EUROPEAN PATTERN RECOGNITIONRENEWABLE ENERGY IMPACTSeptember 2016

WITH FUNDING FROM:PART OF:

THE EUROPEAN PATTERN RECOGNITION PROJECT HAS RECEIVED FUNDINGFROM THE SWEDISH ENERGY AGENCY, THE RESEARCH COUNCIL OF NORWAYAND THE SCIENTIFIC AND TECHNOLOGICAL RESEARCH COUNCIL OF TURKEY(TÜBITAK) IN THE FRAMEWORK OF THE JOINT PROGRAMMINGINITIATIVE ERA-NET SMART GRIDS PLUS, WITH SUPPORT FROM THE EUROPEANUNION’S HORIZON 2020 RESEARCH AND INNOVATION PROGRAMME.

• Unique system solutions for measurement of energy and power quality

• Enabling optimized action to save energy & reduce cost and environmental impact

• Class A measurement instruments & easy to use system software

• Flexible through 100% in-house development

• Access to world-class PQ competence such as Borås University

• Proven products – over 10 years in the market

• Based in Gothenburg, Sweden

METRUMwww.metrum.seMANAGING PARTNER OF THE EPR PROJECT

WITH FUNDING FROM:PART OF:

EXAMPLES OF TODAY’S APPLICATIONS

PATTERN RECOGNITION IS A BOOMINGFIELD WITHIN ARTIFICIAL INTELLIGENCE

”So why not apply it on the electrical grid?”

ARTIFICIAL INTELLIGENCE

MACHINE LEARNING

PATTERNRECOGNITION

WITH FUNDING FROM:PART OF:

CLASSIFIED OBJECTS(TRAINING DATA)

PATTERN RECOGNITIONWORKS IN 4-STEPS

• V, I• Harmonics• RMS values• Peak values• RoCoF• …

a0a1a2a3a4

an

SUPPORT VECTORMACHINE

a0a1a2a3a4

an

BUSH!

DATAACQUISITION PRE-CONDITIONING FEATURE EXTRACTION ANALYSIS

SUPPORT VECTORMACHINE

3-PHASEFAULT

CLASSIFIED OBJECTS(TRAINING DATA)

WITH FUNDING FROM:PART OF:

PATTERN RECOGNITION IS WELL-SUITEDFOR POWER GRID ANALYSIS• Many useful patterns in PQ and energy data• Algorithms handle large data quantities• Large potential to provide:

• New insights and ideas• Solutions to important problems• Optimized network planning & operation• Pro-active maintenance• Framework for future applications• …

WITH FUNDING FROM:PART OF:

EPR BUILDS ON PROMISING RESULTSFROM EARLIER STUDIESPattern Recognition applied on 1642 disturbances from 2network operators.

One-phase fault Phase-to-phase fault Phase-to-phase-to-earth fault

Three-phase fault Transformer start

à Accuracy: ~98 %

WITH FUNDING FROM:PART OF:

DEVELOP, TEST AND VALIDATE METHODS FORAPPLYING PATTERN RECOGNITION TO6 APPLICATION AREAS

USING PATTERN RECOGNITION AS A TOOLFOR GRID OPERATORS AND GENERATORS TO• OPTIMISE THE POWER SYSTEM• ENABLE HIGHER SHARE OF RENEWABLE

ENERGY GENERATION IN THE SYSTEM

PROJECT OBJECTIVE:

PROJECT GOAL:

WITH FUNDING FROM:PART OF:

6 APPLICATIONS FOR PATTERN RECOGNITION =6 WORK PACKAGES

PRO-ACTIVEMAINTENANCE

POWER CONSUMPTIONSTUDIES

LOCAL ENERGY SYSTEMS -

MICROGRIDS

POWER REGULATION BYVOLTAGE CONTROL HOSTING CAPACITY INERTIA SUPPORT

LEAD: METRUMDEMO:

ENERJISA& MÄLARENERGI

LEAD: EMBRIQDEMO:

FALBYGDENS ENERGI

LEAD: GLAVA E.C.DEMO:

GLAVA E.C.

LEAD: MÄLARENERGIDEMO:

MÄLARENERGI

LEAD: ENERJISADEMO:

ENERJISA

LEAD: STRICASE STUDY:

MÄLARENERGI

WITH FUNDING FROM:PART OF:

FACTS & FIGURES

• Partners from industry, research & academic worlds

• >25 project team members

• 4 Demonstrators, 5 locations in Sweden & Turkey• >30 PQ instruments @ 3 distribution companies, 1 solar test site &1 wind farm• Solar PV control system, micro-grid control system, energy data management system• Data from customer & other PQ meters• Data server

• Budget: 1.6 Million EUR

• March 2016 – February 2018

8 main partner companies from 3 countries

WITH FUNDING FROM:PART OF:

HAVING THE RIGHT COMPETENCE ANDKEY END-USERS IN THE PROJECTTECHNOLOGY PROVIDERS

COMPETENCE & TESTING FACILITIES PROVIDERS

END USERS: UTILITIES / DSOs / TSOsValidate goals, deliverables & createend-user acceptence:

Reference groups / advisory board:

… (more to be added)

WITH FUNDING FROM:PART OF:

PRO-ACTIVE MAINTENANCE

DEMO:EnerjisaMälarenergi

à Give earlywarning signals

à Enable pro-activemaintenance ofelectrical grids’components

à Decreaseinterruption costs

Analyse trendsinPower QualitywithPatternRecognition

LEAD: METRUM

WITH FUNDING FROM:PART OF:

POWER CONSUMPTION STUDIES

à Increase powergrids’effectiveness

à Reduce costs forDSOs to upstreamnetworks

à Increase capacity

à Increase flexibility

§ Use P.R. tocategorize events,consumers anddescribe loads’voltagedependency

§ Combine PQ andenergy metering

§ Develop a searchengine for DSOsand others toidentify unwantedevents

DEMO:Falbygdens Energi

LEAD: EMBRIQ

WITH FUNDING FROM:PART OF:

LOCAL ENERGY SYSTEMS - MICROGRIDSà Enable preventive

maintenance,optimise design ofmicro-grids

à Increase solargeneration

à Provide systemservices toDSOs/TSOs

à Increase powergrids’ capacity andflexibility

§ Use P.R. to analysePQ, generation &weather data to:

§ Monitor, detect &predict errors insolar PVgeneration

§ Identify faults inmicro-grids’components

§ Use advanced IT tomanage “smartpower components”in micro-grids

DEMO:Glava Energy Center /Arvika kommun

LEAD: GLAVA E.C.

WITH FUNDING FROM:PART OF:

POWER REGULATION BY VOLTAGE CONTROL

à Optimise & lowerDSOs subscribedcapacity toupstream networkand avoid penalties

à Enable congestionmgmt & peakshaving, increasecapacity

à Enable peakshaving on nationallevel

§ Using PatternRecognition toestimate thevoltagedependency ofloads

§ Use simulationsto estimate thelevel of impactof voltageregulation onloads

DEMO:Mälarenergi

LEAD: MÄLARENERGI

WITH FUNDING FROM:PART OF:

HOSTING CAPACITY

à Provide a toolfor networkplanning

à Increasecapacity forrenewableintegration

§ Develop“hostingcapacity”method§ Investigate KPIs

that limitrenewables’integration§ Identify &

quantifybottlenecks§ Case study

CASE STUDY:Mälarenergi

LEAD: STRI

WITH FUNDING FROM:PART OF:

INERTIA SUPPORT

à Verify TSOsestimation of inertia

à Enable inertiasupport from windfarms based on P.R.technology

à Analyse the impacton inertia fromchanging of theloads

§ Use PatternRecognition toanalysechanges infrequency to§ Assess the level

of inertia in apower system§ Trigger signals

to a wind farmfor inertiasupport

DEMO:Enerjisa

LEAD: ENERJISA

WITH FUNDING FROM:PART OF:

2018

CLEAR MILESTONES,CLEAR DELIVERABLES

PROJECTPLANNING

LITERATURE STUDIES

DATAANALYSIS

SOLUTION DEVELOPMENT & TESTING

1

20172016

DEMOINSTALLATION

CONCLUSIONS &RECOMMENDATIONS

2 3 4

FINAL REPORTING

5 6

WITH FUNDING FROM:PART OF:

DISSEMINATION & KNOWLEDGE SHARING:KEY PART OF THE PROJECT• Active communication

• Seminars, Articles• (Social) media

• Creating a knowledge database• Power quality and other energy data samples• Including object/feature classification• Basis for further research in pattern recognition

• Find us on ERA-Net’s SG+ EXPERA community & working groups:

http://www.smartgridsplus.eu/

• Linkedin: https://www.linkedin.com/groups/8541881

WITH FUNDING FROM:PART OF:

THE EPR PROJECT WILL BRING:ü A platform for development of pattern recognition applications

ü Applications for DSOs, TSOs, industry and energy companies

ü Help tackle present & future challenges (e.g. distributed generation, powerelectronics, electrical vehicles):

1. Keeping STABILITY2. Increasing CAPACITY3. Increasing FLEXIBILITY4. Decreasing COSTS FOR INTERRUPTIONS

(improved power quality and security of supply)

ü Active contribution to the European knowledge community

ü A basis for research on pattern recognition technology in power systems

WITH FUNDING FROM:PART OF:

CONTACT USwww.europeanpatternrecognition.euwww.metrum.se

“3RD GENERATION METERING”

Metrum Sweden AB

Anders Personsgatan 16

416 64 Göteborg, Sweden

Robert OlofssonCTO Metrumrobert.olofsson@metrum.se+46 (0) 70 777 8540

Michiel van AsseldonkEPR Project Managermichiel.vanasseldonk@poyry.com+46 (0) 70 259 6344

WITH FUNDING FROM:PART OF:

THANKS FOR YOUR ATTENTION !!

QUESTIONS ??