceDistribution

Tran

smis

sion

GenSet Wind Photovoltaic Battery

CommercialIndustrial Residential

DistributionSubstation

SmartController

RegionalDispatch/

Aggregation

CentralGenerating

StationsNuclear, Coal and Gas

FuelCell

Storage

Power Electronics Interface

Fully Active Distribution Network

17/40

Communications in power systems

CC – Control CentreSS – SubstationVPP – Virtual Power PlantReg - Regional

Adopted from presentations at • Bi-directional info flow

Power & Information Flow

Sources of uncertainties

• Network– topology, parameters & settings (e.g., tap settings, temperature dependent line ratings) – observability & controlability

• Generation – pattern (size, output of generators, types and location of generators, i.e., conventional,

renewable, storage)– parameters (conventional and renewable generation and storage)

• Load– time and spatial variation in load, load composition, models and parameters,

• Controls– parameters of generator controllers (AVRs, Governors, PSSs, PE interface), network

controllers (secondary voltage controller), FACTS devices and HVDC line controllers

• Contractual power flow (consequence of different market mechanisms and price)

• Faults (type, location, duration, frequency, distribution, impedance)

• Communications (noise, time delays and loss of signals)

18/40

WhatModelling and Control Challenges this new “environment” will result in?

WhatModelling and Control Challenges this new “environment” will result in?

19/40

• Efficient use and reliance on existing and global monitoring data(WAMS) for state estimation, static and dynamic equivalents and control (including real time control)– Efficient data management (signal processing, aggregation, transmission) and ICT

network reliability are essential for both static and dynamic observability– Optimal placement of monitoring devices (PMU) may not be an issue due

to perceived high deployment of those

Challenges - 1

• Modelling requirements for steady state & dynamic studies – large interconnected networks with mixed generation, FACTS and

short/long distance bulk power transfers using HVDC lines – clusters of RES (generation and storage) of the same or different type– LV and MV distribution network cell (DNC) with thousands of RES– demand, including new types of energy efficient and PE controlled loads,

customer participation and behavioural patterns, EV, etc.

20/40

• Design of supplementary controllers based on WAMS to control and stabilise large system (including real-time) or parts of it(which may vary) with uncertain power transfers and load models and stochastically varying and intermittent generation, demand and storage – stochastic/probabilistic control

• Design of new control systems/structure (hierarchical, adaptive, close to real time) for power networks with fully integrated sensing, ICT technologies and protection systems – risk limiting control

• Modelling/analysis of efficient and effective integration of different energy carriers into self sufficient energy module/cell

Challenges - 2

21/40

Sources of “Big Data” in Power System

• SCADA (Supervisory Control And Data Acquisition) systems• WAMS (Wide Area Monitoring Systems)

• Advanced metering devices (“Intelligent/Smart” meters)• Bi-directional communication enabled mobile (e.g. EVs) and

stationary devices (e.g. domestic appliances)• PQ monitoring• Customer surveys• Internet resources (related to network and generation

performance and customer behaviour)• ....

22/40

Data...Questions that need to be asked?• What do we want to learn from collected data and why (for what

purpose)?• How will this facilitate “better” network control and

operation?• Do we know what data we already have and what to do with

them?• Do we already have all the data that we need?• What “additional data” do we need?

• Where and how should we collect the data? Do we have achoice?

• What should be the “sampling” rate and the “quality “ ofadditional data?

• Do we know what techniques to use to efficiently andeffectively process new (and old) data?

Have we answered any of these questions yet, or at least have we

asked them?

If not, isn’t it the time to start doing so?

23/40

Examples of current thinking and potential solutions

Examples of current thinking and potential solutions

24/40

Dynamic equivalent models of Wind farm using probabilistic clustering

1 2 3 4 5 6 7 8 9 100

0.05

0.1

0.15

0.2

0.25

0.3

0.35

No. of equivalent turbines

Pro

babi

lity

Probability of equivalent turbinesWind speed variation inside a wind farm at 15 m/s, 322o

Support Vector Machine

based clustering

25/40

Dynamic equivalent models of Wind farm using probabilistic clustering

0 1 2 3 4 5 6 7 8 9 100

0.2

0.4

0.6

0.8

1

Time (sec)

Act

ive

Pow

er (p

.u)

3 4

0.3

0.35

0.4

Detail modelAggregate model

0 1 2 3 4 5 6 7 8 9 10-5

0

5

10

15

20

25

30

35

40

Time (sec)

Rea

ctiv

e Po

wer

(p.u

)

2 3 40

2

4

6

8

Detail modelAggregate model

P and Q response for Detailed and Probabilistic model at wind speed = 10 m/s, wind direction = 100°

In the case studied, simulation time was reduced by up to 96%.

26/40

• Full converter-connected generator in parallel with a composite load model.

• 7th order dynamic equivalent model of an ADN• P and Q are the outputs, and the V and f the inputs to the model.

Active distribution network models

27/40

Active distribution network models

1 1.5 2 2.5 3 3.5

-5

0

5

10

15

20

activ

e po

wer

(MW

)

Simulated +20% of power (DIgSILENT)median parametersSimulated -20% of power (DIgSILENT)

1 1.5 2 2.5 3 3.50

20

40

60

time (sec)

reac

tive

pow

er(M

Var)

Simulated +20% of power (DIgSILENT)median parametersSimulated -20% of power (DIgSILENT)

P and Q responses at bus B1(grid supply point) for 20% power export/import to/from the grid for the fault at bus 4 (F4) and T1 network composition.

28/40

Active distribution network modelsG2 G3

G1

T2 T3

T1

T4

Load C

Load A Load B

Load GBBus 6

Bus 7 Bus 8

Bus 9

Bus 10 Bus 11

Bus 12

Bus 13

Bus 14

Bus 1-33 kV

P and Q responses for fixed fault

validation with median parameters.

0 1 2 3 4 5-5

0

5

activ

e po

wer

(MW

) TEST 2

grey-box modelfull ADNC model

0 1 2 3 4 5-10

0102030

time (sec)reac

tive

pow

er (M

Var)

grey-box modelfull ADNC model

0 1 2 3 4 5145

150

155

160

165

170

175

180

activ

e po

wer

(MW

)

time (sec)

full ADNC modelgrey-box (median)grey-box (average)

P responses G2.

0 1 2 3 4 580

82

84

86

88

90

92

activ

e po

wer

(MW

)

time (sec)

full ADNC modelgrey-box (median)grey-box (average)

P responses G3.

29/40

Data mining for on-line transient stability assessment (TSA)

98,0%

98,5%

99,0%

99,5%

100,0%

0,02

0,35

0,68

1,02

1,35

1,69

2,02

2,35

2,69

3,02

3,36

3,69

4,02

4,36

4,69

Prob

abili

stic

Pre

dict

ion

Acc

urac

y

Time After Fault is Cleared (s)

For any type and duration of fault at any possible location and any system load level, the Decision Tree predicts the system stability status correctly using 12 input signals from PMUs (rotor angles and speeds) with over 98% accuracy within0.2 s after the clearance of the fault,and it is close to 100% accurate after about 2.5 s.

30/40

Data mining for on line TSA – missing signals, noise, measurement errors

Accuracy of prediction of stable simulations measurement error and noise.

88%

90%

92%

94%

96%

98%

100%0,

02

0,35

0,68

1,02

1,35

1,69

2,02

2,35

2,69

3,02

Acc

urac

y

Time After Fault is Cleared (s)

Case 1

Case 2

Case 3

Case 4

Case 5

Accuracy of prediction of stable simulations with missing signals.

Case The generator from which the signal is missing

1 None2 G43 G94 G11

5 G4, G2, G7, G12, G8, G5, G16, G3, G10, G9, G11

84%

86%

88%

90%

92%

94%

96%

98%

100%

0,02

0,35

0,68

1,02

1,35

1,69

2,02

2,35

2,69

3,02

3,36

3,69

4,02

4,36

4,69

Acc

urac

y

Time After Fault is Cleared (s)

Original Test DatabaseSNR 50 dBSNR 40 dBSNR 30 dB

31/40

PCM can be used to quickly assess the Risk of Instability (RoI). True value = 1.91%; PCM-based value = 1.89%.

Modal estimation with Probabilistic Collocation Method (PCM)

True value

PCM value

32/40

• Large Test System with 2 VSC HVDC lines• Fifty-one uncertain parameters reduced to eight using ranking

Method based on eigenvalue sensitivites• Completed in just 2.44% of the time required for standard

(full-simulation) technique – over forty times faster.

Modal estimation with PCM and reduction of uncertain parameters – large network

33/40

Risk-based Probabilistic Small-disturbance Security Analysis

crit  using deterministic values

distribution of crit  due to uncertainty

reduction in probability of unstable oscillations when considering a forecast power flow P1<P0.

The 95% Probabilistic SecurityMargin is calculated from the trendlines with all lines in service (478MW)and when line 1-30 is out of service(403MW). 75MW is the allowableincrease in NETS NYPS flow beforeP(stab.) < 95%. (“security value” of line1-30)

The deterministic security margin(without uncertainties) is 890 MW withall lines in service. At this pointP(stab.)=58.2% when system variabilityis considered.

34/340

Demand response forecasting Metering Data Customer Surveys General Knowledge of Demand Composition

DDLC Based on Load Classes

Participation of Load Categories in

Different Load Classes

DDLC Based on Load

CategoriesData Processing with Parameter

Uncertainty

Estimation of Dynamic Response of Aggregate Demand at Given Time

Dynamic Signatures/Responses of Different

Load Categories

Field Measurement Lab Measurement Computer Simulations

Data Processing with Parameter

Uncertainty

Probabilistic Dynamic Signatures of Different

Load Categories

Time Dependant Dynamic Response of Aggregate Demand

Probabilistic DDLC

35/40

Demand composition forecasting

36/40

Estimated/predicted/forecasted demand response

Comparison of different a) P and b) Q responses at different times of day (solid line: 03:00; dashed line:

04:00; dash-dot line: 12:00; dotted line: 18:00)

0 0.5 1 1.5 2 2.5 395.5

9696.5

9797.5

9898.5

9999.5100

100.5

t(sec)

P(%

of P

at t

his

hour

)

3:00

Before ShiftingIM+ResistiveResistiveIM

37/40

Reliability of grid supply points vs. Ageing related failure of transformers

ENS of grid supply points for the period 2013 - 2020

Probabilistic study of the relationship between unavailabilityof transformersdue to ageing failureand overall power system reliability.

38/40

Summary Summary

39/40

Future power networks need to be (will likely be) modelled and operated by exploiting possibilities offered by state-of-the-art WAMS, integrated ICT systems and “intelligent” PE devices and using– Non-deterministic & close to real time approaches for (energy)

system control and operation– Stochastic, probabilistic and computer intelligence based

models, data handling and methodologies to minimise the effect of uncertainties and maximise the use of information contained in available data

Summary

The “future” of (electrical)Energy

systems depends/relies onMonitoring, Communications and Control,

i.e., E=MC2

40/40

41/41

Key areas to watch