EssaysonvesselemissionsandexternalitycostsinLasPalmasPort

MilušeTichavskaDepartmentofGeography

UniversidaddeLasPalmasdeGranCanaria

Supervisor:Prof.BeatrizTovar

AthesissubmittedforthedegreeofDoctorofPhilosophy

AmispadresSilviaNaranjoyRadkoTichavsky,amishermanosVladislavyVojtech,amitíaCeciliaysobretodo,aPaulaSánchez‐Cabezudoconmuchocariño.

AgradecimientosNo podría comenzar esta sección sin primero expresar elmás sincero e infinitoagradecimientoamidirectoradetesislaProf.BeatrizTovaryaque,esteprocesode formación no habría llegado a los objetivos propuestos de no ser por suinestimableapoyopersonalyporlaexcelenciaprofesionalquelecaracteriza.Porotraparte,megustaríadargraciasalProf. FranciscoCabrera, alProf.VíctorAraña y a todos los miembros de Instituto para el Desarrollo Tecnológico y laInnovaciónenComunicaciones(IDeTIC)porhabermepermitidoconstruirenpartemitesisensusinstalacionesyencolaboraciónconsuequipodetrabajo.Tengolacertezadequeelambientepositivo,emprendedorydecolaboraciónalosqueestuveexpuesta durante nuestro periodo de colaboración inicial, han resultado sertambién,unfactorclaveeneldesarrolloylaculminacióndeestetrabajo.También me gustaría agradecer el apoyo y la colaboración de investigadoresinternacionalesqueheconocidoalolargodelcamino.EnparticularalDr.DimitrisLekkas,alDr.ArgyriosStasinakisyalSr.DemitrisMemosyelequipoMarineTraffic.TambiénalDr.Jukka‐PekkaJalkanenyalSr.LasseJohansson,miembrosdelGrupodeCalidaddelAireenelInstitutoMeteorológicodeFinlandia;alDr.TristanSmithalequipodeinvestigacióndetransportemarítimodelInstitutodeEnergíaenlaUCL;alDr.ErnestosTzannatosdelaUniversidaddelPireo,alaDra.SuSong,delInstitutode Recursos Mundiales; al Dr. Spiros Papaefthimiou y la Srita. AlexandraMaragkogianni de la Universidad Técnica de Creta. Por otra parte,me gustaríaextenderelmásespecialdelosreconocimientosmisamigosyfamiliaresinmediatos,yaquehanrepresentadounadelasmotivacionesmásgrandeshaciaquemehanllevadoentreotrascosas,alarealizacióndeestegradoacadémico.Por último, doy las gracias por el apoyo financiero y los facilitados por laUniversidaddelasPalmasdeGranCanaria,laAutoridadPortuariadeLasPalmasyMarineTraffic.

AcknowledgementsIcouldnotstartthissectionwithoutfirstlyexpressingmymostsincereandinfinitegratitude to my supervisor Prof. Beatriz Tovar. None of this would have beenpossiblewithoutherinestimablepersonalsupport,andprofessionalexcellence.Moreover, I would like to thank Prof. Francisco Cabrera, Prof. Victor Araña,members of the Institute for Technological Development and Innovation inCommunications (IDeTIC) for allowingme to partly buildmy thesiswithin theirpremises. I am certain that the positive, entrepreneurial and collaborativeenvironment to which I was exposed has represented a milestone within thisprocess.I would also like to acknowledge the support and collaboration of internationalresearchersthatIhavemetalongtheway.ParticularlytoDr.DimitrisLekkas,Dr.ArgyriosStasinakis,Mr.DemitrisMemosand theMarineTraffic team.Also toDr.Jukka‐Pekka JalkanenandMr.Lasse Johansson fromtheAirQualityGroupat theFinnishMeteorologicalInstitute;toDr.TristanSmithfromandtheShippingGroupattheUCL,EnergyInstitute;Dr.ErnestosTzannatosfromtheUniversityofPiraeus,Dr.SuSongfromtheWorldResourcesInstitute;Dr.SpirosPapaefthimiouandMs.AlexandraMaragkogiannifromtheTechnicalUniversityofCrete.Furthermore, I would like to extend the most special of acknowledgements toimmediate friends and family, as they have been one of the strongest driverstowardspersonalsuccessandthecompletionofthisacademicdegree.Finally, I would like to expressmy gratitude for the financial and data supportprovidedbyUniversidaddelasPalmasdeGranCanaria,thePortAuthorityofLasPalmasandMarineTraffic.

ContentsListofTablesListofFiguresIntroducciónyresumengeneral…………………………………………………………………..15Introductionandsummary……………………………………………………………………….…...27Chapter1.Port‐cityexhaustemissionmodel:AnapplicationtocruiseandferryoperationsinLasPalmasPort…………………………………………………………………………37

1.1.Introduction…………………………………………………………………………..……...371.2.Regulationandpracticesforthecontrolofshipexhaustemissionsinports……………………………………………………………………………………………………381.3.Methodology………………………………………………………………………………….41

1.3.1.TheSTEAManditsmaincomponents……………………………….421.3.2.Inputvalues…………………………………………………………………….441.3.3.Modelperformanceanduncertaintyconsiderations………….45

1.4.Casestudy:LasPalmasPort…...………………………………………………………461.4.1.Results……………………………………………………………………………..481.4.2.Geographicalcharacterizationofresults…………………………...55

1.5.Discussionandpolicyimplications…………………………………………………571.6.Conclusionsandfutureresearch………………………………………………….....611.7.References…………………………………………………………………………………….63

Chapter2.Externalcostsofvesselemissionatport:areviewofthemethodologicalandempiricalstateoftheart………………………………………………….67

2.1.Introduction………………………………………………………………………………….672.2.Methodology:ImpactPathwayApproachtoairpollution(abottom‐upapproach)……………………………………………………………………………………………712.3.Literaturereviewonexternalcostsderivedfromairemissionsinshipping:atop‐downapproach…………………………………………………………….762.4.Conclusionandfutureresearch……………………………………………………….842.5.References……………………………………………………………………………………..87

Chapter3.Environmentalcostsandeco‐efficiencyfromvesselemissionsinLasPalmasPort…………………………………………………………………………………………………….91

3.1.Introduction………………………………………………………………………………….913.2.ExternalCostEstimation:MethodologicalAspectsandLiteratureReview……………......……………………………………………………………..……………….953.3.ResultsandDiscussion..…………………………………………….………………….102

3.3.1Externalcostestimation…………………………………………………..1023.3.2.Eco‐efficiencyindicators……………………………………………..…..108

3.4.ConclusionsandFutureResearch………………………………………………....1133.5.References………………………………………………………………………………..….117

ListofTablesChapter1

1.1.STEAMmodelregardingthepropertiesofships………………………………………….451.2. Emissions and operative time related to traffic sectors as predicted by theSTEAM……………………………………………………………………………………………………………491.3.ExhaustemissionsandoperativetimeofcruiseandferryvesselsaccordingtosizeclassesaspredictedbytheSTEAM…………………………………………………………….53Chapter2

2.1. Top‐down external cost estimation derived from air emissions inshipping………………………………………………………………………………………………………….82Chapter3

3.1.Summaryofpreviouspapersonmeasuringexternalcostderivedfromvesselemissionsatport…………………………………………………………………………………………...1003.2.Externalcostfactorsusedinthiscasestudy………………………………………………1053.3.Estimatedexternalcosts(€)……………………………………………………………………..1063.4.Overallporteco‐efficiencyperformance……………………………………………….….1113.5.Porteco‐efficiencyperformancepershippingsector……………..……………...….112

ListofFiguresChapter1

1.1.MaincomponentsoftheSTEAMandtheirinter‐relations…………………...………441.2.EmissionpercentagebyshippingsectorinLasPalmasPort…………….................501.3.OperativepercentagebyshippingsectorinLasPalmasPort……………...............511.4.EmissionpercentageofcruiseandferryvesselsrelativetothetotalemissionsinLasPalmasPort………………………………………………………………………….........................521.5.OperativepercentageofcruiseandferryvesselsrelativetothetotaloperativeinLasPalmasPort..………………………………………………………………………………………....521.6.Seasonalpercentageofemissionsandoperativehoursofcruisevessels,relativetothetotalsaccountedinLasPalmasPort……………………………………………………….541.7.Seasonalpercentageofemissionsandoperativehoursofferryvessels,relativetothetotalsaccountedinLasPalmasPort……………………………………..……..................541.8. Geographical characterization of estimated CO2 at source from passengervesselshotellingandmanoeuvring.Left:2011,right:2012……………………………….57Chapter22.1.ImpactPathwayApproachtoairpollution…………………………………………………71Chapter33.1.NeighbourdistrictsofLasPalmasPort……………………………………………………....943.2.ImpactPathwayApproachtoairpollution…………………………………………………963.3.ShareofemissionsbyshippingsectorinLasPalmasPort…………………………...1043.4.Estimatedlocalexternalcosts(€)……………………………………………………….…...1073.5.MonthlyexternalcostsfromvesselemissionsinLasPalmasPort……............108

14

15

IntroducciónyresumengeneralEn los últimos años, los efectos negativos relacionados con las emisionesatmosféricasderivadasdelcrecimientodeltransportemarítimohangeneradounapreocupacióncreciente.Sehareconocidoquelosbuquesnosólocontribuyenalosefectosnegativosaescalaglobalquesederivandelaelevacióndelatemperatura(cambio climático), sino que también son responsables de efectos negativosexperimentadas en las comunidades locales (Cullinane and Cullinane, 2013).Derivadodeesto,lospuertosestánadoptandodemaneracreciente,herramientasde gestión para hacer cumplir o fomentar el desarrollo sostenible (Lam yNotteboom,2014)ylareduccióndelasemisiones.Ciertamente,ademásdelincrementoenlastemperaturasglobalescausadasporelCO2, la exposición a gases contaminantes derivados de la combustión decombustibles fósiles, como el NOx, SOx, CO, VOC y partículas volátiles (PM1), serelacionademaneracontinuadaconconsecuenciasnegativassobrelasalud.Estosefectosindeseablessepresentantantoacortocomoalargoplazo.Algunosejemplosson: doloresde cabeza,mareos, náuseas, problemas respiratorios, enfermedadescrónicas,ingresosencentrosdesaludymortalidadprematura(Corbettetal.,2007).Lanecesidaddereducirlacontaminacióndelairehasidoampliamentereconocidacomounacuestióndepolíticaenlospuertos.Elcontroldelasemisionesrequierecomounpasoprevioimprescindibledelacapacidaddecuantificarlasydesarrollarinventariosprecisosde lasmismas.Enefecto,disponerde informaciónsobre lasemisiones es necesario para poder evaluar adecuadamente los impactos deproyectosdemejorasportuariasodecrecimientodelaactividaddelaflotaenelpuerto, así como para planificar las estrategias o programas voluntarios demitigación y, ayudar a los responsables políticos en el desarrollo de requisitos

1 Las partículas volátiles o suspendidas son una mezcla de compuestos microscópicos o muypequeñosenformadepequeñaspiezasdepartículaslíquidasysólidassuspendidasenelaire(porejemplohollín,polvo,humoyneblinas).Sucomposiciónsedefineenlosestándaresinternacionales(ISO8178)deacuerdoconlamedidadesudiámetro(10micrasomenosy2.5micrasomenos).

16

normativoseficacesanivelnacionaleinternacionalparareducirlasemisiones.

Efectivamente,elprocesodecombustiónvinculadoalaoperacióndelosbuquesenregionescercanasalacostayenlospuertos,contribuyealincrementoenlosnivelesde exposición de residentes y visitantes de la ciudad portuaria a sustanciaspeligrosas(Tzannatos,2010;Miola,2010;.Castellsetal,2014;TichavskayTovar,2015).Estosedebeprincipalmentealasoperativasdeatraque,maniobra(llegadaysalida)ynavegacióndentrodelpuertoysucanaldeaproximación.Enestesentido,eltiempoquepermaneceelbuqueatracadopuederesultarenunagrancontribuciónalacontaminacióndelaire.Enparticular,cuandoestosutilizansuspropios generadorespara cubrir susnecesidadesde electricidad en atraque.Ademásdeesto,laszonasdeatraqueamenudoselocalizancercadezonaspobladasporloqueelimpactodelasemisionesliberadasdurantelasoperativasdemaniobrayatraquepuedenresultar,desdeunaperspectivalocal,enunamayorproporcióndeexternalidadesnegativasdelasqueresultaríansiesasemisionesfuesenliberadasenmarabierto(GoldsworthyyGoldsworthy,2015).Elretodeidentificarlosperfilesoperativosdelosbuquesatracados,enmaniobrayen navegación de crucero (aproximación al puerto), en conjunción con ladependenciadelasemisionesqueresultadeesosperfiles(lacargarealdelmotorencadaunadeellas),sepuedeabordarsisedisponederegistrosdeposicióndecada buque 2 y de bases de datos complementarias que contienen los detallestécnicosdelanaveydesusmotores.Estogarantizaquelaubicación,lavelocidad,laruta,lasdimensiones,eltipodemotoryelconsumodecombustibledecadabuqueseaconocidoentodomomento;loqueevitalaadopcióndesupuestosacercadelas2EstospuedenobtenerseatravésdelsistemaAIS(AutomaticIdentificationSystem).Paraestatesisseutilizóinformaciónobtenidaatravésdeestesistema.UnaunidadAISconsisteenuntransceptorderadioVHFcapazdeenviaraotrosbuquesyaestacionesreceptorasidentificadas(terrestresysatelitales), laposición,rumbo,velocidad,eslora,tipodebuque, informaciónde identificacióndelbuque,entreotros.LaunidadAISdeabordotransmitelainformacióndemaneraautomáticaysinintervención de la tripulación del buque. Este sistema fue concebido con el objetivo original deayudaralanavegaciónenelseguimientodelosbuquesylaprevencióndecolisionesenelmar,queconel tiempo,haevolucionadohastaconvertirseenunsistemaconunamultituddeaplicacionesadicionales,incluyendolaobtencióndebasesdedatosrealistasparalaestimacióndeemisiones.UntransceptorAIStransmiteregularmente,informaciónestáticaytambiénrelacionadaconlatravesía(cada6minutos),ademásdeinformacióndinámicaconunafrecuenciarelacionadaconlavelocidaddelbuque(2‐10segundos)yestadodenavegación(3mincuandoestáanclado).

17

variables anteriores y, además, facilita el beneficio adicional que se deriva de laposibilidadderepresentarlosresultadosobtenidosenmapasdealtaresoluciónquereflejenlacaracterizacióngeográficadelasemisionesenlasáreasdepuerto‐ciudad.Las metodologías basadas en el sistema de comunicaciones AIS, ya han sidoutilizadasparaestimaremisionesdebuques(Jalkanenetal.,2008)conanterioridad,sinembargo,nuncaantessehabíansugeridocomouninstrumentodeapoyoa laformulación de políticas y medidas correctoras dirigidas a un sector específico(pasajeros)osussubsectores(crucerosytransbordadores)dentrodeuncontextoinsular.Enloqueaestorespecta,elCapítulo1titulado"Port‐cityexhaustemissionmodel:an application to cruise and ferry operations in Las Palmas Port" presenta, porprimeravezenlaliteratura,uninventariodedocemesesdeemisionesdebuquesenpuerto.Esteinventariosehaconstruidoapartirdeunmodelofullbottom‐up3einformaciónrealdetráficodebuquesdurante2011enelpuertodeLasPalmas.EstainformaciónhasidoobtenidaatravésdemensajestransmitidosporelSistemadeIdentificaciónAutomática(AIS).El inventariodeemisionesseharealizadoconelobjetivodeanalizar, tanto lasemisionesde tráficomarítimoengeneralcomo lasoperacionesdecrucerosyferriesenparticular.Losresultadossedescribenportipodecontaminante(NOx,SOx,PM2.5,CO,CO2)ysedesagreganatendiendoaltamañode3Losenfoquestop‐downybottom‐upsonampliamentereconocidosenlaliteraturadeunavariedaddetemasdeinvestigación(Sabatier,1986).Estosincluyenlacuantificacióndelasemisionesalaire(paso necesario para obtener los costes externos) y los costes externos. Un enfoque capta latecnologíadeltransporteenformaagregada(top‐down)yelotroenformadesagregada(bottom‐up)arrojandoresultadosdiferentesdebidoacomplejasinteraccionesentreefectos,estructuraydatos.Tanto en la estimación de emisiones como en la de costes externos, el enfoque top‐down utilizavariableseconómicasagregadasmientrasqueelenfoquebottom‐upconsiderainformaciónrefinadaydesglosada,ensumayoríabasadaenelrendimientotécnico.Laaplicacióndeunouotroenfoquevaríadeacuerdoconelobjetodeestudio.Paralaestimacióndelas emisiones, se utiliza un enfoque top‐down, que se basa en la venta de combustibles, cuandoinformación refinada de tráfico no existe o no está disponible. Sin embargo cuando existe lainformacióndetráfico(obtenidaatravésseguimientodebuques‐vesseltracks‐oescalasenpuerto)yestádisponible;seutilizaunenfoquebottom‐updebidoalaprecisióndelosparámetrosdeentradadelmodelocomo,porejemplo,eltipodebuque,laubicación,eltamañoylosdetallestécnicos.Porúltimo,Miolaetal.,(2010),hablandeunenfoquefullbottom‐upcomoelusodeunenfoquebottom‐uptantoenlacuantificacióndelasemisionescomoenlacaracterizacióngeográficadelosresultados.En cuantoa la caracterizacióngeográficade las emisiones, el nivel dedetalle alcanzado tambiéndepende del enfoque seguido. Así, con un enfoque, bottom‐up se tiene en cuenta informaciónindividualdelosbuquesysuposiciónmientrasqueconunenfoquetop‐downlavaloraciónserealizasin,oconinformaciónparcial,sobrelaposicióndelosbuques(esdecir,laactividadgeográficadeltráfico marítimo se estima con base en una sola ruta de navegación o una célula de actividadgeográficaparticular,conindependenciadequebuquellevaacabolaactividad).

18

losbarcos, al tipodebuqueya laoperativadenavegación(atraque,maniobraonavegacióndeaproximaciónapuerto).Enestecapítulosellegaalaconclusióndeque el tráfico marítimo en general, y el transporte marítimo de pasajeros enparticular,representanuna fuentedecontaminacióndelaireenelpuertodeLasPalmas.Losmapasdeemisioneselaboradosconbaseenlosresultadosconfirmanlaubicación de focos de liberación de emisiones en los muelles asignados aoperacionesdecrucerosyferries.Lasrecomendacionesdepolíticaquesederivandelosresultadosobtenidosalientanallevaruncontrolregulardelasemisionesconelpropósitodeorientareldiseñodeinstrumentosdepolíticaambientaleincentivosbasadosenelmercadoquetenganencuentalosperfilesdecontaminaciónydeoperativaespecíficosdecadasubsector(ferris,cruceros,portacontenedores,graneleros,yotros).Porotrolado,sesugierese lleven a cabo estudios de viabilidad de proyectos que permitan reducir lasemisionescomo,porejemplo,lainstalacióndesistemasdeatraqueautomáticoylaprovisión a los buques de combustibles alternativos (como el gas licuado, LNG)menos contaminantes. Por último, pero nomenos importante, que se analice laviabilidad de instalaciones que permitan el suministro a buques atracados deenergíaeléctricadesdetierra(evitandoasíqueutilicensuspropiosgeneradores),dando prioridad a su instalación en losmuelles de atraque habitual de aquellossectores o sub‐sectores que, tras confirmarse por un estudio de dispersiónatmosférica,exposicióneimpacto,presentenunperfilcontaminanteconmayoresefectoslocales.Los resultados del Capítulo 1 no sólo describen perfiles de contaminantes yoperativa en el puerto de Las Palmas si no que, también ponen en valor lasmetodologíasdemedicióndeemisionesbasadasendatosAIS.Enparticular,cuandose acompañan demodelos de calidad del aire, estimaciones de su impacto y deestudios económicos para abordar el diseño de medidas correctivas dirigidas asubsectores específicos en el transportemarítimo (como cruceros y ferries). Losresultados y recomendaciones de política de este estudio pueden ayudar en laadecuaciónomejoradelapolíticaexistenteenelpuertodeLasPalmas,pudiendo

19

sertambiéndeutilidadaotrasciudadesportuarias,seancontinentalesoinsulares,bajocondicionesdeltráficomarítimosimilares.Lasemisionesdegasescontaminantesrelacionadasconlospuertos,comocualquierexternalidadnegativa;reflejanuncosterealprocedentedeunaactividadeconómicaydanlugaraunresultadoquenoesóptimo.Dehecho,lapoblaciónquehabitalasciudades portuarias experimenta consecuencias perniciosas derivadas de ladegradaciónenlacalidaddelaireasociadoaltráficomarítimo(Corbettetal.,2007).Estasemisionesresultanenexternalidadesnegativasqueafectanazonasurbanasyrurales y que pueden sermonetizadas como costes externos. En este sentido, elCapítulo 2 titulado "External costs of vessel emissions at port: A review of themethodological and empirical state of the art" presenta una revisión de lasmetodologíasexistentesenlaactualidadparaestimarelcostedelasexternalidades(coste externo4en adelante) derivado de las emisiones de los buques en puertocomopasoprevioalavaloraciónqueserealizaenelCapítulo3.Además,seconstataquelaliteraturaempíricaqueestimaemisionesdegasescontaminantesderivadasde buques en puerto es reciente. Sus orígenes se encuentran en 2009, año depublicacióndelprimerartículo.Deentonceshastaahora,ysegúnelconocimientodelosautores,soloexistennueveartículossobreeltemaentrelosquesecuentaelCapítulo3deestatesis.Según larevisiónefectuadade la literaturarelativaa lasmetodologíasexistentesparaestimarel costeexternode lapolucióndelaire, en laactualidad,esposibleestimarecosteexternosibiensuestimaciónesfuentedeincertidumbreporquelamismaestácondicionada,principalmente,porlimitacionesmetodológicasylagunasenelconocimientodisponible.ElImpactPathwayApproach(IPA)5esidentificado

4 Son aquellos costes impuestos a la sociedad que, sin actuación o intervención política, no sontenidosencuentaporlosdistintosusuarios,enestecaso,delospuertos.Enestatesiselinterésrecaeen la componentemedio ambiental del coste externo, que incluye los costes relativos a la salud,costesmateriales,dañosenlabiosferayriesgosalargoplazo. 5 Esta metodología para evaluar las externalidades ambientales derivadas de los ciclos decombustiblesutiliza larutade impacto (deahísunombre)queabarcadiferentesetapas,desdelaemisiónde loscontaminantes,dispersiónyconcentración,elcálculode los impactosenunidadesfísicasylavaloracióneconómica.

20

como lametodologíabottom‐up6máscompletay lamejorprácticasugeridaenelcálculo de costes externos derivados de las emisiones liberadas al aire. Ha sidoampliamente adoptada, entre otros, por los principales estudios Europeos sobrecostes externos de transporte (CAFE, BeTa, HEATCO y NEEDS). La complejidadmetodológica y los recursos económicos, implícitos en el uso de lametodologíabottom‐up IPA ha resultado en una aceptación generalizada, en la literaturaempírica de estimación de coste externo derivado de los barcos en puertos, delenfoquetop‐downyelusodefactoresdecosteporpaísoregión,obtenidosdelosprincipalesestudioseuropeos(BeTa,CAFEyNEEDS).ElCapítulo2concluyequelosinventariosdeemisionesyloscostesexternosdelosnueveestudiosdepuertosencontradosenlaliteratura,yrevisadosenestecapítulo,son significativamente diferentes y difíciles de comparar debido a variacionesmetodológicas, supuestos asumidos en las estimaciones, categorías de coste yfactoresdeemisiónutilizados,entreotros.Portanto,sedefiendequeesdesumaimportanciarevisarestasdiferenciasparaidentificarelmejorenfoqueycuálessonlos inconvenientesdeutilizar lasegundamejoralternativacuandonoexisteotraopción.Estofavorecequeelmejorenfoquetermineimponiéndoseloquenosóloharía más comparables los diferentes estudios sino que, lo que es aún másimportante,redundaríaenunamayorprecisióndelasestimacionesalgo,porotraparte,devitalimportanciasiestasestimacionesvanaservir,asuvez,debaseparalaestimacióndecostesexternos.Enestesentido,larevisióndeliteraturaefectuadaseñalaquelasdiferenciasdecalidadenlainformacióndetráficoqueseutilizapararealizar las estimaciones son notables y dignas de mencionar. Ciertamente, lostrabajos revisados que utilizan un enfoque bottom‐up 7 en la estimación de

6Enlaestimacióndecostesexternostambiénespreferibleelenfoquebottom‐upsobreeltop‐downporque permite una evaluación precisa, basada en información detallada, posibilidades dediferenciaciónyunamejorprecisiónenlosresultadosobtenidos(costesexternosmarginales).Sinembargo,estareconocidoqueelusodeesteenfoqueimponerequisitoscostososycomplejosparaobtener los costes externos. Por lo tanto, se sugiere, y está ampliamente acepado, el uso de unenfoquetop‐downcuandonosepuedenrealizarestudiosbottom‐uponoestándisponibles.Dehecho,comosepresentaenelCapítulo2deesta tesis, la literaturasobreestimacióndecostesexternosdebidoalasemisionesdelosbuquesdelpuertosebasa,exclusivamente,enelusodefactoresdecosteyvariableseconómicasagregadas(enfoquetop‐down). 7Generalmentehablando,enlaestimacióndeemisionespuedenutilizarsedosenfoquesprincipales:top‐downybottomup.Elprimeroconsisteenestimarlasemisionesdemodoindirectoapartirdelasestadísticasdeventadefuelmientrasqueelsegundoutilizadatosdeactividaddelaflotadebarcos.

21

emisionesnosiemprereconocen,deformaclara,queestánutilizandolasescalasenpuerto como fuente de informaciónde tráficomarítimo, ni describen el nivel dedetalledelasposicionesdebuquesquecontienenlasbasesdedatosutilizadas.Porel contrario, normalmente hacen referencia exclusivamente a la elección de unenfoquemetodológicobottom‐upbasadoeninformacióndetráfico8.Por otra parte, la revisión efectuada muestra que el enfoque metodológicorepresentativoparaestimarlasemisionesdebuquesenpuerto(pasoprevioparaestimar los costes externos), es un enfoque bottom‐up basado en las escalas enpuerto.Enestesentidoydebidoalmayorniveldeprecisiónquepodríaobtenerseatravésdelusodeinformacióndeposicióndebuques(vesseltracks)sesugieresuuso;evitandodeestamaneralanecesidaddeutilizarvalorespromedio(distanciayvelocidad)y,deserposible,quesehagasiguiendounenfoquefullbottom‐up9.Porúltimoyenrelaciónconlaestimacióndeloscostesexternos,todoslosartículosrevisadossiguenunenfoquetop‐down.Estoseatribuyealosrequisitoscostososycomplejosparaobtenercostesexternosdesdeunenfoquebottom‐up.Porotraparte,lafaltadeestudiosquemodelenladispersiónatmosféricadeemisionesdebuques,complicaelescenariometodológicoresultando,portanto,enunaampliaaceptacióndelusodefactoresdecosteporpaísoregión(top‐down).Con base en lo anterior, el Capítulo 2 también señala la necesidad de mejorasmetodológicas y sugiere la realización de estimaciones más refinadas (tanto deemisionesdebuquescomodecostesexternosderivados)yaqueestobeneficiaríalacalidadde la informaciónnecesariaparaalimentara lasmedidasdepolíticaquepodríandiseñarseparacontribuirainteriorizarloscostesexternosestimados.Porúltimo,unavaloraciónbottom‐up(IPA)específicasobreemisionesdebuques(aún

Elniveldeprecisióndelasestimacionesesmayorenelenfoquebottom‐upporloquesiemprequeseaposibleeselquedeberíaserutilizado.Paramásdetalle,véasenotaalpie3. 8Losmodelosbottom‐upparaestimaremisionespuedenbasarseendatosrealesobtenidosatravésde AIS o en estadísticas de escalas en puerto. En este segundo caso el investigador no tieneinformaciónrealdelaruta,velocidad,yotrosdatosdelbarcoporloquetienequehacerunbuennúmerodesupuestosquereducenlacalidaddelaestimación.Paramásdetalle,véasenotaalpie3. 9 El enfoque full bottom‐up utiliza un enfoque bottom‐up tanto para la cuantificación de lasestimacionescomoparalacaracterizacióngeográficadelosresultados.Paramásdetalle,véasenotaalpie3.

22

nollevadaacabo),sesugierecomoinvestigaciónfuturaaunqueporelmomento;losresultadosobtenidosutilizandofactoresdecostedeBeTa(únicoinformedisponible,que presenta factores de coste dedicados a las emisiones de buques en puerto)proporcionen una primera aproximación cercana a la magnitud real de costesexternosderivadosdelasemisionesdebuquesenpuerto.YaqueenelCapítulo1sesugiereque,conelfindepermitirlainternalizacióndeldañoderivadodelasemisionesylaconsiguientemejoradelbienestarpúblico; lainvestigaciónsobreemisionesdebuquestambiéndeberíaabordarlavaloracióndecostesexternosderivados,estatesiscompletaelprocesoconelCapítulo3titulado“Environmentalcostsandeco‐efficiencyfromvesselemissionsinLasPalmasPort”EnestecapítuloseextiendelainvestigaciónrealizadaenelCapítulo1(inventariodeemisiones)alaestimacióndeloscostesexternoseindicadoresdeeco‐eficienciaenelpuertodeLasPalmas.En definitiva, en el Capítulo 3, se estiman los costes externos derivados de lasemisionesdegasespresentadasenelCapítulo1deestatesisquefueronobtenidasapartirdevariablesdesagregadastantotécnicascomodetráfico.Portanto,estecapítulo contiene el primer trabajo en la literatura que sugiere la estimación yestimaloscostesexternosapartirdeuninventariodegasesobtenidoatravésdeunenfoque full‐bottomupbasadoendatosAIS10cerrandoungapen la literaturaycontribuyendoaunamejorametodológica,necesariaparaobtenerresultadosmásprecisos. Por otra parte, y ya en términos de costes externos, las estimacionesrealizadas siguenun enfoque top‐down, como todos los trabajos existentes en laliteraturaeidentificadosenelcapítuloanterior,sibienadiferenciadelostrabajospublicadoshastaahoralasestimacionesrecogidasenesteCapítuloreflejantodoslosposiblesumbralesexistentes(altosybajos)defactoresdecostedisponiblesenBeTa,CAFEyNEEDS.

10Comoya seha comentado, el enfoquebottom‐up basado endatosAIS es elmejorde entre losactualmentedisponiblesporqueeliminalaincertidumbreylimitacionesdescritasquesederivandelosinventariosdeemisionesquesiguenunenfoquebottomupbasadoenlasescalasenpuerto.

23

Adicionalmente,enesteCapítulotambiénsecalculanindicadoresdeeco‐eficiencia.Estos indicadores son una herramienta valiosa para promover el desarrollosostenible.Suusosebasaenelconceptodecrearmásbienesyserviciosmediantelareduccióndelimpactoambientalrelacionadoconlaproduccióndelosmismos.Entérminos generales, los indicadores de eco‐eficiencia se utilizan para medir ygestionar el crecimiento ecológico mediante la comparación del rendimientomedioambiental y económico entre los diferentes sectores económicos, laidentificacióndepolíticassusceptiblesdemejorayelseguimientodelastendenciasdeeco‐eficienciaeneltiempo(UNESCAP2009).En la actualidad los puertos, y en relación con las emisiones, tienen el objetivocomún de crearmecanismos institucionales para para reducir la contaminaciónatmosférica y el cambio climático, entre otros, mediante el inicio de estudios,estrategiasyaccionesparasupervisarymejorarlacalidaddelaire.Conelpropósitode promover la necesidad primaria del desarrollo sostenible, los puertos (comomuchasempresas),comienzanaexplorarnuevasformasdegestiónquepermitanlaintegracióndelagestiónambientalenlaeconomíalocalylasociedad(Coto‐Millánetal.2010).Entreellasseencuentran,principalmente,elcontroldelosimpactosambientales a través de estrategias de gestión ambiental; la medición deldesempeño (eco‐eficiencia) a través de la valoración ambiental (emisiones) enrelación con los factores económicos (producción)11 y, finalmente; apoyando eldiseñodeinstrumentosdepolíticaquetenganlosindicadoresdeeco‐eficienciaencuenta.La eco‐eficiencia, como un indicador de rendimiento, proporciona informaciónvaliosaalsistemaportuarioparamejorarsuposicióncompetitiva(Coto‐Millánetal.,2010).Dehecho,elrendimientofinancierodelospuertosesclaveparaconvertirseen un importante centro de negocios, pero no es suficiente para garantizar susostenibilidad.Paraasegurarestoúltimo,debeabordarse tambiéneldesempeñoambientalysocial,entreotros,mediantelarecopilacióndeinformaciónsobrelosimpactosambientalesyeldesempeñoparareflejarsusituaciónglobal(Coto‐Millán11Losindicadoresdeeco‐eficienciapodríandefinirse,comosehaceenestatesis,comolarelaciónentrelosimpactosdelservicio(costesexternos)yloquesehaproducido(toneladas,pasajeros,etc.).

24

et.2010).Porestarazón,enesteCapítuloseobtienenindicadoresdeeco‐eficiencia(costedeemisiones/producciónrelacionada)tantoparaelpuertoensuconjuntocómo, y esto es también una novedad en la literatura, por sector de transportemarítimo (buques de contenedores, cruceros, petroleros, entre otros). Estosindicadores de eco‐eficiencia definidos por sector se sugieren como posiblesindicadores de desempeño ambiental y económico de los distintos sectores quepuedenserútileseneldiseñodepolíticasportuarias.Ensíntesis,atravésdeestecasodeestudio,nosólosecuantificaelcosteexternoeindicadoresdeeco‐eficienciaasociadosalasemisionesdebarcosenelpuertodeLasPalmasen2011sinoquetambién semuestra la utilidadde estasmedidas comoposiblesherramientasdeapoyoaotrasAutoridadesPortuariasygobiernoslocales.Paraconcluir,elCapítulo3sugierecomoinvestigaciónfutura,laestimacióndeloscostes externos e indicadores de eco‐eficiencia siguiendo un enfoque bottom‐up(IPA)queestébasadoentreotrascosas,eninformaciónrefinadadeconcentracióndegasesypartículascontaminantesenlaatmósferaycondicionesmeteorológicaslocales. Esto es de especial interés ya que a pesar de las incertidumbresmetodológicas (también existentes en IPA), su esta aceptado en la literatura deestimación de costes externos siendo además considerado como el enfoquemetodológicoqueconunmayorrangodeprecisiónensusresultados,podríaserutilizadoeneldiseñodepolíticaambientaldeemisionesatmosféricasapesardeque la metodología IPA no fue originalmente desarrollada con ese propósito.Adicionalmente, y ya que las estimaciones realizadas no incluyen fuentesalternativasdeemisionesenelpuerto,sesugierenfuturasmejorasdelosresultadosmediantelainclusióndelasemisionesgeneradasentierraylosefectosderivadosenlasaluddelosmarinerosyprofesionalesdelmar. Enresumen,esta tesiscomprendeunenfoquesistemáticohaciaelanálisisde lasemisionesatmosféricasyloscostesexternosgeneradosporeltráficomarítimoenelPuertodeLasPalmasenelaño2011.Losresultadostienencomoobjetivoreflejarel rendimiento del puerto de Las Palmas teniendo en cuenta preocupacionessociales,económicasyambientales.Elobjetivodeesteenfoqueesapoyarunmodelode gestión portuaria, que extienda el concepto de gestión portuaria hacia el

25

conceptodegestiónsostenibleensumásampliaacepción:sostenibilidadeconómica,social y medio ambiental. Asimismo, los resultados de los indicadores de eco‐eficienciapretendenfacilitarfuturosanálisiscoste‐beneficioysuposibleusoparala valoración de instrumentos de política tendentes a reducir las emisionesderivadasdelosbuquesenLasPalmas,ciudadqueacogeunacantidadconsiderablederesidentesyturistas.Finalmenteesteestudiotambiéncontribuyealaliteraturareciente de estimación de las emisiones de los buques en puerto, cálculo de loscostesdeexternosderivadosyeco‐eficienciadescribiendo,atravésdeestecasodeestudio, la utilidad de estas medidas como herramientas de apoyo a otrasAutoridadesPortuariasylosgobiernoslocales.Parafinalizaresteresumen,cabeseñalarqueeltextodelostrescapítulosdequeconsta está tesis y que se presentan a continuación, corresponden a trabajoscientíficosque;obienseencuentranactualmenteenrevisión(Capítulo2y3)enrevistasdeimpacto(JCR)o,comoeselcasodelCapítulo1,yahansidopublicados(Tichavska, M., Tovar, B., (2015), publicado en Transportation Research Part A:Policy and Practice, 78, 347‐360. Abstracted/indexed in: ISI web of knowledge.Impactfactor(referidoa2013):2,525.RevistaclasificadaenelprimercuartildelascategoríasdeEconomíayTransporte).ReferenciasCastells, S.M., Santamarı́a, J.J.U.,MartıńezDeOsés, F.X. (2014),Manoeuvring and

hotellingexternalcosts:enoughforalternativeenergysources?Marit.PolicyManage.41(1),42–60.

Corbett,J.J.,Winebrake,J.J.,Green,E.H.,Kasibhatla,P.,Eyring,V.,Lauer,A.(2007),Mortalityfromshipemissions:aglobalassessment.Environ.Sci.Technol.41(24),8512–8518.

Coto‐Millán,P.,Mateo‐Mantecon,I.,DomenechQuesada,J.L.,CarballoPanela,A.andPesquera, M.A. (2010), Evaluation of Port externalities: The ecologicalFootprint.InCoto‐Millán,P.,Pesquera,M.A.,andCastanedo,J.(Ed.),Essaysonporteconomics.SpringerScience&BusinessMedia.SpringerHeidelbergDordrecht,London,NewYork.

Cullinane,K.,andCullinane,S.(2013),Atmosphericemissionsfromshipping:Theneedforregulationandapproachestocompliance.TransportReviews,33(4),377‐401.

Jalkanen, J. P., Brink, A., Kalli, J., and Stipa, T. (2008), State of the Art‐AIS basedemission calculations for the Baltic Sea shipping. Hrvatski meteorološkičasopis,43(43/1),177‐180.

26

Lam, J.S.L., Notteboom, T., (2014), The greening of ports: a comparison of portmanagementtoolsusedbyleadingportsinAsiaandEurope.Transp.Rev.34(2),169–189.

Miola,A.,Ciuffo,B.,Giovine,E.,Marra,M.(2010),Regulatingairemissionsfromships.TheStateoftheArtonMethodologies,TechnologiesandPolicyOptions.JointResearch Centre Reference Report, Luxembourg, EUR24602EN, 978‐92,ISBN.

Sabatier, P. A. (1986) Top‐down and bottom‐up approaches to implementationresearch:acriticalanalysisandsuggestedsynthesis.Journalofpublicpolicy6(1),21‐48.

Tzannatos,E.(2010),ShipemissionsandtheirexternalitiesfortheportofPiraeus‐Greece.Atmos.Environ.44(3),400–407.

UNESCAP(2011),Eco‐efficiencyindicators:MeasuringResource‐useEfficiencyandthe impact of Economic Activities on the Environment. United NationsPublication,EnvironmentandDevelopmentDivision,Bangkok,Thailand.

Tichavska,M.,Tovar,B.,(2015)Port‐cityExhaustEmissionModel:anapplicationtoCruise and Ferry operations in Las Palmas Port. TransportationResearchPartA:PolicyandPractice,78,347‐360.

Goldsworthy, L. and Goldsworthy, B. (2015), Modelling of ship engine exhaustemissions in ports and extensive coastal waters based on terrestrial AISdata–AnAustraliancasestudy.EnvironmentalModelling&Software,63,45‐60.

27

IntroductionandsummaryInrecentyears,negativeeffectsrelatedtoairemissionsderivedfromthegrowthofshippinghave increasingly raised concern. It hasbeen recognized thatoperativevessels do not only contribute to negative effects on a global scale ‐risingtemperatures in the climate system‐ but also to hazardous consequencesexperienced in local communities (Cullinane and Cullinane, 2013). Indeed,managementtoolsatportarebeenincreasinglyaddressedtoenforceorencouragegreendevelopment(LamandNotteboom,2014)andtheabatementofemissions.InadditiontotheincreaseofglobaltemperaturescausedbyCO2;theexposuretocombustiongases(NOx,SOx,CO,VOC)andParticulateMatter(PM12)suspendedintheatmospherearecontinuouslylinkedtoshort‐termandlong‐termconsequencessuch as headaches, dizziness, nausea, coughing, laboured breathing, cronicaldiseases,lungcancer,hospitaladmissionsandprematuremortality(Corbettetal.,2007).Theneedtoabateairpollutioniswidelyrecognizedasapolicyissueinportsandharbours. The control of atmospheric emissions requires the ability to quantifythese and to develop accurate emission inventories for ports. Indeed, emissioninformation is necessary to properly assess the impacts of port improvementprojectsorthegrowthinshippingactivity,aswellastoplanmitigationstrategiesorvoluntaryprogramsandtoaidpolicymakerstowardsthedevelopmentofeffectiveregulatoryrequirementsatnationalandinternationallevels.

Fuel combustion fromoperative vessels also contributes to coastal emissions, attimesincreasingexposurelevelstohazardoussubstancesonresidentsandvisitors(Tzannatos,2010;Miola,2010;Castellsetal.,2014;TichavskaandTovar,2015).Thismainlyresultsfromhotelling,manoeuvringandcruisingoperativesofvesselsatport.

12ParticulateMatter (PM) is associatedwith tinypiecesof solidor liquidparticles suspendedasatmosphericaerosol.Itscompositionisdefinedininternationalstandards(ISO8178)accordingtodiametermeasures(10micrometresorless,and2.5micrometresorless).

28

In this respect, time spent at anchormay result into a large contribution to airpollution.Particularlywhenusingdieselgenerators tocoverelectricityneeds. Inaddition to this, berths are often located near populated areas so the impact ofemissionsreleasedduringhotellingandmanoeuvringmodesmaylargelyresultinlocal effects than those emissions released at open sea (Goldsworthy andGoldsworthy,2015).The challenge of identifying operative profiles of ships at berth (hotelling),manoeuvring and normal cruising navigation in conjunction with the emissiondependency on engine load can be addressed with ship position records anddatabases containing ship technical and engine details, respectively. Thisguaranteeslocation,speed,route,dimensions,engineandfuelconsumptionforeachship 13 is always acknowledged, providing the additional benefit of visualizingresultsthroughhigh‐resolutionmapspresentingthegeographicalcharacterizationof emissions in port‐city areas. AIS‐based methodologies have been alreadyintroducedtoestimatevesselemissions(Jalkanenetal.,2008).Nevertheless,theyhadneverbeenpresentedasaninstrumenttoassistpolicydesignandcorrectivemeasuresofaspecificshippingsector(passenger)andsubsectors(cruiseandferry)withinanislandcontext.Inthisrespect,Chapter1titled“Port‐cityexhaustemissionmodel:anapplicationtocruiseandferryoperationsinLasPalmasPort”presentsforthefirsttimeatwelve‐monthvesselemissionsinventoryinport,builtfromafullbottom‐up14modeland13 These parameters (also used in this thesis) can be obtained through the AIS (AutomaticIdentification System). An AIS unit is equipped with a VHF radio transceiver enabling thetransmissionofinformationfrom/amongshipsandreceivingstations(terrestrialandsatellite).TheAIS‐boardautomaticallyunit transmits informationwithout the interventionof thecrewsuchas:positionofthevessel,course,speed,dimensions,shiptypeandsometechnicaldetailsamongothers.Thissystemwasdesignedwiththeoriginalaimtoassistnavigationinthemonitoringofshipsandthepreventionofcollisionsatsea,whichovertime,hasevolvedintoasystemwithamultitudeofadditionalapplications,Theseinclude,obtainingRealisticbasedataforestimatingemissions.14Top‐downandbottom‐upapproachesarewidelyrecognizedinavarietyofresearchsubjectsoverliterature(Sabatier,1986).Theseincludethequantificationofairemissions(requiredsteptoobtainexternal costs) and external costs. Each approach captures transportation technology in anaggregated(top‐down)ordisaggregatedform(bottom‐up)reflectingdifferencesinresultsduetocomplexinterplaysbetweenpurpose,structureanddatainput.Inbothemissionandexternalcostestimation,top‐downapproachesuseaggregatedeconomicvariableswhilebottom‐upapproachesconsiderrefinedanddisaggregatedinformation,mostlybasedontechnicalperformance.

29

AISshiptracksaddressingvesseltrafficingeneralandcruiseandferryoperationsinparticular.Emissionassessmentisbasedonafullbottom‐upmodelandmessagestransmittedbytheAutomaticIdentificationSystemduring2011inLasPalmasPort.ResultsaredescribedasabreakdownofNOx,SOx,PM2.5,COandCO2,accordingtoshipclasses,operativetypeandtime.Itisgenerallyconcludedthatvesseltrafficingeneral andpassenger shipping inparticular are a sourceof air pollution inLasPalmasPort.Emissionsmapsconfirm locationofhotspots inquaysassigned forcruiseandferryoperations.Policyrecommendationsencourageregularmonitoringofexhaustemissionsandmarket‐basedincentivessupportedbydetailsonpollutingandoperativeprofiles.On theotherhand, feasibility studiesaresuggested forautomatedmooring,LNGbunkering facilities and also shore‐side energy services, prioritizing berthing ofshippingsectors(orsub‐sectors)withthehighestshareofexhaustemissionsoncetheir local effects have been confirmed by a dispersion, exposure and impactassessment.Results in Chapter 1 do not only provide operative and polluting profiles in LasPalmas Port but also suggest the possible value of AIS based methodologies.Particularly,whenaccompaniedwithairqualitymodelling, impactandeconomicstudies to address the design of corrective measures for specific sub‐sectors inshipping,ascruiseandferry.ResultsandpolicyrecommendationsofthisstudymayalsosupportadequacyorimprovementofexistingpolicyinLasPalmasPort,beingacasetransferabletoport‐cityareasandislandsundersimilartrafficconditions.Theapplicationofapproachesvariesaccordingtothesubjectofstudy.Foremissionestimation,atop‐down approach that is based on fuel sales is used when refined traffic information is notavailable.On the other hand, a bottom‐up approachbasedon traffic information (obtained fromvesseltracksorportcalls)isusedwhenavailable,duetotheaccuracyofinputparameterssuchasshiptype,location,sizeandtechnicalparticulars.Finally,afullbottom‐upcategoryisdescribedbyMiolaetal.,(2010),astheuseofabottom‐upapproachforboth,thequantificationofemissionsandthegeographicalcharacterizationofresults.Regarding the geographical characterization of emissions, the level of detail achieved is alsodependentontheapproachfollowed(bottom‐upandtop‐down).Thus,withabottom‐upapproach,individualinformationofvesselsanditspositionareconsideredwhilewithatop‐downapproachvaluationisbasedwithout,orwithpartialinformationonthepositionofvessels(i.e.thegeographicalactivityofshippingisestimatedbasedonasingleshippingrouteoraparticulargeographicactivitycell,nomatterwhichvesselcarriesouttheactivity)..

30

Port‐related exhaust emissions, as any negative externality, reflect a real costaccruing from an economic activity and lead to a suboptimal outcome. Indeed,population located in port‐cities experience air quality detriments associated tovessel traffic and the atmospheric concentration of air pollution (Corbett et al.,2007).These result inurbanandruralexternalities15that canbemonetisedasexternalcosts.Inthisrespect,Chapter2titled“Externalcostsofvesselemissionsatport:Areviewofthemethodologicalandempiricalstateof theart”presentsareviewofexistentmethodologiescurrentlyused toestimateexternalitycosts16fromvesselemissions in shipping and in harbours. The empirical literature that estimatesexternalcostsfromvesselemissionsatportisrecentanddatesbackto2009,whenthefirstrelatedpaperwaspublished.Fromtherevisedresearchandtothebestofourknowledge,includingChapter3ofthisthesis,only9paperswerefound.Basedonthereview,theestimationofexternalitycostsissourceofuncertainty.Itisconditionedamongothers,bymethodologicaluncertaintiesandinformationgapsonavailableknowledge.TheImpactPathwayApproach(IPA)isconsideredasthemostcomprehensivebottom‐up17methodologyandthebestpracticeforcalculatingexternalcostsderivedfromairemissions.Ithasbeenwidelyadopted,amongothers,overmajorEuropean studies (CAFE,BeTa,HEATCOandNEEDS).Methodologicalcomplexityandcostlyresourcesareimpliedintheresearchpathwayofbottom‐upstudiesthataddressshippingandports.Forthisreason,andasafirstapproximation

15Accountedurbanexternalitiesincludehealth issues, increased mortality rates and the degradation of built environment. Rural externalities relate to crop damages. 16Costsimposedtosocietythatwithoutpoliticalactionorinterventionarenottakenintoaccountby the related users, in this case, of ports. In this thesis, the related interest relies in theenvironmental component of externality costs, which includes the derived costs of mortality,morbidity,thedegradationofbuiltenvironmentandthelossofcrops.17 In the case of external costs a bottom‐up approach is also preferred as it enables a refinedassessmentbasedondetailedinformation,differentiationpossibilitiesandanimprovedprecisioninderivedresults(marginalexternalcosts).Nevertheless,costlyandcomplexrequirementsarealsorecognized to obtain external costs from a bottom‐up approach. Thus, the use of a top‐downapproachissuggestedandwidelyacceptedwhenbottom‐upstudiescannotbeperformedorarenotavailable. Indeed, aswe present in Section 3, literature on harbour external costs due to vesselemissionsisexclusivelybasedontheuseofcostfactorsandaggregatedeconomicvariables(top‐downapproach).

31

toestimates, ithasbeenwidelyacceptedtofollowatop‐downapproachanduseper‐unit cost factorsobtainedmostly frommajorEuropeanstudies (BeTa,CAFE,NEEDS).Chapter 2 concludes that emission inventories and estimated costs from theavailable harbour studies are significantly different and complicated to compareduetomethodologicalvariations,assumptions,costcategories,selectedemissionfactorsandothers.Itisparamounttoreviewthesedifferencestohighlightthebestapproachandthedrawbackwhenasecondbestalternativeisapplied.Inthissense,thereviewremarks thatprecisiondifferenceson traffic informationused for theestimationsarenoteworthy.Availableliteraturedoesnotalwaysspecifyportcallsastheirsourceoftrafficinformationnordescribethelevelofdetailaccountedfromshipmovementsbutprovideanoveralldescriptionofactivity‐based(bottom‐up)methodology to estimate emissions. Moreover, our review has shown that therepresentativeapproachusedtoestimateemissionsatport(asaprevioussteptoestimateexternalcosts), isabottom‐upapproachbasedonportcalls.Duetotherefined accuracy of obtained results, we encourage the use of a full bottom‐upapproachand frequentlyupdatedvessel tracks, avoiding in thisway theneedofusingaveragevalues(i.e.distanceandspeed).Finallyandregardingtheestimationofexternalcosts,everystudyfollowedatop‐downapproach.Thisisattributedtocostly and complex requirements to obtain external costs from a bottom‐upapproach. Moreover, the lack of dispersion modelling practices not widelyundertaken in shipping complicates thismethodological scenario.Thus, enablingthewideacceptanceofatop‐downapproachinexternalcostestimationsbasedoncountryorregioncostfactors.Basedontheabove,Chapter2alsosuggestsmethodologicalimprovementsandthepossible achievement of refined estimations (of vessel emissions and derivedexternal costs) in ports and shipping as these may benefit the quality of inputinformation needed to feed policymeasureswhich contribute to internalize theexternalcostestimated.Finally,an integratedassessment(IPA)specific tovesselemissionshasnotbeenyetaddressedintheavailablestudiesandissuggestedasfutureresearchalthoughfornow,theobtainedresultsinBeTaprovideameaningful

32

insight to the magnitude of costs associated to vessel emission externalities,specificallybecauseitistheonlyavailablereportsofar,whichpresentscostfactorsdedicatedtoseaports.Sinceitissuggestedthatinordertoenabletheinternalizationandimprovementofthepublicwelfare,researchonvesselemissionsshouldalsoaddressthevaluationof economic costs derived from vessel emissions shipping, Chapter 3 titled“Environmentalcostsandeco‐efficiencyfromvesselemissionsinLasPalmasPort”extendsthevesselemissionresearchinChapter1totheestimationofexternalcostsandtheeco‐efficiencyperformanceofLasPalmasPort.Firstly,itobtainsexternalitycostsofvesselemissionsfromdisaggregatedvariablesasindividualvesseltracksandtechnicaldetails.Thisapproacheliminatesthedominantuncertaintiesreportedby previous vessel emission inventories (based on port calls) used to estimateexternality costs and fills the gap of methodology improvement, necessary toachieve more accurate results. Secondly, and in terms of externality costs, thisharbour study presents the available lower and upper thresholds of top‐downestimatedcostsavailableinBeTa,CAFEandNEEDS.Additionally, inthisChapterderivedeco‐efficiencyparametersareobtained.Eco‐efficiency indicators are considered as a valuable tool to promote sustainabledevelopment.Itsuseisbasedontheconceptofcreatingmoregoodsandservicesbyreducing the related environmental impact. Generally speaking, eco‐efficiencyindicators are used to measure and manage green growth by comparingenvironmental/economic performance among different economic sectors, byidentifying policy areas for improvement in achieving economic benefit and, bytrackingeco‐efficiencytrendsovertime(UNESCAP,2009).At present, in ports and towards air emissions, its common aim is to createinstitutionalmechanismstoabateairpollutionandclimatechange,amongothers,byinitiatingstudies,strategiesandactionsthatmonitorandimproveairquality.Topromotetheprimaryneedofsustainabledevelopment,ports(likemanycompanies),starttoexploremanagementphasesthatenabletheintegrationofenvironmentalmanagementintolocaleconomyandsociety(Coto‐Millánetal.2010).Namely,the

33

controlenvironmentalimpactsthroughenvironmentalmanagementstrategies;themeasurement of eco‐efficiency performance by valuating environmental(emissions)witheconomicfactors(production)18,andatlast;supportthedesignofpolicyinstrumentsthattakethelaterindicatorsintoaccount.Eco‐efficiency,asaperformanceindicator,providesportsystemswithinformationofvaluetoimprovetheircompetitivepositionwhenundertakingtheiractivitywithbusiness‐oriented criteria (Coto‐Millán et al. 2010). Indeed, the financialperformanceofports iskeytobecominganimportantcentreofbusinessbutnotenoughtoguaranteetheirsustainability.Toensurethis,environmentalandsocialperformance must be addressed among others, by collecting information onenvironmentalimpactsandperformancetoreflectitsoverallstatus(Coto‐Millánetal. 2010). For the latter reason, in this chapter, eco‐efficiency parameters areobtained(environmental/productionperformanceofvesselemissions)ingeneralshippingandalso(asaliteraturenovelty)pershippingsub‐sector(container,cruise,tankers, among others). These eco‐efficiency parameters are suggested, as anindicatorofenvironmentalandeconomicperformancetobeconsideredforpolicyuse in port‐cities. Summarizing, results respond to the research question of theeconomicimpactandenvironmental/productionperformanceofvesselemissionsin Las Palmas Port, describing through the case study, the utility of thesemeasurements(externalcostandeco‐efficiencyindicators)assupporttoolstoPortAuthoritiesandlocalgovernments.Toconclude,Chapter3suggeststhatfutureresearchalsoaddresstheseindicatorsby followinganintegratedapproachbasedamongothers,onrefined informationfrom pollutant concentration and local meteorological conditions. This is ofparticularinterestsince,indespiteofmethodologicaluncertainties(alsoexistentinIPA),theuseofthisapproachisacceptedinliteratureforestimatingexternalcostsand it remains so far, as themostaccurateapproach tobeused in thedesignofenvironmentalpolicytoaddressatmosphericemissions.Also,andsinceadditionalsources of emissions at port were not included in this study we suggest future18Thatis,eco‐efficiencyindicatorscouldbemeasured,aswedointhisthesis,astheratiobetweentheimpactsoftheservice(externalcosts)andwhathasbeenproduced(ton,passengers,andsoon).

34

improvements of results by including land‐based sources of emissions and thederivedeffectsonsailorsandmaritimeprofessionals.In summary, this thesis comprises a systematic approach and analysis of airemissions and externality costs from vessel traffic in Las Palmas Port. ResultsattempttoindicateperformanceofLasPalmasporttowardssocial,economicandenvironmental concerns. Aim of this approach is to support an environmentaloperationmodel,which extends value‐basedmanagement exploring relations ofeconomic and ecological capital efficiency. Also, eco‐efficiency results aim tofacilitate future cost‐benefit analysis used for evaluating abatement policyinstruments in Las Palmas, where a large population of residents and visitingtourists are continuously hosted. Finally this study, also contributes to recentliterature of vessel emissions, externality costs and eco‐efficiency by describingthroughthecasestudy,theutilityofthesemeasurementsassupporttoolstoPortAuthoritiesandlocalgovernments.In should be noted that the text from the threemain chapters in this thesis areadaptations of scientific papers that have either been already published in JCRjournals ranked in the first quartile of the categories ofEconomyandTransport(Chapter1,inTransportationResearchPartA:PolicyandPractice,Tichavska,M.,Tovar,B.,2015)orthathavebeensubmittedandcurrentlyunderreview(Chapter2and3).ReferencesCastells, S.M., Santamarı́a, J.J.U.,MartıńezDeOsés, F.X. (2014),Manoeuvring and

hotellingexternalcosts:enoughforalternativeenergysources?Marit.PolicyManage.41(1),42–60.

Corbett,J.J.,Winebrake,J.J.,Green,E.H.,Kasibhatla,P.,Eyring,V.,Lauer,A.(2007),Mortalityfromshipemissions:aglobalassessment.Environ.Sci.Technol.41(24),8512–8518.

Coto-Millán, P., Mateo-Mantecon, I., Domenech Quesada, J.L.,Carballo Panela, A. and Pesquera, M.A. (2010), Evaluation of Port externalities: The ecological Footprint. In Coto-Millán, P., Pesquera, M. A., and Castanedo, J. (Ed.), Essays on port economics. Springer Science & Business Media. Springer Heidelberg Dordrecht, London, New York.

35

Cullinane,K.,andCullinane,S.(2013),Atmosphericemissionsfromshipping:Theneedforregulationandapproachestocompliance.TransportReviews,33(4),377‐401.

Jalkanen, J. P., Brink, A., Kalli, J., and Stipa, T. (2008), State of the Art‐AIS basedemission calculations for the Baltic Sea shipping. Hrvatski meteorološkičasopis,43(43/1),177‐180.

Lam, J.S.L., Notteboom, T., (2014), The greening of ports: a comparison of portmanagementtoolsusedbyleadingportsinAsiaandEurope.Transp.Rev.34(2),169–189.

Miola,A.,Ciuffo,B.,Giovine,E.,Marra,M.(2010),Regulatingairemissionsfromships.TheStateoftheArtonMethodologies,TechnologiesandPolicyOptions.JointResearch Centre Reference Report, Luxembourg, EUR24602EN, 978‐92,ISBN.

Sabatier, P. A. (1986) Top‐down and bottom‐up approaches to implementationresearch:acriticalanalysisandsuggestedsynthesis.Journalofpublicpolicy6(1),21‐48.

Tzannatos,E.(2010),ShipemissionsandtheirexternalitiesfortheportofPiraeus‐Greece.Atmos.Environ.44(3),400–407.

UNESCAP(2011),Eco‐efficiencyindicators:MeasuringResource‐useEfficiencyandthe impact of Economic Activities on the Environment. United NationsPublication,EnvironmentandDevelopmentDivision,Bangkok,Thailand.

Tichavska,M.,Tovar,B.,(2015)Port‐cityExhaustEmissionModel:anapplicationtoCruise and Ferry operations in Las Palmas Port. TransportationResearchPartA:PolicyandPractice,78,347‐360.

Goldsworthy, L. and Goldsworthy, B. (2015), Modelling of ship engine exhaustemissions in ports and extensive coastal waters based on terrestrial AISdata–AnAustraliancasestudy.EnvironmentalModelling&Software,63,45‐60.

36

37

Chapter1Port‐cityexhaustemissionmodel:AnapplicationtocruiseandferryoperationsinLasPalmasPort1.1.IntroductionFerryandcruise,sharepositiveeffectsandeconomicbenefits inportsandcities.However negative impacts, including air pollution, also relate to engine exhaustemissionswhileoperatingatport(Castellsetal.,2014;Changetal.,2014).Shippingactivityandthepropagationofexhaustgasesresultingfromthecombustionoffuelshaveasignificantimpactonairqualityinport‐cityareas.Mostimportantly,however,harmful ship emissions into the air have been addressed as a risk factor forcardiovascular,respiratoryconditionsorevenhumandeath(Corbettetal.,2007).Asaresultofthis,complianceandenhancementofemissionregulationinshippinghasbeenpursued.Policymakersneedthesupportofmethodologieswhichwillreliablyinformthemonhowmuch,where,howandwho releases emissions inorder todecideonaneffective regulatory framework for the improvement of air quality and thereductionsofgreenhousegasesemissions.Datascarcityanduncertaintyhasledtoawidespreaduseofmethodologiesforestimatingtheexhaustemissionsinshipping.Still,theutilizationofnewtechnologieswithreliabledataretrievingcapabilitiesonvesseltrafficquestionusefulnessofthemethodologiesproposedsofar(Miolaetal.,2010).Morespecifically,AIS‐assistedemissioninventoriescanbeeffectivelyusedtoassesstheimpactofshippinginportareas(Ngetal.,2012).Thechallengeofidentifyingoperativeprofilesofshipsatberth(hotelling),manoeuvringandnormalcruisingnavigation in conjunction with the emission dependency on engine load can be

38

addressedwithshippositionrecordsanddatabasescontainingshiptechnicalandengine details, respectively. This offers the ability to model the geographicalcharacterizationofemissionsthroughhigh‐resolutionmapsinport‐cityareas.AIS‐basedmethodologieshavebeenalreadyintroducedtoestimateshippingemissions(Jalkanenetal.,2008),buttheyhaveneverbeenpresentedasaninstrumenttoassistpolicydesignandcorrectivemeasuresofaspecificshippingsector(passenger)andsubsectors(cruiseandferry)withinanislandcontext.Thus,themaincontributionofthispaperistopresentevidenceontheapplicationofAIS‐basedmethodologiesto assess exhaust emissions of cruise and ferry services according to ship sizeclasses,timeandshipactivityphases(i.e.hotelling,manoeuvringandcruising).The results of this study aim at the improvement of the current environmentalpolicy inLasPalmasPort, aswell as inother islandportsexperiencinga similarshippingactivity.Thestructureofthispaperisasfollows:Section1.2providesanoverviewonregulationandpracticesforthecontrolofshipexhaustemissionsinports.Section1.3describesemissionestimationinshippingtothenpresentthefullbottom‐upShipTrafficAssessmentModel(STEAM)usedinthiscasestudy.Section1.4presentsresultsforemissionestimationasabreakdownofshipactivityphase,typeandtimeandshipsizeclasses.Also,thegeographicalcharacterizationofresultsis described through a selection of high‐resolution maps. Discussion and policyimplications are presented in section 1.5, followed by conclusions and futureresearchrecommendationsinsection1.6.1.2. Regulation and practices for the control of shipexhaustemissionsinportsCurrentregulationseekstoreduceemissionsfromshipsthroughtheintroductionof minimum fuel quality standards and the implementation of new abatementtechnologies. The InternationalMaritimeOrganization (IMO)has addressed shippollutionundertheMARPOLconvention.Theregulationofairpollutionbyshipswas defined in MARPOL Annex VI, first adopted in 1997 and enforced in 2005includingaprogressivereductionofSOxandNOxandindirectlyParticulateMatter

39

(PM)inEmissionControlAreas(ECA).MARPOLAnnexVIistheonlyglobalregimethatclearlyaddressesthecontrolofairemissionsfromships.

The European Union (EU) has also expressed its concerns about the impact oftransport on air quality through the Strategy for Sustainable DevelopmentpublishedonitsWhitePaperonTransportPolicy(Gemeinschaften,2001),leadingto the establishment of stringent sulphur regulation for marine fuels throughdirectives:2012/33/EU,2005/33and1999/32.Accordingtothese,allpassengershipsoperatingonscheduledservices toor fromanyEUport shouldnotexceed1.5%sulphurlimitandallvesselscallingatanEUportshoulduselowsulphurfuel(lessthan0.1%)orashore‐sideelectricityfacilityduringportstayslongerthantwohours.Inaddition,withtheframeworkofIMOregulations,MARPOLAnnexVIsetsamaximum 0.1% sulphur for all ship operations in ECAs from 2015, which withregard to European waters are currently limited to the Baltic Sea, the EnglishChannelandtheNorthSea.ItshouldbealsonotedthattheexpressedEUwillingnesstounilaterallywidentheenforcementofsulphurrestrictions toallEuropeanseafacescomplianceconstraints inrelation to theUnitedNationsConventionontheLawoftheSea1982(UNCLOS)towhichtheEUissignatory.ThereiscurrentlynolegalbasisfortheEUtoexerciseextra‐territorialjurisdictionandthisislikelytogivenon‐EUstatesand industrialbodiesgrounds for challengingemissions reductionmeasuresadoptedbytheEUformaritimetransport(Miolaetal.,2010).As ports constitute the nodes of maritime transport where all shipping routesultimately converge, theyareparticularlyexposed to theburdenof shipexhaustemissions.Therefore,inresponsetothisproblemandbesidestheprovisionsoftheIMO and EU framework, they have been collectively or individually active inadoptingvoluntarymeasures,whichaimatimprovingtheairqualityandachievingemission reductions of greenhouse gases (CO2). Thesemeasures either take theformofofferingeconomicincentives(i.e.environmentallydifferentiatedportdues)ortheundertakingofinfrastructuralinvestments,whichencourageshipoperatorsto make use of environment friendly services (i.e. shore‐side electricity, LNGbunkering, automatedmooring systems, andothers). For further information on

40

greenshippromotionthroughmajorEuropeanportsandtheuseofenvironmentalindices,thereaderisreferredtoGibbsetal.,(2014).Atpresent,thereisnopreferenceforaspecificenvironmentallydifferentiatedportcharging system, although it should be noted that the relevant EU proposal(COM/2013/295) in action 8 suggests that “to encourage a more consistentapplication of environmentally differentiated port infrastructure charges, theCommissionwillproposeprinciples forenvironmentalchargingandpromotetheexchangeofgoodpracticesby2015”inordertoabateairemissionsandaddressthetechnologicalalternativesavailableforthis.Withregardtoinfrastructuralportfacilities,shore‐sideLNGbunkering19isevidentin Norway (i.e. Kristiansund, Mongstad, Bergen), Sweden (i.e. Stockholm), theNetherlands (i.e. Rotterdam, Antwerp, Amsterdam) and Belgium (i.e. Zeebruge)whilstthefeasibilityofprovidingshore‐sideelectricityserviceshasbeenstudiedinvariouslocations(Tzannatos,2010)andisalreadyofferedtoshipsinthewestcoastof the USA. In summary, a positive evolution of exhaust emissions related toshippinghascertainlyledtoregulatorystringencyandtechnologicaldevelopmentof abatement options. Nevertheless and regardless their existence, adoptionfeasibilityofabatementalternativesbyportsandfleet,isoftenlimitedbycosts,alack of incentive regulation (Buhaug et al., 2009) and diversified approachestowards air emissions. For instance, in Spain,monitoring and abatement actionsaddressedbyportstowardsairemissionshavebeensofar,mostlyfocusedonland‐basedandnotsea‐basedactivities(PuertosdelEstado,2011). Indeed, ifexistent,diverseapproachescanbeidentifiedwithinsustainabilityreports.Ononehand,sea‐based emissions fromshipping are referredas an externality related to theportcommunityandnottothePortAuthority(monitoringonly, land‐basedemissionsfacingpiers or near parking and transit areaof trucks); on the other hand, PortAuthorities as Port of Cartagena, do aim to reduce sea‐based emissions by

19 World map with LNG bunkering activities in ports is accessible online athttp://www.lngbunkering.org/.Websitelaunchedin2014bytheInternationalAssociationofPortsandHarbours.

41

monitoring, in line with the present study, three operative scenarios: cruising,manoeuvring,andberthingofvessels.Inthisrespect,approachestosea‐basedemissionsbyportsmightevolvepositivelyover time. Particularly sincemarket value and competitiveness of ports‐cities isincreasinglystrengthenedbysustainablecontributionsinaccordancetoregulatoryframeworkand retrofittingof global fleet (Merk,2013;Leeat al., 2014). Indeed,management tools at port are also addressed to enforce or encourage greendevelopment (Lam and Notteboom, 2014) and the abatement of emissions. Forinstance, resolving berth and quay‐crane allocation issues considering fuelconsumption and speed reduction (Du et al., 2011; Hu et al., 2014) starts to beaddressedinliterature(Zisetal.,2014).Inrelationtothis,trialscarriedoutbyBPandMaerskwithtankersshowedpromisingresults,withsavingsupto27%infuelconsumptionforsomejourneys,andaveragesavingsbetween12%and20%(Gibbsetal.,2014).1.3.MethodologyOver theyears, thequantityandgeographicalcharacterizationofemissionshavebeenconsideredforvaluationinmaritimetransport.Bottom‐up(Tzannatos,2010;Paxian et al., 2010) and top‐down (Endresen et al., 2007) methodologicalapproaches have been applied over time, to quantify emissions. A bottom‐upapproachisreferredtocalculationsbasedonfleetactivity.Thiscanbedonebyusingport calls andestimatedvesseloperativeor, throughvessel tracksand real timeoperative.Ontheotherhand,atop‐downapproachisreferredtoestimationsbasedonfuelsalesstatistics20.Regardingthegeographicalcharacterizationofemissionsand the levelofdetailachieved, this isalsodependenton theapproach followed(bottom‐up and top‐down). Hence, with a bottom‐up approach, individualinformationofvesselsanditspositionaretakenintoconsiderationwhilewithatop‐down approach valuation is based without, or with partial information on the20TichavskaandTovar (2015)shows that the representativeapproach inharbourstudies is thebottom‐up approach althoughmost of them are based on port calls and theoretical operative ofvessels.This,despitetheaccuracylevelofdetailobtainedfromvesseltracks(high‐definitiontrafficinformation that avoids operative assumptions in‐port and is based on real and not theoreticaloperativeofvessels).

42

positionofvessels(i.e.thegeographicalactivityofshippingisestimatedbasedonasingleshippingrouteoraparticulargeographicactivitycell,nomatterwhichvesselcarriesouttheactivity).Finally,afullbottom‐upapproach,describedbyMiolaetal.,(2010), istheuseofbottom‐upapproachesboth, foremissionestimationandthegeographicalcharacterizationofresults.Datascarcity,andassumptionsinliteratureresultinanopendebateonadequacyofapproachesandcontextsanalysedsofar(Miolaetal.,2010).Buhaugetal.(2009)hasmadeanattempt tohomogenizeresults fromdifferent studies.Uncertaintieswhen comparing results, confirm the need for the so‐called, full bottom‐upapproach21, by integrating a great extent of traffic data (vesselmovements, portcalls)andtechnicalcharacteristicsofvessels.Technicalinformation,shipactivitiesand geographical distribution of maritime traffic were, until a few years ago,retrieved from average information (i.e. Du et al., 2011). Today, if not in a totalextent,improvementsinaccuracycanbeachievedasaresultoftheintroductionofinnovativetechnologiesforvesselmonitoringatsea.

AIS‐basedinventoriescanbeeffectivelyusedtoassessshippingemissionsinportareas.Thechallengeofidentifyingoperatingprofilesofshipsatberth,manoeuvringand normal cruising navigation and, emission relation to engine load can beaddressedwith position records and commercial databases containing technicalandenginedetails.Thisguaranteeslocation,speed,route,dimensions,engineandfuelconsumption foreachship isalwaysacknowledged,providing theadditionalbenefit of visualizing results through high‐resolution maps presenting thegeographicalcharacterizationofemissionsinport‐cityareas.1.3.1.TheSTEAManditsmaincomponentsEmissionresultsinthiscasestudy(NOx,SOx,PM2.5,COandCO2,)areestimatedbythe STEAM, model specifically designed to assess emissions based on shippingactivityandinformationprovidedbytheAutomaticIdentificationSystem(AIS)22.21Thisistheapproachfollowedbythemodelusedinthiscasestudy:theSTEAM.22TheAISwasconceivedasanavigationalaidforshipmonitoringandcollisionavoidanceatsea.AccordingtotheIMOregulation19.2ofSafetyOfLifeatSeas(SOLAS),anAIStransceivershallbe

43

Themodellingofemissionsisbasedonbasicprinciplesofshipdesign,includingthemodelling of the propelling power for each vessel in terms of its speed. TheinstantaneousvelocityisobtainedfromtheAISsignal,whichisalsousedtoidentifythe ship.Also, engine loads during voyages are determined with reasonableaccuracybasedontheratioofshipspeedandthecalculatedresistancethattheshipisrequiredtoovercomeataspecifiedspeed(Jalkanenetal.,2009).The effect of waves on the consumption of fuel and on the emissions to theatmosphereisalsomodelled.Atitscurrentstate,theprogramtakesweathereffectsintoaccountbyapplyingadditionalpowerrequirementsinbadweatherareaswithhighvaluesofsignificantwaveheight.ThewaveheightdataisobtainedfromtheWAve Model (WAM) further described in Thomen et al., (1994) and Tuomi,(2008).Theadditionalpowerrequirementdependsonparametersdescribingthethree‐dimensionalstructureofthehullandthedirectionofthewaves(Jalkanenetal., 2008).Also, shipping routes and speed changes are included specifically andthereisnoneedtoguesswhichroutesshipsmaytakeduringthevoyage.Themodelalsoincludesanenhancedmodellingofthepowerconsumptionofauxiliaryengines,whichdependonshiptypeanditsoperationmode(Jalkanenetal.,2012).Model components are presented in Figure 1. Input data and output results arepresented in the uppermost and lowest row of rectangles. Arrows describe theinformationflowofthemodel,dependencybetweenfactorsinadditiontodottedandsolidarrowsusedforvisualclarity.Coloursdenotevariablecategoriesincludedinthemodel.

equippedineverysea‐goingshiplargerthan300grosstonsandeverypassengervesselirrespectiveof size. Its system regularly transmits static and voyage‐related information every 6minutes inaddition to dynamic informationwith a frequency related to the vessel´s speed underway (2‐10seconds)andnavigationalstatus(3min.whenanchored).Forthiscasestudy,dynamicinformationfromtheAISwasusedfortheestimations(seesection1.3.2).

44

Figure1–MaincomponentsoftheSTEAMandtheirinter‐relations

Source:adaptedfromJalkanenetal.(2012).1.3.2.InputvaluesAshipdatabaseofover50,000vesselparticulars(overa thirdpartof theglobalfleet)andAISpositionrecordsdefineinputvaluesfortheSTEAM.Ship database considers: physical properties, engine particulars and variablespresentedonTable1.Databaseholdsinformationamongothers,onexperimentallydetermined emission factors, installed abatement techniques, shaft generators,specific fuel oil consumption, fuel type and sulphur content of fuel formain andauxiliaryengines.Informationwasprovidedmainly,by:IHSFairplayshipregister,enginemanufacturers,localauthoritiesandshipowners.For this case study, AIS data gathered from coastal stations was provided byMarineTraffic.Thisresultsinadataflowofthousandsofdynamicinformationpership, per year. Longitude, Latitude, a unique nine digit MaritimeMobile ServiceIdentity(MMSI)ofvessels,NavigationStatus(atanchor,underwayusingengine,notunder commandamongothers), Speedover ground,Course (relative to truenorthto0.1°),Heading(0to359degrees)andTimestampinCoordinatedUniversalTime(UTC)integratedynamicfieldsfromAISdatatransmittedbypassengervesselsduring2011inLasPalmasPort(areadefinedbyboundingcoordinates:LATfrom28°to28,45°,LONfrom‐15,60°to15°)

45

Table1‐STEAMmodelregardingthepropertiesofshipsIdentification Physicalproperties Mainengineproperties

Auxiliaryengineproperties

Shipname Length ME,Fuelsulphurcontent AE,installedkWIMOregistrynumber Breadth ME,abatementtechnique NumberofAEMMSIcode Draught ME,SFOC AE,FueltypeShiptype Buildyear ME,design AE,Fuelsulphur

contentGrosstonnage Designspeed ME,model AE,SFOCDeadweighttonnage Numberofcabins ME,stroketype AE,abatement

technique Hulltype ME,rpm NumberofME ME,installedkW ME,Fueltype1 ME,Fueltype2 MeasuredEFsNote: MMSI = Mobile Maritime Service Identity, ME=Main engine, AE=Auxiliary engine, rpm=crankshaftrevolutionsperminute,SFOC=SpecificFuelOilConsumption,MeasuredEF=ExperimentalvalueforemissionfactorsofNOx,SOx,COandPM.Source:adapted from Jalkanenetal. (2009).Additionaldata fields, like theexistenceofbulbousbow,propellerdetailsandcargocapacityarealsoused(seeJalkanenetal.,2012).1.3.3.ModelperformanceanduncertaintyconsiderationsAs described in Johansson et al., (2013), the model has been able to predictaggregateannualfuelconsumptionofacollectionoflargemarineshipswithameanprediction error between 9 and 15%. Still, uncertainties regarding properties ofshipsandAISinformationshouldbeconsideredaswellaslimitationsforauxiliarypowerpredictionsanditsrelevanceforshipemissionsoccurringinportareas.Forinstance,indespiteoftheconsiderationofnavigationresistancethroughwaves,power prediction during voyages may be also uncertain due to the neglect ofadditionalenvironmentaleffectsinestimations(e.g.wind,currents,seaicecoversandothers).Inadditiontothis,insufficientinformationonempiricalperformanceofenginesandchemicalcompositionofexhaustgasesandparticulatematter,non‐existentinrecentliteratureandnotconsideredinthemodelshouldbealsotakenintoconsideration.Ship routes and a high‐resolution activity map can be expected from the AISmessages and their location signal transmitted. Still, the temporal and spatialcoverageoftheAISwilldependonperformanceofterrestrialreceivingstationsandequipment installed in vessels (AIS transponder, GPS, and navigation tools). An

46

accurate modelling of engine power usage during hotelling and manoeuvringoperationsinharbourareasrequiresafrequentupdateofdata(severaltimesperminute)asthespeedofvesselsmaychangeconstantly.Temporalgapsandalowfrequency of messages may be the cause of significant inaccuracies. Also,uncertaintiesrelatedtotheauxiliaryenginepowerusageonboardvesselscouldbesignificant. Commercial databases may offer an incomplete representation ofinstalled auxiliary engine power, which must be augmented with data fromclassificationsocieties, fleetownersorenginemanufacturers.This information isoften unavailable and must be estimated based on existing knowledge, likeextensive vessel boarding programs (Starcrest, 2011). In contrast to the mainengine power predictions, there is no accurate, generally available model forauxiliaryengineusageofshipsasitmayvaryinaccordancetotheircargohandlinggear,needforheating,cooling,pumpingandadditionalusesofenergy.AnoverviewoftheSTEAMhasbeenpresentedinthissection,forextensivedetails;thereaderisreferredtoJalkanenetal.,(2009),Jalkanenetal.,(2012)andJalkanenetal.,(2013).1.4.CaseStudy:LasPalmasPort.TheCanaryIslandsareoneofSpain'sseventeenAutonomousCommunitiesandanoutermost region of the European Union. Located in the Atlantic Ocean, thiscollection of seven islands is at 115 kilometres of distance from the northwestAfricancoastat1,200kilometresfromthenearestmainlandportinsouthernSpain(Cadiz).TheAutonomousCommunityoftheCanaryIslandsisadministratedbytwoprovinces,LasPalmasandSantaCruzdeTenerife.GranCanariaandTenerife,bothcapital islands, are the main transport nodes connecting the archipelago withmainlandSpainandothercountries.MainportsintheCanaryIslands,LasPalmasPort (located in Gran Canaria) and S.C. Tenerife Port (located in Tenerife) aremanagedbydifferentPortAuthorities23.In2011,cargotransportedintheseports,summeduptomorethan88%oftheCanaryIslandstotalfreight.

23AdetailedanalysisoftheportmanagementmodelinSpainisbeyondthescopeofthispaperbutitcouldbefoundinRodriguez‐ÁlvarezandTovar(2012).

47

TheadvantagesofhavinggoodportconnectivityhaveaccruedtotheCanaryIslandsonce one of their ports became an international hub24(Tovar et al., 2015). LasPalmasPortisamajorlogisticplatformbetweenEurope,AfricaandAmericaanditoffersmanyadvantagestoocean‐goingvesselssuchasarecognizedtechnicalandcommercial maritime community and competitiveness in supplies and repairservices.Itslocationbetweenmaincommercialtraderoutesmakesitacargohub(over 19 million tons from loading, unloading and transhipments). Moreover,passengertraffic,withover908,000passengers in2011 isgrowingsteadilyovertime.To meet the maritime transportation demand in the Canarian archipelago(passengers or passengers and goods), ferry routes are offered in a daily basisthroughdirectorscaledservices.Huboperationsaresetinbothmaincanariesports.Direct connections are regularly offered from Gran Canaria, and Tenerife indirectiontootherCanaryIslands:Fuerteventura,Lanzarote,LaGomeraandSpanishmainland (Huelva). In addition to the regular ferry services, passenger numbersaccountedincruiseoperationsintheCanaryIslandshaveincreasedsteadilyupto1,600,000withamajorparticipationofCarnival(49%)andRoyalCaribbeancruiselines(23%),(EDEI,2011).AccordingtoEPPE,passengershareofLasPalmasPortincreased in over 20%with a total of 1,605,531 passengers in 2013. Sustainedmarketgrowthincreasestheneedtoidentifyandmeasureenvironmentalimpactsgeneratedbypassengertraffic,particularlyinlocationswhereactionstomitigatethesearenotbeingpursuedasitisinmanyotherharboursintheworld.Aswe statedbefore, cruisepassengersnumbers and cruise servicesdeployed inCanarianportsalso increaseover time. Indeed,LasPalmasPortholdsoneof thelargestsharesincontinuousgrowth.Forinstance,in2011,197portvisits,418,184cruisepassengers(22%fromthetotalshareofthearchipelago)andupto57cruiseships were accounted. In 2013, a growth in numbers can be noticed with the425,267 cruise passengers reported, while a remarked increase can be noticed

24Drivenmainlybycontaineroperations,thetransshipmenttrafficinLasPalmasPorthasreachedaratecloseto69%ofthetotalnumberofhandledcontainers,whereasTenerifeportfocusitscontainertrafficmerelyonthedomesticmarket.

48

duringthefirstquarterof2014withatotalof511,248.Thiscanbeattributedtotheconcentration of ships with hub operations. In fact, further expansion of cruiseoperationscouldbeexpectedasrecruitmentofnewvesselsandthehostingofnewhubservicesovertheyearalsorelate‐inadditiontowhatwereferinpreviouslines‐to socio‐political stability, currency, infrastructure improvement and recognizedqualityonlandservices,attributesacquiredandalsorecognizedinthisport.Emissionsestimatedforpassengertrafficarepresentedinthefollowingsectionasa breakdown of hotelling, manoeuvring cruising operations and size classes. Tofollow, geographical characterisationof resultswill be addressedby a sampleofhigh‐resolutionmaps.1.4.1.ResultsAccordingtoterrestrialAISdatareceivedduring2011,3,183uniqueshipsentered,navigatedandanchoredinLasPalmasPortduring2011.Fromthese,1,228ofthem(39%)areclassifiedascargo,329(10%)ascontainer,612(19%)astankersand,seventy‐fourvessels(2%)aspassengervessels.Enginerequirementsandoperativeneedsamongshippingsectorsarediverse,particularlyinaport‐citycontextwherethe speed and load rates of engines will constantly change according torequirementsofportsandvessels.Indeed,theenergydemand,theengineloadsandresultingemissionsfromdifferenttypeofvesselsmayvary.Tosetanexamplewemayremarkhowlargecruisevessels(withmorethanathousandair‐conditionedcabins), will most probably demand more energy and contribute with moreemissionsthanacargocarrierwhenatport.Also,theregularandtightschedulesofferryvesselsmayincreasetheloadlevelofenginesduringacceleration‐decelerationandhencealsoincreasetherelatedemissionshare(Jalkanenetal.,2009).Inordertobetterunderstandthesedifferences,emissionresultsforNOx,SOx,PM2.5,CO, CO2 and operative time spent at hotelling, manoeuvring and cruising arepresentedinTable2forninedifferentshiptypes.Shiptypesanditsoperativetype

49

andtimeareanautomatedclassificationfrominformationtransmittedbytheAIS25.FromTable2,anoticeableshareofemissionsrelatedtolocal(NOx,SOx,CO,PM2.5)andglobaleffects(CO2)areseenforpassenger,containerandtankervesselswhilecategories as others and unknown (mostly small sized ships) contribute with aminimumamountofemissionsindespitetheconsiderablenumberofhourstheyspend at port. Shipping sectors identified in this research as the ones with thehighest level of emission contribution (passenger, container, and tanker vessels)have been also identified as shipping categories with representative emissionsharesbyotherstudiesasStipaetal.,(2007);DeMeyeretal.,(2008);Howittetal.,(2010);Eijgelaaretal.,(2010);BerechmanandTseng,(2012);Ngetal.,(20129;Kallietal.,(2013);Jalkanenetal.,(2013);Johanssonetal.,(2013).Table2‐Emissionsandoperativetimerelatedtotrafficsectorsaspredicted

bytheSTEAM

Note:Fuelconsumptionandemissionsreleasedwhileatberthareincluded.ComplianceofIMOandEUregulatoryframeworksformarinefuelsisassumedandalsoincludedintheseestimations.TimespentinHotelling(H),Manoeuvring(M),Cruising(C)anditsTotal(T)isexpressedinhoursEmissionresultsforallcategoriesarepresentedasarelativepercentagefromthetotalsatport,inFigure2.Thisfigurebetterreflectsshipemissionscontributionatthesourceandenablestheviewofremarkedshares.Particularly,thepassengersubsectorsareidentifiedasthemainsourceofemissionswiththeexceptionofCO.Inaddition to this, major figures of emissions are allocated in container vessels,followedcloselybytankers.Moreover,figure3showsthatovera35%ofSOxanda

25DetailsoncodesandcategoryofvesselstransmittedbytheAISarereferredintheITU‐R,(2010).Unknowncategoryreferstofailurecasesofidentification(nostaticmessagereceived,noconnectiontonationalMMSIdatabasesavailable,smallvesselsnottransmittingavalidIMOregistrynumber).

50

30%ofPM2.5resultfromthepassengersectorwhilealmosta50%ofresultingNOx,COandCO2deriveboth,frompassengerandcontainerships.

Figure2–EmissionpercentagebyshippingsectorinLasPalmasPort

Ontheotherhand,Figure3reflectshoursofhotelling,manoeuvringandcruisingasarelativepercentageofoperativetimeatport.Inthisregard,andasitwasstatedbefore, the activity levels of other and unknown categories are noticeable.Specificallywhencomparedtotheirrelativelowemissionshare.Theexplanationforthismightrelyonsmallsizedenginesandalowdemandofenergyfromleisurenavigation.Ontheotherhand,differencesamongoperativetime(5%)comparingpassengervesselstotherest,shouldbenoted.Thisissignificantlylow,whilecargo,container and tanker categories together represent almost a 45% from the totalactivity of vessels at port, However, when the profile is observed by type ofoperativetheroleofcruisingisclearlyrepresentativeforthepassengersectorwhilecargo,containersandtankersreflectaratherbalancedoperativeprofile.

51

Figure3–OperativepercentagebyshippingsectorinLasPalmasPort

According to terrestrial AIS data received during 2011, seventy‐four uniquepassengershipsentered,navigatedandanchoredinLasPalmasPortduring2011.Fromthese, fifty‐sevenvessels(75%)areclassifiedascruiseshipsandtwelveofthem(19,7%)asRoPaxvesselsoperatingferryservices.Consideringpollutingandoperative differences among shipping sectors, it is also of interest to analysedifferenceswhencategorizedasthesub‐groupsthatregardthiscasestudy(cruiseand ferry). In order to address this, Figure 4 and 5 reflect results as a relativepercentageofthetotalemissionsandoperativetimeatport. Itcanbenotedthatcontributionofferriesissignificantlylargeinoverallemissions.ParticularlySOx,COandPM2.5(relatedtoseverehealthconsequences)andCO2(relatedtogreenhousegaseffectsandglobalwarming)whiledifferencesonsharesofoperativetimearealsonoticeable.Ferryvesselsreflectanoticeablepercentageofcruising,similartocargo sectors (seeFigure2and3)while in the caseof cruisevessels, this isnotrepresentative.

52

Figure4–EmissionpercentageofcruiseandferryvesselsrelativetothetotalemissionsinLasPalmasPort

Figure5–Operativepercentageofcruiseandferryvesselsrelativetothe

totaloperativeinLasPalmasPort

Shipsize(weight)isalsocrucialtofuelconsumptionandemissionestimation.Table3presentsresultsforcruiseandferryasatotalofpassengercategories.Accordingtoresultsandintermsofemissionstheroleofthelargestsizeclasses(from60ktto80kt)isrepresentativeforcruiseoperations,whileforferryservicesthisismainlyallocatedbetween10and30kt.

53

Intermsofpollutantscommonlyrelatedtolocaleffects(NOx,SOxandPM2.5),thelargestshareofresultsisattributedtoferryvesselsbetween10kt‐30kt,andcruisevesselswithaGTbetween30kt‐45ktandover80kt.Ontheotherhand,emissionsrelatedtoglobaleffectsallocate,mostly,underthelargestclassesofcruisevesselsandferriesbetween10and30kt.Regardingoperativedetails,Table3alsoenablesadetailedanalysisofcruiseandferries. In this respect, although the relative importance of operative typesmeasuredinhours,isthesame(hotelling,withthelargestshareofhoursfollowedbycruisingandmanoeuvring)eachsubsectorhasparticularities.Thatis,thatcruisevessels spend more time in hotelling (89%) than ferries (81%) being resultsoppositewhenreferredtocruising(9%forcruiseand17%forferries).Operativeprofiles of vessels are consistent with the overall nature of the passenger sub‐sectors.Thismeanscruisingandhotellingsharesderivedfromloading/unloadingoperationsandwaitinginharbourareaswhilepassengersgoashore,inthecaseofcruise.Inthecaseofferries,tightschedules,frequentservicesandalowturnaroundtimeatport.Table3‐Exhaustemissionsandoperativetimeofcruiseandferryvessels

accordingtosizeclassesaspredictedbytheSTEAM

Note:Fuelconsumptionandemissionsreleasedwhileatberthareincluded.ComplianceofIMOandEUregulatoryframeworksformarinefuelsisassumedandalsoincludedintheseestimations.

LasPalmasPort(2011)

NOx[Ton]

SOx[Ton]

PM2.5[Ton]

CO[Ton]

CO2[Ton]

H[Hours]

M[Hours]

C[Hours]

T[Hours]

GTBelow4kt 2 1 0 0 96 10 1 3 14GTBetween4kt‐10kt 9 2 0 1 474 1575 4 17 1596GTBetween10kt‐20kt 3 1 0 0 133 55 2 10 67GTBetween20kt‐30kt 7 2 0 1 355 148 6 29 183GTBetween30kt‐45kt 26 10 2 2 1261 371 13 50 433GTBetween45kt‐60kt 22 9 2 2 1003 195 9 32 236GTBetween60kt‐80kt 74 39 8 9 3529 1002 31 179 1212GTover80kt 16 13 3 3 831 301 15 63 379

TotalCruise 158 75 16 18 7683 3657 79 384 4120GTBelow4kt 0 0 0 0 21 199 1 1 202GTBetween4kt‐10kt 57 13 3 6 2590 3046 45 593 3685GTBetween10kt‐20kt 588 316 65 47 27446 10854 329 2433 13616GTBetween20kt�