lable at ScienceDirect

Journal of Cleaner Production xxx (2014) 1e15

Contents lists avai

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

The dynamics of the Swedish biofuel system toward a vehicle fleetindependent of fossil fuels

Alessandro Sanches-Pereira*, Maria F. GómezDepartment of Energy Technology, KTH Royal Institute of Technology, Brinellvägen 68, 100 44 Stockholm, Sweden

a r t i c l e i n f o

Article history:Received 8 August 2013Received in revised form7 March 2014Accepted 8 March 2014Available online xxx

Keywords:BiofuelsBioenergy systemsSupply chain dynamicsTransport sectorSweden

* Corresponding author.E-mail addresses: perei@kth.se, lelle9@me.com (A

http://dx.doi.org/10.1016/j.jclepro.2014.03.0190959-6526/� 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Sancheindependent of fossil fuels, Journal of Clean

a b s t r a c t

The objective of this study is to present an analytical framework monitoring the development of theSwedish biofuels system, to evaluate its impacts on the achievement of 10% of renewable fuels by 2020,and to identify development patterns in order to establish a vehicle fleet independent of fossil fuels by2030. The methodological approach relies on systems thinking approach and uses causal loops diagramas a guide to our analysis. The results show that policymakers have to decide among different pathwaysof growth based on internal- and external-resources of the Swedish biofuel system as well as to applyeither supply or demand pressures onto the system in order to accomplish a vehicle fleet independent offossil fuels by 2030.

� 2014 Elsevier Ltd. All rights reserved.

1. Introduction

The development of renewable transport fuels in Sweden hasbeen on the country’s agenda since the first international oil priceshock in 1973, which has helped to establish the building blocks forthe growing renewable fuels in the country, especially biofuels(Hillman and Sandén, 2008; Ulmanen et al., 2009; McCormick et al.,2012). In 2012, the final energy use in the Swedish domestictransport sector amounted to around 92 TWh, of which renewablefuels comprised 8.52 TWh. The share of renewables in the domestictransport sector has more than tripled in the last decade, from 2.5%in 2000 to approximately 9% in 2012. In a business as usual scenario(BAU), Sweden may exceed its 2020 target of 10% renewable fuelsby 2%, with renewables accounting to approximately 12 TWh(Swedish Energy Agency, 2009, 2011; 2012a, 2012b, 2012c, 2013a;Newman, 2010; Svenska Bioenergiföreningen, 2012, 2013). Yet,the Swedish government’s vision is more ambitious and thecountry is determined to have a vehicle fleet independent of fossilfuels by 2030, which means decreasing from today’s share value of91% fossil fuels to less than 50% in the final energy use for domestictransport in the country by 2030 (Government of Sweden, 2010).This study aims to present an analytical framework to: (i) monitorthe development of the Swedish biofuels system, (ii) evaluate its

. Sanches-Pereira).

s-Pereira, A., Gómez, M.F., Ter Production (2014), http://d

impacts on the achievement of the mandate of 10% renewable fuelsby 2020, and (iii) identify development patterns so as to establish avehicle fleet independent of fossil fuels by 2030.

The study relies on secondary data collected from governmentalinstitutions and statistical databases. Academic publications, re-ports from important stakeholders, and personal communicationsare used to complement and validate processed information, whichis fed into the Consideo Modeler software platform to design causalloops. Through this, we intend to match the potential energy use inthe Swedish domestic transport sector with the country’s desiredtargets. Our methodological steps include (i) identification of thedevelopment trend in the current Swedish biofuel system usingforecast analysis against chosen scenarios, (ii) calibration of targetscenarios using cross-analysis between forecasting and backcastinganalysis, (iii) description of the Swedish biofuel system using acause and effect model, and (iv) identification of reinforcing orbalancing effects within the Swedish Biofuel system.

2. The Swedish biofuel system

Currently, biodiesel is the frontrunner among the renewableenergy carriers used in the Swedish transport sector. In 2012, itscontribution to the renewable energy mix used in domestictransport is approximately 44%, closely followed by bioethanolwith around 28%, renewable electricity (i.e., generated by biomass-,hydro- and wind-power) with 18%, and upgraded biogas with near10% (Swedish Energy Agency, 2009, 2011; 2012a, 2012b, 2012c,

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Table 1Potential scenarios of energy consumption for the Swedish domestic transportsector in 2030.

Energy carrier BAU Efficient (EF) Fuel-switching(FS)

Fossil fuels 97.4 TWh 90% 21.9 TWh 40% 21.9 TWh 31%Biofuels 7.9 TWh 7% 25.8 TWh 47% 39.7 TWh 56%Electricity 3.2 TWh 3% 7.0 TWh 13% 9.1 TWh 13%Total 108.5 TWh 100% 54.7 TWh 100% 70.7 TWh 100%

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e152

2013a, 2013b; Svenska Bioenergiföreningen, 2012, 2013). The totaluse of renewable fuels has increased from 2 TWh in 2000 to morethan 8 TWh in 2012. However, the development trends vary for thedifferent energy carriers reflecting changing directions in Swedishpolicies. Understanding the allocation of renewable energy carrierscan provide an outlook of their role in meeting the mandate of 10%renewable fuels in the Swedish domestic transport sector by 2020.One interesting fact observed in the current contribution ofrenewable fuels is that renewable electricity has maintained itsamount of energy provided rather stable (e.g., oscillating around1.5 TWh) despite losing its position as the key renewable energycarrier during the period between 2000 and 2012. In fact, renew-able electricity provides slightly less energy today than it did in2000. Nevertheless, electricity is expected to gradually regain itsposition as an important energy carrier in the long run (e.g., 2050),particularly in passenger cars (Olsson and Hjalmarsson, 2012).

According to Sköldberg et al. (2010) the amount of electricityneeded by 2030 to supply a growing fleet of electric vehicles (EVs)would increase from today’s 3 TWh up to 9 TWh (Sköldberg et al.,2010). Also, the expansion of EVs pushes the development of con-ventional vehicle toward more efficient engines with lower emis-sions. This creates a competition where everyone would like tooffer an attractive product. For example, “green cars” in Swedencurrently have emissions bellow 120 g of CO2 per kilometer andthey are expected to get even lower (e.g., “super-green cars” haveemissions bellow 50 g). Assuming a BAU scenario e in which thefinal energy use in the Swedish domestic transport sector wouldaccount for about 108 TWh by 2030 and that EVs are not going tostimulate electricity generation in the median run (e.g., before2030) e and supposing the share of renewable electricity is keptunder current values, the amount of renewable electricity by 2030would be around 1.6 TWh or 17% of the renewable energy used fordomestic transport in Sweden (Sköldberg et al., 2010). Hence, theremaining demand of 83% would need to be covered by biofuels,which corroborates their role as key energy carriers in achieving avehicle fleet independent of fossil fuels.

Under a more optimistic perspective, Sköldberg et al. (2010)present two new scenarios that involve substantial changes in theSwedish energy mix used in the domestic transport sector. The firstscenario, called Efficient (EF), assumes that vehicles in general aremore energy efficient by 2030. It also assumes that the amount offossil fuels decreases from 90% to 40% of the total energy use in thetransport sector in the same period. In this context, the amount ofenergy used in Sweden would be cut by almost 50% when it iscompared to the reference scenario BAU. This means that roughly30 TWh or 54% of energy use in Swedish domestic transport sectorwould be covered by renewable energy, in which biofuels stillremain the key energy carrier with about 88% share of the totalrenewable energy used in the country’s domestic transport. Thesecond scenario is based on a fuel switching (FS) approach, inwhichit is assumed that future vehicles keep the current efficiency. In thiscase, it is crucial increasing the shares of biofuels and electricity inorder to reach the same level of fossil fuel use as in the EF scenario.Again, biofuels play a significant role toward the vision of achievinga vehicle fleet independent of fossil fuels by 2030. In this scenario,biofuels would account to about 89% share of the total renewableenergy used in the domestic transport (Sköldberg et al., 2010).Table 1 summarizes these potential scenarios for 2030.

Each one of these potential scenarios ratifies biofuels asimportant energy carriers that can help Sweden to achieve a vehiclefleet independent of fossil fuels. It is important to notice that noneof these scenarios provides information about the contribution ofbioenergy carriers in the final supply. Therefore, there is no indi-cation on how diversified and well-distributed could be thecontribution of bioenergy carriers in the system in order to build up

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

security and reliability levels. It should be noted that in anincreasingly carbon-constrained world, we assume multi-carrierenergy systems based on renewables as more reliable energy sys-tems than fossil fuel based systems. In addition, its flexibility wouldincrease the likelihood of meeting the country’s desired target of avehicle fleet independent of fossil fuels.

2.1. Swedish biofuel system’s dynamics

In other to understand the contribution of bioenergy carriers,our study uses a system thinking approach. System dynamicsthinking began in the mid-1950s as a tool to improve under-standing of complex industrial processes. This approach utilizesdrivers within the system, which are represented as informationefeedback processes, to model a feedback structure closely tied tosystem performance. Since then, it has been applied to solve a widerange of problems, from industrial and organizational to policy-oriented issues (Sterman, 2000; Bush et al., 2008).

Bioenergy systems present a higher complexity than fossil fuelsystems since they have a larger stakeholder networks and industryassociations (Börjesson et al., 2009). Bioenergy systems can be seenas a classic system dynamics case and their problems tend to sharethe following four important characteristics (Bush et al., 2008):

� They are dynamic, which implicates change over time.� They involve multiple stakeholders with diverse interests,ranging from farmers, entrepreneurs, and consumers to poli-cymakers, whose interactions can impact the system’s overallperformance.

� They present interdependencies with different systems andprocesses, which are critical to their overall performance (e.g.,competition among food, fiber and fuel production).

� It can be challenging to identify communication issues, struc-tural components, and leverage points.

Therefore, using a system dynamics model provides useful in-formation about the future development of the Swedish biofuelsystem. It can help to identify underlying structural componentsand its interactions, which could be used as leverage points to steerand control the system’s development. Fig. 1 illustrates the devel-opment of bioenergy contribution per type of biofuel to theSwedish domestic transport in the last decade, which is the resultof the involvement of both the national and local levels in breakingdependence on oil (Swedish Energy Agency, 2009, 2011, 2012a,2012b, 2012c, 2013a, 2013b; Svenska Bioenergiföreningen, 2012,2013; McCormick et al., 2012). Bars indicate the amount in TWhof renewable energy used in the domestic transport sector from2000 to 2012. Lines represent the biofuel share in percentage or itscontribution to the energy supplied in the same period.

Fig. 1 shows that the increment of renewable energy between2000 and 2012 relies entirely on the bioenergy contribution. It alsoshows that bioethanol was the key bioenergy carrier in 2000 with60% contribution in the total bioenergy used (e.g., 0.26 TWh) by theSwedish vehicle fleet in that year. Yet, the behavior of the system

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Fig. 1. Development of bioenergy contribution per type of biofuel to the Swedish domestic transport.

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 3

changes from 2005 forward, which led biodiesel to become the keybioenergy carrier with 54% of contribution as compared to 34%from bioethanol and 12% from upgraded biogas in 2012 (SwedishEnergy Agency, 2009, 2011, 2012a, 2012b, 2012c, 2013a, 2013b;Svenska Bioenergiföreningen, 2012, 2013). The figure also sug-gests that bioethanol contribution could decrease further if thecurrent behavior continues. Another interesting detail observed inthe figure is the symmetric shape between the bioethanol andbiodiesel data, which indicates a strong relationship between theirsystemic behaviors. One way to validate this assumption is toidentify the very same pattern in the development of the Swedishfleet and to confirm whether what is happening is a fuel substitu-tion process or general efficiency improvements in vehicles.

2.2. Composition development of the Swedish vehicle fleet

Nowadays, the Swedish vehicle fleet accounts to more than 7million vehicles in use, inwhich the composition can be assumed as62.78% personal cars, 7.86% heavy-duty vehicles (e.g., trucks), 0.20%buses, and 29.19% motorcycles, mopeds e a lightweight motorizedbicycle that can be pedaled as well as driven by a low-poweredgasoline engine e and other vehicle categories (i.e., tractors,trailers, snowmobiles, etc.). This study focused only on personalcars, heavy-duty vehicles, and buses that are fueled by gasoline,diesel, bioethanol, and gas fuels (e.g., admixture of natural gas andupgraded biogas). They represent around 71% of the vehicles in usein Sweden (TransportNäringen i Samverkan, 2010; Trafikanalys,2012, 2013). Other fuel categories such as electric and hybrid ve-hicles are not included in the analysis because of their smallnumbers and low impact on the relationship between the bio-ethanol and biodiesel fuels during the last twelve years.

Data sets related to the Swedish fleet composition revealedsimilar symmetric behavior occurring between the numbers ofgasoline and diesel vehicles on the data sets related to personalcars and heavy-duty vehicles. However, the same behavior doesnot occur on data connected to buses despite the fact that the

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

numbers of vehicle running on gasoline persist on waning(TransportNäringen i Samverkan, 2010; Trafikanalys, 2012, 2013).In this particular case, the application of gas fuel is increasingrapidly and pushing the country’s production of biogas (SwedishEnergy Agency, 2012d).

Our analysis shows that gasoline vehicle numbers aredecreasing and being replaced; validating that what is happening isa fuel substitution process in the fleet. This is a result of many policyinstruments based on low-carbon emissions, which have beenaffecting the development of the transport sector in Sweden duringthe last few years (Swedish Energy Agency, 2011). Important in-formation is that dependency on diesel is increasing. In fact, heavy-duty vehicles already are highly dependent on it. Hence, a shortageof diesel would impact Sweden’s capability of bioenergy generationsince several components of the bioenergy system are highlydependent on road transport and heavy-duty vehicles, which areresponsible for machinery operation in the forestry sector andtransport of raw materials from forests to fuel factories and ofbiofuels to heating plants (Swedish Energy Agency, 2011). In thiscontext, being capable of using low-admixture of biofuels withoutrequiring technical adaptations in vehicles is an important factornot only for maximizing biofuel penetration in the market but alsofor guaranteeing energy security, reducing fossil fuel dependencyby whatever means possible, and meeting sooner the target of 10%of renewable fuels in the domestic transport sector.

Without a doubt the composition development of the Swedishfleet has a direct influence on how biofuels are being consumed. Forexample, the decrease of gasoline vehicles in the fleet compositionsince 2005 has redirected the bioethanol delivering pathways. Forthe last ten years its 5% blended formwas the common outlet but ithas lately switched mainly to E85, which is a biofuel with 85%denatured bioethanol used by the flex-fuel vehicles in the country.

E85’s role has fast increased since 2005 as a result of the Na-tional Climate Policy and the government’s commitment to elimi-nate fossil fuel dependency by 2030 (Government of Sweden(2010); European Union, 2003). Another important component is

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Table 2Adjusted potential scenarios of energy consumption for the Swedish domestictransport sector in 2030 and bioenergy contribution per type of biofuel.

Energy carrier BAU Efficient (EF) Fuel-switching(FS)

Bioenergy 20.2 TWh 19% 25.8 TWh 47% 39.7 TWh 56%Bioethanol 7.4 TWh 7% 9.4 TWh 17% 14.4 TWh 20%Upgraded biogas 3.9 TWh 4% 5.1 TWh 9% 7.8 TWh 11%Biodiesel 8.9 TWh 8% 11.3 TWh 21% 17.5 TWh 25%

Fossil fuels 85.1 TWh 78% 21.9 TWh 40% 21.9 TWh 31%Electricity 3.2 TWh 3% 7.0 TWh 13% 9.1 TWh 13%Total 108.5 TWh 100% 54.7 TWh 100% 70.7 TWh 100%

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e154

the fact that the national association for the automobile industrygave its support to the initiative. However, the trend abruptlychanged in 2009. A glitch in the behavior e a sudden contributiondecrease e shows that consumers reacted rapidly to changes inpolicy and price fluctuations on fuels (Swedish Energy Agency,2009, 2012b; 2013b). In July 2009, the government removed thepremium given for clean vehicles purchasing that lead to a sharpdecline in vehicle sales in that year (Swedish Energy Agency, 2011).Also, biofuel sales are highly dependent on the relative price offossil fuels. For example, bioethanol consumption in Sweden isattractive until it costs up to around 74% of the gasoline price perliter (Pacini and Silveira, 2011). In addition, 2009 was a peculiaryear for bioethanol consumption. In this particular year, the bio-ethanol average cost of 1 L amounted to around 80% of the gasolineprice per liter (Svenska Petroleum och Biodrivmedel Institutet,2013). As a result, it abruptly reduced the E85’s attractiveness.However, the systemic behavior has been stabilized and its currentcondition follows the trend prior to the glitch leading E85 tobecome the main bioethanol outlet by 2011 (Swedish EnergyAgency, 2009, 2012b; 2013b).

In the case of biodiesel, the development of the fleet compo-sition has influenced a fast growing consumption of the 5% blendof biodiesel, which corroborates not only the fuel substitutionprocess in the fleet but also the fact that the Swedish dependencyon diesel is increasing (Swedish Energy Agency, 2009, 2012b;2013b). It is important to address that efficiency improvementsin vehicles are important, especially on reducing emissions, butthey have not yet been influential in the consumption trend ofbiofuels.

Currently, the admixture of biofuels in Sweden is still 5% pervolume of fuel but the Government allows blending of bioethanolin gasoline up to 10% and up to 7% of FAME-based biodiesel and upto 15% of HVO-based biodiesel in diesel since May 2011 (SwedishEnergy Agency, 2013b; Government of Sweden, 2011;Transportstyrelsens, 2012; European Union, 2009). The nationalassociation for the automobile industry has developed a list of carmodels that can run on 10% bioethanol (E10) (BIL Sweden, 2011).For diesel vehicles there is no need for a list containingmanufacturerecommendations since the technical admixture limit allows up to20% biodiesel blending before considering modification to dieselengines.

Despite low numbers in the fleet composition, gas fuel vehiclesalongside flex-fuel vehicles have gained momentum since 2005. Infact, they present the highest growth rate in the last decade whencompared with gasoline and diesel vehicles. In Sweden, gas fuelvehicles run with an admixture ratio of 40% natural gas and 60%upgraded biogas since 2009 forward (Swedish Energy Agency,2009, 2012b; 2013b; Sanches-Pereira and Gómez, 2013).

Besides meeting policy targets, other important factors havecontributed in fostering upgraded biogas as vehicle fuel. Itsmanageable integration into natural gas networks already in placeand the fact that upgraded biogas is recognized as a reliable energysource provide a strong incentive especially in the south andsouthwest regions of the country (Baltic BiogasBus, 2012). Around50% of the produced biogas in 2011 was upgraded and used asvehicle fuel (Swedish Energy Agency, 2012d). In Sweden, the biogasused as transport fuel increased from 0.28 TWh in 2007 to0.81 TWh in 2012 (Swedish Energy Agency, 2009, 2012b, 2013b).The usage has more than doubled in a four-year period. Although2007 established biogas as the leading fuel by the Swedish gasvehicle fleet according to the volume used, the leading fuel in termsof energy provided was still natural gas by a lowmargin of less than1% over biogas. It is the year 2008 that sets the turning point. In thatyear, upgraded biogas provided around 60% of the energy used bythe Swedish gas fleet, holding since then the leading position in

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

terms of both volume and energy provided (Swedish EnergyAgency, 2012b).

2.3. Development trends of the Swedish biofuel system

The correlation between the information from the bioenergysystems dynamics and from the fleet composition developmentwithin the Swedish domestic transport sector made clear thatbiofuels influence significantly the country’s ability to reach itsdesired targets. Our analysis shows that the continuous growth ofbiofuel use in transport ratifies the effectiveness of policies in place,which are fostering the fuel substitution process rather than gen-eral efficiency improvements in vehicles. Important driving forces,such as the manageable integration of biofuels into distributionnetworks already in place (e.g., blended fuels) and the fact thatbiofuels are recognized as a reliable energy source, provided astrong incentive to this very growth. However, current forecastscenarios do not present how diversified and well-distributedbiofuels would be by 2020 and 2030 and how reliable theirdevelopment trends are (Sköldberg et al., 2010). Aiming atanswering that, target scenarios were reexamined and adjustedusing the results from the correlation so as to match the currentdevelopment trend of the Swedish biofuel systems. The adjustedset of data estimates that by the year 2020 the final energy use inthe Swedish domestic transport sector would account for around101 TWh, from which biofuels would represent approximately14 TWh. In this context, biofuel contributionwill already exceed theSwedish target of 10% share of renewables in a BAU scenario. Notethat the target would be met without even considering therenewable electricity input that would raise further this share andwithout double counting the bioenergy produced from waste,residues, non-food cellulosic material, and lignocellulosic material(European Commission, 2012). It is also important to mention thatadjusted scenarios do not take into account future impacts of thecap regarding the amount of biofuel derived from food crops at 5.5%of renewable transport fuel target in the European Union, whichwas recently agreed at the European Parliamentary EnvironmentCommittee (ENVI) on July 2012.

In this context, the adjusted set of data validated the very strongrelationship between the systemic behaviors of bioethanol andbiodiesel. However, it contradicted the trendline showing thatbioethanol contribution would diminish in the future. In fact, thedevelopment trend shows that bioethanol would keep on playingan important role and its contribution to the Swedish biofuel sys-tem would account to roughly 33%, against biodiesel with around54% and upgraded biogas with near 13% in the year 2020. As aresult, for the year 2030, the bioenergy contribution varies from19% e instead of the pessimistic 7% assumed by Sköldberg et al.(2010) e to a highly optimistic 56% share of the total forecastedenergy use in the Swedish transport sector. Table 2 presents theadjusted potential scenarios in 2030 and the contribution per typeof biofuel. The table also shows that the energy system’s reliability

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 5

to meet the desired targets would increase because of the diver-sified biofuel options resulting frommore even distribution of theirshares.

Fig. 2 illustrates the development trends, in which lines corre-spond to the amount of energy provided in TWh per type of biofueland displayed in the right y-axis. Bars depict the energy use in theSwedish domestic transport sector and their values are displayed inthe left y-axis. Highlighted bars illustrate the development trendsand their respective adjusted values for the year 2020 and 2030scenarios.

Note that this forecast was produced using best-fit curved linebased on polynomial regression for each one of the parameters. Thecoefficients of determination (R2) ranged from 0.94386 to 0.99789,which indicates a good fit of trendline to the estimated data. Inaddition, the results are based on current admixture level of bio-fuels in Sweden, which is 5% per volume of fuel. Hence, increasingthe blending would have a directly proportional impact on thedynamics of the Swedish biofuel system. It would most likely raisebioenergy contribution’s values and, at the same time, change thebioenergy carriers’ diversification and contribution. For example,biodiesel’s contributionwould increase faster since its blending hasthe advantage of allowing up to 20% biodiesel before consideringmodification to the diesel engine and thus can increase its sharefaster than other biofuels e especially in a situation in whichblending can be assumed as an important strategy that can enableSweden meeting its targets sooner.

At first sight, the biofuel system’s dynamics and its developmenttrends reflected the effectiveness of Swedish policies to fostergrowth of bioenergy use in transport. Our estimative shows agrowing, diversified, and well-distributed energy system. However,it is important to point out that a larger use of bioenergy dependson resource availability and establishment of effective productionchains so as to substantiate building up security and reliabilitylevels within the system itself. Thus, the interaction among struc-tural components of the system and their causal links are key ele-ments to understand whether there is a gap between desiredtargets and the current development of biofuel use in the transportsector or not.

Fig. 2. Development trends of t

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

3. Causal loops within the Swedish biofuel system

In order to better understand the interactions among structuralcomponents of the Swedish biofuel system and their behaviorpatterns, we utilized in this study Causal Loop Diagram (CLD) as amodel for representing cause and effect relationships within thesystem. The diagram is an analytical tool commonly used in systemdynamics to increase understanding and facilitate the analysis ofcomplex systems enabling the identification of potential actions tomodify the system in order to achieve desired goals.

One advantage of the CLD is that it illustrates interactions thatare difficult to be verbally described, such as circular chains of causeand effect (Kirkwood, 1989). Despite the fact that CLDs cannot beused as a conceptualizing tool since they lack the organizing powerrequired to represent decision-making processes, they can be usedas an analytical tool to provide an overview of a given system’sdevelopment (Morecroft, 1982). Hence, CLD is an important tool: (i)to capture the structure of a system, (ii) to provide an overview ofloop configurations so as to analyze systemic development, and (iii)to highlight important feedback loops that have the potential tosolve an identified problem (Binder et al., 2004). In our case, thediagram helps us to visualize and analyze interactions in order tounderstand whether there is a gap between desired targets and thecurrent development of biofuel use in the transport sector or not. Infact, causal loops are used not only to indicate relationships thatexist among components but also to indicate their influences oneach other by establishing reinforcing or balancing effects(Sterman, 2000; Kirkwood, 1989).

A CLD is formed by structural components and arrows thatindicate causal links or direct connections between components. Apositive sign indicates a reinforcing effect. For example, a change inthe structural component “A” produces a change to component “B”in the same direction. As a result, an incremental change happeningin “A” creates an incremental change in “B”. In contrast, a negativesign shows a balancing effect, in which a change in “A” produces achange in “B” in the opposite direction. In this case, an incrementalchange in “A” generates a decline in “B”. Note that arrows areunidirectional and they imply direction of causation and not a time

he Swedish biofuel system.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e156

sequence, in which a causal link from structural component “A” tocomponent “B” simply means that when “A” increases, so does “B”instead of something needs to happen first on “A” before occurringin “B” (Kirkwood, 1989).

The next analytical step is to define the causal links, which in-cludes reinforcing or balancing decisions. Meaning, it is necessaryto classify the connections as positive (þ) or negative (�). There is adifference between descriptive modeling that just visualizesalready known dynamics and explorative modeling that revealsfeedback loops. The chosen modeling platform e Consideo Modelere allows two ways of weighting these connections in order toexplore unknown influences onto structural components. The firstis absolute qualitative weighting that uses percentage of themaximum influence that one component has onto another. Thesecond is relative qualitative weighting by comparison of in-fluences between components, which uses the default strengthlevels in the software platform such asWeak (10 points),Middle (17points), and Strong (25 points) influences. In this case, the thicknessof the links indicates the degree of influence between two com-ponents. Note that neither of the two weighting methods allowstructural component to exceed the maximum received influenceof 100% or 100 points. Despite the fact that a single componentcannot exceed the overall maximum level of received influence, itcan influence more than one structural component at maximumlevel. During the design phase of the CLD, this study utilized theabsolute qualitative weighting approach to verify and validate thesystem’s development patterns and structural components beforerunning the model to analyze the 2030 scenarios.

In addition to measuring the intensity of the influence, it isessential to weight its delay. The delay is expressed in terms of timeneeded by a structural component to respond or adjust to anincoming influence and is visualized in the diagram as crossinglines on the arrows. No crossing line illustrates Short Term, one lineexemplifiesMedium Term and Long Term is pictured as two crossinglines. For example, let us assume that structural components “A”,“B”, and “C” are respectively Fleet Size, Fuel Demand, and BiofuelDemand. The cause and effect chain among them means that anincrease in the size of the Swedish fleet would increase in themedium run the country’s fuel demand, which will in a shortperiod of time directly influence the biofuel demand. The delaybetween Fleet Size and Fuel Demand is based on the reasoning thatan increment or decline on the fleet size e relation between newregistrations and decommissioning of old cars e gradually changesthe fuel demand. It is also a reinforcing influence, whichmeans thatdecreasing the fleet size will reduce fuel demand and vice-versa.Once the fuel demand has increased or decreased, the biofuel de-mand quickly responds to it because biofuels are consumed in bothblended and pure forms.

3.1. Causal loop diagram of the Swedish biofuel system

“Essentially, all models are wrong, but some are useful” (Box andDraper, 1987). In reality, no model can fully capture thecomplexity of a system, including all causal links and their directand indirect effects. For that reason, Min and Zhou (2002) suggestthat, in order to reduce the uncertainties between reality andmodel, boundaries need to be clearly defined. By doing so, it isimportant to have a clear vision of the system’s structural compo-nents. They have to replicate the identified behavior along thenetwork in such a way that it is as close to reality as possible,reflecting key characteristics and dimensions of a system in an“easy-to-understand” fashion (Min and Zhou, 2002). The higher thesystem’s complexity, the greater the number of structural compo-nents required for exemplifying it and vice-versa. This is illustratedby the twenty structural components in the CLD of the Swedish

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

biofuel system, which generalize specific events associated withbiofuel demand and, at the same time, take into considerationpatterns of development that characterize specifically the Swedishbiofuel system.

Our CLD was based on the bioenergy market model proposed byJeffers et al. (2013) and on the bioenergy systems diagram proposedby Sanches-Pereira (2012), which were adapted to the Swedishcontext. The idea behind was to interchange between supply ca-pacity and perceived demand in order to identify structural com-ponents that can create desirable outcomes within the Swedishbiofuel system (Jeffers et al., 2013; Sanches-Pereira, 2012). Notethat we used the adjusted data set related to energy use in theSwedish domestic transport sector in order to present potentialdevelopment patterns of the Swedish biofuel system instead ofproviding a precise value. Therefore, there is an exchange of pre-cision in return for tracking ability of development patterns withinthe system to manage its complexity. Fig. 3 illustrates our CLD andpresents the selected structural components and their causal linkswithin the Swedish biofuel system.

The following analytical step was to identify causal link changesover time and to understand how structural components behave inresponse to these changes. By identifying these componentchanges, one has the possibility of eliminating a problem orimproving the system’s performance or quality (Kirkwood, 1989;Sterman, 2000). Note that removing structural components mayincrease system’s efficiency and adding structural components maywell improve its quality.

With causal link changes over time defined, the study canidentify feedback loops, which can be described as a closed path orsequence of cause and effects. The aim of this step is to betterunderstand the system’s development. While structural compo-nents are the building blocks of a system, the feedback loops showhow a component indirectly influences itself and the systemthrough other components in the loop (Richardson & Pugh III, 1981;Kirkwood, 1989). There are two types of feedback loops. A positive(þ) or reinforcing feedback loop can be illustrated as a “snow-balling” effect, which can lead to system’s growth or decline at anever-increasing rate. However, reinforcing loops can be misleading.They can be perceived as a slow process during its early stages ofdevelopment, but then this process speeds up exponentially. As aresult, it can lead to late reactions and ineffective decision-making.A negative (�) or balancing feedback loop provides stability to asystem or regulate its dynamics in order to meet a set goal. It isimportant to mention that balancing loops can be very powerful,which can give to the system a quite rigid structure that is averse tochanges (Kirkwood, 1989).

Our CLD presents six feedback loops within its structure, threereinforcing feedback loops and the remaining balancing feedbackloops. Table 3 presents the type of feedback loops and theircorrespondent structural components.

In short, the reasoning behind feedback loops behavior is asimple cause and effect relationship. For example, the first feedbackloop is formed by a bidirectional relationship between two struc-tural components, Biofuel Production versus Biofuel Imports. Thisstraightforward relationship means that, on the one hand,increasing the Swedish biofuel production would reduce biofuelimports; therefore, it reinforces in return the development of thelocal biofuel production. On the other hand, increasing biofuelimports would cause the opposite effect by reinforcing furtherimports through reducing the contribution of local biofuels in theSwedish system. The remaining feedback loops increase theirnumber of structural components, thus their complexity. As aresult, the largest feedback loop is formed by nine structuralcomponents: Biofuel Production, Biofuel Imports, Biofuel Inventory,Biofuel Price, Land Transfer to Biofuel Crop, Biofuel Crop Land Area,

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Fig. 3. Causal loop diagram of the Swedish biofuel system.

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 7

Biofuel Crop Yield, Local Raw Material and Raw Material Imports. Therelationship among them is unidirectional. That means thatincreasing biofuel production would reduce biofuel imports butbiofuel inventory could still increase because of local supply. Thiswould reduce the biofuel price and influence biofuel crop yieldsand local supply of raw material. Consequently, it would increaseraw material imports to balance local biofuel production andmaintain the country’s biofuel inventory.

Reinforcing loops are important because they dictate the sys-tem’s growth patterns. As a result, understanding them give usinsight on how the Swedish biofuel system would cover thecountry’s demand for fuel andmeet the desired targets. In this case,the reinforcing feedback loops illustrate two distinctive andpolarized growth patterns. The first pattern is an internal-basedgrowth pattern that would greatly rely on exploring further theSwedish local resource potential. This pattern increases localdevelopment, raises local raw material and biofuel reliability andquality, and improves the country’s energy security and indepen-dency. Yet, it also increases land transfer to biofuel crops, which canreduce agricultural and forest areas, and the local resource poten-tial becomes a limiting factor for further development of the sys-tem. The second pattern is an external-based growth pattern thatheavily relies on raw material imports to increase local productionand/or biofuel imports to boost the national biofuel inventory. Inthis case, the limiting factor would be based on global markets andon commodities flows, which in recent years were characterized byshortages, increased volatility, and large price swings (SwedishEnergy Agency, 2012e). In contrast, balancing loops regulate the

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

system’s growth patterns or harmonize them in order to meet a setgoal. Understanding both types of feedback forces gives us anoverview of which structural components within the Swedishbiofuel system would strongly influence its development (e.g.,growth patterns) toward a set goal, as illustrated in next sections.

3.2. Supply and demand pressures

Supply pressure is based on all direct and indirect impacts suf-fered by the structural component Biofuel Production. Conversely,the demand pressure relies on biofuel inventory reinforcement tomeet fuel demands. This pressure is the sum of all direct and in-direct impacts onto the structural component Biofuel Inventory. Itshould be noted that all structural components are important andtheir influences (i.e., direct and indirect impact) on componentsBiofuel Production and Biofuel Inventory. Intuitively, the structuralcomponents Fleet Size and Biofuel Price are perceived as compo-nents that can steer the system’s development based on demandand supply influences. Yet, the correlation among the six feedbackloops, which is a counterintuitive analysis, points out threedifferent structural components e Biofuel Imports, Local Raw Ma-terial, and RawMaterial Imports e instead of structural componentsFleet Size and Biofuel Price. The analysis indicates that these threecomponents are drivers for the system’s oscillation or balance be-tween external and internal growth patterns since the interactionamong them is what defines a feedback loop as reinforcing orbalancing forces (see Section 3.1). An important aspect of identi-fying these three structural components is the fact that they

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Table 3Feedback loops within the Swedish biofuel system.

Feedback loops

Reinforcing a. Biofuel Production / (�) Biofuel Imports / (�) BiofuelProductionb. Biofuel Production / (þ) Biofuel Inventory / (�) BiofuelPrice / (þ) Land Transfer to Biofuel Crops / (þ) Biofuel CropLand Area / (þ) Biofuel Crop Yield / (þ) Local RawMaterial/ (�) RawMaterial Imports/ (þ) Biofuel Productionc. Biofuel Production / (�) Biofuel Imports / (þ) BiofuelInventory / (�) Biofuel Price / (þ) Land Transfer to BiofuelCrop / (þ) Biofuel Crop Land Area / (þ) Biofuel CropYield / (þ) Local Raw Material / (þ) Biofuel Production

Balancing a. Raw Material Imports / (�) Land Transfer to BiofuelCrops / (þ) Biofuel Crop Land Area / (þ)Biofuel CropYield / (þ) Local Raw Material / (�) Raw Material Importsb. Biofuel Production / (þ) Biofuel Inventory / (�) BiofuelPrice / (þ) Land Transfer to Biofuel Crops / (þ) Biofuel CropLand Area / (þ) Biofuel Crop Yield / (þ)Local RawMaterial / (þ) Biofuel Productionc. Biofuel Production / (�) Biofuel Imports / (þ) BiofuelInventory / (�) Biofuel Price / (þ) Land Transfer to BiofuelCrop / (þ) Biofuel Crop Land Area / (þ)Biofuel CropYield / (þ) Local Raw Material / (�) Raw MaterialImports / (þ) Biofuel Production

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e158

corroborate the relevance of including the sustainability criteria forbiofuel production and/or imports in the EU’s Renewable EnergyDirective as a mechanism to lead changes in the system. The sus-tainability criteria recognize the link between resource supplyalong with production and demand factors. The goal is to maintainthe system’s integrity and to avoid unreasonable burdens to localproducers and the environment (European Union, 2003). Hence, itis clear that meeting the desired targets needs a strategic combi-nation of both reinforcing and balancing forces in order to controlrelationships among structural components of the biofuel systems,to speed up local production, and to stabilize the country’s in-ventory so that diversification and security of supply are guaran-teed within the system.

In order to evaluate supply and demand influences onto po-tential scenarios, we use the software platform to generate insight-matrixes. Our chosen timeframe defines the year 2014 as short-term, the year 2020 as medium-term, and the year 2030 as long-term. The insight-matrix is a useful tool because it does not onlyplace the structural component’s strength of influencing the chosencomponent into a broader perspectivee either Biofuel Inventorywhen analyzing the demand pressure or Biofuel Production whenanalyzing the supply pressure e but it also compares its impactover time as a reinforcing or balancing influence within theSwedish biofuel system (Sanches-Pereira, 2012).

The insight-matrix illustrates the sum of influences or impactsonto a chosen component. The calculation of the impacts is fairlysimple. Let the CLD (N, g) indicates the Swedish biofuel system. Inwhich a set of nodes N ¼ {A,.,n} refers to structural componentsand a real-valued n � n matrix g, where gij represents causal linksbetween these components using values of 0 or 1. The idea is thattwo structural components are either connected or they are not. Ifgij¼ 1 then i is linked to j, otherwise gij¼ 0. The causal link indicatesa pathway between two structural components. For example, threestructural components e “A”, “B”, and “C” e are interconnectedthrough causal links weighting 100% each. Arbitrarily, 1 multipliesdirect impulses and 0.1 multiplies indirect impulses. Thus, thedirect influence of “A” onto “B” is 100% and the indirect influence of“A” onto “C” is 10%. Consequently, a causal loop comprising thesethree structural components implies an indirect influence of “A” toitself equal to 1%. Note that influences can be classified as rein-forcing (þ) and balancing (�). The sum of all direct and indirect

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

impacts that structural components would have onto selectedcomponent throughout causal chains is plotted in the x-axis of theinsight-matrix. As a result, the x-axis illustrates the relationshipstrength between the selected structural component and others.The further a component is located to the right, the stronger itsforce is on impacting the selected structural component. To the left,it is the opposite. Finally, the algorithms calculate the change of theimpact over time. For short-term impact the value of the influenceis divided by 4, for medium-term impact by 2, and for long-termimpact by 1. The notion is to understand the feedback loop influ-ence over time and the results are plotted in the y-axis of theinsight-matrix (Lorentz and Neumann, 2012; Neumann, 2013). Thatmeans that the orientation on the y-axis and its quadrants show theinvolvement of the structural component in feedback loops and itsinfluence as either reinforcing or balancing forces (Neumann,2013). Fig. 4 presents the insight-matrix configuration.

The analysis using the insight-matrix not only presents thestructural components’ influence in the scenarios but also theirbehavior under different pressures (i.e., either supply or demandpressures). Note that a structural component positioned in theinsight-matrix indicates only its potential impact in comparison toother components under a defined pressure and nothing else(Neumann, 2013). Selected structural components are indicated inthe matrix as: Biofuel Imports exemplified by the number 1, LocalRaw Material represented by the number 2, and Raw Material Im-ports characterized by the number 3.

4. Insight-matrix results and discussion

The scenarios are presented under a context of demand pressure,in which the network needs to reinforce the Swedish biofuel in-ventory for meeting the desired targets. Fig. 5 introduces theinsight-matrix based on the current development of biofuel use inthe transport sector and how the selected structural componentsaffect the Swedish biofuel system until 2014. The strongestcontributor to its growth is the component Biofuel Imports. Theother two components e Local Raw Material and Raw MaterialImports e overlap each other indicating that their contributioncorresponds to approximately the same pressure onto the system.As a result, the insight-matrix indicates that in response to a rapidincrease of demand for biofuels such as raising admixture values,the Swedish biofuel system in the short-run would hold on tobiofuel imports so as to expand the system.

This scenario closely emulates the current dynamics within theSwedish biofuel systems and shows that all structural componentsapply reinforcing pressure onto the system. Despite reaching themandate of 10% renewables in the domestic transport sector by2020, Sweden may rely almost entirely on biofuel imports. As amatter of fact, biofuel imports already play an important role inmeeting the mandate and they may continue to do so. Fig. 6 illus-trates the Swedish biofuels imports and exports volumes in theregion from 2006 to 2012 (Dahlbacka, 2009; Swedish EnergyAgency, 2012e).

Fig. 7 pictures the insight-matrix for the 2020 scenario, inwhichall three structural components remain as strong reinforcingpressures. The matrix shows that, over time, the influence strengthof the component Biofuel Imports onto the Swedish biofuel in-ventory increases three-fold. Despite securing the status of maininfluence to the country’s inventory, its contribution to system’sgrowth does not increase at the same rate. The other two structuralcomponents, Local Raw Material and Raw Material Imports, upholdthe overlapping position and their reinforcing pressure onto thebiofuel system remains unaltered. This matrix also shows that theSwedish biofuel production would benefit from a growing marketdemand, which is exemplified by the six-fold increase of the

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Fig. 4. Insight-matrix configuration.

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 9

influence strength of the two structural components e Local RawMaterial and Raw Material Imports e over the inventory. Anotherinteresting result is that the scenario analysis corroborates thedevelopment trend discussed in section 2. It confirms that biofuelimports, bioethanol in this specific case, would keep on playing an

Fig. 5. Insight-matrix results of the deman

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

important role in the development of the Swedish biofuel systemtowardmeeting the desired targets of 10% of renewable fuels by theyear 2020.

Fig. 8 illustrates the insight-matrix for the 2030 scenario. Aninteresting fact is that structural components’ influences display

d pressure in the short-term scenario.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Fig. 6. Swedish biofuels imports and exports volumes between 2006 and 2012.

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e1510

no deviation in their values despite varying the amount of esti-mated bioenergy use in the potential scenarios described inTable 2 (i.e., BAU with 20.2 TWh, EF with 25.8 TWh, and FS with39.7 TWh). Therefore, all scenarios present the structuralcomponent Biofuel Imports as the strongest contributor. In fact, thesystem’s growth would almost entirely rely on imports to keep onexpanding because its influence onto the structural componentBiofuel Inventory is more than double of the other components’influences. Fig. 8 also shows that component Local Raw Materialdecreases its reinforcing pressure significantly between 2020 and2030 but it still remains as one of the system’s growing factors.Raw Material Imports, on the other hand, changed its behaviorfrom reinforcing to balancing pressure. Therefore, structuralcomponent Raw Material Imports would provide stability to theSwedish biofuel system in the long-run. Its balancing pressure

Fig. 7. Insight-matrix results of the demand

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

would regulate the system in order to meet biofuel demands andthis indicates that local resource potential of biomass is eithercloser to meet its biophysical limits and cannot enhance the localbiofuel production any longer or there is a lack of local manage-ment capability and infrastructure to harness it further. As a result,the local biofuel production based on raw material imports be-comes the key factor on balancing the system around 2030 aftergaining momentum in 2020. Hence, any potential scenario guidedby demand pressure would heavily rely on raw material imports inthe long-run to increase local production and/or biofuel imports toboost the national biofuel inventory. In this case, the Swedishbiofuel system would be heavily dependent on global markets by2030.

Unfortunately, the country’s current policy framework is notstrong enough to trigger changes within the Swedish biofuel

pressure in the medium-term scenario.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Fig. 8. Insight-matrix results of the demand pressure in the long-term scenario.

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 11

system in order to change the pattern guided by a demand pressure.The reason behind this is not only a lack of infrastructure andmanagement capability related to local production, especiallyregarding 2nd generation biofuels, but also a need for strongerpolicy instruments to trigger changes.

Fig. 9. Biofuel production and consump

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

Finally, the scenarios are presented under a context of a supplypressure, in which the system’s dynamics need to increase the localbiofuel production for meeting the desired targets. Interestingly,this scenario does not mirror the current dynamics within theSwedish biofuel systems. For example, the actual growth of the

tion in Sweden from 2006 to 2012.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e1512

Swedish biofuel production flattened following the European trenddue to low domestic biofuel prices versus high feedstock prices inthe last couple of years (Flach et al., 2013). Fig. 9 presents the bio-fuel production and consumption in Sweden from 2006 to 2012(Swedish Energy Agency, 2009, 2011, 2012a, 2012b, 2012c, 2013a;Svenska Bioenergiföreningen, 2012, 2013). Note that all upgradedbiogas is consumed by the local gas vehicle fleet. Therefore, the redlines representing upgraded biogas production and consumptionoverlap each other (in the web version). It is also important tomention that current upgraded volumes of biogas in Sweden arenot enough to cover local demand. As a result, natural gas is used tocomplement vehicle gas supply.

Fig. 10 shows the insight-matrix for the 2014 scenario under acontext of a supply pressure. Firstly, it shows that only one structuralcomponent applies reinforcing pressure onto the system and theother two are borderline components between reinforcing andbalancing forces. Clearly, the strongest contributor to a potentialsystem’s growth is the component Local Raw Material. The othertwo components e Biofuel Imports and Raw Material Imports e areprone to having a power position within the Swedish biofuel sys-tem that can produce fluctuations in the system between theexternal and internal growth patterns mentioned previously. Theyalso indicate a hindering aspect to the system’s expansion (i.e., lowdomestic biofuel prices versus high feedstock prices), if policymeasures are not used to protect local raw material supply againstcompetition resulting from imports.

Fig. 11 displays the insight-matrix for the 2020 scenario, inwhich Local Raw Material remains as the strongest contributor tobiofuel production in Sweden but its strength does not vary. Thefigure points to a steady but slower growth rate of the Swedishbiofuel system in themedium-run. Biofuel Imports and RawMaterialImports increase their balancing pressure onto the system indi-cating a very powerful influence on providing stability to the sys-tem based on imports. Once again, it is shown that without policy

Fig. 10. Insight-matrix results of the suppl

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

measures the system would keep itself unchanged. It would keepits quite rigid structure based on the demand pressure instead ofshifting to a supply driven system.

Fig. 12 presents the insight-matrix for the 2030 scenario, inwhich Local Raw Material becomes the strongest influence on thesystem’s stability. The matrix also shows a powerful infrastruc-ture and management capability set to harness the local resourcepotential and convert it into biofuels. In this context, Raw Ma-terial Imports has a role between backup and growth factorswithin the Swedish biofuel system. The structural componentBiofuel Imports changes its behavior from balancing pressurewithin the system to reinforcing pressure but its influence ontostructural components diminishes in the long-run. In this case,the component Biofuel Imports becomes a backup for the Swedishbiofuel system. Once more, the insight-matrix illustrates thatwithout policy measures the system would follow the samedevelopment pattern presented previously in Fig. 11. The de-pendency on biofuel imports corroborates the assumption thatthe Swedish biofuel system would be heavily dependent onglobal markets by 2030. As a result, the Swedish biofuel pro-duction based on local raw material becomes the key factor forbalancing the system. Therefore, any potential scenario guided bysupply pressure would heavily rely on local resources to meet thedesired targets. In this case, the Swedish biofuel system would beheavily dependent on new policy measures to boost local biofuelproduction in the short-run (e.g., increase investments on 2ndgeneration biofuels production) in order to guarantee a supplydriven system by 2030.

Although our analysis can oversimplify the problem, it buildsknowledge and validates gathered information through a variety ofsources. Hence, it is important to address that follow-up research isnecessary to match the potential energy use in the Swedish do-mestic transport sector presented in our study with the Swedishresource potential and availability of local bioenergy carriers.

y pressure in the short-term scenario.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

Fig. 11. Insight-matrix results of the supply pressure in the medium-term scenario.

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 13

5. Concluding remarks

Bioenergy contribution will likely reach beyond the 10%renewable fuel target in 2020 even if the current development

Fig. 12. Insight-matrix results of the suppl

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

pattern is upheld. Nevertheless, our analysis identified thatmeeting the 2020 target may rely almost entirely on biofuel im-ports. As a result, the Swedish biofuel system will continuedepending on biofuel imports, which makes it periodically exposed

y pressure in the long-term scenario.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e1514

to increased competition and pricing pressures. The fact that the settarget for 2020 is estimated to be exceeded ratifies the effectivenessof the Swedish policymeasures in place. In fact, the country’s policyframework not only recognizes biofuels as a reliable energy sourcebut also explores their manageable integration into distributionnetworks as blended fuels (McCormick et al., 2012; Grönkvist et al.,2013). Yet, the current framework is not sufficient to guarantee theestablishment of a vehicle fleet independent of fossil fuels by 2030.In fact, the BAU scenario indicates that fossil fuels will endure as theleading energy carrier in the Swedish domestic transport sectoraccounting to around 78% of the final energy use (see Table 2).

Biofuel imports already play an important role in meeting thedesired targets and this role is bound to continue in the future. In2011, for example, 55% of the bioethanol and 60% of the biodieselused in Sweden was imported mostly from France and Lithuaniarespectively (Swedish Energy Agency, 2012e). Unfortunately, thecurrent policy framework is not strong enough to trigger changeswithin the Swedish biofuel system in order to change this pattern.The reason for this is the fact that the system itself is guided by aquite rigid structure and averse to changes based on current de-mand pressure instead of a supply pressure. Note that supplypressure relies on increasing local biofuel production to foster bio-fuel use and demand pressure depends on biofuel inventory rein-forcement to meet fuel demands. This reliance on imports showsnot only a lack of infrastructure and management capability relatedto local production, especially regarding second generation bio-fuels, but also a need for stronger policy instruments to triggersystemic changes.

Our analysis also indicated another hindering factor that is in-vestment downplaying caused by high level of uncertainty withinbiofuel systems. In general, bioenergy systems are still seen as riskyby traditional and well-established investors, and this tends toprevent long-term investments in bioenergy infrastructure (i.e.,bio-refineries or other biofuel production plants). The higher theuncertainty, the lower the investments made in bioenergy systemsinfrastructure. One of the current causes for uncertainty is relatedto the fact that robust and clear sustainability criteria, especiallyrelated to land use changes, have not yet been established forbiofuel systems. These criteria could balance the interests ofdifferent stakeholders at national or international levels, which inturn affects the confidence on biofuel availability and increasesrisks when it comes to quantity, quality, price, as well as demandvolumes across the main supply sources. For example, the previ-ously mentioned cap regarding the amount of biofuel derived fromfood crops at 5.5% was set to avoid competition with food and fiberproduction and deals with indirect changes in land use and biofuelstandards that would affect the biofuel system as a whole.Regardless whether the directive proposal enters into legal force ornot, this may still create uncertainties regarding the future devel-opment of the Swedish biofuel system. Therefore, the system isalready influenced and this uncertainty makes investors questionthe credibility of existing infrastructure on facing such a challenge.

Despite the aforementioned challenges to the future of biofuelsin Sweden, there is no doubt that the Swedish biofuel systemswerebuilt on government support and incentives. Reliance on suchsupport does not only keep the system stable but it is also a tool inchanging the course of its development so as to turn challenges intoopportunities andmeet the Swedish government’s vision on havinga vehicle fleet independent of fossil fuels by 2030 withoutcompromising security of supply. To conclude, our results showthat policymakers have to decide on how to steer the system’sdevelopment not only by combining different pathways betweenexternal-based or internal-based growth but also deciding onapplying either supply or demand pressures. These choices have tobe taken in a very short-run in order to shape the development of

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

the Swedish biofuel system away of the potential BAU scenario. Bydoing so, the system would avoid failing against the target ofestablishing a vehicle fleet independent of fossil fuels by 2030.

Acknowledgments

The authors gratefully acknowledge valuable comments fromcolleagues at the Nordic Consortium for Systems Analysis (NOR-COSA), our research fellows at KTH e especially Maria Xylia e, andthe participants of the 4th International Workshop: Advances inCleaner Production. We would also like to thank Lucas Elias for thesupport and caring.

References

Baltic BiogasBus, 2012. Baltic Sea Region Biogas Infrastructure Overview: Existingand Planned. Baltic BiogasBus, Stockholm.

BIL Sweden, 2011. I vilka personbilsmodeller kam man tanka E10-bränsle? BILSweden, Stockholm.

Binder, T., Vox, A., Belyazid, S., Haraldsson, H., Svensson, M., 2004. Developingsystem dynamics models from causal loop diagrams. In: Proceedings of the22nd International Conference of the System Dynamics Society, July 25e29.Oxford, System Dynamics Society.

Börjesson, P., Ericsson, K., Di Lucia, L., Nilsson, L.J., Åhman, M., 2009. SustainableVehicle Fuels e Do they exist? Lund University, Lund.

Box, G.P., Draper, N.R., 1987. Empirical Model-Building and Response Surfaces.Wiley.

Bush, B., Duffy, M., Sandor, D., Peterson, S., 2008. Using system dynamics to modelthe transition to biofuels in the United States. In: IEEE SMC InternationalConference on System of Systems Engineering. IEEE.

Dahlbacka, B., 2009. Sweden Biofuels Annual 2009. USDA Foreign AgriculturalService, Stockholm.

European Commission, 2012. Renewable energy progress and biofuels sustain-ability. European Commission, Brussels.

European Union, 2003. Directive 2003/30/EC of the European Parliament and of theCouncil on the Promotion of the Use of Biofuels or Other Renewable Fuels forTransport. Official Journal of the European Union, Brussel.

European Union, 2009. Directive 2009/30/EC of the European Parliament and of theCouncil of 23 April 2009 Amending Directive 98/70/EC as Regards the Specifi-cation of Petrol, Diesel and Gas-oil and Introducing a Mechanism to Monitorand Reduce Greenhouse Gas Emissions. Official Journal of the European Union,Brussels.

Flach, R., Bendz, K., Krautgartner, R., Lieberz, S., 2013. EU Biofuels Annual 2013.USDA Foreign Agricultural Service, Hague.

Government of Sweden, 2010. The Swedish National Action Plan for the Promotionof the Use of Renewable Energy in Accordance with Directive 2009/28/EC andthe Commission Decision of 30.06.2009. Regeringskansliet, Stockholm.

Government of Sweden, 2011. Drivmedelslag (2011:319). Regeringskansliet, Stock-holm. Retrieved 17.12.12, from: http://62.95.69.15/cgi-bin/thw?%24%7BHTML%7D¼sfst_lst&%24%7BOOHTML%7D¼sfst_dok&%24%7BSNHTML%7D¼sfst_err&%24%7BBASE%7D¼SFST&%24%7BTRIPSHOW%7D¼format%3DTHW&BET¼2011%3A319%24.

Grönkvist, S., Peck, P., Silveira, S., Åkerman, J., Larsson, M., Khedkar, P., 2013. PolicyInstruments Directed at Renewable Transportation Fuels e An InternationalComparison. The Swedish Knowledge Centre for Renewable TransportationFuels, Göteborg.

Hillman, K.M., Sandén, B.A., 2008. Exploring technology paths: the development ofalternative transport fuels in Sweden 2007e2020. Technological Forescating &Social Change 75 (8), 1279e1302.

Jeffers, R.F., Jacobson, J.J., Searcy, E.M., 2013. Dynamic analysis of policy drivers forbioenergy commodity markets. Energy Policy 53 (1), 249e263.

Kirkwood, C.W., 1989. System Dynamics Methods: A Quick Introduction. ArizonaState University, Temp.

Lorentz, U., Neumann, K., 2012. Assessing the effectiveness of policy measures withthe help of qualitative modeling. In: Berlin Conference on the Human Di-mensions of Global Environmental Change, 5e6 October, 2012, pp. 1e11. Berlin.

McCormick, K., Bomb, C., Deurwaarder, E., 2012. Governance of biofuels for trans-port in Europe: lessons from Sweden and the UK. Biofuels 3 (3), 293e305.

Min, H., Zhou, G., 2002. Supply chain modeling: past, present and future. Computers& Industrial Engineering 43 (1e2), 231e249.

Morecroft, J.D., 1982. A critical review of diagraming tools for conceptualizingfeedback system models. Retrieved 04.03.13, from System Dynamics Society.Dynamica Digital Archive: http://www.systemdynamics.org/dynamica/articles/81/5.pdf.

Neumann, K. (2013, March 19). Personal Communication. Information about thealgorithm used to design the insight-matrix given by the Executive andShareholder of Consideo Company.

Newman, M., 2010. Networks: An Introduction. Oxford University Press, Oxford.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019

A. Sanches-Pereira, M.F. Gómez / Journal of Cleaner Production xxx (2014) 1e15 15

Olsson, L., Hjalmarsson, L., 2012. Policy for Biomass Utilisation in Energy andTransport Systems: The case of biogas in Stockholm, Sweden. World RenewableEnergy Forum 2012. WREF, Denver.

Pacini, H., Silveira, S., 2011. Consumer choice between ethanol and gasoline: lessonsfrom Brazil and Sweden. Energy Policy 39 (11), 6936e6942.

Richardson, P., Pugh III, A.L., 1981. Introduction to System Dynamics Modeling withDynamo. Pegasus Communications, Westford.

Sanches-Pereira, A., 2012. Assessment of Sustainability within Bioenergy SupplyChain Management. PhD Dissertation. UNICAMP State University of Campinas,Campinas.

Sanches-Pereira, A., Gómez, M.F., 2013. Towards a cleaner vehicle fleet: the dy-namics of the Swedish biofuel system. In: Giannetti, B.F., Almeida, C.M.,Bonilla, S.H. (Eds.), 4th International Workshop e Advances in Cleaner Pro-duction: Integrating Cleaner Production into Sustainability Strategies. UNIP, SãoPaulo.

Sköldberg, H., Löfblad, E., Holmström, D., Rydén, B., 2010. Ett fossilbränsleober-oende transportsystem år 2030: Ett visionsprojekt för Svensk Energi ochElforsk. Elforsk, Stockholm.

Sterman, J., 2000. Business Dynamics: Systems Thinking and Modeling for a Com-plex World. McGraw-Hill, Irwin.

Svenska Bioenergiföreningen, 2012. Fakta om bioenergi: Biokraft. Retrieved12.11.12, from Svebio: http://www.svebio.se/biokraft-0.

Svenska Bioenergiföreningen, 2013. Fakta om bioenergi: Biokraft. Retrieved02.08.13, from: http://www.svebio.se/biokraft-0.

Svenska Petroleum och Biodrivmedel Institutet, 2013. Utveckling av försäljningsprisför bensin, dieselbränsle och etanol. Retrieved 30.05.13, from: http://spbi.se/statistik/priser/.

Swedish Energy Agency, 2009. Transportsektorns energianvändning 2008. Statensenergimyndighet, Eskilstuna.

Please cite this article in press as: Sanches-Pereira, A., Gómez, M.F., Tindependent of fossil fuels, Journal of Cleaner Production (2014), http://d

Swedish Energy Agency, 2011. Energy in Sweden 2011. Statens energimyndighet,Eskilstuna.

Swedish Energy Agency, 2012a. Statens energimyndighet. Retrieved 18.10.12, from:http://www.energimyndigheten.se/sv/Statistik/Energilaget/.

Swedish Energy Agency, 2012b. Transportsektorns energianvändning 2011. Statensenergimyndighet, Eskilstuna.

Swedish Energy Agency, 2012c. Kortsiktsprognos över energianvändning ochenergitillförsel 2012e2014, Våren 2012. Statens energimyndighet, Eskilstuna.

Swedish Energy Agency, 2012d. Produktion och användning av biogas år 2011.Statens energimyndighet, Eskilstuna.

Swedish Energy Agency, 2012e. Analys av marknaderna för biodrivmedel: Aktuellamarknadsfrågor för första och andra generationens biodrivmedel. Statensenergimyndighet, Eskilstuna.

Swedish Energy Agency, 2013a. Kortsiktsprognos över energianvändning ochenergitillförsel 2012e2014, Våren 2013. Statens energimyndighet, Eskilstuna.

Swedish Energy Agency, 2013b. Transportsektorns energianvändning 2012. Statensenergimyndighet, Eskilstuna.

Trafikanalys, 2012. Fordon 2011. Retrieved 29.11.12, from: http://www.trafa.se/sv/Statistik/.

Trafikanalys, 2013. Fordon 2012. Retrieved 19.04.13, from: http://www.trafa.se/sv/Statistik/.

TransportNäringen i Samverkan, 2010. Miljöarbete i transportnäringen: en kortsammanställning över miljöfordon och drivmedel. TransportNäringen i Sam-verkan, Stockholm.

Transportstyrelsens, 2012. Fordonsregler. Retrieved 17.12.12, from: http://www.transportstyrelsen.se/sv/Vag/Fordon/fordonsregler/E10/.

Ulmanen, J.H., Verbong, G.P., Raven, R.P., 2009. Biofuels developments in Swedenand the Netherlands: protection and socio-technical change in a long-termperspective. Renewable and Sustainable Energy Reviews 13 (6e7), 1406e1417.

he dynamics of the Swedish biofuel system toward a vehicle fleetx.doi.org/10.1016/j.jclepro.2014.03.019