文献2：Knowledge brokering on emissions modelling in Strategic Environmental (2)

Knowledge brokering on emissions modelling in Strategic Environmental Assessment of Estonian energy policy with special reference to the LEAP modelPiret Kuldna a ,⁎,Kaja Peterson a ,Reeli Kuhi-Thalfeldt a ,ba Stockholm Environment Institute Tallinn Centre,Lai 34,Tallinn 10133,Estonia bTallinn University of Technology,Ehitajate tee 5,Tallinn 19086,Estoniaa b s t r a c ta r t i c l e i n f o Article history:Received 22December 2014Received in revised form 4June 2015Accepted 4June 2015Available online 11June 2015Keywords:Strategic Environmental Assessment Knowledge brokering LEAPEnergy policyStrategic Environmental Assessment (SEA)serves as a platform for bringing together researchers,policy devel-opers and other stakeholders to evaluate and communicate signi ficant environmental and socio-economic effects of policies,plans and programmes.Quantitative computer models can facilitate knowledge exchange between various parties that strive to use scienti fic findings to guide policy-making decisions.The process of facilitating knowledge generation and exchange,i.e.knowledge brokerage,has been increasingly explored,but there is not much evidence in the literature on how knowledge brokerage activities are used in full cycles of SEAs which employ quantitative models.We report on the SEA process of the national energy plan with re flections on where and how the Long-range Energy Alternatives Planning (LEAP)model was used for knowledge broker-age on emissions modelling between researchers and policy developers.Our main suggestion is that applying a quantitative model not only in ex ante ,but also ex post scenario modelling and associated impact assessment can facilitate systematic and inspiring knowledge exchange process on a policy problem and capacity building of par-ticipating actors.©2015Elsevier Inc.All rights reserved.1.IntroductionSEA is a type of impact assessment to integrate environmental con-cerns into strategic decision-making with the aim to promote sustain-able development.In a decision-making process,impact assessments can serve various and simultaneous functions:information generation,debate and deliberation,attitude shifting and complexity structuring (Hugéet al.,2011).Farrell et al.(2001)also emphasise the communica-tive role of assessment processes as bridges between science and policy.In the energy sector,where impact assessment deals with complex technical and structural issues,uncertainties and multiple impacts,quantitative models are widely used as planning and analysis tools for addressing these challenges.While modelling itself is not driving the policy,nor deciding the political feasibility of targets,quantitative models can offer structured insight into areas of uncertainty (Strachan et al.,2009;Mundaca et al.,2010),help understand the interactions and promote discussion (Jebaraj and Iniyan,2006)and support various approaches to future studies (Finnveden et al.,2003).Van Daalen et al.(2002)have summarised four contributions of quantitative models to the environmental policy-making life cycle:1)eye-opener,2)visualiserof alternative future scenarios,3)vehicle for inspiring political consen-sus,and 4)assistant in identifying concrete policy decisions and poten-tial policy outcomes.However,the use of models in policy development is not always self-evident.Brugnach et al.(2007)suggest that both the modelling and the policy-making communities should contribute to the integration of modelling information into policy formulation.For modellers,this involves promoting models as tools of communication,learning and exploration,and for policy-makers,this suggests a need to view modelling as a tool for informing the public and scienti fic com-munity in an uncertain world.SEAs can serve as platforms for enhancing such knowledge ex-change,where information is not simply transferred from researchers to decision-makers,but developed and co-produced interactively with actors involved (Sheate and Partidário,2010;Fazey et al.,2012).Assess-ments are also more likely to in fluence decision-making if the decision-makers are sharing and acquiring not just information but knowledge (information that has been processed through learning)(Sheate and Partidário,2010).Similarly,model generated output alone may not be enough for decision-making.However,knowledge exchange is often conducted on an ad-hoc basis,based on ‘what seems to work ’(Reed et al.,2014).In impact assessments,Partidário and Sheate (2013)sug-gest not simply hope that the process of facilitating knowledge genera-tion and exchange,i.e.knowledge brokerage,will happen,but explicitly design it in order to support constructive and collaborative planningEnvironmental Impact Assessment Review 54(2015)55–60⁎Corresponding author.E-mail address:piret.kuldna@seit.ee (P.Kuldna)./10.1016/j.eiar.2015.06.0010195-9255/©2015Elsevier Inc.All rightsreserved.Contents lists available at ScienceDirectEnvironmental Impact Assessment Reviewj o u r n a l h o me p a g e :ww w.e l s e v i e r.c o m /l o c a t e /e i a rprocesses and ultimately more sustainable outcomes.At the same time impact assessors need to adopt aflexible,adaptive and learning ap-proach themselves too,as the decision-making processes are often long and unpredictable(Kørnøv and Thissen,2000).This is supported by Runhaar and Driessen(2007)who argue that SEA studies will obvi-ously have a greater impact when they areflexible and tailored to the actual evolvement of policy processes,rather than adhering to strict, standardised and detailed procedures.Likewise,knowledge brokering activities in SEA should also depend on the needs and expectations of actors involved.Knowledge brokering has been categorised in several ways.For example,Ward et al.(2009)identify three models of knowledge brokering:knowledge management,linkage and exchange,and capaci-ty building,which represent common roles performed by knowledge brokers.In environmental modelling,Krueger et al.(2012)highlight the convening,communicating,mediating and translating roles of knowledge brokers.Turnhout et al.(2013)differentiate three knowl-edge brokering repertoires:supplying,bridging and facilitating,illus-trating the relation between knowledge production and use.Michaels (2009)specifies six knowledge brokering strategies to environmental policy problems and settings and based on her work,Jones et al. (2013)set out six functions of knowledge intermediaries.These strate-gies/functions are listed in order of increasing intensity of relationship building and commitment of resources:•informing(disseminating research results to policy developers,limit-ed exchange between the producers and users of the knowledge);•consulting/linking(advising on problems delineated by parties seeking counsel,linking policy developers with the expertise needed for a par-ticular policy area);•matchmaking(bringing together individuals who can contribute to policy-making and who would not otherwise meet,e.g.from another discipline or with different types of knowledge);•engaging(involving other parties in discussions that are framed by one party,to provide knowledge on an as-needed basis);•collaborating(parties jointly frame the process and negotiate the sub-stance of issues,may need significant resources of time and money);•building adaptive capacity(parties jointly direct interactions and dis-cussions of problems with multiple dimensions in which learning is considered,to improve the ability of parties in handling multiple and emerging issues).Depending on the decision-making issue,all or some of these strat-egies may be appropriate at different points of the policy cycle.For in-stance,Jacobson et al.(2005)have shown that consulting can be an effective strategy for interactive knowledge transfer to enhance the use of research-based knowledge in decision-making.Sheate and Partidário(2010)demonstrate the role of engagement and capacity building in strategic planning and assessment.In order to distinguish between different knowledge brokerage strategies related to environmental policy problems,we take the typol-ogy of Michaels(2009)as a starting point and examine the knowledge exchange process on emissions modelling between researchers and pol-icy developers in the national energy plan SEA.In our study the model is LEAP(Long-range Energy Alternatives Planning model)which is a soft-ware tool for energy-environment policy analysis and impact predic-tion,developed at the Stockholm Environment Institute(Heaps,2008; ).LEAP can be used for both demand and supply side energy modelling,including for building and comparing scenarios to meet the energy demand in all sectors of an economy and account for energy and non-energy sector greenhouse gas(GHG) emission sources and sinks.The literature shows that most of the LEAP-model applications have been at the national level,primarily driv-en by policy requirements:for developing low-carbon economies, future sustainable energy and climate policies,and renewable energy development scenarios or to reduce CO2emissions(e.g.,Tao et al., 2011;Yophy et al.,2011;Roinioti et al.,2012;Özer et al.,2013;Park et al.,2013).At the city level LEAP has been applied for urban energy saving and emissions reductions(e.g.,Lin et al.,2010)and globally for developing forward-looking assessments of energy challenges (Nilsson et al.,2012).In Estonia,the LEAP model has been used for the national level ex ante SEA of Energy Plan2020and in ex post scenario modelling and associated impact assessment for the preparation of En-ergy Plan2030+.The paper is structured as follows.Section2describes the knowledge needs in the context of Estonian energy policy and knowledge generation with LEAP in the national energy plan. Section3describes the study methodology.In Sections4and5,the study results are presented and discussed,respectively.Thefinal section draws conclusions based on the studyfindings.2.Test case contextThe knowledge needs for formulating a national development plan for the energy sector(hereinafter:national energy plan)primarily in-clude information on available resources and policy measures needed to meet European Union(EU)and national climate and energy policy targets and to ensure energy rmation on environmental, economic and social impacts of energy policy measures as well as im-pact assessment procedural knowledge is also essential.In Estonia,the energy sector is dominated by one primary energy source:oil shale. As a domestic fossil fuel,oil shale provides for approximately70%of the country's total primary energy supply.Although oil shale increases energy security of Estonia(the country's overall level of energy import dependence is approximately15%),the energy sector is Estonia's most environmentally impactful sector.The majority of the country's GHG emissions come from oil shale extraction and combustion,which are the main drivers behind the national economy's high carbon intensity values(OECD/IEA,2013).The national energy plan is the highest-level strategic document on the Estonian energy sector which elaborates on development visions and objectives and describes measures for achiev-ing them.As such,the document formulates national energy policy,and its targets and measures have to be reflected down to the lower level development plans that implement this policy.The authority responsi-ble for national energy plan preparation is the Ministry of Economic Af-fairs and Communications(MoEAC,hereinafter:policy developers).The first Estonian energy sector national development plan was approved by the Estonian Parliament in1998and has now been reviewed three times,combined with SEA according to the EU Directive SEA Directive 2001/42/EC.The second and third authors of the current paper together with other energy experts(hereinafter:researchers)from the SEI Tallinn Centre(SEIT,a non-governmental research institute)carried out the SEA for Energy Plan2020upon being selected by the MoEAC through public procurement.The guidelines for SEA on the website of the Minis-try of the Environment do not outline specific requirements on SEA tool choice or application(Ministry of the Environment,2009).However, there has been a certain history of the use of quantitative models in SEAs for national energy plans since Energy Plan2020(Table1).2.1.The use of LEAP model in the SEA of the national energy planIn the Energy Plan2020,the application of LEAP–a new model for Estonian national energy planning–was proposed by the SEA re-searchers who had the LEAP modelling expertise.In particular,we pro-posed to model the most important emissions of energy sector and to find out which electricity and heat production scenarios have the lowest level of carbon dioxide(CO2)and sulphur dioxide(SO2).Since LEAP is a scenario-based modelling system,it was possible to work closely with the policy developers responsible for the preparation of the national en-ergy plan throughout the development of alternative scenarios.Issues regarding how detailed the model should be,what sectors should be modelled,and which parameters and input data to use were negotiated56P.Kuldna et al./Environmental Impact Assessment Review54(2015)55–60with them.The alternative scenarios in Energy Plan2020were policy scenarios which are appropriate to use if the future can be affected through strategic action(Finnveden et al.,2003).The policy changes of scenarios were characterised by different shares of fuels for electricity and heat generation and respective production capacities,based on the national targets in the energy sector.The alternative scenarios for elec-tricity and heat production werefirst drafted by the researchers during the SEA scoping stage and then specified in the impact assessment stage based on information gained from the policy developers.The scenarios were analysed according to the implications of different futures,on the basis of environmental,social and economic criteria proposed by both the researchers and policy developers,ing the backcasting approach(Robinson,2003).Out of the total26criteria,the LEAP model was used to develop projections for CO2and SO2emissions of the scenarios(two criteria),EcoSense model for calculating the full costs of electricity production technologies(1criterion)and qualitative multi-criteria analyses for assessing the impact of the scenarios against the rest24criteria(SEI Tallinn,2009).During the emissions modelling stages,the main role of the policy developers was to provide informa-tion on the planned energy supply and demand data,verify modelling assumptions and the set of scenarios.For data visualisation and communication in LEAP,one can present and share model assumptions and results through Excel,Word and PowerPoint.In Energy Plan2020,the LEAP model was used to visualise scenarios and to compare emissions values.The results were presented by the researchers in working meetings with the policy developers,in legally mandated public consultations of the SEA and at a series of public energy forums organised by the MoEAC that introduced national energy plan scenarios and implementation measures to a wider public.In the communication process,the main role of the policy developers was to assist in the interpretation of the quantitative modelling results of the emission scenarios.2.2.Energy Plan2030+In the autumn of2012,MoEAC and Estonian Development Fund (EDF)(hereinafter together or separately referred to as policy devel-opers)launched the updating of Energy Plan2020.The new plan,Energy Plan2030+,addresses a longer time horizon and wider scope issues than its predecessors.Its overall aim is to ensure energy supply in accor-dance with European Union climate and energy policy objectives while improving the state of the environment and long-term competitiveness of Estonia.The plan establishes strategic objectives for power engineer-ing,heat production,transportation,housing and domestic fuel produc-tion until2030and with an outlook to2050.Based on analyses and modelling results,Estonia's preferred energy supply scenario is selected in the plan(MoEAC,2015).Thus,at the initiation of the process,the pol-icy planning addressed a problem that can be characterised as a moder-ately structured policy problem(Hisschemöller and Hoppe,1995; Turnhout et al.,2007;Runhaar and Driessen,2007):there was an agree-ment of necessity for a new energy plan and a broad policy goal was set, but the means to reach the goal most effectively and efficiently were un-certain.Along with updating the plan,ex post evaluation of the imple-mentation of the previous national energy plan was carried out.While the need for mid-term and/or ex post evaluation of a sectoral develop-ment plan is stipulated in government regulations(Government of the Republic,2007),no explicit legal requirements specify that sectoral de-velopment plan impacts arising from implementation must be assessed during this process.The ex ante SEA of Energy Plan2030+was conduct-ed by the EDF(a state-run public institution),who coordinated together with the MoEAC the preparation of the national energy plan.As the LEAP model was applied to the impact assessment of the previous na-tional energy plan,the policy developers also considered it a candidate tool for the SEA of the new plan.In summary,knowledge exchange process between SEA experts, modellers and policy developers can evolve at any modelling stage within SEA,such as:1)while deciding on modelling objectives and scope at the SEA scoping stage,2)during model design and quantitative modelling(as scenarios are built and while model inputs are deter-mined)and3)in ex post evaluation and modelling(Table2).3.MethodologyIn order to reflect on where and how a knowledge brokerage ap-proach was applied to the emissions modelling process in the SEA and in the preparation of the new energy plan,we use Michaels' (2009)framework of strategies which has specifically been devel-oped for responding to environmental policy problems or settings (described in the Introduction).Our focus is on expert knowledge of researchers and policy developers,excluding knowledge generat-ed from other stakeholder and public participation.The study was divided into two parts:1)Analysis of SEA process for Energy Plan2020and ex post identifica-tion of knowledge brokerage strategies that the authors applied dur-ing the SEA.The analysis was based on information from the secondTable1Estonian national energy plans and main tools used in SEAs.Title of the national energy plan Year ofparliamentaryapprovalMain tools used in the SEANational Long-TermDevelopment Plan for theFuel and Energy Sector until2005and vision until20181998[No SEA was applied]National Long-Term Development Plan for the Fuel and Energy Sector until 2015,including SEA 2004Qualitative assessmentmatrices(Fischer,2007)—scoring of impact significanceNational Development Plan for the Energy Sector until2020 (Energy Plan2020), including SEA 2009LEAP(Heaps,2008)—modelling of CO2and SO2emissions in energy supplyand consumption scenariosEcoSense(Schleisner,2000;Finnveden et al.,2003)—calculation of full costs(including both productionand external costs)ofelectricity productionscenariosQualitative assessmentmatrices—scoring of impactsignificanceNational Development Plan for the Energy Sector until 2030,with outlook to2050 (Energy Plan2030+), including SEA Scheduled forapproval in thefirst half of2015AirViro(SMHI,2010)—modelling of air emissionsSimaPro(PRé,2013)—modelling of impacts onhealth,ecosystems,resourceuse and climate changeIndicator sets(de Ridder et al.,2007)—measuring changes inimpactTable2Main stages of knowledge exchange process between SEA experts,modellers and policydevelopers,on the example of national energy plan and application of LEAP.Stage of policyplanningStage of SEA Stage of modellingInitiation Scoping Deciding on modellingobjectives and scopeDevelopment of draftpolicy documentImpact assessment Model design,quantitativemodelling andcommunication of resultsMid-term and/or expost evaluation ofpolicy documentMid-term and/or ex postevaluation of impacts(SEA follow-up)Ex post modelling57 P.Kuldna et al./Environmental Impact Assessment Review54(2015)55–60and third authors of the paper(who were respectively the main SEA expert and the LEAP modeller of the ex ante SEA),and on the ex ante SEA report(SEI Tallinn,2009).2)Ex post scenario modelling and preparation for the SEA of EnergyPlan2030+.Ex post modelling was the re-assessment of LEAP pro-jections made for Energy Plan2020in order tofind out to what ex-tent LEAP model results had realised and to identify possible reasons for deviation.It was guided by research and initiated by the authors without a formal government mandate,thus not being part of the of-ficial SEA process.Although in the ex ante SEA report for Energy Plan 2020,ex post modelling was not directly proposed as a form of mon-itoring,the updating of the national energy plan was known to be started soon,and thus it was an appropriate time to follow up on the SEA modelling outcomes.The additional aim of ex post modelling was to interact with the policy developers for continuing the use of LEAP in updating the energy plan and associated SEA(collaborating strategy of knowledge brokerage).The policy developers,in turn, were interested in identifying the most suitable model(s)for conducting an impact assessment of the new national energy plan.In order to discuss the ex post modellingfindings and offer opportu-nity for policy learning,a workshop was arranged by the researchers.Besides collaborating strategy,we did not elaborate any other knowledge brokerage strategy in advance to be employed in the process,but according to Michaels's(2009)framework,it could be anticipated that elements of other strategies can also be adopted.Michaels(2009)suggests for fundamental decision-making which involves opportunities to rethink how to approach a policy domain, collaborating as the primary brokering strategy,and for addressing moderately structured problems capacity building strategy.Develop-ment of the national energy plan can be characterised as belonging to both of these problem and decision-making types.4.ResultsWe identifiedfive primary strategies and their combinations which we used for brokering modelling knowledge between the researchers and policy developers in the development of Energy Plan2020and in the preparation of Energy Plan2030+(Table3).In Energy Plan2020, these strategies were identified upon reflection after the SEA process. In ex post modelling of Energy Plan2020and in the preparation of Ener-gy Plan2030+one strategy was chosen beforehand and others speci-fied in the course of action.The most frequently used knowledge brokerage strategies with re-gard to modelling in the national energy planning process were consult-ing and collaborating.Both approaches were applied formally(as an expert work in the SEA,in the working groups and Advisory Board of Energy Plan2030+)as well as informally(initiated by the researchers). Consultations were also part of the formal SEA process.The researchers brought in knowledge about modelling and environmental impact of power and heat generation while the policy developers contributed with knowledge on available and future energy production capacities and technology.High-involvement strategies,collaborating and engag-ing,were adopted by the researchers when an agreement with the pol-icy developers was sought on issues such as modelling objectives or model design and scenarios.For example,in order to discuss and agree on the ex ante and ex post modelling objectives,face-to-face com-munication and meetings with the policy developers were organised. Engaging strategy refers to the involvement of policy developers in the ex post modelling workshop,which was set up and moderated by the re-searchers.At this workshop,the researchers presented the ex post modelling results which demonstrated higher actual CO2and SO2emis-sions levels than those modelled in the ex ante SEA.The reasons for this were mainly related with larger scale electricity export than expected and with the postponement of the shutdown of old oil-shale production units.After the presentation,the policy developers were given an op-portunity to interpret the modellingfindings and discuss how the out-come can be used for the national energy plan renewal.The policy developers also described their overall expectations towards modelling of energy sector impacts,such as integration of regional energy market trends and impacts into the model and analysing the effects of policy measures on energy consumption.Informing strategy was appropriate to apply at the end of ex ante SEA process,when major changes to the modelling were no longer neces-sary.However,communication with the policy developers at this stage was still two-way and consequently yielded information for both sides,in particular with regard to interpretation of modelling re-sults and uncertainties.Capacity building was our main strategy in learn-ing from ex post modelling,since in the national energy policy such a re-assessment of emissions scenarios with the same model and comparing the forecast with present situation was carried out for thefirst time. Consulting and collaborating strategies were again employed when the researchers were invited to participate in Advisory Board and thematic working groups for Energy Plan2030+,during which the potential util-ity of LEAP model was discussed several times,initiated both by the re-searchers and the policy developers.Matchmaking was the only strategyTable3Knowledge brokerage(KB)strategies,examples and outcomes in relation to LEAP during the national energy plan process.National energy plan Primary KBstrategies inrelation toLEAPExamples of application of KB strategies SEA cycle stage OutcomesEnergy Plan 2020ConsultingandcollaboratingInteractions with policy developers for settingmodelling objectives and building scenarios in LEAPScoping and impactassessmentAgreement to project CO2and SO2emissions in electricity andheat production scenariosInforming Visualisation of emissions scenarios at workingmeetings and public energy forums.Impact assessment Approval of modelling results and the SEA reportDescribing modelling process,assumptions,uncertainties and outcomes in the SEA report.Ex post modelling Collaborating Interaction with policy developers for clarifyinginformation needs and setting ex post modellingobjectivesEx post(follow-up)evaluation of thenational energy planAgreement to extend the modelling time period from2030until2040and to use revised statistical data on CO2and SO2emissions for the past time periodEngaging Workshop with policy developers to discuss modellingoutcomes of electricity scenarios and modelling needsfor the new energy planAgreement to conduct ex post modelling also for heat scenariosCapacitybuildingLearning from ex post scenario modelling Joint understanding of external factors and uncertainties thatare difficult to predict in long-term emission scenariosEnergy Plan 2030+ConsultingandcollaboratingDiscussions in working groups and Advisory Board onmodelling needs and LEAP possibilitiesScoping Intention to use LEAP for the analyses of scenarios;however,the intention did not realise58P.Kuldna et al./Environmental Impact Assessment Review54(2015)55–60。