SETIS

SETIS magazineNo. 13 – November 2016

Energy SystemsModelling

10001001000100110101000101110100111010100101101000001000100010010001001101010001011101

ISSN 2467-382X

2

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

3

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Contents

4. EditorialbyJanNill,PolicyOfficeratDirectorate-GeneralforClimateAction

5. SET-Plan Update

8. DavidConnolly,coordinatoroftheH2020project“HeatRoadmapEurope", talks to SETIS

11. TheEUReferenceScenario2016

15. MarcOliverBettzüge,directoroftheInstituteofEnergyEconomicsattheUniversity ofCologne(ewi),talkstoSETIS

17. Abetterlifewithahealthyplanet:pathwaystonet-zeroemissions

20. AlistairBuckley,co-authorof‘AreviewofenergysystemsmodelsintheUK: Prevalentusageandcategorisation’talkstoSETIS

22. Energysystemmodellingintheindustry:theEDFR&Dperspective

27. METIS:thenewshort-termenergysystemmodelexploredbytheDirectorate-General forEnergy

30. Theimportanceofopendataandsoftwareforenergyresearchandpolicyadvice

34. TheNordicETP2016

37. OSeMOSYS:opensourcesoftwareforenergymodelling

39. Sharedexperiencesinintegratedenergysystemsmodelling

42. MarkO’Malley,directoroftheInternationalInstituteforEnergySystemsIntegration, talks to SETIS

4

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Policy-makersneeduptodateinformation,meaningfulfiguresand

analysisontheimpactofpolicymeasures.Energysystemsmodelling

canprovidethemwithallofthis.ThisismyexperienceasaEuropean

Commissionpolicy-makerwhohasusedenergysystemmodelling

forsevenyears.Atleastthreemodellingchallengesremain:energy

marketchanges,modelcombinationsandtransparency.

Up to date information: Energy system modelling needs to be

basedonthelatesttrendsand,unlikemacromodelling,ithasthe

opportunitytobeinformedbyrecentdata.AnexampleistheEU

ReferenceScenario2016(seeCantonetal.),whichprojectsenergy,

transportandgreenhousegas(GHG)emissiontrendsto2050.

Meaningful figures:Minus40%in2030,60%in2040and80%in

2050.ThisistheEU'sGHGemissionreductionpathway(compared

to1990)derivedbyenergysystemandnon-CO2 emission modelling

fortheCommission'sLow-CarbonEconomyRoadmapin2011.Itisa

goodexampleofhowmodellinghasinformedpolicy-makersinsetting

theGHGtargetfortheEU's2030climateandenergyframework.

Policy impact analysisisthemostdifficulttask.Howtoappropri-

atelyreflectexistingpolicyinstrumentsandtheirinteractions?How

tosimplifytheessenceoffuturepoliciesforpolicyscenarios?Itis

herethatthe“system”componentofenergysystemmodellingis

mostimportant.Forexample,apossiblefuturecarbonpricetrajectory

resultingfromtheinterplayofthelegallydeterminedamountofEU

EmissionTradingSystemallowancesandthechangingconditions

ofenergysupplyanddemandcanonlybegeneratedbyamodel

whichcoversalltheseelements.

Three challenges:First,energymarketschangeprofoundly.Supply

actorshavemultipliedandelectricitymarketdynamicshavechanged

withthepolicy-leddiffusionofrenewables,whileinterconnectionsare

becomingmoreimportant.Thesetrendsaresettocontinueandwill

bereinforcedwiththeriseofenergystorageanddemandresponse.

Thisisachallengeinparticularforenergymodelsofwhichthebasic

structurehasoftenbeendevelopedintimesofpublicmonopoliesor

ofoligopolisticcompetitionoflargesuppliers.Second,interactions

betweentheenergysystemandotherpartsoftheeconomyare

ofincreasingpolicyrelevance.Thedebateonthesustainabilityof

theincreasinguseofbiomassisonlyoneexample.TheEU'sGHG

effort-sharingtargetscouldonlybeproperlyanalysedbycombining

energysystemmodelsandmodelswhichcovertheagriculture,

forestryandwastesectors.Howtobestoperatesuchcombinations

toensurerobustandtimelyanalysesremainsachallenge.Third,

despitesignificant improvements,combiningcomplexmodelling

withtransparencyremainsachallenge,andstakeholders’demands

areincreasinginthisrespect.

MycolleaguesandIlookforwardtoseeinghowexistingandnew

energysystemmodelsaddressthesechallengeswhilecontinuing

toprovidequantitativeinformationtopolicy-makersthatisupto

dateandpolicyrelevant.

By Dr Jan Nill, European Commission, Directorate-General Climate Action

Editorial

Dr Jan Nill works at the European Commission, Directo-

rate-General Climate Action. He has been responsible for

climate policy-related EU energy modelling from 2009 to

2016. Currently he works as policy officer in the unit CLIMA

C.2 Governance & Effort Sharing, mainly on the Effort Sharing

Regulation Proposal on binding annual emission reductions

by Member States from 2021 to 2030 and the monitoring

of energy and greenhouse gas projections. Jan holds a PhD

in economics from the University of Kassel.

5

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Energy System Modelling

•• Somebackgroundontheenergymodellingactivitiescarriedout

attheJointResearchCentre(JRC),theEuropeanCommission’s

scienceandknowledgeservice, isavailableattheEuropean

CommissionScienceHub,togetherwiththeresultingpublications

datingbackto2005.

•• In2013,theJointResearchCentrepublisheda reportonthe

JRC-EU-TIMESmodel:Assessingthelong-termroleoftheSET-

Planenergytechnologies.Themainobjectiveofthisreportwas

topresentthemaininputsandassumptionsusedintheJRC-

EU-TIMESmodel,developedbytwoformer1JRCinstitutes:the

InstituteforProspectiveTechnologicalStudies(IPTS)andthe

InstituteforEnergyandTransport(IET).Themodelisdesigned

toanalysetheroleofenergytechnologiesinmeetingEurope’s

energyandclimatechange-relatedpolicyobjectives.Itmodels

theuptakeanddeploymentoftechnologyanditsinteractionwith

theenergyinfrastructure,includingstorageoptions,inanenergy

systemsperspective.

•• InAugust2014theDirectorate-GeneralforEnergylauncheda

publictenderaimedatdevelopinganewtool,METIS,tomodelthe

Europeanenergysystem,properlycustomisedtotheEuropean

1 TheDGJRCisorganisedinDirectoratesasofJuly2016.

Commissionneeds.METIS isexpectedtoaccuratelysimulate

themainaspectsoftheEuropeanenergysystemandbecali-

bratedwithdatafromthecurrentEUenergysystem(coveringall

28MemberStates).TheEuropeanCommissionwillusethisto

exploreandanalysetheeffectsofdifferentpoliciesandtrends

attheregional,nationalandEuropeanlevelsbyrunningseveral

scenariosfordifferenttimehorizons.Themodellingeffortwill

focusmainlyontheelectricity,gasandheatsectors,bothfor

theshort-termandthemedium-tolong-term.Thecontractwas

awardedinDecember2014toaconsortiumledbyArtelys.

•• TheJRCorganisedanexpertworkshopon“Addressingflexibilityin

energysystemmodels”inDecember2014.Theobjectiveofthe

workshopwastogatherexpertsfrommodellingteamsdealing

withtheseproblemsfromdifferentperspectives,rangingfrom

energysystem-widetodetailedsectoralenergymodels,inorder

toshareandcomparemodellingapproachesandresults,and

identifygapsandpotentialsolutions.

•• Followingtheworkshopon“Addressingflexibilityinenergysys-

temmodels”,in2015theJRCpublishedareportonAddressing

flexibilityinenergysystemmodelsinwhichitsummarisedthe

presentationsandfindingsfromthe2014workshop.

•• Also in 2015, the JRC published the JRC-EU-TIMES report

SET-Plan UpdateThe European Strategic Energy Technology Plan (SET-Plan) aims to transform the way we produce and use energy in the EU, with the goal of achieving EU leadership in the development of technological solutions capable of delivering 2020 and 2050 energy and climate targets.

Energy system models allow us to understand the impact, and thus consider the ‘design', of changes in the energy system. This is increasingly important for an energy system in transition that should absorb increasing levels of intermittency whilst meeting the objectives of security, sustainability and competitiveness and placing the consumer at the centre. The following is a non-exhaustive chronological overview of some selected actions taken to support the development and use of energy system models in EU energy planning, in addition to a more general look at recent actions in support of the SET-Plan.

©iStock/jeka 1984

6

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

BioenergypotentialsforEUandneighbouringcountries.This

reportwasthefirstinaseriesofreportsonlow-carbonenergy

technologiespotentials,andaddressedthequantificationofcur-

rentandfuturebiomasspotentialcontributiontodecarbonisation

pathwaysoftheenergysystem.Thedatasetsproducedare

inputintotheJRC-EU-TIMESmodeltoanalysethemaindrivers

offuturebiomassusewithintheenergysystems.

•• InApril2015,theEuropeanCommissionissuedacallfortenders

foraStudyontheMacroeconomicsofEnergyandClimatePol-

icies.Thismajorproject,awardedtoaconsortiumledbyCam-

bridge EconometricsandincludingE3Modelling and Trinomics

aspartners,iscurrentlyongoingandwillextendthecapability

oftwoglobalenergy-economy-environmentmodelstogivea

fuller impactassessmentofthepoliciesdesignedtopromote

energyefficiencyandthetransitiontoalow-carboneconomy.

Thetwomodelshavebeenchosentorepresenttwoverydiffer-

enttraditionsineconomics:post-Keynesianmacro-econometric

modelling(theE3MEmodel)andComputableGeneralEquilibrium

modelling(GEM-E3).

•• InDecember2015theExecutiveCommitteeoftheEuropean

EnergyResearchAlliance(EERA)agreedtolaunchaJointPro-

gramme on Energy Systems Integration(ESI).Asub-programme

onmodellingaimstodevelopintegratedenergysystemmodels

thatcapturethestrongphysical,economicandregulatoryinter-

actionsthatexistwithinenergysystemsandthatfullyutilise

increasingvolumesofdata.

•• InFebruary2016,theMEDEASprojecthelditskick-offmeeting.

FundedunderHorizon2020,thisprojectaimstouseopensource

softwaretodesignanewenergy-economymodelforthefuture

EUtransitiontoalow-carbonenergysystem.

•• OnJune30,2016theEU’sInnovationandNetworksExecutive

Agency(INEA)organisedaworkshoponEnergySystemModelling

withtheobjectiveofbringingtogetherthefourH2020projects

fundedunderthetopic“LCE21–2015:Modellingandanalysing

theenergysystem,itstransformationandimpacts”toidentify

possiblesynergiesand/oroverlaps.ApartfromMEDEASpro-

jecttheother3awardedprojectsoftheLCE21-callareREEEM,

REFLEX and SET-Nav.

•• InJuly2016,theECpublisheditslatesteditionoftheEURef-

erenceScenario2016,whichprojectsenergy, transportand

greenhousegasemissionstrends intheEUupto2050.The

ReferenceScenarioisaprojectionofwhereourcurrentsetof

policiescoupledwithmarkettrendsarelikelytolead.TheEUhas

setambitiousobjectivesfor2020,2030and2050onclimate

changeandenergy,sotheReferenceScenarioallowspolicy-mak-

erstoanalysethelong-termeconomic,energy,climatechange

andtransportoutlookbasedonthecurrentpolicyframework.

•• AlsoinJuly2016theJRCpublishedanewissueoftheGECO

2016:GlobalEnergyandClimateOutlook.RoadfromParis,which

examinestheeffectsongreenhousegasemissionsandenergy

marketsofareferencescenariowherecurrenttrendscontinue

beyond2020;oftwoscenarioswheretheIntendedNationally

DeterminedContributionshavebeenincluded;andofa2°Csce-

narioinlinewithkeepingglobalwarmingbelowthelimitsagreed

in internationalnegotiations.Thereportpresentsanupdated

versionofthemodellingworksupportedbytheEuropeanCom-

mission’sDirectorate-GeneralforClimateAction(DGCLIMA)in

theUNFCCCnegotiationsthatresultedintheParisAgreement

oftheCOP21inDecember2015.

•• InAugust2016,theJRCpublishedatechnicalreportlayingout

themodellingapproachthat is implementedinthePOTEnCIA

modelling tool (PolicyOrientedTool forEnergyandClimate

ChangeImpactAssessment).Thismodelwasdevelopedbythe

JRC’sformerInstituteforProspectiveTechnologicalStudies(IPTS)

toassesstheimpactsofalternativeenergyandclimatepolicies

ontheenergysector,underdifferenthypothesesaboutsurround-

ingconditionswithintheenergymarkets.

•• InSeptember2016, theJRC,DGRTDandtheUnitedStates

DepartmentofEnergyorganisedanexpertworkshopon“Under-

standingtheWater-EnergyNexus:IntegratedWaterandPower

SystemModelling”,whereapproximately70EuropeanandUSsci-

entistsfromacademia,governmentandindustryinvolvedinpower

systemmodellinggatheredinordertocompareandexchange

state-of-the-artmodellingmethodologiesandbestpractices,

identifyinggapsandpotentialsolutions.Thediscussionstookinto

accountmodellinganddata-relatedmethodologicalaspects,with

theirlimitationsanduncertainties,aswellaspossiblealternatives

tobeimplementedwithinpowersystemmodels.

General SET-Plan related news and activities from JRC/SETIS

•• TheJointResearchCentrepublishedanumberofreportsin2016.

InadditiontothereportscoveredinthelastSET-Planupdate,

theJRChaspublishedareporttitledMapping regional energy

interestsforS3P-Energy,themaingoalofwhichwastocarryout

afirstidentificationofregionswithcommonenergytechnology

interestsaccordingtotheirsmartspecialisationstrategies.

•• OnJuly202016,theCommissionpresentedasetofmeasures

toacceleratetheshifttolow-carbonemissionsinallsectorsof

theeconomyinEurope.ThepackagewillhelpMemberStates

prepareforthefutureandkeepEuropecompetitive.Itispartof

theEU’sstrategyforaresilientEnergyUnionwithaforward-look-

ingclimatepolicy.

•• TheEuropeanParliamentadoptedtheEUStrategyforHeating

and Coolingataplenarysessionon13September2016.The

resolutionrecognisesthehugeuntappedpotentialofusingrecov-

erableheatanddistrictheatingsystemsandthefactthat“50%

7

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

ofthetotalEUheatdemandcanbesuppliedviadistrictheating”.

•• InhisStateoftheUnionAddressinSeptember2016,European

CommissionPresidentJean-ClaudeJunckerhighlightedthat

smarterenergyusecombinedwithambitiousclimateactionis

creatingnewjobsandgrowthinEuropeandisthebestinvest-

mentinEurope’sfutureandinthemodernisationoftheEuropean

economy.

•• InthecontextoftheprocesstowardsaSET-PlanIntegrated

RoadmapandActionPlan,organisations(universities,research

institutes, companies, public institutions and associations)

involvedinresearchandinnovationactivitiesintheenergyfield

are invitedtoregister intheEuropeanenergyR&I landscape

database,whichaimsatfacilitatingpartnershipsandcollabo-

rationacrossEurope.Registrationisopentostakeholdersfrom

theEUandH2020associatedcountries.Organisationsareable

toindicatetheirareaofactivityaccordingtotheenergysystem

challengesandthemes,as identified intheSET-Plan process

towardsanIntegratedRoadmapandActionPlan.Thedatabase

ispubliclyavailableontheSETISwebsite.

•• DuringthelastSET-PlanSteeringGroup meeting in September,

fouragreementsonstrategictargetsandprioritieswereendorsed

bytheSET-PlanSteeringGroupandrelevantstakeholders.The

agreedDeclarationsofIntentconcerntheKeyActions1&2,and

9and10oftheIntegratedSET-PlandedicatedtoEurope“Being

n°1inrenewables”regardingoceananddeepgeothermalenergy,

“Renewingeffortstodemonstratecarboncaptureandstorage

(CCS)intheEUanddevelopingsustainablesolutionsforcarbon

captureanduse(CCU)”and“Maintainingahighlevelofsafety

ofnuclearreactorsandassociatedfuelcyclesduringoperation

anddecommissioning,whileimprovingtheirefficiency”.Themost

recentSteeringGroupmeetingtookplaceinBrusselsOctober19.

•• The9thSET-PlanConference ‘EnergyUnion:towardsatrans-

formed European energy system with the new, integrated

Research,InnovationandCompetitivenessStrategy’ istotake

placeinBratislava,Slovakiaon30November-2December2016.

•• TwoJRC-organisedside-eventsaretobeheldinthemarginsof

theSET-PlanConference.Thefirstisaworkshoptopresentthe

recentfindingsandinputsofSETIStotheStateoftheEnergy

Union reportanditsaddedvaluefortheoverallprogressofEU

innovationintheenergysector.Thisworkshopwillalsopresent

theTechnologyInnovationMonitoring(TIM)tooldevelopedby

theJRC.ThesecondworkshopwilldealwithFundinginnovative

low-carbonenergydemonstrationprojectsinthecontextofthe

NER300programme.

8

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

What are the main insights that we aim to achieve through the development of energy system models?

Energysystemmodelshelpusunderstandtheimpactofmaking

changestotheenergysystembeforewemakethem.Theinsights

tendtovarysignificantlydependingonthetime-horizoninquestion.

Somemodelsfocusontheshort-term(yearsahead)sotheyusually

modelexistingtechnologieswithinexistingfinancialframeworks,

suchasamodelthatanalyseshowtodispatchapowerplanton

theelectricitymarketswehavetoday.Othermodelsfocusmoreon

thelong-term(decadesfromnow),sotheycanprovideinsightsfor

moreradicalchangestothetechnologies,institutions,andmarkets

wehavetoday.Forexample,thesemodelswilloftenincludewave

powerorpower-to-gas,bothofwhicharenotevencommercially

availablerightnow.EnergyPLANisprimarilydesignedtoanalysethe

large-scaleintegrationofrenewableenergyandenergyefficiency,

basedontheSmart Energy Systemsconcept.Renewableenergystill

providesarelativelysmallamountofourenergytoday,soanalysing

‘large-scaleintegration’requiresalong-termperspectiveovermany

decades.EnergyPLANisthereforefocusedonradicalchangesto

ourenergysystemcomparedtotoday,but italsosimulatesthe

energysystemonanhourlybasistoaccountforintermittencyfrom

renewableenergy.

Tell us a little about the EnergyPLAN model and its energy system analysis procedures.

EnergyPLANisprimarilyasimulationmodel,but italso includes

someoptimisation. Iwouldequatethe ‘user’ofEnergyPLANto

a‘designer’:theuserdesignsanenergysysteminEnergyPLANin

termsofdemands,capacities,efficiencies,andcostsandonceitis

complete,theusersimulateshowthatenergysystemperforms.

However,tocarryoutthesimulation,theusermustalso instruct

themodelhowto‘optimise’itsdecisionsduringeachhourofthe

simulation.Inotherwords,theoptimisationtellsthesimulationhow

tomakeitsdecisionsduringeachhouroftheyear.

ThemostcommonoptimisationweuseinEnergyPLANiscalledthe

‘technicaloptimisation’,wherethemainobjectiveistoreducethe

energyconsumedduringthesimulation.Alternatively,theusercan

usean ‘economicoptimisation’wherethemodelwill reducethe

costoftheenergysystemduringthesimulation.Itisimportantto

notethattheoptimisationonlyreferstotheoperationoftheenergy

systemduringeachhourandnottothe‘design’oftheenergysystem.

Inotherwords,thecapacityofwindturbinesinyourenergysystem

willnotbealteredduringtheeconomicoptimisation,buttheway

thosewindturbinesoperateeachhourmaybe.

DavidConnollyCoordinatoroftheH2020project“HeatRoadmapEurope”andoneofthedevelopers oftheEnergyPLANmodel

TALKS TO SETIS

©iStock/artJazz

9

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

How does the EnergyPLAN model compare with other models; what are its distinguishing features?

IwoulddefineEnergyPLAN’snicheinthemixofmodelsthatcurrently

existas:itsimulatesallsectorsoftheenergysystemonanhourly

basisaftertheyhaveundergoneradicalchanges.Itcandothisdue

toacombinationofthefollowingkeycharacteristics:

•• Itcansimulateradicalchangesforrenewableenergyandenergy

efficiency,since itconsidersallmajortechnologiesthatexist

today(includingdistrictheating)aswellastechnologieswhich

arenotcommerciallyavailableyet,suchashydrogenproduction,

biomassgasification,carboncapture,andelectrofuels.

•• Themodelconsiderstheentireenergysystem,includingelec-

tricity,heating,cooling,industry,andtransport,sotheimpactof

changingtheheatsectorisreflectedintheothersectorsalso.

•• EnergyPLANisanhourlymodelsoitensuresthatdemandand

supplyarealwaysmetonhourlybasisacrosstheelectricity,

districtheating,andgasnetworks.

•• Itaccountsforsynergiesacrossallsectorsonanhourlybasis

whenintegratingrenewableenergy,whichisbasedontheSmart

Energy Systemconcept.Itisveryimportanttoconsiderthesesyn-

ergieswhenquantifyingtheimpactoffutureduetotheadditional

flexibilitythatthesesynergiescreateforintermittentrenewables

likewindandsolar.

How does the EnergyPLAN model contribute to the design of energy planning strategies?

Itquantifiestheimpactofimplementinglarge-scalepenetrationsof

renewableenergyandenergyefficiency,usuallyintermsofenergy,

emissions,andcosts.Byquantifyingtheimpact,wecanoftenreveal

thatsomedecisionsaremuchmoreorlesssignificantthanpolicy-

makersrealise.AverygoodexampleofthiscomesfromourHeat

RoadmapEuropework.Initially,policy-makersthoughtthatdistrict

heatingwasveryexpensive,especiallyduetotheconstructionof

thepipesinthestreets.However,byquantifyingthis,wehavebeen

abletodemonstratethatdistrictheatingischeaperthannatural

gasinmanycountries.Evenmoresurprising,duringthiscalculation

wefoundoutthatthepipesinthegroundareoneofthesmallest

costsforadistrictheatingscheme,eventhoughtheyarethemost

visiblesincetheyrequireconstructiononthestreets.This isvery

©iStock/zlikovec

10

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

importantforpolicy-makers:forexample,recentlywewereadvising

a localmunicipalityabouttherolloutofdistrictheating intheir

city.Theywerefocusingonthecostofthepipesinthestreetsince

theyassumedthiswouldhavethemost influenceontheoverall

economicviabilityoftheproject.However,afterquantifyingthe

breakdownofthecostforthem,theycouldseethatthepriceof

theheatsupplyhadamuchbiggerinfluencethanthepriceofthe

pipes,sowerecommendedthattheyfocustheireffortsonsecuring

alowandstableheatsupplyprice.Thisisaveryspecificexample,

butinmoststudiesEnergyPLANchangesperceptionslikethisona

broaderenergy-systemscale.Forexample,ithaspreviouslybeen

usedtodemonstratehow100%renewableenergysystemshave

comparablecoststofossil-fuelbasedenergysystems,whichcan

befoundat:www.SmartEnergySystem.eu.

A key objective of EnergyPLAN is to aid in the design of 100% renewable smart energy systems. How will it achieve this?

Ourresultstodateindicatethatthekeyto100%RenewableEnergy

andtheSmartEnergySystemsconceptisintegratingthevarious

sectors:electricity,heating,cooling,industry,andtransport.Historically

thesesectorshaveevolved individuallyfromoneanother:power

plantsproducingelectricity,boilerscreatingheat,andcombus-

tionenginesprovidingtransport.Weneedtoremovethis‘sectoral

approach’andmovetowardsan‘energysystem’approach,since

thiswillcreatemanynewopportunitiesforbothenergyefficiency

andrenewableenergyintegration.

Let’staketheelectricityandheatsectorsasanexample,sincemany

EUcountrieshavealreadystartedconnectingtheseinrecentdecades.

Ifthesesectorsaredesignedinisolationthenthepowerplantswill

onlyproduceelectricity,but ifthesetwosectorsaredesignedin

combinationwithoneanother,thenitisverylikelythatcombined

heatandpower(CHP)plantswillbemosteconomical.Apowerplant

hasanefficiencyof30-50%forelectricitygeneration,whereasaCHP

planthasanefficiencyof80-90%forelectricityandheatproduction

together.Hence,thereisoftenasignificantimprovementinenergy

efficiencybyreplacingapowerplantwithaCHPplant,something

wequantifiedforfiveEUcountriesintherecentSTRATEGOproject:

thesecountriesareCroatia,CzechRepublic,Italy,Romania,andthe

UnitedKingdom.

Similarly, ifwetrytooptimisetheintegrationofrenewableelec-

tricitywithasolefocusontheelectricitysector,thenwewilllimit

oursolutionstothosethatexistwithintheelectricitysectorsuchas

interconnection,demand-sidemanagement,batteries,andpumped

hydroelectricstorage.However,ifweoptimiseacrosstheelectricity

andheatsectorstogether,thenwewillbeabletousecheaper

alternativesforthe integrationofrenewableelectricitysuchas

heatpumpsandthermalstorage.WealreadyseethisinDenmark,

wherelarge-scaleelectricboilersareintegratingmorewindpower

viathermalonthedistrictheatingnetwork.Thisisoftenacheaper

solutionsincethermalstorageisapproximately100timescheaper

thanelectricitystorage,soweoftenuseEnergyPLANtoquantify

howmuchadditionalwindpowerwecanaccommodateduetothe

connectionbetweentheelectricityandheatsectors.

EnergyPLANalsoconnectscooling,industry,andtransportwiththe

electricityandheatsectorstoidentifysynergiesthatincreaseenergy

efficiencyandrenewableenergy.Byusingthissectoralapproach,

100%renewableenergysystemsbecomemoreeconomicallyviable

andthusmorelikelytobeimplemented.

How does your model accommodate new technologies and new research and development?

Wetrytoreleaseanewversionofthemodelevery6monthson

thewebsite.Updatesareverycloselylinkedtotheresearchprojects

thatweareinvolvedinandexistingtechnologieswithinEnergyPLAN

areregularlyupdatedifweidentifyanewconsiderationinoneof

theseprojects.Newtechnologiestendtobe includedovertime

ratherthanallatonce.Forexample,power-to-gasoriginallybegan

asanadditionalelectricitydemandforhydrogenproduction,butas

welearnedmoreaboutthetechnology, itevolvedintoindividual

componentsintheprocesssuchaselectrolysers,hydrogenstorage,

carboncapture&recycling,andbiomassgasification.

David ConnollyDavid Connolly is an Associate Professor in Energy Planning at Aalborg University in Copen-

hagen, Denmark. His research focuses on the design and assessment of 100% renewable

energy systems, with a key focus on the integration of intermittent renewables (such as

wind and solar power), district heating, electric vehicles, and the production of electrofuels/

synthetic fuels for transport.

11

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Approach

TheEuropeanCommission’spolicydecisionsareunderpinnedby

thoroughanalysesandimpactassessments.Whendevelopingand

implementingtheEnergyUnionStrategy,theCommissionusesa

widerangeofmathematicalmodelsandtoolstoexplorepolicy

proposalsandevaluatetheirpotentialenergy,transport,economic,

socialandenvironmentalconsequences.

TheEUReferenceScenarioisoneoftheEuropeanCommission’skey

analysistoolsusedinthecontextoftheEnergyUnion.Itisupdated

regularlyasitprojectstheimpactofcurrentEUpoliciesonenergy

andtransporttrendsaswellaschangesintheexpectedamount

ofgreenhousegasemissions.Itprovidesprojectionsonafive-year

periodupuntil2050fortheEUasawholeandforeachEUcountry.

Itisnotdesignedasaforecastofwhatislikelytohappeninthe

future. It ratherprovidesabenchmarkagainstwhichnewpolicy

proposalscanbeassessed.

On20July,theEuropeanCommissionpublisheditslatestReference

Scenario:theEUReferenceScenario2016(REF2016).Withtheactive

participationofnationalexpertsfromallEUcountries,theEuropean

Commissionworkedinpartnershipwithamodellingconsortiumled

bytheNationalTechnicalUniversityofAthenstodevelopREF2016,

makinguseofarangeofdifferentmodels.

Theprojectionsarebasedonasetofassumptions, includingon

populationgrowth,macroeconomicandoilpricedevelopments,

technology improvements,andpolicies.Regardingpolicies,pro-

jectionsshowtheimpactsofthefull implementationofexisting

legallybinding2020targetsandEUlegislation.Assuch,theyalso

showthecontinuedimpactpost2020ofpoliciessuchastheEU

EmissionsTradingSystemDirective(includingtheMarket Stability

Reserve),theEnergyPerformanceofBuildingsDirective,Regulations

on ecodesign and on CO2emissionstandardsforcarsandvans, as

wellastherecentlyrevisedF-gasRegulation.Suchpoliciesnotably

influencecurrentinvestmentdecisions,withimpactsonthestock

ofbuildings,equipmentandcars,whichhavelong-lastingeffects

post-2020onGHGemissionsorenergyconsumption.

TheEUReferenceScenario2016

12

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Results

REF2016issetuptomeetthebindingenergy and climate targets

for2020,thelatterbeingachievedasaresultofexistingpolicies.

However,itshowsthatcurrentpoliciesandmarketconditionswill

deliverneithertheEU’s2030targetsnorthelong-term2050objective

of80to95%greenhousegas(GHG)emissionreductions.Overall

GHGemissionsdecreaseby26%in2020,35%in2030and48%

in2050.GHGemissionsfromsectorscoveredbytheEffortSharing

Decisionareprojectedtodecreaseby16%in2020andby24%

in2030below2005levels,lessthanemissionsinsectorscovered

bytheEU Emission Trading System.In2020,therenewableenergy

share(RES)ingrossfinalenergyconsumptionreaches21%,while

in2030itincreasesslightlyfurther,reaching24%.Inaddition,the

energyefficiency2020non-bindingtargetisnotmetinREF2016,the

scenarioprojectingareductioninprimaryenergysavings(relativeto

the2007baseline)of18%in2020,and,respectively,24%in2030.

The EU’s energy production is projected to continuetodecrease

fromaround760Mtoein2015toabout660Mtoein2050.The

projectedstrongdeclineinEUdomesticproductionforallfossilfuels

(coal,oilandgas)coupledwithalimiteddeclineinnuclearenergypro-

ductionispartlycompensatedbyanincreaseindomesticproduction

Figure 1: Projection of key policy indicators

712

16

2124

27 31

-18

-24

-7-15

-26-35

-42

-48

-80

-70

-60

-50

-40

-30

-20

-10

0

10

20

30

40

50

1990 1995 2000 2005 2010 2015 2020 2025 2030 2035 2040 2045 2050

RES share inGross FinalEnergyConsumption

Primary energysavingscompared to2007 baselineprojection(% change)

Total GHGemissions (%change to 1990)

ofrenewables.Biomassandbiowastewillcontinuetodominatethe

fuelmixofEUdomesticrenewableproduction,althoughtheshare

ofsolarandwindintherenewablemixwillgraduallyincrease.

The EU’s import dependencyshowsaslowly increasingtrend

overtheprojectedperiod,from53%in2010to58%in2050.RES

deployment,energyefficiencyimprovementsandnuclearproduction

(whichremainsstable)counteractsthestrongprojecteddecrease

intheEU’sfossilfuelproduction.

The EU power generation mixchangesconsiderablyoverthe

projectedperiodinfavourofrenewables.Before2020,thisoccurs

tothedetrimentofgas,asastrongRESpolicytomeet2020targets,

verylowcoalpricescomparedtogasprices,andlowCO2 prices do

nothelpgastoreplacecoal.After2020,thechangeischaracterised

byfurtherRESdeployment,basedonmarketconditions,butalsoa

largercoaltogasshift,drivenmainlyinanticipationofincreasing

CO2prices.VariableRES(solarandwind)reacharound19%oftotal

netelectricitygenerationin2020,25%in2030and36%in2050,

demonstratingthegrowingneedforflexibilityinthepowersystem.

Theshareofnucleardecreasesgraduallyovertheprojectedperiod

despitesomelifetimeextensionsandnewbuilt,from27%in2015

to22%in2030.

Source: PRIMES, GAINS

13

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Primary energy demand andGDPcontinuetodecouple,whichis

consistentwiththetrendsobservedsince2005.Energyefficiency

improvementsaremainlydrivenbypolicyupto2020andbymarket/

technologytrendsafter2020.Withregardtothefuelmixinfinal

energydemand,thereisagradualpenetrationofelectricity(from

Figure 2: Evolution of final energy demand by fuel (Mtoe – above, shares – below)

20%intotalfinalenergyuse in2005to28%in2050).This is

becauseofgrowingelectricitydemandascomparedtootherfinal

energyuseandtosomeelectrificationofheating(heatpumps)and

toalimitedextentofthetransportsector.

Solids

Oil

Gas

Electricity

Heat (from CHPand DistrictHeating)

Other

0

200

400

600

800

1,000

1,200

1,400

2005 2010 2015 2020 2025 2030 2035 2040 2045 2050

5%4%

20%

24%

42%

5%

10%

4%

22%

23%

36%

4%

10%

5%

25%

22%

35%

3%

11%

5%

27%

22%

34%

2%

11%

5%

28%

22%

32%

1%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2005 2020 2030 2040 2050

Source: PRIMES

14

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Cristina MohoraCristina Mohora received a Master’s Degree in Financial and Monetary Policies from the Academy of Economic Studies in Bucharest and a PhD in Economic Modelling from Erasmus University Rotterdam. After holding an academic position at the Academy of Economic Studies in Bucharest and a research position at Université Libre de Bruxelles, she joined the European Commission in 2008. Between 2008 and 2010, Cristina has been involved in the energy system modelling work at the Directorate-General Energy and Transport. Since 2010 she has been responsible for the modelling work coordinated by the economic analysis unit of the Directorate-General Mobility and Transport.

Jan NillDr Jan Nill works at the European Commission, DG Climate Action. He has been one of the coordinators of the EU Reference Scenarios 2013 and 2016. Currently he works as policy officer in the unit CLIMA C.2 Governance & Effort Sharing. Jan holds a PhD in economics from the University of Kassel.

Joan Canton Joan Canton is an Economic Analyst at the European Commission’s Directorate-General for Energy, focusing on the modelling of energy systems, supporting the preparation of the Commission's Impact Assessments on climate and energy issues, as well as on monitoring the implementation of the Energy Union Strategy. Before working for DG Energy, he worked in DG Climate Action and in DG Economics and Financial Affairs. He holds a PhD in econom-ics from the University of Aix-Marseille and has worked as an Assistant Professor in the Economics Department of the University of Ottawa (Canada).

Investment expendituresforpowersupplyincreasesubstantially

until2020drivenbyREStargetsanddevelopments,butslowdown

thereafter,until2030,beforeincreasingagainfrom2030onwards

notablyduetoincreasingETScarbonpricesreflectingacontinuously

decreasingETScapbasedonthecurrentlinearfactor.Newpower

plantinvestmentisdominatedbyRES,notablysolarPVandwind

onshore.Investmentexpendituresindemandsectorsoverthepro-

jectedperiodwillbehigherthaninthepast.Theynotablypeakin

theshorttermupto2020,particularlyintheresidentialandtertiary

sectors,asaresultofenergyefficiencypolices.

Energy system costsincreaseupto2020.Largeinvestmentsare

undertaken,drivenbycurrentpoliciesandmeasures.Overall,in2020

energysystemcostsconstitute12.3%ofGDP,risingfrom11.2%

in2015,alsodrivenbyprojectedrisingfossilfuelprices2.Despite

furtherfossilfuelpriceincreases,between2020and2030theshare

remainsstableanddecreasesthereafter,asthesystemreapsbenefits

fromtheinvestmentsundertakeninthepreviousdecade(notably

viafuelsavings).Inthisperiod,theshareofenergysystemcostsin

GDPisgraduallydecreasing,reachinglevelscloseto2005by2050.

2 Totalsystemcostsincludetotalenergysystemcosts,costsrelatedtoprocess-CO2 abatement and non-CO2GHGabatement.

15

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

distinctivefeatureofourmodellingapproachisthestrongemphasis

oneconomictheoryalongsideadeepunderstandingoftherelevant

technologies.Thus,theinsightsgeneratedfromourmodelsreveal

importantfindingsabouteconomicinterdependenciesandeffects

ontopofmeretechnology-basedanalysis.

How do you ensure the robustness of your simulation tools?

Robustness isensuredbyconsistency-checks,back-testingand

economicreview.Consistency-checksverifythatmodelresultsare

internallyconsistent,e.g.,energybalancesheetsarecorrect,notech-

nologicalboundariesareinfringedetc.Back-testingrunsthemodel

withhistoricaldata,whichisaviablewayforidentifyingpossible

shortfallsofthemodel.However,duetofundamentaldifficulties

withaccurateback-testing inacomplexenergyenvironment,we

alsoaddwhatwecall “economicreviewofthemodels”,namely

checkingmodelsandmodelresultswithrespecttotheirfitwith

economictheoryandobservedandforeseeablemarketbehaviour.

Tell us a little about ewi Energy Research & Scenarios and the work that you do.

ewiEnergyResearch&Scenariosisanon-profitorganizationfocus-

singonappliedeconomicresearchonenergymarketsandenergy

policy.Wehaveateamofabout35people,manyofthemsimul-

taneouslypursuingtheirPhDattheUniversityofCologne.Besides

conductingresearchprojects,wealsoofferresearchanddevelopment

supportaswellaseconomicadvicetogovernment,organisations

andcompanies.Thus,weregularlyprovidedecision-makerswith

soundquantitativesupportbasedonourstrongeconomicand

modellingexpertise.

What role do energy system models play in your research?

Energysystemmodelsareattheanalyticalcoreofourresearch.

Werun,andcontinuouslyimprove,modelsofglobalfuelmarkets

aswellastheEuropeanelectricity,gas,andheatmarkets.The

MarcOliverBettzügeDirectoroftheInstituteofEnergyEconomicsattheUniversityofCologne(ewi)andPresidentof theSupervisoryBoardofewiEnergyResearch&Scenarios.

TALKS TO SETIS

16

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

How can energy systems modelling contribute to a suc-cessful energy transition in Europe?

Energy systems models are important analytical tools to assess

potentialmarketdevelopments,includingtheirreactiontocertain

politicalmeasures.Hence,decision-makers intheenergydomain

mayusesuchmodelstoobtainamoreprofoundinformationbase

fortheirdecisionsandactions.Importantly,however, itshouldbe

stressedthatmodelstypicallygeneratescenarios–notforecasts.

Hence,itisimportantfordecision-makers,andthegeneralpublic,

toadequatelyinterpretthemeaningofscenariosbeforecomingto

conclusionsabouttheirimplications.Therefore,wehavedesigned

ourmodelsas“anti-blackboxes”,andwedevotealotoftimeand

efforttosupplyingtransparencyandinterpretationalongsideour

scenarioanalyses.

Marc Oliver BettzügeDr Marc Oliver Bettzüge has been a professor of economics, in particular energy economics, and Head of the Chair of Energy Economics - Department of Economics - at the University of Cologne since 2007. He is also Managing Director and Chairman of the Management Board of the Institute of Energy Economics at the University of Cologne (EWI). Professor Bettzüge has been a member of the German Parliament’s Study Commission on Growth, Wellbeing and Quality from 2011 to 2013. In addition he plays an active role in various committees and advisory boards.

What has your research revealed to be the most urgent issues facing the European energy system?

Thereisofcourseadifferencebetweenurgencyandimportance.From

aneconomicperspective,themosturgentissueinelectricityisthe

increasinggeographicimbalancebetweensupplyanddemandinthe

Europeanelectricitysystem,exacerbatedbyaratherslowexpansion

ofthegridandaninadequateconfigurationofbiddingzones.Forthe

gassupplysystem,ourmodelssuggestthaturgentdecisionsaround

Nord Stream 2havewide-rangingpoliticalramificationswhichshould

betransparentlytakenintoaccount.Withrespecttoimportance,our

modelsconsistentlyshowthattheinsufficientalignmentofEUand

nationalenergypoliciesleadstoinefficientandineffectiveoutcomes

withrespecttomitigatingCO2-emissionsinEurope.Thus,afunda-

mentaloverhaulofthepoliticalapproachtoenergyandclimatepolicy

wouldbeveryreasonablefromaneconomicperspective.

©iStock/anyaberkut

17

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Energy transitions are not new

Theworld’senergysystemisenteringamajortransition.Transitions

havehappenedbefore.Thenineteenthandtwentiethcenturiessaw

growthintheuseofcoal,andthenoilandnaturalgaswhenthe

moderncombustionenginetookoff.Butsincearoundthe1970s,

whilstthere’sbeenplentyofgrowthintheuseofenergy,themixof

fuelshasbeenrelativelystatic.Morerecently,withadditionalconcerns

overlocalandatmosphericemissions,theworldisseeingimpressive

ratesofgrowthinwindandsolarpower–aneweraoftransition.

Shell Scenarios help to navigate uncertainties about the future

Shellhasbeenusingscenarioplanningforover40yearstohelp

deepenitsstrategicthinking.Thescenarioshelpdecision-makers

toexplorethefeatures,uncertainties,andboundariesshapingthe

futurelandscape,andtoengagewithalternativepointsofview.

Ourscenariosconsiderlong-termtrendsineconomics,geopolitical

shiftsandsocialchangeaswellastechnologicalprogressand

theavailabilityofnaturalresources.Theyarebasedonplausible

assumptionsaboutfuturedevelopment,andincludetheimpactof

differentpatternsofindividualandcollectivechoices.

Shell’s Energy Scenarios are underpinned by quantitative modelling

Shell’sWorldEnergyModel(WEM)providesarigorousquantitative

frameworktounderpinthelogicofourscenarios.TogetherwithShell’s

GlobalSupplyModel,theWEMisacoretoolexploringalternative

evolutionsofenergydemandindifferentcountriesandindifferent

sectors,helpingtomaintainsystemconsistency,toexplorethemost

significantfactorsinpolicy,technologyandconsumerchoices,and

toexaminetheimpactsinonepartoftheworldmadebyshiftsin

another.

Shell’s latestScenariospublication,“ABetterLifewithaHealthy

Planet.PathwaystoNet-ZeroEmissions,”takesthemostoptimistic

featuresofour2013“NewLensScenarios”–MountainsandOceans

–andcombinesthemwithindividuallyplausiblefurthershiftsinpolicy,

technologydeployment,circumstances,andeventsthatmightmove

theworldontoanew,evenlower-emissiontrajectory,resultingin

net-zeroemissionsonatimescaleconsistentwithglobalaspirations.

Future energy demand will at least double

Thisworkstartsbyattemptingtoquantifythemagnitudeoffuture

energydemand.Asweconsiderthefuturedevelopmentofecono-

mies,andassumesignificantenergyefficiencyimprovements,we

estimatethatanaverageofabout28,000kWhofprimaryenergy

perpersonisapproximatelyrequiredtosupportthedecentquality

oflifetowhichpeoplenaturallyaspire.

Andifweassumeafuturepopulationofaround10billionpeopleby

theendofthecentury,andmultiplyitby28,000kWhpercapita,we

seethattheglobalenergyneedwouldbeabout280trillionkWha

year–roughlytwicethesizeofthecurrentenergysystem.

Abetterlifewithahealthyplanet:pathwaystonet-zeroemissions

©iStock/Rawpixel Ltd

18

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Fossil Approximately 50% electrification of end use.With Carbon Capture and Storage

EmergingNet-ZeroEmissionsWorld

2015

Hydrocarbons alongside renewables

Acrosstheenergysystem,it’slikelythatdifferentdegreesofdecar-

bonisationandenergyefficiencywillbeachievedatdifferentpaces,

indifferentplaces,andindifferentsectorsoftheeconomy.

Toarresttheaccumulationofgreenhousegasesintheatmosphere,

theworldwilleventuallyneedtoseeoverallemissionstodropto

net-zero.Inanet-zeroemissionsworld3withadecentqualityoflife

enjoyedbythemajorityofthepopulation,renewableenergieswill

dominate,andtogetherwithnuclearcouldmakeaboutthreequar-

tersoftheenergysupplycarbonneutral.Butrenewablesprimarily

produceelectricity,whichcurrentlycountsforlessthanone-fifthof

energyuse.Theproductionofchemicalsandplasticswouldcontinue

torelyonfeedstockfromoilandgas,andwherehightemperatures

3 Aworldinwhichtheamountofcarbonreleasedisbalancedbyanequivalentamountsequesteredoroffset.

ordenseenergystoragearerequired–suchasinmanyindustrial

processeslike iron/steel/cementmanufactureorheavyfreightor

airtransport,wewillseethecontinuedneedforhydrocarbonfuels.

Electrification is key for low CO2 and high efficiency

Inordertoachievebothlowemissionsandhighefficiency,theelec-

tricitymarketsharewillneedtogrowfromonefifthoftheenergy

consumedtoatleastahalf.Electrificationneedsbeparticularlyhigh

inhouseholdsandservicesectors,butneedstoextendfurtherinto

othersectorssuchasfoodprocessingandlightmanufacturing.For

passengertransport,hydrogenfuelcellandelectricdrivesshould

becomecommon,whileforaviation,shippingandfreighthydrocar-

bonswilllikelyremainimportant.

Foraworldwithwidespreadprosperity,theenergysystemwilldoubleoverthecourseofthis century.

ENERGYSOURCE GAS OIL COAL BIOENERGY NUCLEAR SOLAR WIND OTHER

2015 21% 31% 28% 11% 5% 0.5% 0.5% 3%

Net-Zero emissionsworld 9% 7% 9% 15% 8% 30% 12% 10%

Figure 3: Plausible energy mix in an emerging net-zero emissions world

Source: Shell analysis

19

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Carbon Capture and Storage is indispensable

It’simportanttonotethatanet-zeroemissionsworldisnotneces-

sarilyaworldwithoutanyemissionsanywhere.It’saworldwhere

remainingemissionsareoffsetelsewhere inthesystem.Toboth

mopupremainingemissionsandprovideopportunitiesfor‘negative’

emissions,theworldwillneedwidespreaddeploymentofcarbon

captureandstorage(CCS).

Incentivising the transition

Achievinganet-zeroemissionsworldatpacewillrequiresignificant

developmentsinnewtechnologydeployment;industrial,agricultural

andurbanpractices;consumerbehaviour;andpolicyframeworks

whichshape,incentiviseormandatethesetransitions.Itwillalso

entailhighlevelsofcollaborationbetweenpolicymakers,businesses

andcivilsociety.

Governmentsneedtoprovidefinancialincentivesviacarbonprices

ortaxesforavoidingemissionsandremoveenergysubsidieswhere

theystillexist.Thisallowsthemarkettofindtheoptimalenergy

mixatlowestcosts.

Sensible measures for progress towards a net-zero emissions world

Analysingthelikelyevolutionofdemandacrosskeyareasofthe

economy,somethingofalogicalorder-of-priorityofactionsemerges:

1. Stimulatingefficiencymeasuresandextendingelectrification

acrosstheeconomywhereverandwheneverpossible;

2. Sustainingmomentumofrenewablesgrowth,particularlysolar

PVandwind,andmaximisingtheabilityofthegridtohandle

theirintermittency;

3. Acceleratingtheswitchfromcoaltogastoimmediatelyreduce

powersectoremissionswhileensuringsupplytomeetdemand

–awayofkeepingcumulativeemissionstoaminimumduring

thetransition;

4. Improvingbuildingsandcityinfrastructuretolowerenergyservice

demandsignificantly;

5. Acceleratinggovernment-directedeffortstopromotelow-carbon

technologiesandinfrastructures,includingnuclear,CCS,hydro-

gentransport, responsiblebioenergyandsustainableforestry,

agricultureandland-usepractices.

Concluding remarks

WehopethatourlatestScenariosworkwillhelpbuildsharedinsights

andperspectivesamongbusinesses,nationalgovernmentsandcivil

societymorebroadly.Asharedunderstandingwouldnotonlyaccel-

eratethenear-termactionstoreduceCO2emissions,butalsothe

deeperstructuraltransformationsrequiredtosustaindecarbonisation

andeconomicgrowthinthelongerterm.

Formoreinformationvisit:www.shell.com/scenarios

Note: Shell Scenarios are part of an ongoing process used in Shell for

40 years to challenge executives on the future business environment.

We base them on plausible assumptions and quantification, and

they are designed to stretch management to consider even events

that may be only remotely possible. Scenarios, therefore, are not

intended to be predictions of likely future events or outcomes and

investors should not rely on them when making an investment

decision with regard to Royal Dutch Shell plc securities. While we

seek to enhance our operations’ average energy intensity through

both the development of new projects and divestments, we have

no immediate plans to move to a net-zero emissions portfolio over

our investment horizon of 10-20 years.

Wim ThomasWim Thomas is Shell’s Chief Energy Advisor and also leads the Energy Analysis Team in Shell’s Global Scenario Group. He has been with Shell for over 30 years. Wim is also Chairman of World Petroleum Council UK National Committee, a Distinguished Fellow of the Institute of Energy Economics Japan, and a former chairman of the British Institute of Energy Eco-nomics in 2005. He holds a postgraduate degree in Maritime Technology, Delft University, the Netherlands.

20

What are the main insights that we aim to achieve through the development of energy systems models?

Inthepast,themainuseofacademicandpolicy-ledenergymodels

wastounderstandtheflowsoftraditionalenergyresourcesinthe

contextsofvalueformoneyandsecurityofsupply.Morerecently,the

emphasisofacademicmodellinghasbeentounderstandlow-car-

bonenergytransitionsand,indoingthis,themodelsneedtolook

muchfurtherintothefutureandtryandpredicthowthesystemwill

evolveandwhatconsequencessuchanevolutionwillhave.However

Ithinkthattherealchallengewhenwedevelopmodelsistounder-

standwhotheintendedaudienceisandwhetherthataudienceis

listening.Fromanacademicpointofview,theendgoalistypically

togetyourmodelusedinapolicycontext.However,policy-makers

oftentypicallyhavealreadyinvestedintheirownmodelsand,as

wefoundasaspillerfromourresearch,intermsofgettingused,

thelanguage,cultureandaddedvalueofamodelisasimportant

asthetechnicalaccuracyorscope.

How do these models help to balance uncertainty in the energy system?

Wereviewedarangeofacademicandpolicymodelsofenergy

systemsbut,intermsofbalancinguncertaintyintheenergysystem,

akeyinteractionistheuseofarangeofdifferentmodelsbysystem

operators.Shorttermenergysecurityneedstobenegotiatedalong-

sidethetransitiontoamoresustainableenergysystem,sohighly

accurateandnumericaloperationalmodelsneedtobeinconversa-

tionwithfuturescenario-basedmodels.IntheUK,theNationalGrid

“GoneGreen”scenariomodelisagoodexampleofhowthisisdone

AlistairBuckleyco-authorof‘AreviewofenergysystemsmodelsintheUK:Prevalentusageandcategorisation’

TALKS TO SETIS

©iStock/suriya9

21

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

inpractice.“GoneGreen”sitsattheinterfacebetweenthesystem

operatorNationalGridandtheirday-to-dayoperationalmodelling,

withpolicy-ledtransitionalmodelsheldbytheUKgovernment

energydepartment(nowaspartofthedepartmentforBusiness,

EnergyandIndustrialStrategy)andacademicenergymodels.Itis

citedformallyandinformallyacrossthesestakeholdersandhasa

bigroleindiscussionsaroundtransitionsintheUKenergysystem.

You recently conducted a review of energy system models in the UK. What were the main findings from this review?

Wefoundthatthemajorityofpublicationsofmodellingresultscame

fromonlyaveryfewdifferentmodelsandthatthepolicydocuments

citethesamemodelsastheacademic literature.This isagood

thing,asitmeansthattheacademicandpolicycommunitiesare

joinedup.FromourresearchIthinkit’sfairtosaythatthereisn’t

thesameinternationaltravellingofenergysystemsmodellingasin

otherscienceandtechnologyfields.Themodelsthatwefoundcited

intheUKscienceandpolicy literatureweremainlyhome-grown.

Itwouldbeinterestingtodoacomprehensivestudytoseewhere

modelshavetravelledinternationallyandhowthishasimpacted

onenergypolicyinthosecountries.

What are some of the limitations of existing energy system modelling tools?

I thinkoneofthemajorchallengesisthe integrationofqualita-

tiveresearchfromthesocialsciencesaroundscenariosandpolicy

changes.It’sallverywellhavinghighlyaccurateandgranularenergy

flowmodelsfordifferentfuturegenerationmixscenariosbutifthe

transitiontothesescenariosisentirelydependentonpoliticalfactors

atboththeEU,nationalandlocallevelthenthemodeliskindof

irrelevant.Ithinkinvestmentinnewapproachestothedemocrati-

sationofmodellingwithparticipationfromkeystakeholderswould

beveryinteresting.WehaveattemptedthisinaUK-basedresearch

projectandfoundcomputationalenergymodelstobeimpenetrable

bymostofthesestakeholders.

Alistair BuckleyAlastair Buckley is a senior lecturer in the department of physics at the University of

Sheffield. His research investigates the integration of solar PV into future energy systems

from technological and socio-technical viewpoints. His Sheffield-based research team has

developed real-time PV power monitoring for the UK transmission network.

Based on your review, do you have any recommendations regarding the optimisation of modelling capacity?

Ithinkthatopeningupallenergybaseddataacrosstheacademic,

policyandsystemoperatorcommunitieswouldresult inastep

changeinintegrationofthedifferentmodellers.Accesstodatais

akeyconstraintinmodellingandopendatasourceswouldallow

validationofdifferentmodelsacrossdifferentscenarios.This,inturn,

wouldresultinawidervarietyofstakeholderstouseawidervariety

ofmodels.Ithinkthiswouldbehighlybeneficial.

©iStock/Madedee

22

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

EnergysystemmodellingisakeytoolforEDFR&D.It isusedto

evaluatethe impactofenergypolicies(renewabledeployment,

EUETS),torecommendbusinessstrategiesandanalysebusiness

opportunitiesbasedonevolutionsintheenergy/powersystemsand

tocontributetothepublicdebate.

Multi-energyanalysisisgainingincreasedattention,asmoreinter-

actionsbetweenelectricity,heatandcold,andgassystemscre-

atepromisingopportunitiesfordecarbonisation(for instance,by

developingmoreefficientusagesorsharingflexibilitiesforabetter

integrationofvariablerenewables).Somemajorchallengesneed

tobeaddressed:modellingtheseinteractionsiscomplex,andcan

leadtooverlycomplexmodelsor,onthecontrary,totheuseof

significantsimplifications.Thesesimplificationsmustbemade

carefully,especiallywhenmodellingpowersystems,astheycan

easilydistorttheresultsandleadtoapartialunderstandingofthe

system’scomplexity.

Morespecifically,energysystemmodellingisoftenlinkedtoone

typeofmodelling,namelymodelsbasedonTIMES,representingthe

interactionsbetweenseveralenergyvectors.Thesemodelsmakeit

possibletosimulateandoptimisedecarbonisationscenariosover

severaldecades,whichmakesthemhighly interesting.However,

mostTIMESmodelsdonotcurrentlyallowforadetailedenough

Energy system modelling intheindustry:theEDFR&Dperspective

©iStock/FooToo

2323

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

representationoftheelectricitymix. Inparticular, theygenerally

cannotgivespecificinsightsontheneedsforflexibilityrelatedtothe

growingpenetrationofwindandsolarsources(forexample,asthese

modelsdonotusehourlystepsormulti-scenarioapproaches,rules

ofthumbareneededtodecideontherequiredpeakingcapacity-an

approachthathasobviouslimitations).

EDFR&Dhasledamajoreffortinrecentyearstostudytheimpli-

cationsofvariousenergyscenariosontheelectricpowersystem.

Variousapproachesweretestedanddevelopedin-house.Oneof

these,usedin“TechnicalandeconomicanalysisoftheEuropean

electricitysystemwith60%RES”,4consistsofusinga“chainof

models”(insteadofasinglemodel),witheachmodelinthechain

makingitpossibletostudyandgraspvariouskeyimpactsofwind

andsolarintegrationinpowersystems.

TheCONTINENTALmodel(seeLangrenéetal)5isthemainstepin

themodellingchaindescribedbelow.Theinputdataandhypothe-

ses(suchastheCO2price,demandlevel,etc.)comefromenergy

4 Alain Burtin, Vera Silva, “Technical and economic analysis of the European electricity system with 60% RES”, EDF R&D, June 2015.

5 Nicolas Langrené, Wim van Ackooij, and Frédéric Bréant, “Dynamic Constraints for Aggregated Units: Formulation and Application”, IEEE transactions on Power Systems, Vol 26, no. 3, August 2011.

scenarios,whicharesometimesestablishedusing largeenergy

systemsmodels(suchastheEDFR&DMadone/TIMESmodel,or

theJRCmodel(seeSimoesetal)),6whichthenconstitutethefirst

stepofthemodellingchain.TheCONTINENTALmodel’s5outputscan

alsobefedintoothermodules/models,constitutingthelaststeps

inthechain,asdescribedbelow.Thefollowingparagraphsdescribe

inmoredepththecoremodelandthesub-modulesdeveloped.

Inordertostudytheimpactofwindandsolar,thecorepowersystem

model(CONTINENTAL)needstofeaturesomeminimumcharacter-

isticstoprovidecredible insights.These“minimal/recommended

requirements”havebeenwelldiscussedintheresearchcommunity

inrecentyears,andthemostoftencitedare:

•• Hourly base And multi scenariosofdemandandvariable

generation:assessingtheneedforbackupgenerationimplies

beingabletotakeintoaccountextremeeventsthatcanhappen

overafewhours,incertainyears.Theuseofaverageprofiles

(forinstance,onlypeakandoff-peaksteps,ononeaverageday

permonth)cannotcapturesuchevents.

6 Sofia Simoes, Wouter Nijs, Pablo Ruiz, Alessandra Sgobbi, Daniela Radu, Pelin Bolat, Christian Thiel, Stathis Peteves, “The JRC-EU-TIMES model, Assessing the long-term role of the SET-Plan Energy technologies”, JRC Scientific and Policy Report, 2013.

Investment / hourly dispatchDetailed description of VGLocationofVG

Loadfactors(with resolution1horlower)

VGforecasterrors

Input dataDemandtimeseries

Investmentcosts

Generationdynamic constraints

Fuelscosts

CO2 price

Networktransfercapacities

Generationmix

Interconnection loadfactors

Generationloadfactors

VGcurtailment

Market prices andgeneration costs

Plantrevenues

Reservesandflexibilityadequacy

Frequencystability

Investment loop

CONTINENTALMODEL

Madone/TIMES model

Dynamicsimulationmodel

Flex Assessment

Economical analysis

© Burtin and Silva

Figure 4: EDF chain of models

24

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Figure 5: European load duration curve of demand and net demand with 60% RES (left) and structure of the generation mix with and without wind and PV generation (right)

•• Multi-zone modelling(i.e.modellingexchangesbetweencoun-

triesintheEuropeansystem):nationalstudiesoftenusea“single

zone”model,usingroughapproximationsforthelevelandprices

ofimportsandexport.AstheEuropeangridgetsmoreintegrated

andinterconnected,theimpactofthesesimplificationsshouldbe

re-assessedregularly.TheCONTINENTALmodelmakesitpossible

tofeatureaEuropeanmulti-zonemarket.

•• Water reservoirs and pump storage management:Hydro

resourcesandexistingpumpstoragearekeyprovidersofflexibil-

ity–mostmodelsrelyonsimplifiedrulesofthumbandadeter-

ministicvisiontodispatchtheseenergy-constrainedresources.

Suchsimplificationsmightunder-oroverestimatetheroleof

theseresources,whileastochasticmodellingofhydro(such

astheoneusedinCONTINENTAL)givesamorerealisticview.

•• Detailed modelling of thermal unit constraints:dynamic

constraintssuchasminimumon/offtime,start-upcosts,mini-

mumstablegeneration,etc.areimportantwhenestimatingthe

systemflexibility–modellingtheseconstraintsgenerallyimplies

verystrongincreasesinproblemcomplexityandcomputingtime.

Itmightthereforenotbepossibletoalwaysmodelthem,but

sensitivitiestotheseparametersshouldbeconsidered.

Additionally,suchmodelsshouldmakeitpossibletoanalysethe

needfordifferenttypesofback-upfossilgeneration(e.g.theshare

ofcombinedcycleversusopencyclegasturbines)– intheEDF

modellingchain.Thesocalled“investmentloop”makesitpossible

toestablishaleastcostback-upgenerationfleet,respectingapre-

definedadequacycriteria(a3/yhoursLossofLoadExpectation).

Figure5givesanexampleofhowthermalgenerationwouldevolve

fromaEuropeansystemwith0%windandsolartoonewith40%.

However,asalreadymentioned,addingtoomanyfeaturesatonce

inonesinglepowerdispatchmodelmightnotalwaysbeagood

option:thecomputingtimeis likelyto increasesharply,butalso

(andperhapsmore importantly) itmight limitthepossibilityto

understandallthephenomenaatplayinahighRESsystem(when

modelsbecomeoverlycomplex,thereisariskthattheybecomefor

mostpeopleamysteriousblackbox,insteadofatoolthatallowsa

betterunderstandingbyengineersandeconomists).

No Wind & SolarTotal: 444 GW

With Wind & SolarTotal: 352 GW

European thermal installed capacity

89GW71GW

34GW

89GW78GW

250GW

86GW99GW

0

50

100

150

200

250

300

GW

1000 2000 3000 80007000600050004000 Nuclear Coal CCGT OCGT

600

500

400

300

200

100

-

-1000

is The need for baseload generation

is reduced by 160 GW

700

With 40% wind and PV the need forbackup generation increases 60 GW

0 2

12

10

8

6

4

2

04 6 8 10 12

2013 Scenario 60% RES

Upw

ard

mar

gin

(GW

)

percentiles 2.5 - 97.5%

14 16 18 20 22 24 0 2

12

10

8

6

4

2

04 6 8 10 12

Upw

ard

mar

gin

(GW

)percentiles 25 - 75%percentiles 47.5 - 52.5%

© Burtin and Silva

25

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Figure 6: Day-ahead operating in 2013 and the simulated operating margin

for the scenario “60% RES” for France during summer

Therefore, for its2015study4, EDFR&Ddevelopedadditional

sub-modulesusingoutputdatafromthepowersystemmodelto

analysespecificpowersystemissues,suchastheneedforopera-

tionalmarginscomparedtotheavailableflexibility(“FlexAssessment”

–seeFigure6),orthebehaviouroffrequency(“Dynamicstability

module”),withoutmodifying(andmakingmorecomplex)thecore

powersystemmodelitself.

TheCONTINENTALmodelandthevarioussub-modulesmakeit

possibletosimulateoneyearperiods.So,itisnotpossibledirectly

toproposeevolutionscenarios,forwhichTIMESmodelsarebetter

suited.Itwould,however,bepossibletoextendthe“chainofmodels”

andtoback-feedinformationintotheTIMESmodel(suchas,for

example,abettervisionoftherequiredthermalback-upgeneration).

Interactionbetweenmodelsinthiswaycanbeagreattooltobuild

decarbonisationscenariosthattakeintoaccountbothmulti-energy

interactionsandthestrongspecificitiesofelectricalpowersystems.

The“chainofmodelling”discussedhere isoneexamplewhere

modellinghasprovidedsignificant insights intothe implications

andchallengesofahighRESEuropeanscenariofortheelectricity

system.4EDFR&D,alreadywithalonghistoryofenergymodelling,is

continuingtoworkonenergysystemmodelling,inanefforttokeep

increasingourunderstandingofthecurrentandfuturechallenges

oftheelectricityandenergysystems.

89GW71GW

34GW

89GW78GW

86GW99GW

0

50

100

150

200

250

300

GW

1000 2000 3000 80007000600050004000 Nuclear Coal CCGT OCGT

600

500

400

300

200

100

-

-1000

is The need for baseload generation

is reduced by 160 GW

0 2

12

10

8

6

4

2

04 6 8 10 12

Time (hours )

2013 Scenario 60% RES

Upw

ard

mar

gin

(GW

)

percentiles 2.5 - 97.5%

14 16 18 20 22 24 0 2

12

10

8

6

4

2

04 6 8 10 12

Time (hours)

Upw

ard

mar

gin

(GW

)

14 16 18 20 22 24

percentiles 25 - 75%percentiles 47.5 - 52.5%

© Burtin and Silva

©iStock/goce

26

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Paul Fourment Paul Fourment graduated in energy engineering from the Ecole Polytechnique (Paris). He

has worked for three years at EDF R&D as a research scientist on renewable integration in

the European Power System. His work deals in particular with the power system’s flexibility

needs, the articulation between nuclear and renewable technologies, and the cohabitation

between global and local systems.

Dr. Vera Silva Dr. Vera Silva is the director of the EDF R&D research program on “Energy systems and

markets” and a senior researcher at EDF R&D in the field of “operation of electricity systems

and markets”. She has 18 years’ experience in the power systems industry and before joining

EDF in 2009 she worked as a research assistant at the Control & Power research group of

Imperial College London and at the University of Manchester.

Miguel Lopez-Botet Miguel Lopez-Botet is the manager and technical leader of a project team responsible for

“Analysis of future energy mixes” within EDF R&D. His main research topics are generation

expansion planning and operation, the integration of wind and solar PV generation and the

value of flexibility and ancillary services.

Timothée Hinchliffe Timothée Hinchliffe is project manager on energy storage economics at EDF R&D. His main

focus over the past five years has been the assessment of energy storage needs through

the use of modelling, and considering the competition with other flexibility levers such as

interconnections or demand response. Timothée holds a Master’s in Electrical engineering

from Supelec.

27

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

METIS is a researchprojectoftheEuropeanCommission’sDirecto-

rate-GeneralforEnergy(DGENER)forthedevelopmentofenergy

simulatorsoftwarewiththeaimtofurthersupportENER’sevi-

dence-basedpolicy-making.Itisdevelopedbyaconsortium(Artelys,

IAEW,ConGasandFrontierEconomics)aspartofHorizon2020and

iscloselyfollowedbyDGENER.

Software description

Unlikeothersimulators,METIS will be owned and operated by DG

ENER,withthesupportoftheCommission’sin-housescienceand

knowledgeservice,theJointResearchCentre.Theintentionisto

haveanin-housetoolthatcanquicklyprovideinsightsandrobust

answerstocomplexeconomicandenergyrelatedquestions,focusing

moreontheshort-termoperationoftheenergysystemandmarkets.

METISwasused,alongwithPRIMES,intheimpactassessmentof

theMarketDesignInitiative.

METISisanenergymodelthatcovers,withhighgranularity(geo-

graphical,timeetc.),thewholeEuropeanenergysystemforelectricity,

gasandheat.Initsfinalversionitshouldbeabletosimulateboth

systemandmarketoperationfortheseenergycarriers,onanhourly

levelforawholeyearandunderuncertainty(capturingweather

variationsandotherstochasticevents).METISworkscomplementary

tolong-termenergysystemmodels(likePRIMESandPOTEnCIA)as

itfocusesonsimulatingaspecificyearingreaterdetail.

METIShasamodularstructurethatmakes iteasytoextendthe

softwarethroughtheadditionofnewmodulesortheadjustmentof

existingones.Themodelrunsareperformedbysoftwarededicated

to large energy system optimisation7.Allcomponentsandmodules

aremanagedbyaplatform8providingacommonframeworkand

setofinteroperablelibraries.

Althoughintendedtobeadetailedoutput-tool,significantemphasis

isalsoplacedonitsuser-friendlinessandfastoperability.Theend

goalofMETISisthatitcanbeusednotonlybyexpertmodellers,

butalso(trained)policy-makersandanalysts.

WiththefirstversionofMETIShavingbeendelivered inJanuary

2015,newversionsareexpectedtobedeliveredgraduallyoverthe

nexttwoyears,includingadditionalmodellingcapabilitiesrelated

toelectricityandgasmarkets,heatanddemandsidemodelling.

METIS Studies

Paralleltothesoftwaredevelopment,theconsortiumwillbeproducing

studiesusingMETIS.Theseareintendedtobetechnicalstudiesof

around50pagesinlength,fullyexploitingtheavailablecapabilities

oftheMETISsoftware.Thescopeofthestudiesisthreefold:

(a) InvestigatetopicsthataredeemedimportantforDGENER,

7 Crystal Optimization Engine,propertyofArtelys.8 Artelys Crystal Platform, propertyofArtelys.

©iStock/drical

METIS:thenewshort-termenergysystemmodelexploredbythe Directorate-GeneralforEnergy

28

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

providingquantitativeresultsassociatedwiththeimpactofthe

examinedpoliciesoraspectsoftheenergysystem;

(b) Presentthecapabilityandappropriatenessofthesoftwareto

addresspolicyquestionsofinteresttoDGENER;

(c) ProvidereadytemplatesforDGENERinordertoperformsimilar

studiesinthefuture.

Which policy questions can METIS answer and which ones can’t it?

Uponfinaldelivery,METISwillbeabletoansweralargenumberof

questionsandperformhighlydetailedanalysesoftheelectricity,gas

andheatsectors.Itwillbepossibletotackleanumberoftopicswith

METISforthewholeEUand/orforspecificregions/MemberStates

(thelistbelowisindicative):

•• TheimpactsofmassRenewableEnergySourceintegrationon

theenergysystemoperationandmarketsfunctioning(forone

orallsectors);

•• Modellingofelectricityandgasmarketsunderdifferentmarket

designs;

•• Modellingofelectricityandgasflowsbetweenzones;

•• Cost-benefitanalysisofinfrastructureprojects,aswellasimpacts

onsecurityofsupply;

•• Generationadequacyanalysis;

•• Studyingthepotentialsynergiesbetweenthevariousenergy

carriers(electricity,gas,heat);

•• Whatisthecost/savingofaspecificmeasureforagivenyear?

•• Impactofnewenergyusages(e.g.electricalvehicles,demand

response)onthenetworkreinforcementandgenerationcosts.

Figure 9: METIS module architecture

Figure 8: Power model

Figure 7: Gas model

Source: METIS

Source: METIS

Source: METIS

29

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

OntheotherhandMETISisnotdesignedtoanswer(atleastinthe

initialstage)thefollowingquestions(again,thelistisindicative):

•• Anytypeofprojectionforthewholeenergysystem;

•• Optimal investmentplanning(capacityexpansion)fortheEU

generationortransmissioninfrastructure9;

•• Impactsofmeasuresonnetworktariffsandretailmarkets;

•• Short-termsystemsecurityproblemsfortheelectricityandgas

system(requiringapreciseestimationofthestateofthenetwork

andpotentialstabilityissues);

•• Flow-basedmarketcouplingandmeasuresontheredesignof

biddingareas.

9 TheplannedversionofMETISwillincludesomecapacityexpansioncapability,abletooptimisethecapacityofcertaintransmissionandgenerationassets.FutureversionsofMETISmayhaveadditionalcapabilities.

Transparency

METISwillbefullytransparentconcerningthemodellingtechniques

applied,withthefinalgoalofbeingabletooffertherelevantsource

codeandnon-commercialdatainputs.Furthermore,alltechnical

documentationandstudiesproducedwillbemadeavailable.

Fortransparencyreasons,alldeliverablesrelatedtoMETIS,including

alltechnicalspecificationsdocumentsandstudies,areintendedto

bepublishedonthewebsiteofDGENER10.

10 Onceoperational,theenvisagedlinkisexpecttobethefollowing: https://ec.europa.eu/energy/en/data-analysis/energy-modelling/metis

Kostis Sakellaris Kostis Sakellaris joined the Economic Analysis Unit of the European Commission’s Direc-torate-General for Energy in 2012. He has been working in the modelling team since then, heavily involved in a number of energy-related impact assessments and projects (METIS, Market Design, 2030 Energy and Climate Framework, Reference Scenario). Prior to that, he worked as an electricity market expert in the Markets and Competition Unit of the Greek Regulatory Authority for Energy. He graduated as a mathematician in Athens, Greece and holds an MSc in Mathematical Finance from the University of British Columbia, Canada.

Figure 10: Performing studies with METIS – screenshot from user interface

Source: METIS

30

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Giventheuncertaintyandcomplexityoftheenergysystem,quanti-

tativemodelsarevitaltoolstoexplorealternativescenariosandhelp

guidepublicpolicy.Yetmostmodelsanddataremaininscrutable

“blackboxes”–whethersmalleconometricmodelsorlargelinear

optimisationmodelswithhundredsofthousandsofinputvariables.

Incontrasttoclosedmodels,“open”modelsimplythatanyonecan

freelyaccess,use,modify,andsharebothmodelcodeanddatafor

anypurpose(OpenKnowledgeFoundation,2015).11Inthisarticlewe

arguewhyenergydataandmodelsurgentlyneedtobecomeopen;

discussthekeyreasonswhymanyarecurrentlynot;andpropose

somenextstepsfortheenergyresearchcommunity.

11 OpenKnowledgeFoundation,2015.OpenDefinition2.1-OpenDefinition-DefiningOpeninOpenData,OpenContentandOpenKnowledge[WWWDocument].URLhttp://opendefinition.org/od/2.1/en/(accessed3.15.16).

Why models and data should be open

Giventhecriticalguidancethatenergymodelsanddataprovideto

decision-makers,theyshouldbemadeopenandfreelyavailableto

researchersaswellasthegeneralpublic,forfourreasons:

1. Improved quality of science.Transparency,peerreview,repro-

ducibilityandtraceabilityleadtohigherqualityscience.Yetthese

principlesarealmostimpossibletoimplementwithoutaccessto

modelsanddata(DeCarolisetal.,2012;Nature,2014).12Human

errorisinevitableunderpressuretodeliver,andmodelmistakes

canhaveprofoundimplications.Forexample,theReinhart-Rogoff13

12 DeCarolis,J.F.,Hunter,K.,Sreepathi,S.,2012.Thecaseforrepeatableanalysiswithenergyeconomyoptimizationmodels.EnergyEconomics34,1845–1853.doi:10.1016/j.eneco.2012.07.004;Nature,2014.Journalsuniteforreproducibility.Nature515,7–7.doi:10.1038/515007a

13 Herndon,T.,Ash,M.,Pollin,R.,2014.Doeshighpublicdebtconsistentlystifleeconomicgrowth?AcritiqueofReinhartandRogoff.Camb.J.Econ.38,257–279.doi:10.1093/cje/bet075

Theimportanceofopendata andsoftwareforenergyresearchandpolicyadvice

©iStock/John Foxx

31

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

spreadsheeterrorarguablyskewedtheinternationaldebateon

austerity(Herndonetal.,2014)14.Suchincidentsserveaswarnings

againstpoorprogrammingpracticessuchasalackofauditing,

aswellasclosedmodelsanddata:itwasonlythroughsharing

thespreadsheetthattheerrorswerediscovered.

2. More effective collaboration across the science-policy

boundary.Betterandmoretransparentscienceitselfoughtto

enablebetterpolicyoutcomes.Academicpeerreviewroutinely

doesnotcheckmodelarithmeticanddatavalidity,andsoa

separateprocessofqualityassuranceisrequired.Whilemostly

absentfromacademicpractice,this isoftenimplementedas

aformalprocedure ingovernment(e.g.,DECC,2015)15.Unlike

academics,governmentsoftenmodelfornumbersratherthan

insight.Thespecificnumberscanbeofgreatsocietal impor-

tance,suchasthelevelatwhichtosetsubsidiesorthecostof

specificpolicies.Often,themostimportantaspectisthequality

ortransparencyof inputdata,ratherthanthenoveltyofthe

modellingmethodology.Inlargedatasetsusedingovernment

decision-making,traceabilityandreferencingcanbecomemajor

problems,ascivilservantsdevelopingmodelsanddataareoften

nottrainedscientists.Openlyavailable,collaborativelydeveloped

datasetsandreferencemodelswouldallowtheburdenofthis

worktobesharedmorewidely,andacrossbothacademiaand

government.

3. Increased productivity through collaborative burden shar-

ing.Collectingdata,formulatingmodelsandwritingcodeare

resource-intensive,whileresearchfundingandtimearescarce

resources.Societybenefits if researchersavoidunnecessary

duplicationandlearnfromoneanother.Individualresearchersgain

moretimetospendonpressingresearchquestionsratherthan

redundantworkonmodelordatasetdevelopment.Furthermore,

researchonlymattersifit isseenandused,andopen-access

publishinghasbeenshowntoincreasereadershipandcitations

(McCabeandSnyder,2014)16.Sinceopenlysharedcodeordata

ismorelikelytobeknowntoothers,itismorelikelytobeused

andfurtherimproved.Thisbenefitstheoriginalresearcherthrough

peerrecognitionandacademiccredit,andmovestheresearch

communityasawholeforward.

4. Profound relevance to societal debates.Reengineeringtheenergylandscapewillaffecteveryone,producingwinners

andlosers.Abalancedsocietalandpoliticaldebaterequires

transparentargumentsbasedonscientific justifications,but

14 Herndon,T.,Ash,M.,Pollin,R.,2014.Doeshighpublicdebtconsistentlystifleeconomicgrowth?AcritiqueofReinhartandRogoff.Camb.J.Econ.38,257–279.doi:10.1093/cje/bet075

15 DECC,2015.QualityAssurancetoolsandguidanceinDECC[WWWDocument].URLhttps://www.gov.uk/government/collections/quality-assurance-tools-and-guidance-in-decc(accessed6.2.16).

16 McCabe,M.J.,Snyder,C.M.,2014.IdentifyingtheEffectofOpenAccessonCitationsUsingaPanelofScienceJournals.EconomicInquiry52,1284–1300.doi:10.1111/ecin.12064

escalatingconcernaboutreproducibilityinsomefieldsisshaking

publicconfidenceinscientificresearch(Goodmanetal.,2016)17.

Finally,besidesthepracticalconsiderationsoutlinedabove,there

remainstheethicalargumentthatresearchfundedbypublic

moneyshouldbeavailabletothepublicinitsentirety.

Why they are (mostly) not open

Despitethesearguments,weseefourmainreasonswhyclosed

modelsanddatamayremainattractiveandrationalinsomecases:

1. Thereisarangeofvalidethicalandsecurityconcerns,particu-

larlywithdata.Researchersmayhaveaccesstodatacontaining

commercialsensitivitiesorpersonal information(particularly

relevantwhenmovingtowardsmoredecentralisedsmartgrids

withtheirfocusonindividualhouseholds).

2. Openlysharingdetailsofmodels,analysisanddatacancreate

unwantedexposure.Flawedcodeordatacandiscreditresearch

resultsandcauseembarrassmenttotheirauthors,butonlyifthey

arevisible.Somemayalsofearthatinexperiencedresearchers

willuseanopenmodeloropendatatoproduceflawedanalysis

thatreflectspoorlyonitsoriginalauthors.

3. It istime-consumingtowrite legibleandreusablecode,track

processingsteps,writedocumentationandrespondtofeature

requests.Becausemodelanddatasetdevelopmentare large

investments,itisoftenrationalforresearchersandinstitutions

tomaintain“tradesecrets”tocompeteforthird-partyresearch

funding:aclassicalcollectiveactionproblemwhereindividual

actorsaretrappedinasuboptimalnon-cooperativeequilibrium.

4. Finally,there issimple institutionalandpersonal inertia,often

alongsidecomplexanduncoordinatedinstitutionalsetups.

Whileunderstandablefromtheperspectiveof individualactors,

collectivelytheseengenderasenseofmistrustincomplex,impen-

etrablemodelsandenigmaticdatasets.Forexample,theEuropean

CommissionfacedcriticismforusingtheproprietaryPRIMESmodel

todeliverkeyresultsfor itsEnergyRoadmap2050(Helmetal.,

2011).18Moresignificantly,theUK’sdecarbonisationwasarguably

delayedforyearsbymodelsthatunderestimatedthescaleofthe

challengeduetoopaqueandheroicallyoptimisticcostassumptions

foronshorewind(HouseofLords,2005).19

17 Goodman,S.N.,Fanelli,D.,Ioannidis,J.P.A.,2016.Whatdoesresearchreproducibilitymean?ScienceTranslationalMedicine8,341ps12–341ps12.doi:10.1126/scitranslmed.aaf5027

18 Helm,D.,Mandil,C.,Vasconcelos,J.,MacKay,D.,Birol,F.,Mogren,A.,Hauge,F.,Bach,B.,vanderLinde,C.,Toczylowski,E.,Pérez-Arriaga,I.,Kröger,W.,Luciani,G.,Matthes,F.,2011.FinalreportoftheAdvisoryGroupontheEnergyRoadmap2050.

19 HouseofLords,2005.TheEconomicsofClimateChange.Vol.II.2005,HL12-IIof2005-06,QQ407-408.

32

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

What needs to be done

Individualresearchersandresearchgroupsmustunderstandthe

practicalitiesofopencodeanddata.Theserangefromissueslike

consideringtheintendedtargetaudienceandchoicessuchaslicensing

anddistributionchannels.Pfenningeretal.(2016)20giveguidance

specificallyforenergyresearch.Moreimportantly,theenergyresearch

communityasawholeneedstomoveforwardonseveralfronts:

1. Work towards reducing parallel efforts and duplication of

work.Thereshouldbebettercoordinationbetweendifferentmodellingefforts.Thiscanincludethedevelopmentofcommon

codebases,commondatasets,communitystandardstoensure

interoperability,andcoordinatedeffortstoenablethird-party

verificationofmodel-basedresults.

2. Increase transparency and reproducibility.Communityefforts

towardstestedanddocumentedcodepackagesforspecifictasks

canserveanimportantpurpose.Butone-offanalysescreated

forspecificpapers,orcodethatiswrittenwiththeunderstanding

thatitwillneverbemadepublic,maybepoorlydocumentedand

structured,meaningitsreleasewouldbeoflimiteduse.

20 Pfenninger,S.,Schmid,E.,Wiese,F.,Hirth,L.,Davis,C.,DeCarolis,J.F.,Fais,B.,Krien,U.,Matke,C.,Momber,I.,Müller,B.,Pleßmann,G.,Quolin,S.,Reeg,M.,Richstein,J.C.,Schlecht,I.,Shivakumar,A.,Staffell,I.,Trön-dle,T.,Wingenbach,C.,2016.Benefits,challengesandsolutionsforopenenergymodelling.OpenEnergyModellingInitiativeWorkingPaper.URLhttps://openmod-initiative.github.io/openmod-working-paper/(accessed1.7.16).

3. Change incentives and bring aboard different stakeholders. Theenergyresearchcommunityandspecificallytheemerging

openmodellingandopendatacommunitiesmustengagewith

otherstakeholderstoensureinstitutionalandacademicrecog-

nitionforopenenergymodels,andtostarttacklingtheharder

problemsthatfollow.Openandtransparent research isnot

currently incentivised: infact,theopposite isoftenperceived

asadvantageousforscientificcareeradvancement.Changing

theseincentiveswill requireeffortsnotonlyfromresearchers

themselvesbutalsofromtheiremployers,fromgrantagencies,

andotherstakeholderslikepublishers(Noseketal.,2015).21

Giventheimportanceofrapidglobalcoordinatedactiononclimate

mitigationandtheclearbenefitsofsharedresearcheffortsand

transparentlyreproduciblepolicyanalysis,thecommunitystillhas

muchworkahead.

21 Nosek,B.A.,Alter,G.,Banks,G.C.,Borsboom,D.,Bowman,S.D.,Breckler,S.J.,Buck,S.,Chambers,C.D.,Chin,G.,Christensen,G.,Contestabile,M.,Dafoe,A.,Eich,E.,Freese,J.,Glennerster,R.,Goroff,D.,Green,D.P.,Hesse,B.,Humphreys,M.,Ishiyama,J.,Karlan,D.,Kraut,A.,Lupia,A.,Mabry,P.,Madon,T.,Malhotra,N.,Mayo-Wilson,E.,McNutt,M.,Miguel,E.,Paluck,E.L.,Simonsohn,U.,Soderberg,C.,Spellman,B.A.,Turitto,J.,VandenBos,G.,Vazire,S.,Wagenmakers,E.J.,Wilson,R.,Yarkoni,T.,2015.Promotinganopenresearchculture.Science348,1422–1425.doi:10.1126/science.aab2374

©iStock/maxkabakov

33

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Joseph DeCarolis Joseph DeCarolis is an Associate Professor in the Department of Civil, Construction, and Environmental Engineering at North Carolina State University. His research is focused on the interdisciplinary assessment of technologies and public policies that promote long-term energy sustainability. He is particularly interested in developing robust decision-making strategies for climate mitigation by conducting analysis with technology-rich energy system optimisation models.

Sylvain QuoilinSylvain Quoilin obtained a European Doctorate in Mechanical/Energy Engineering in 2011 at the University of Liege (Belgium), where he then became lecturer. He is now a scientific officer at the Joint Research Centre of the European Commission, focusing on energy policy support and on the modelling of future European energy systems.

Iain StaffellIain Staffell is a lecturer in Sustainable Energy at Imperial College London, holding degrees in Physics, Chemical Engineering and Economics. His research centres around decarbonising electricity, ranging from the economics of battery storage and nuclear power to modelling the integration of renewables into electricity systems.

Dr Stefan Pfenninger Dr Stefan Pfenninger is a member of the Climate Policy Group at ETH Zurich, Switzerland. He holds degrees in environmental science, environmental policy, and environmental engi-neering. His research interests span the challenges and opportunities in the clean energy transition, particularly the modelling of futures with high shares of wind and solar energy.

Dr Lion Hirth Dr Lion Hirth is a post-doctoral researcher at the think tank MCC, a fellow at the PIK research institute, and director of the consulting firm Neon. His research interest lies in the econom-ics of wind and solar power. He has been concerned with open energy data in a number of projects and maintains the open-source power market model EMMA.

34

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

TheNordicEnergyTechnologyPerspectives(NETP)seriesassesses

howtheNordicregioncanachieveacarbon-neutralenergysystem

by2050.NETP2016marksthesecondeditionintheseries,thefirst

havingbeenpublishedin2013,andpresentstechnologypathways

towardsanear-zeroemissionNordicenergysysteminadditionto

in-depthscenariostailoredtoinformpolicy-makingintheregion.

TheanalysisconductedinNETP2016ispresentedaroundtheNordic

CarbonNeutralScenario(CNS),whichcallsforan85%reductionin

emissionsby2050(from1990levels).22Toachievethistarget,three

macro-levelstrategicactionsareelaborated.Thefirstofthesecalls

fortheplanningandincentivisationofaNordicelectricitysystem

that issignificantlymoredistributed, interconnectedandflexible

thanatpresent.Thesecondcallsfortheaccelerateddevelopment

oftechnologythatwillincreasethedecarbonisationoflong-distance

transportandtheindustrialsector.Finally,thethirdstrategicaction

22 TheNordic4°C(4DS)entailsa42%reductionandservesatthebaseline.

aimstotapintothepositivemomentumofcitiestostrengthen

nationaldecarbonisationandenergyefficiencyeffortsintransport

andbuildings.

Achieving a carbon-neutral energy system

TheNordiccountrieshavealreadydecarbonisedaspectsoftheir

energysystems,havingdecoupledCO2emissionsfromGDPgrowth

overtwodecadesago.Howeverthisprocesswillhavetopickupin

paceiftheCNSistobeachieved.Policyandtechnologyinnovationwill

becrucialinthisregard.Thepoliciesandtechnologiesimplemented

todatehavealreadycapturedthemostcost-effectiveopportunities

toweakenthelinkbetweeneconomicgrowthandemissions,leaving

greaterchallengesinsectorswhereprogresshasbeenmoredifficult.

TheCNSrequiresadramaticchangeinthecompositionofprimary

energysupply,coupledwithaggressiveenergyefficiencypoliciesthat

TheNordicEnergyTechnologyPerspectives2016

©iStock/moodboard

35

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

substantiallyreducedemand.Underthescenario,bioenergysurpasses

oilasthelargestenergycarrier,withtotaldemandforbiomassand

wasteincreasingfromalmost306millionMWhin2013toover444

millionMWhin2050.However,themostdramatictransformation

oftheNordicpowerandheatingsystemwillcomefromthecom-

binationofadeclineinnuclearandasignificantbuild-outofwind

power,resultingingenerationthatfarexceedsdomesticdemand,

evenwhenreducednucleargenerationisfiguredin.

Withanincreasefrom7%ofelectricitygenerationin2013to30%

in2050,windwilldisplacefossilandnuclear.Whilethetransition

ofheatingnetworksfromfossilfuelstoheatpumpsandelectric

boilerswilladdflexibilitytoanintegratedpowerandheatsystem,

theincreaseinwindgenerationwillputnewdemandsonhowthe

electricitymarketisorganised.Hydropowerwillbeincreasinglyval-

uableforregulatingthemarket,butwillnotsufficeonitsown.The

increaseinvariabilitywillrequirebalancingthroughacombination

offlexiblesupply,demandresponse,storageandelectricitytrade.

Increasedtradewillreducesystemcostsandenhanceflexibility,but

longleadtimesinsettingupinterconnectorsandstrengtheningthe

gridmaydelayachievingfullpotential.

Industrial sector decarbonisation the greatest challenge

The60%reductionintheCO2intensityofindustrycalledforinthe

CNSwillrequireaggressiveenergyefficiencycombinedwithother

measures,suchasswitchingfuelandfeedstockto lower-carbon

energymixes,andthedeploymentof low-carboninnovativepro-

cesses,includingCCS.Increasedinternationalcooperationwillalso

berequired,forexamplethroughinternationalcarbonpricingor

energyperformanceauditingmechanisms,asthesewillplayakey

roleinmitigatingtherisksofthelow-carboninvestmentsneeded

todecarbonise industry, therebyreducingpotential impactson

competitiveness.

AchievingtheCNSwillrequirea10%increaseininvestmentsover

thatneededforthe4DS22targetintheperiodfrom2016to2050,

representinganadditionalinvestmentofaboutEUR298billion.23

Thegreatestrelativeinvestmentincreasesarerequiredinbuildings

andindustry,withanincreaseof47%requiredinthefiveindustrial

sectorsanalysed,whichtogetheraccountfor80%ofthetotalfinal

energyusebyindustryintheNordicregion.Thisrepresentsacumu-

lativeinvestmentofaroundEUR27billion,mainlyassociatedwith

energyefficiencyimprovementsandthedeploymentoflow-carbon

innovativeprocesses.AtEUR179billion,thelargestshareofaddi-

tionalcumulativeinvestmentisaccountedforthebytransportsector.

23 US$333billion.

Radical transformation of transport

Transport,whichcurrentlyaccountsforalmost40%ofNordicCO2

emissions,deliversthegreatestemissionreductionintheCNS.Trans-

portrequiresadramaticemissionsslash–fromabout80million

tonnesofCO2in2013tojustover10milliontonnesin2050.This

targetcanbeachievedthroughathree-pronged‘avoid-shift-improve’

strategyofreducingtransportactivity(avoid),shiftingtomoreeffi-

cientorlesscarbon-intensivetransportmodes(shift)andadoption

ofmoreefficientorlesscarbon-intensivetransporttechnologiesand

fuels(improve).Improvementstotechnologiesandfuelswillplay

thelargestroleinthetransformationoftransport,largelybecause

avoidandshiftstrategieshavealreadybeendeployed.

Inthefaceofsteadilyrisingdemandfortransportservices, the

successoftaxationandsubsidyapproaches inpowerandheat

generationwillprovideasolidfoundationforsimilarlyassertive

policiesintransport.Consequently,transport’soverallenergyusein

theCNSwilldecreasebyover20%comparedto2000,despitea70%

increaseinoverallpassengerandfreightactivity.Underthescenario,

electricityaccountsfor10%offinalenergyuseintransportin2020,

butthankstothehighpowertrainefficiencyofelectricmotors,elec-

tricity’sshareoftransportactivityismuchhigher:64%ofroadand

railpassengerkilometresand42%ofroadandrailfreightactivity.

Furthermore,theCNSrequiresatriplingofthecurrentrateofimprove-

ment inspaceheatingenergy intensityofNordicbuildings.This

willbeachievedprimarilythroughthedeepenergyrenovationof

existingbuildings,whichwillconstitute70%oftheNordicstockin

2050.Energyefficiencygainsinbuildingscanunlockbiomassand

electricityforuseinothersectors,avoidinginfrastructureinvestments

inpowerandheatandCO2emissionsintransportandothersectors.

Nordic Energy Research

NordicEnergyResearchisanintergovernmentalorganisation

supportingandcoordinatingsustainableenergyresearchinthe

Nordicregion.Itistheplatformforcooperativeenergyresearch

andpolicydevelopmentundertheauspicesoftheNordicCouncil

ofMinisters.NordicEnergyResearch’sgovernancestructureis

closelyconnectedtoboththenationalpoliticalsystemsofthe

fiveNordiccountriesaswellastheintergovernmentalNordic

system.Thiscreatesaconstantinteractionbetweenresearch

strategies,resultsandkeyissuesonthepoliticalagenda.For

moreinformation,seeNordicEnergyResearch’sstrategy.

36

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Recommendations

TheNETP2016stipulatesthatgovernmentswillneed,individually

and inacoordinatedmanner, toplaya leadrole instimulating

actionstoachievethesettargets.Specifically,actionsinfourkey

areasareidentified.Governmentswillneedtostrengthenincentives

for investmentandinnovation intechnologiesandservicesthat

increasetheflexibilityoftheNordicenergysystem.Furthermore,

effortswillberequiredtoboostNordicandEuropeancooperation

ongridinfrastructureandelectricitymarkets.

Itwillalsobenecessarytoensurethelong-termcompetitivenessof

Nordicindustrywhilereducingprocess-relatedemissions.Forthis,

governmentswillhavetoacttoreducetheriskof investment in

low-carbonindustrialinnovationsandusepublicfundingtounlock

privatefinanceinareaswithsignificantemissionreductionpotential

butalowlikelihoodofindependentprivatesectorinvestment.

Finally,governmentsintheregionwillhavetoactquicklytoaccel-

eratetransportdecarbonisationbyusingprovenpolicytoolssuchas

congestioncharges,differentiatedvehicleregistrationtaxes,bonus-

malusregimesandalteredparkingfees.Atthesametimetheyshould

stepupinvestmentsincycling,publictransportandrailnetworks.

Implementationoftheseshort-termpolicyrecommendationswill

createtheframeworkconditionsfortheambitiouspartwayoutlined

bytheCNStobeachieved.

©iStock/Wavebreakmedia Ltd

37

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

OSeMOSYS: opensourcesoftware forenergymodelling

OSeMOSYSisafree,opensourceandaccessibleenergysystems

modelgenerator. Itcangeneratesmallvillageenergymodelsto

globalmulti-resourceintegratedassessmenttools.Itcanbeused

toassessenergysupplysecurity, investmentoutlooks,andGHG

mitigationstrategies.

Ingeneralterms, itcalculateswhat investmentstomake,when,

atwhatcapacityandhowtooperatethemtomeetthesaidpolicy

target(s)atthelowestcost.

Itisthereforeusedtodevelopmodelstoinformpolicydesignandis

partofthelargestHorizon2020LCE21energymodellingresearch

effortREEEM.org.Furthermore,itisusedtounderpinselectedoutputs

oftheDGEnergyInsightEnergy.orgthinktank,andisusedbynational

governmentsintheECandbeyondformediumtolong-termplanning.

OutsideofEurope,itwasusedfortheWorldBankandtheUnited

NationsEconomicCommissionforAfrica.Thereinthedevelopment

(andtradebetween)theelectricitysectorofeveryAfricancountry

wasanalysed.Intheformer,thefocuswasunderstandingthecli-

materesilienceofthesystemunderdifferentfutures.Inthelatter,

thescaleofinvestmentfortheworld’sfastestgrowingcontinent

wasquantified.AsimilareffortforSouthAmericaisbeingusedto

understandthatcontinent’s infrastructuredevelopment.Models

generatedhavebeenbroadanduseful.

OSeMOSYS.orgwaslaunchedatOxfordUniversity in2011ata

UKEnergyResearchCentre (UKERC)meetingandincludedco-authors

fromUniversityCollegeLondon(UCL),theUnitedNationsIndustrial

DevelopmentOrganization(UNIDO),UniversityofCapeTown(UCT),

StanfordUniversity,thePaulScherrerInstituteandothers.Itwasin

responsetotheobservationthatallcountriesneedtoassessthe

quantitativeevolutionoftheirenergysectorsduetoenergy’shighly

strategicroleindevelopment.Atthetimetherewerenoopensource

optimisingenergysystemmodelgeneratorsavailabletodoso.All

©iStock/mindscanner

38

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

aspectsofthetoolareopen,thisincludesthecode,themathematical

programminglanguageandthesolversused.

CurrentlyadministeredbyKTH(theRoyalInstituteofTechnology,

Sweden)itsuptakeinanacademicsettingisaccelerating.Scoresof

universities,academicpapersandagrowingnumberofgovernments

aretakingthetooluptouseinacademicandreal-worldanalysis.

Thisbodeswell,asthecommunitycontributingtoitsdevelopment

bothgrowsandaddscriticallyacademicallyreviewedadvances.

Importantly,italsoensuresthatthereareamultitudeof‘service

providers’shouldcommissionedstudiesneedtobeundertaken.

Italsofeaturesaspartofabroader initiative(called Optimus

Community) ledbytheUN.Furtherstarterdata-setsandoffthe

shelfmodelsarebeingdevelopedtohelpmakeopenrevieweddata

availabletobebuiltonrapidly. Initialdatasetsbeingdeveloped

includeallEU,AfricanandSouthAmericancountries.Theaimisto

covertheglobewithpeer-reviewedopenaccessdata.

Whilethere isagrowingcommunityofusersandapplications,a

specialfocusisEurope.FocusingontheSET-Plan,attheheartof

theREEEM.orgprojectareintegratedEuropeanenergysystemmod-

els.OneofthetoolsbeingdevelopedisanOSeMOSYSgenerated

model.Thatmodel(togetherwithamoredetailedMARKAL-TIMES

modeldevelopedbytheUniversityofStuttgart)willdeterminethe

costoptimaltechnologypathwaytomatchsupplywithdemandin

allEUcountriesintechnologicaldetail.Itwillprovidethebackbone

toatailoredevaluationoftheimpactofSET-Plantechnologies.

Todoso,informationfromseveralothermodelsisbeingintegrated.

TheOSeMOSYS-generatedmodelwillfocusonbeinganopen-source

engagementtool.Itwillreplicateandhighlightthekeyunderlying

dynamicsoftheintegratedEuropeanenergysystem.(Thiswillbe

complementedbye-learningtoolstobuildcapacitiesandshare

expertisebasedontheassessmentsperformedinthisproject.)Itwill

enableanswerstoquestionslikewhatresearchfundingandincreased

investmentcostwouldberequiredtomeetSET-Planobjectives,in

additiontosettingoutdetailedsub-targets-andtheirimplications.

ThemodelgeneratorcanbedownloadedfromOSeMOSYS.org.

Therein,resources,papers,datasetsandcodeinthefreemathemat-

icalprogramminglanguageGNUMathProgcanalsobedownloaded.

Recentlythecodehasbeentranslatedintomorethanonelanguage

(or‘technology’)andiscurrentlyavailableinGAMS(apopularlan-

guageamongsteconomists)andPython(awidelyusedopensource

language).YoucansignuptotheOSeMOSYSnewsletteronthe

website,wheretoolsarealsoavailabletodownload.Furthermore,in

thenextfewweeksanewinterfaceistobereleased,andaglobal

summerschoolistobelaunched.

Mark HowellsMark Howells directs the division and holds the chair of Energy Systems Analysis (KTH-dESA)

at the Royal Institute of Technology in Sweden. His group leads the development of some of

the world's premier open source energy, resource and spatial electrification planning tools;

he has published in Nature Journals; coordinates the European Commission's think tank

for Energy; is regularly used by the United Nations as a science-policy expert; and is a key

contributor to UNDESA's ‘Modelling Tools for Sustainable Development Policies'.

©iStock/boygovideo

39

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Sharedexperiences in integrated energy systems modelling

Akeychallengeinachievingasuccessfultransitiontoalow-carbon

Europeisimplementingthecorrectsuiteofpolicymeasuresthatare

basedonrobustevidence.Todaypolicy-makersacrossEuropedraw

onintegratedenergysystemmodelstoinformlongrangeclimate

mitigationandenergypolicychoices.EstablishedEuropeanmodels

suchasPRIMES,TIMES,MESSAGE,EnergyPLANandnewermodels

suchasPOTEnCIAandOSeMOSYSconsiderallmodesofenergy

(electricity,heatingandtransport)acrossallsectorsoftheeconomy

inanintegratedfashion,ratherthantreatingindividualmodesin

isolationwhichcanleadtopoorlyinformedinsights.

Ourresearchinintegratedmodellingstartedwithaspecificfocuson

thewiderenergysysteminIreland.WeusetheTIMESintegratedmod-

el,24whichisatechno-economicoptimisationframeworkdeveloped

overthepast40yearsthroughtheInternationalEnergyAgency’s(IEA)

EnergyTechnologySystemsAnalysisProgram(ETSAP).Themodel

allowsuserstogeneratefutureenergysystempathwaystomeet

energyneedsatleastcost,subjecttouserdefinedconstraints.TIMES

optimisesforenergyservicedemands(i.e.theutilitywegetfrom

24 TheTIMESIntegratedEnergyModelofIrelandwasfundedbytheEnvironmentalProtectionAgency(EPA)andSustainableEnergyAuthorityofIreland(SEAI).

©iStock/Varijanta

40

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

energyuse)suchaslighting,heating,passengerkilometres,tonnes

ofsteelandcementetc.This issignificantbecauseasasociety

wedon’tintentionallyuseenergy,butratherhaverequirementsfor

mobility,lighting,goodsetc.Inall,TIMESconsidersawiderangeof

over1,300technologiesinthetimeframeto2050fromlightbulbs,

cars,fridges,heaters,boilers,powerplants,biorefineriesetc.

Early inourresearchwerecognisedanumberof limitationsto

ourmodellingtechniquesandidentifiedkeyareasthatrequired

improvement.Oneoftheseareaswashowintegratedmodelslike

TIMESdealtwithvariablerenewablegenerationsuchaswindand

solarpower.Manyintegratedmodelshaveasimplifiedtemporaland

technicalresolutionofthepowersysteminordertokeepproblems

computationallymanageable,howeverthiscomeswiththetrade-off

ofpoorrepresentationofvariabilitywithinthemodels.Toresolvethis

issue,wedevelopedsoft-linkingtechniquestolinktheenergysystem

modeltodedicatedpowersystemmodels.Thisallowsustoleverage

thestrengthofhigh-resolutiontechnicalandtemporalpowersystem

models.Indoingthiswecouldaccountforgreatertemporalresolution

(15minuteorhourlysimulations)andalsocaptureimportanttechnical

characteristicsofthepowerplantssuchasramprates,startcosts

etc.WerecentlyexpandedthesetechniquestoincludethefullEU28

powerandgassystemswithwaterasournexttargetfordevelopment.

Thegeographicalexpansionoftheresearchwaspartiallydrivenby

theneedtomodelgreater interconnectedmarkets(bothgasand

electricity)withintheEUandalsotounderstandthedistributionof

effortfordecarbonisationacrossallEUMemberStates.Soft-linking

techniqueshastheadvantagethatitallowsustoverifythetechnical

robustnessofsimulationsbutcomeswiththechallengethatanextra

modelmustbemaintainedanditrequiresmodellerjudgementon

feedbacktotheenergysystemmodel.

Anotherchallengewastheintegrationofland-useandagriculture

intoourmodels.IrelandisuniqueinEuropeasover30%ofGHG

emissionscomefromagriculture.Thisresearchrequiredustowork

closelywithagriculturalscientiststodevelopaframeworkwhere

wecouldaccountfortheseemissionsandmodeltheinteractions,

particularlyforlandusecompetition,betweentheenergysectorand

agriculturalsector.Ourcurrentresearchonlanduseandagriculture

hasanimportantfocusontheroleofbioenergyandtheimplications

ofindirectlandusechange(ILUC)andrecentamendmentstothe

RenewableEnergyDirective(socalled ‘ILUCdirective’).Whilethe

scienceofILUCisatanearlystageourinitialresultspointtoincreased

costsfordecarbonisationwhenILUCisconsidered.

©iStock/chombosan

41

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

Dr Paul Deane Dr Paul Deane is a research fellow with the Energy Policy and Modelling Group in University College Cork in Ireland. He has been working in the energy industry for approximately 15 years in both commercial and academic research. His research activities include integrated energy systems modelling to assess holistic pathways to low carbon energy futures. Paul is also a member of the Insight_E group which is a European, scientific and multidisciplinary think-tank.

LikemanyresearchgroupsacrossEurope,wehaveseenourinte-

gratedmodelsexpandinsizeandcomplexity.Complexityisunavoid-

ableinsuchlargemodelsduetothemulti-dimensionalandintricate

natureofenergysystems,butcomplexityhastobebalancedwith

theinherentuncertaintyin longrangeinputssuchasfuelprices

andmacroeconomicestimations.Thechallengeofmakingmodels

computationallymanageablehasoftenforcedusto lookatour

simplificationsandheuristicsandaskthequestion,“arewemaking

ourmodelsbetterorjustgettingthewronganswerquicker?”Arecent

focusofourresearchistryingtounderstandwhatlevelofcomplexity

isappropriateinlongtermmodelsgiventheuncertaintyininputs,

andtryingtounderstandhowthisvalueofcomplexitydiminishes

aswelookintothefuture.Wehavefounditbeneficialtoexplore

multiplepathwaysandseekoutcommonalitybetweenpathways

ratherthanfocusondeterministicsolutions.

High-performancecomputingoffersexcitingpossibilitiesforfurther

developmentofintegratedmodelling,howevermanyofthecurrent

architectureprocessesarechallengingtoparallelise.Projectslike

‘BEAM-Me’inGermanyareinvestigatingthepotentialforhigh-per-

formancecomputingtoenhanceenergysystemmodels,anditwill

beinterestingtoseewhatdevelopmentsoccur.

Aboveallwehavelearnedthatmodellingthefutureisahumble

scienceandgreatcaremustbetakennottoconfusemodelinsights

forpredictions.Humanbehaviour,economicvolatilityandtechnol-

ogyreadinessarebutasmallsectionofelementsthathavebig

influenceonresultingpathwaysfrommodels.Wemustbeaware

thattheboundariesoftheenergysystemdon’tstopattheendof

thepipelineorcable;theyextendintoourlives,communities,and

wellbeing.Currentmodellingeffortsprimarilyhaveatechno-eco-

nomicfocus,howeverthechallengeofdecarbonisation,andmore

recentlythegreaterlevelofdecarbonisationrequiredbytheParis

Agreementwillrequireourmodellingcommunitytolookoutwardto

otherdisciplinestoinformpathwaysthatweasasocietyarewilling

totravelontogether.

©iStock/SundayMorning014

42

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

What does Energy System Integration involve and what value does it bring?

Energy Systems Integration (ESI) is the process of coordinating

the operation and planning of energy systems across multi-

ple pathways and/or geographical scales to deliver reliable,

cost-effective energy services with minimal impact on the

environment.

Energysystemshaveevolvedfromindividualsystemswith little

ornodependencies intoacomplexsetof integratedsystemsat

scalesthatincludecustomers,cities,andregions.Thisevolutionhas

beendrivenbypolitical,economic,andenvironmentalobjectives.As

wetrytomeetthegloballyrecognisedimperativetoreducecar-

bonemissionsthroughthedeploymentoflargerenewableenergy

capacitieswhilealsomaintainingreliabilityandcompetiveness,

flexibleenergysystemsarerequired.Thisflexibilitycanbeattained

throughintegratingvarioussystems:byphysically linkingenergy

vectors,namelyelectricity,thermal,andfuels;bycoordinatingthese

vectorsacrossotherinfrastructures,namelywater,data,andtrans-

port;byinstitutionallycoordinatingenergymarkets;and,spatially,

byincreasingmarketfootprintwithgranularityallthewaydownto

thecustomerlevel(Figure11).

ESIisamultidisciplinaryarearangingfromscience,engineering,and

technologytopolicy,economics,regulation,andhumanbehaviour.

ItisthisfocussimultaneouslyonbreadthanddepththatmakesESI

suchachallengingandexcitingarea.

ESIisoneofseveralglobalsocialandengineeringtrendsthatwill

shapethesolutionstothekeychallengesofthenextdecades:

resourcestress,climatechange,megacities,urbanisation,cyberse-

curity,andinfrastructureresilience.ESIisanumbrellaconceptthat

encompassesactivitiestackledinthecontextofsmartgrids(grid

MarkO’MalleyDirectoroftheInternationalInstituteforEnergySystemsIntegration

TALKS TO SETIS

©Thinkstock

43

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

modernisation)andsmartcities.However,thesetwoapproachesare

morelimited,withonefocusedonasingleenergyvector(electricity)

andtheotherlimitedingeographicalscaletoacity–sotheymay

missimportantopportunitiesthatcanarisebyconsideringallenergy

vectorsandallscales.

The value of ESI is in coordinating how energy systems

produce and deliver energy in all forms to reach reliable,

economic, and or environmental goals at appropriate scales.

Analysis and design of integrated energy systems can inform

policymakers and industry on the best strategies to accom-

plish these goals.

What are the principal objectives of the European Energy Research Alliance, Joint Programme on Energy Systems Integration and the International Institute for Energy Sys-tems Integration?

TheimportanceofESIisbeingrecognisedglobally.Mostsignificantly,

ESIisacentralthemerunningthroughtheEuropeanCommission’s

StrategicEnergyTechnologyPlan(SET-Plan)IntegratedRoadmap.

It isalsoacentralthemeoftheCleanEnergyMinisterialanda

majorresearchthemewiththeU.S.DepartmentofEnergynational

laboratorycomplex.

InFebruary2014theUSDepartmentofEnergy,NationalRenew-

ableEnergyLaboratoryandPacificNorthwestNationalLaboratory

co-hostedaninvitationonlyworkshoponESI inWashingtonDC.

Therewere40seniorlevelattendeeswith29fromtheUS,10from

EuropeandonefromChina.Theworkshopwasdesignedtovalidate

theimportanceofESIasanemerginginterdisciplinaryscientificarea

andgaugetheappetitefortheestablishmentofaninstitute–the

InternationalInstituteforEnergySystemsIntegration(iiESI).Itwas

agreedbyallparticipantsthatESIisanimportantandemergingarea

andthatforminganorganisationsuchasiiESIwasverypositiveand

timely.TheroleandvalueofiiESIinfosteringinternationalcollabora-

tion,stimulatingthesharingofknowledgeandprovidingindependent

analysiswasrecognisedbyall.TheindependenceofiiESIwasseen

asafundamentalcharacteristic,inparticularwithrespecttovaluing

ofparticulartechnologies/solutionsdeployedintheenergysystem.

iiESIasaformalorganisationcameintobeinginJuly2016asa

globalinstituteaimedattacklingthechallengesofenergysystems

integrationthroughglobalcollaborationandeducation.Formalising

iiESI as a global, member-driven organisation of leading ESI

scholars and practitioners provides a structure for leveraging

each other’s experiences and expertise, coordinating research

agendas, and sharing best practices from around the world.

Theestablishmentofaformal institutewillallowthegroupto

expandandgrowtomeetthechangingneedsoftheESIcommunity.

Figure 11: Energy Systems Integration

Source: iiESI

44

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

ThereisalsoanewEuropeanEnergyResearchAlliance(EERA)Joint

Programme(JP)inESI.AnEERAJPiscreatedbyinterestedorgan-

isationsthatdefineajointresearchagendaforatopicincludedin

theSET-Plan.EERAJPscoordinateresearchbasedonthepartic-

ipating institutionsownresources. Inaddition,theJPcanobtain

supplementaryfundingfromnationalorEUsources.Theaimis

tograduallyincreasetheamountofdedicatedfundingtotheJPs.

ThiswillallowaJPtowidenanddeepencoordination.EERA JP ESI

seeks to bring together research strengths across Europe to

optimise our energy system, in particular by benefiting from

the synergies between heating, cooling, electricity, renewable

energy and fuel pathways at all scales.Theenergyelementsof

thewaterandtransportsystemarealsoincluded,asistheenabling

dataandcontrolnetworkthatenablestheoptimisation.TheEERAJP

ESIisdesignedtodevelopthetechnicalandeconomicframework

thatgovernmentandindustrieswillneedtobuildthefutureefficient

andsustainableEuropeanenergysystem.

What are the main ESI research challenges and how can they be overcome?

InMarch2015,ImperialCollegeLondonhostedaniiESIworkshop,on

ESIResearchChallenges.Thiswasattendedby38expertsfromEurope,

USA,Africa,Asia,RussiaandAustralia.Thedisciplinesrepresented

rangedfromEngineering,Economics,SocialSciences,Mathematicsand

Physics,andindustrywasalsorepresented.Notsurprisinglyoneofthe

mainoutcomesoftheworkshopwasthe“needtocombineeconomic,

social,andpoliticalperspectiveswithengineeringknowledge”.The

needforeducationanddisseminationfeaturedstrongly.Intryingto

identifyresearchchallengeshowever,theneedforcleardefinitionsof

ESIandof“optimality”inanintegratedenergysystemwasapparent.

Therewaslittleornoconsensusontheoptimalityissuedespitesome

followupteleconferencesandemailexchanges.

EachenergysystemwillapproachESIfromadifferentstartingpoint

(e.g.,anurbanareainthedevelopedworldwillhaveadifferent

approachcomparedtoaruralareainthedevelopingworld).It is

crucialtodefinethegeographicalscopeaswellasthecomponents,

theboundaries,andtheinfluenceofthesurroundings.Forexample,

renewableintegrationisthedrivingforceofESIinmanyregions,but

notall.Insomeregions,themaindriversareincreasedcombined

heatandpower(CHP),increasedefficiency,ashiftfromcoalgen-

erationtonaturalgas,orsimplyelectrification.Differentincentives,

decision-makingprocesses,andaccesstocapitalduetolocation

orscalewillresultinverydifferentenergysystemsandapproaches

toESI(e.g.,agovernmentcaninvestinhigh-voltagetransmission,

whileindividualswillnot).Aseachenergysystemdevelops,itwillbe

necessarytoconstantlyre-evaluatethesysteminordertoassess

howitisbestcoordinated.

DevelopingcoordinatedsystemsthroughESIanalysisrequiresa

properunderstandingofthedifferentactors involved,alongwith

theirmotivations,their incentives,andtheinformationtheyhave

accessto.Fromawhole-systemperspective,theactors ineach

energydomaintendtoactontheinformationtheyhaveinways

thatmaximisebenefitsfortheirdomain,butnotfortheentireenergy

system.Forexample,eachuserconsumesbasedontheirown

requirements,eachmarketvaluescertainfinancialoutcomes,and

eachgovernmentservesitsownsocialorpoliticalmotivations–but

theremaybenocoordinationacrossthesedomainstodetermine

thebestoptionforallactorsinvolved.Pooroutcomescanpotentially

arisefromthislackofinformationand/orcoordination,andmaynot

bemonetaryinnature;apoorlyexecutedenergytransitioncould

resultinenergysystemsthatlacktechnicalintegrity,socialequity,

and/orpoliticalacceptability.

TheconsiderationsthatgovernESIarenumerousandcomplex,

andtheoutcomesandtheirvaluecanbedifficulttodefine.Oneof

thefirststepstodeterminethisvalueistodefineasetofrobust

metricsspanningtheengineeringandsocialsciences(e.g.,financial

impacts,emissionscosts, resiliency,publichealthconsiderations,

socialutility,etc.)tomeasureandhighlightthevariousbenefits.

Anysetofdefinitionsormetricswillhavetobeflexibleenoughto

accommodateawiderangeofcircumstances.Metricsalsoneedto

besimpleenoughtoallowforanoverallholisticunderstandingof

howthedifferentaspectsinteract.

The main outcome of the London Workshop is the need for the

global research community to adopt a common and clearly

understood common language and consensus on the scope

of the ESI. This is needed before a detailed interdisciplinary

research roadmap for ESI can be articulated with confidence.

BecauseESIisabroadtopicthatincludesalltypesofenergysources

andend-useapplications, it ishelpfultocategoriseexamplesof

ESIintoafewareas.HereweprovideseveralexamplesofESIthat

havebeenorganised intothree“opportunityareas”:streamline,

synergise,andempower.

Streamlinereferstoimprovementsmadewithintheexistingenergy

systembyrestructuring,reorganising,andmodernisingcurrentenergy

systemsthroughinstitutionallevers(i.e.,policies,regulations,and

markets)or investment in infrastructure. Increasingtheflexibility

ofenergyendusehaspotentialsystem-widebenefitsandcould

createnewmarketsforproductsandservices.However,capturing

thesebenefitswillrequireproperregulatoryandmarketstructures,

newoperationalandplanningparadigms,physicalenergynetwork

characteristics,anintegratedcommunicationssystem,andsuitably

flexibleend-useproducts.Manyofthesearecurrentlylackinginthe

45

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

existingenergysystemandrequireasystem-wideunderstanding

todeliverpragmaticandsustainablesolutions.Developingmore

integratedenergysystem-widepolicieswillenablebettermanage-

mentofuncertainties.

Moreintegratedenergynetworksandproperfunctioningreal-time

locationalmarketswillrewardcapacityandflexibility.Inaddition,the

removalofinstitutionalbarriersbetweendistributionandtransmission

systemswillallowbetterintegrationofdistributedresourcesand

facilitateregional integration.Byprovidingstandardisedrequire-

ments,updatedinterconnectionandinteroperabilitystandardsand

gridcodeswillstreamlinetheenergysector.

Investmentintheappropriateinfrastructurewithintheintegrated

energysystemwill improveflexibility.Expansionoftheelectrical

transmissiongridwillenableflexibilitybyaggregationacrossscales.

Pipelineinfrastructureisrequiredtoincreasethepenetrationofbio

and/orsyntheticfuels.Investmentindatainfrastructurewillenable

consumerstomorefullyparticipateintheenergysystemandwill

improveenergynetworkoperationsthroughforecastingandanalytics.

SynergisedescribesESIsolutionsthatconnectenergysystems

betweenenergydomainsandacrossspatialscalestotakeadvantage

ofbenefitsinefficiencyandperformance.Todate,thecouplingof

heatandelectricitysectorshasfocusedonthesupplyside(e.g.,CHP)

forfuel-savingpurposes.However,atthesystemlevel,itsinherent

inflexibilitycanleadtosub-optimaloverallsystemperformance.A

goodexampleofthisiswindcurtailmentinChina,whichisinpartdue

totheinabilityofphysicallyinflexibleCHPplantstoreduceelectricity

productionwhileprovidingheat.ESIsolutionsthatintegrateheat

storageintotheCHPplantarebeingdevelopedandindicateashift

fromthesupplysidetothedemandside(e.g.,electricalheatingof

water,thermalstorageinbuffersandheatpumps). It ispossible

tocapitaliseon“virtualstorage”wheretheflexibility inonepart

ofthesystem(e.g.,heat,transport,water,etc.)canbeintegrated

with,forexample,theelectricitysystem,andusedinasimilarman-

nertoelectricitystorage.Thisvirtualstoragecanbesignificantly

cheaperthandedicatedstorage,asitdoesnotrequirelargecapital

investment–butitdoesrequireamoreintegratedenergysystem.

Demandmanagement(e.g.,controllingheatingandcoolingloads)

technologiescurrentlybeingdeployedanddevelopedare inpart

leveragingthisvirtualstorage.However,ESIproposesthatitisata

grandscalewherefuel,thermal,water,andtransportsystemswill

besystematicallyplanned,designed,andoperatedasflexible“virtual

storage”resourcesfortheelectricitygrid(andviceversa).Thereis

alsothepotentialtousethenaturalgasfuelgridtocreateenergy

storagethroughthe“power-to-gas”concept.

EmpowerreferstoESIactionsthatincludetheconsumer,whether

throughtheirinvestmentdecisions,theiractiveparticipation,ortheir

decisionstoshiftenergymodes.Investmentsinenergyefficiencyare

increasinglyrecognisedasacost-effectivewaytoreduceenergy

demandandcanleadtosystem-widebenefitsthatincludeupstream

capitalandoperationalsavings.Fromanoverallenergysystempoint

ofview,energyefficiencyatthelevelofanindividualbuildingmay

beinconflictwiththeflexibilitythatthedemandsidecanprovideto

thegrid.Energyefficiencyimprovementsortargetsalsocontributeto

broadersocialandpolicygoals,notablymacro-economicefficiency,

©iStock/taka4332

46

S E T I S M a g a z i n e N o v e m b e r 2 0 1 6 - E n e r g y S y s t e m s M o d e l l i n g

industrialproductivity,publicbudgetbalance,securityofsupply,and

healthbenefits.Thisbuilding-levelinvestmentneedstobemade

bytheconsumer.Theformerlytotallyseparatedsectorsoftrans-

portandelectricitymaybecomemoreintegratedthroughplug-in

electric(hybrid)vehiclesandcarbatteries,buttheconsumerneeds

toacceptthismodeoftransport.Thepotentialinsomeregionsfor

thermalgridshasbeenraised,butquestionsremainastohowlarge

theyshouldbe,howbesttointegratethemintotheelectricitygrid,

and,importantly,howconsumerrequirementswillbeensuredand

whetherconsumerswillacceptthem.

What is the role and main requirements of modelling in ESI?

ModellingplaysacriticalroleinESIresearch.Modellingisameans,

notagoalinitself.

ESIismostvaluableatthephysical,institutional,andspatialinter-

faces,wherethereareinteractionsandnewchallengesandoppor-

tunitiesforresearch,demonstration,anddeploymenttoreapits

commercialandsocietalbenefits.Thereforetheseinteractionsmust

beunderstood,quantified,analysedandthensolutionsdesigned

anddeployed.Asthesystemsarecomplex,typicallydistributed

withphysical,economicandregulatoryaspects,itisonlypossible

toinvestigatethemeffectivelyandatreasonablecostbyusinggood

models.Thesemodelsneedtofocusontheinterfacesandwillneed

torepresentallmajorenergyproducingandconsumingsectors

withsufficienttemporalandgeographicalgranularitytobeableto

trulyrepresenttheESIchallengesandopportunities.Ofparticular

importanceisuncertaintyinoperationsandininvestmenttimescale,

whichneedstobecaptured.Theneedforhighqualitydatacannot

beoveremphasised.Modelsareonlyasgoodasthedatathatis

usedtotunemodelparameters,validatemodels,developscenarios

andinputdatasetsetc.

Mark O’Malley Mark J. O’Malley received B.E. and Ph.D. degrees from University College Dublin, Ireland, in 1983 and 1987, respectively. He is the Full Professor of Electrical Engineering at University College Dublin, is the Director of the University College Dublin Energy Institute and Electricity Research Centre and is a member of the Royal Irish Academy. He is also the Director of the International Institute of Energy Systems Integration and the coordinator of the European Energy Research Alliance Joint Programme in Energy System Integration. His research area is energy systems integration in particular grid integration of renewables.

Thesemodelsallowustoaddressunanticipatedfeedbacksinthe

system,identifyefficientstrategies,evaluatepossiblemarketdesign

andpolicies,etc.Modellingisneededtounderstandhowtoachieve

costeffective integrationofenergysectors,whatarethemost

promisingnewpathwaysandtechnologiesandhowthesystem

performancemaychangeunderdifferentscenariosandpolicies.

Modellingthereforeneedstosimulatethephysicalsystemaswell

astheenergymarket,regulatoryframework,underlyinguncertainty

inweatherandlonger-termresourcesandallthewaytoconsumer

behaviour,andhowtheactors’decisions(operationalandinvestment

decisions)affecttheperformanceofthephysicalsystem,andhow

regulationaffectstheactors’decisions.

Anextremelywidesetofdiversemodelsdoexist.However,focus

typicallyisonsectorsandenergycarriers,individually.Overallenergy

sector(oreconomywide)modelsexist,butoftenlacktechnicaldetail,

crucialtoaccountforthevariabilityofrenewableenergysources

suchaswindandsolarphotovoltaic.ThescopeofESImodelsneeds

tobelargerthanthatoftraditionalmodels.Theidealmodelwould

includealltheabove-mentioneddimensions,thephysicsaswellas

themarket,butthisisneitherfeasiblenorpractical.As a result,

the challenge is to develop a suite of models than can be

used together. Preferably this should be made in a way that

enables much better co-operation between model developers

and users across the globe. Well-defined interfaces between

models, open source code and high quality open source data

would help to avoid duplicate effort.Differenttypesofmodels

areneededfordifferentquestions:simulationandoptimisation,

shorttermandlongterm,physicalandmarketmodels.

48SETIS

LD-AD

-16-003-EN-N