lable at ScienceDirect

Journal of Cleaner Production 17 (2009) S18–S26

Contents lists avai

Journal of Cleaner Production

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

European trade of forest products in the presence of EU policy

Robert Lundmark*, Anna MansikkasaloLuleå University of Technology, Economics Unit, SE-971 87 Luleå, Sweden

a r t i c l e i n f o

Article history:Received 22 May 2008Received in revised form22 January 2009Accepted 22 January 2009Available online 5 March 2009

Keywords:BiomassBioenergyWhite PaperRES-EForest fuelEU trade

* Corresponding author. Tel.: þ46 920 49 23 46; faE-mail address: robert.lundmark@ltu.se (R. Lundm

0959-6526/$ – see front matter � 2009 Elsevier Ltd.doi:10.1016/j.jclepro.2009.01.010

a b s t r a c t

Increasing the share of renewable energy is of principal concern for the EU energy policy. A number ofpolicies have been adopted, and, in part, been implemented by the EU member countries. An increasingshare of renewable energy implies an increasing utilisation of biofuels in general and of forest-basedbiomass in particular. However, in the EU, the endowment and uses of forest-base biomass are diversesuggesting that an increasing trade would become necessary in order to cost effectively increase theutilisation of forest-based biomass. The purpose of this study is to, in the presence of EU energy policy,quantify and analyse possible trade levels for forest fuels in the EU. Particularly, the consequences ontrade after implementing the White Paper and the RES-E Directive are analysed. Investigating theEuropean trade in forest fuels is important for understanding how industry sectors in the EU will beaffected by the policies. The results suggest that the implementation of the White Paper and the RES-EDirective will increase the trade in forest fuels, resulting in total trade increases of up to 67 percent.Furthermore, the national net trading levels are possible to derive. Depending on policy implementationthe results differ – a country that was net importing given the White Paper implementation can insteadbe net exporting when applying the RES-E Directive. The fact that the policy implementations willincrease the trade may mitigate potential industry problems to secure the needed inputs. On the otherhand, the integration of countries increases the possibility for some industries to increase theirproduction even more, possibly strengthening any input scarcity problems. It is therefore not possible togenerally conclude if a more integrated European forest fuel market, and hence an increased Europeanforest fuel trade, will mitigate industry problems to secure their needed inputs or not.

� 2009 Elsevier Ltd. All rights reserved.

1. Introduction

Trade liberalisation, economic growth and increasing environ-mental concern have, amongst other issues, created ample subjectmatter for trade analysis. Major issues arising from the growingimportance of forest product trade include affects on inputcompetitive industries, biodiversity protection, global warming,forest sustainability and the hardening of trade blocs (see e.g. [29]for issues regarding world trade in forest products). Furthermore,an increasing demand for forest-based biomass, not least from theenergy sector, has several implications on the direction andmagnitude of their trade. The increasing demand originates froman overall economic growth but is also, in many aspects, driven bypolicy decisions. In the European Union (EU) there is a desire toincrease the share of renewable energy in its energy mix (of whichforest-based biomass constitutes a significant part). As a conse-quence, both EU and national energy policies have been

x: þ46 920 49 20 35.ark).

All rights reserved.

implemented that support the use of forest-based biomass forenergy purposes. For instance, in 1999 approximately 46 percent ofthe biomass used for energy in EU19 originated from the forestrysector [2]. With an increasing demand, imports have become analternative supply source for many users and the traditionally local/regional markets have started to transform towards a more inter-national orientation [20]. The purpose of this study is to developa numerical simulation model of the European trade in forest-basedbiomass in order to assess possible changes in the trade patternsarising from the implementation of different EU energy policies.The development of European trade in forest-based biomass mightimprove the possibility to cost effectively reach EU energy policygoals. Consequently, the analysis of European forest-based biomasstrade is an important input for understanding how selected parts ofthe European economies will be affected by the EU policies aimedat promoting the use of renewable energy sources.

Figs. 1–3 depict the development of net exports of wood fuels,chips & particles and roundwood, respectively, in 1994, 2000 and2006 for the EU countries listed. On average there is more variationin the trade volume of roundwood than in the trade volumes ofchips & particles and wood fuel. However, a clearer trend towards

-1.2-1.0-0.8-0.6-0.4-0.2

00.20.40.6

Fran

ce

Hun

gary

UK

Est

onia

Cze

ch R

epub

lic

Lat

via

Spai

n

Slov

akia

Slov

enia

Pola

nd

The

Net

herl

ands

Port

ugal

Ger

man

y

Bel

gium

Den

mar

k

Finl

and

Aus

tria

Swed

en

Ital

y

Net

exp

ort

of w

ood

fuel

(Mili

on m

3 solid

)

Net export of wood fuel in 2000 (million m3solid)Net export of wood fuel in 1994 (million m3solid)

Net export of wood fuel in 2006 (million m3solid)

Net exporting countries

Net importing countries

Fig. 1. Net export of wood fuel for 1994, 2000 and 2006 for European countries. Source: FAO (2008).

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26 S19

increasing or decrease the net trade can be observed for chips &particles and wood fuel within a number of countries compared tofor roundwood. In fact, the countries can be divided into twocategories according to their net exports. The first group of coun-tries shows evidence of an increasing or decreasing net trade. Forwood fuel these countries are Czech Republic, Spain, Denmark,Finland and Italy; for chips & particles Germany, Latvia, Estonia,Portugal, Denmark, Spain, Finland, Sweden and Italy and; forroundwood Poland, Spain, Austria and Finland.

The other group of countries exhibits either a stagnant trend ora shift in the trade direction. Italy, Slovakia, Latvia, France, Estonia andSweden have improved their relative trade balances in roundwoodmoving towards a higher degree of self-sufficiency, while Hungary,Slovakia and Estonia exhibit the same pattern for wood fuel. For chips& particles the net trade has increased for all countries, with theexception for France, the Netherlands and Belgium, suggesting thatthe domestic markets depend to a higher degree on trade.

2. Trade theory and modelling

In general, the magnitude and direction of the trade flows offorest-based biomass are determined by several factors, forinstance: (1) geography (e.g., distance); (2) size of the economies(e.g., demand); (3) character of forest endowments (e.g., supply)and; (4) policies (e.g., tax, subsidies, regulations). Classical tradetheory prescribes that trade occurs because there are differencesamong trading partners in their relative costs of production. Inforest products, empirical research based on theoretical andapplied work done by Vanek [47] and Leamer [35] suggests that theprincipal factor determining the direction and magnitude of nettrade is the size of a country’s forest endowment relative to the sizeof its economy [9,40].1

Trade modelling approaches for evaluating effects of policiesusually fall into three categories: computable general equilibrium,spatial partial equilibrium, and non-spatial partial equilibrium.Each approach has its own particular uses, advantages, and disad-vantages. Computable general equilibrium (CGE) models provideanswers to questions about policies at the level of whole sectorsand economies. An advantage of CGE models is that they include allsectors of an economy thus providing an analytical framework ofhow a specific policy might affect both prices and quantities of allaffected markets. Second, the effects of trade changes can bemeasured in terms of aggregate national economic welfare. Theprimary weakness of CGE models is their lack of model resolution.

1 Various measures of forest endowments, net trade and size of the economyhave been used in the literature. For instance, Leamer [35] suggests that whenmeasuring resource endowments it is not so much the extent of the resource, e.g.,forest areas, but the flow of goods and services that can be derived from theresource, e.g., volume of timber harvested, that matters.

Due to tradeoffs between model accuracy and size on the one handand model resolution on the other, the effects at the sub-sectorlevels are usually not revealed. This means that it is difficult toevaluate the effects of policies on specific products.

Partial equilibrium models assume that the feedback from sec-toral changes to aggregate variables is negligible thus allowinga more detailed analysis of the affect a specific policy might have oncommodity markets. Spatial partial equilibrium models have beenused to study the effects of trade levels of forest product (e.g.,[10,11,14,27,28]). Indeed, partial equilibrium models are the mostcommonly used approach to forest sector analysis. Partial equilib-rium modelling, often based on the seminal work of Samuelson[42] and Takayama and Judge [46], is characterised by the endog-enous determination of factors such as trade flows and prices,conditional on exogenous economic activity outside the sector suchas changes in different policies (see, for example [1,15,33]).However, often the partial equilibrium models are only brieflydescribing European forest fuel markets and their interdepen-dencies. However, Lundmark [37] develops a partial equilibriummodel for the Swedish forest cluster in which the interdepen-dencies of various forest product markets are incorporated. Thus,using a partial equilibrium analysis may be more appropriate thanapplying a general equilibrium model (for other general equilib-rium models besides the GTAP, see, for example [4,5,13,36]).

3. The European Forest Trade Model

The European Forest Trade Model (EFTM) is a static numericalmodel focusing on the trade levels in 19 European countriesimplemented using GAMS (General Algebraic Modeling System).2

Fig. 4 presents the sectors and their connection addressed in theEFTM (bold lines represents the flows included in the model). In theEFTM the industry sectors optimal input combination is where thetechnical rate of substitution equals the input price ratio. The EFTMassumes that a specific country uses its domestic input supply aslong as the domestic price is lower than the import price before.

Numerical calculation of an economic equilibrium requires thechoice of functional forms for production possibilities and prefer-ences. In applied modelling, CES functions (including Leontief andCobb-Douglas specifications as sub-cases) are amongst the mostcommon. Such functions have certain mathematical properties(regularity) that ease the numerical analysis but are still flexibleenough to allow for appropriate representation of economicbehaviour. The model is formally described in Appendix.

The EFTM uses calibrated CES specifications for the cost,demand and production functions [7,8], in which each industry is

2 GAMS is a high-level modelling system for mathematical programming andoptimisation.

-3

-2

-1

0

1

2

3

Ger

man

y

Lat

via

Est

onia

Fran

ce

Cze

ch R

epub

lic

Slov

akia

Pola

nd

Hun

gary

UK

Slov

enia

The

Net

herl

ands

Port

ugal

Den

mar

k

Spai

n

Aus

tria

Bel

gium

Finl

and

Swed

en

Ital

y

Net

exp

ort

of c

hips

&pa

rtic

les

(Mill

ion

m3 so

lid)

Net export of chips & particles in 2000 (million m3solid)

Net export of chips & particles in 1994 (million m3solid)

Net export of chips & particles in 2006 (million m3solid)

Net exporting countries

Net importing countries

Fig. 2. Net export of chips & particles for 1994, 2000 and 2006 for European countries. Source: FAO (2008).

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26S20

assumed to use two inputs (production factors) in their productionprocess making the inputs substitutes. For instance, for the energysector (the heat and power industry and the refined woodfuelindustry) this implies that the input mix of by-products and loggingresidues is weakly separable from other fuel inputs (such as coaland oil). This can partly be motivated by the policy requirementthat a certain share of the production should be based on biomass.The elasticity of substitution3 is set between 0.5 and ten dependingon industry and country (e.g. [41,43,49]). Since no reliable estimateshave been found it is assumed that the elasticity of substitution isrelated to the size of the national market, i.e., the produced andconsumed quantities. It is thus assumed that a larger marketindicates that the production techniques are more developed andthat the substitution between inputs is technically easier. Fig. 4indicates which inputs the different industry sectors use and canthus substitute between. In addition to the technical possibilities tosubstitute between inputs, their relative price is another determi-nant of the chosen input mix (e.g. [31,32,48]).

The Baseline scenario outlines the European trade in forest-basedbiomass without any EU energy policies implemented. In thisscenario the heat, power, refined woodfuel and pulp industry mustproduce exactly the benchmark output level, using two inputs whendoing so. The industry sectors will choose their input combinationsaccording to the inputs’ marginal product and their relative price.The industry sectors strive to minimise their total cost of productionhence their optimal input combination is where the technical rate ofsubstitution equals the input price ratio. Countries will demandrelatively large quantities of the inputs used in their dominateproduction and if the domestic supply of these inputs is less thanneeded the countries can import the remaining quantity. Thedifference between the domestic quantity supplied and thedomestic input demand indicates a country’s net import level.

In the refined woodfuel industry due to high quality restrictionsand therefore a dominating use of by-products (mainly sawdust), itis not likely that the producers can perform major substitutionsbetween by-products and logging residues. As a consequence, theelasticity of substitution between by-products and logging residuesis set relatively low compared for instance with the heating andpower sector. For the largest producers; Austria, Denmark, Finland,Italy, Poland and Sweden it is set to 1.25 and for the remainingcountries, which all have a relatively small production of refinedwoodfuel it is set to 0.9 [22]. For the heat industry, Sweden, Austria,Finland, Germany and Latvia which all produce forest fuel based

3 The elasticity of substitution measures the degree of substitutability betweenany pair of production factors.

heat above the EU19 country average are assumed to have anelasticity of substitution of ten. The remaining countries areassumed to have a substitution elasticity of five. For the powerindustry, Austria, Denmark, Finland, Germany, Italy, theNetherlands, Sweden and the UK have the highest forest fuel basedpower generation. For this reason it is assumed that they will beable to substitute more easily between by-products and loggingresidues. As a result, the elasticity of substitution is set to ten. Forthe remaining countries the elasticity of substitution is set to five.The elasticity of substitution between roundwood and by-productsfor the pulp industry is set to three [43]. Some mills clearly prefer tobuy pulpwood, others prefer to use wood chips while some millsare indifferent [12,34]. Thus, whatever the reason might be theelasticity of substitution cannot be assumed to be as high as in theheat and power industries. In contrast, it cannot be assumed to beas low as in the refined woodfuel industry since some plants clearlysubstitute between the inputs. Table 1 summarises the assump-tions made regarding the elasticity of substitution.

3.1. Benchmark values

The initial (or benchmark) values are summarised in Tables 2–4.In Table 2 only net production is considered. Due to insufficientstatistics, the production of heat was estimated using informationon plant capacity and average operation time [2,20]. No refinedwoodfuel production statistics are available for the Czech Republicand Belgium. For these countries it is assumed that the productionof refined woodfuel amounts to five percent of total heat and powerproduction.

Table 3 presents the benchmark values for the production andprice of by-products, logging residues and roundwood. Theproduction of inputs is determined exogenously. The data obtainedon by-product production from FAOStat is reduced by 40 percent toreflect that FAOStat has a broader definition of by-productscompared to this study. Statistics on the production of loggingresidues are estimated based on harvested volumes of roundwood.

Statistics on the prices of industrial by-products as well aslogging residues were collected from Alakangas and Vesterinen [3]and EUBIONET II [20], and complemented with statistics presentedin Alakangas and Vesterinen [2]. The data for logging residues andby-product prices have been collected from different sources,which all use different years. A majority of the data presented willbe for the year 2005 but some figures are for 1999 (in 2005 yearsvalue). The price of roundwood is estimated based on export values.

The benchmark values for the consumption of inputs are pre-sented in Table 4. It is assumed, for the calculation of the bench-mark values, that the heat and power industry uses an equal

-15-13-11-9-7

-5-3-1135

Lat

via

Est

onia

Fran

ce

Ger

man

y

Slov

akia

Hun

gary

Cze

ch R

epub

lic

Den

mar

k

UK

The

Net

herl

ands

Slov

enia

Pola

nd

Port

ugal

Bel

gium

Spai

n

Ital

y

Aus

tria

Finl

and

Swed

en

Net export of roundwood in 2000 (million m3solid)

Net export of roundwood in 1994 (million m3solid)

Net export of roundwood in 2006 (million m3solid)

Net exporting countries

Net importing countries

Net

exp

ort

of r

ound

woo

d(M

illio

n m

3 solid

)

Fig. 3. Net export of roundwood for 1994, 2000 and 2006 for European countries. Source: FAO (2008).

By-products

Logging residues

SawlogsPulpwood

Pellets, etc

By-products

SawlogsPulpwood

Logging residues

By-

prod

ucts

Hea

t

Saw

logs

Hea

t

Pulp

woo

d

By-

prod

ucts

Log

ging

res

idue

s

Rou

ndw

ood

By-products

By-products

Export

Pulp Industry

Import

DomesticForest

Heat IndustryPower Industry

SawmillIndustry

RefinedWoodfuelIndustry

WoodboardIndustry

Fig. 4. Forest fuel flows of focus. Sources: Ericsson and Nilsson [19] and Lundmark and Soderholm [38].

Table 1Elasticity of substitution by country and industry sector.

Heatindustry

Powerindustry

Refined woodfuelindustry

Pulpindustry

Austria 10 10 1.25 3Belgium 5 5 0.9 3Czech

Republic5 5 0.9 3

Denmark 5 10 1.25 3Estonia 5 5 0.9 3Finland 10 10 1.25 3France 5 5 0.9 3Germany 10 10 0.9 3Hungary 5 5 0.9 3

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26 S21

amount of logging residues and by-products. The same is assumedfor the pulp industry but for roundwood and by-products. Therefined woodfuel industry is assumed to use ten percent loggingresidues and 90 percent by-products.

The energy efficiency for the heating sector is assumed to be 85percent for Sweden, Austria, Finland, Germany and Latvia. For theremaining countries is set to 75 percent. The average energy effi-ciency for the power industry is approximately 22 percent [26]which is used for all countries. The material efficiency in the refinedwoodfuel industry in Sweden, Austria, Denmark, Finland, Italy andPoland is set to 80 percent [17]. For the remaining countries thematerial efficiency is set to 70 percent due to the fact that many ofthese countries, such as Spain, have relatively small markets forrefined woodfuels [39].

Italy 5 10 1.25 3Latvia 10 5 0.9 3Netherlands 5 10 0.9 3Poland 5 5 1.25 3Portugal 5 5 0.9 3Slovakia 5 5 0.9 3Slovenia 5 5 0.9 3Spain 5 5 0.9 3Sweden 10 10 1.25 3United

Kingdom5 10 0.9 3

Source: Wibe [49]; Sisak [43]; Rehn [41] and EUBIONET II [20].

3.2. Scenario 1: the White Paper implemented

An indicative goal of the White Paper is to have a biomassenergy consumption of 8.5 percent [21]. Stated differently, thisimplies that the EU would have to produce 1570 TWh of biomassenergy. When the White Paper target was set in 1997 the biomassbased energy was 523 TWh suggesting an increase of 1047 TWh.Furthermore, it was expected that forest fuels and agriculturalresources should contribute to one third of the necessary increase.

Table 2Benchmark production levels by country and industry.

Country Heat Power RWF Pulp

GWh m3seb(1000)

GWh m3seb(1000)

GWh m3seb(1000)

tonnes(1000)

Austria 3699 1739 1746 821 858 406 1932Belgium 815 386 513 241 132 63 531Czech

Republic423 199 593 279 50 24 753

Denmark 2133 1003 1834 862 905 429 1Estonia 1359 639 23 11 34 16 69Finland 4050 1904 10,182 4786 791 374 11,134France 1350 635 1371 644 84 39 2504Germany 31,500 14,805 3900 1833 429 203 2879Hungary 119 56 678 319 22 10 1Italy 140 66 1913 899 753 356 515Latvia 7200 3384 6 3 112 53 1Netherlands 2916 1381 1836 863 84 39 117Poland 1800 846 768 361 572 271 1042Portugal 2008 951 1264 594 84 39 1949Slovakia 475 223 3 1 17 8 609Slovenia 131 61 90 42 0.6 0.3 153Spain 2149 1017 1353 636 95 45 1973Sweden 16,758 7876 6614 3109 4140 1960 12,108United

Kingdom5 3 1867 877 5 2 341

EU19 79,029 37,171 36,554 17,180 9167 4339 38,612

Sources: Own calculations based on Alakangas and Vesterinen [3]; Alakangas andVesterinen [2]; EUBIONET II [20]; EUROSTAT [23] and FAO (2008).

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26S22

In this case no distinction between forest fuels and agriculturalresources is made. It is therefore assumed that the increase isequally distributed between forest fuels and agricultural resources.This suggests that the utilisation of forest fuels will increase by 32percent on a national wide basis. The 32 percent increase is set forthe heat, power and refined woodfuel industries in total. Thus, theindividual industry productions can vary. However, although theindividual industry productions can vary it has an upper bound notallowing an industry to increase its output production with morethan 110 percent, mimicking a limitation in short-run capacity.

Table 3Benchmark production and price of inputs by country and input.

Country By-productsa,b Log

0000 m3 (GWh) V m3 solid�1 (V GWh�1) 000

Austria 4522 (9551) 31 (15) 249Belgium 633 (1337) 26 (12) 837Czech Republic 1334 (2818) 7 (3) 278Denmark 88 (185) 19 (9) 158Estonia 2400 (5059) 19 (9) 107Finland 11,149 (23,546) 19 (9) 918France 1709 (3609) 32 (15) 614Germany 4980 (10,518) 17 (8) 992Hungary 138 (291) 25 (12) 547Italy 432 (912) 19 (9) 524Latvia 2631 (5557) 6 (3) 231Netherlands 769 (1625) 32 (15) 160Poland 2316 (4891) 36 (17) 559Portugal 1406 (2969) 18 (9) 213Slovakia 719 (1518) 18 (9) 175Slovenia 107 (226) 11 (5) 349Spain 2880 (6083) 15 (7) 260Sweden 11,100 (23,443) 26 (12) 17,United Kingdom 1276 (2695) 13 (6) 158EU19 50,588 (106,843) 20.5 (9.6) 68,

a Production data are own calculations based on EUBIONET II [20]; FAOSTAT [24] andb Price data from Alakangas and Vesterinen [3]; Alakangas and Vesterinen [2]; EUBIONc The price of logging residues for Slovenia was not possible to obtain. For simplicity, thi

are similar in both countries although Slovakia generally has a more developed forest fu

3.3. Scenario 2: the RES-E Directive implemented

The RES-E Directive sets the EU target of renewable power to22.1 percent by 2010 [16]. The RES-E Directive states that themember states shall adopt and continuously publish nationalindicative targets for future consumption of electricity producedfrom renewable energy sources in terms of a percentage of totalelectricity consumption. The RES-E Directive presents nationaltargets on renewable electricity without specifying the sharegenerated by forest fuels. In 2005, the average EU share of forestfuels was 29 percent which is assumed constant even afterimplementing the RES-E Directive. The RES-E Directive onlyconcerns electricity thus the only industry directly affected by thepolicy implementation in scenario 2 is the power industry, whichhas to increase its utilisation of forest fuels according to the indi-cated indicative targets in Table 5.

4. Results

Fig. 5 summarises the resulting net import of by-products underthe scenarios. In the baseline scenario the largest net importers ofby-products are Italy, the United Kingdom and Latvia. The countrieshaving the largest net export volumes are Sweden, Finland andEstonia.

With the White Paper implemented (scenario 1) the total forestfuel based energy in a country has to increase by 32 percent. Fig. 5indicates that Sweden will still have the largest net export of by-products compared with the baseline scenario. For Sweden, theimplementation of the White Paper results in an increase in theSwedish production of heat, while the power and refined woodfuelindustries decrease their production. Due to the conditions speci-fied in the EFTM, this implies that Sweden is relatively good atproducing heat. Furthermore, Sweden will use more by-productsand logging residues, compared with the baseline scenario, in orderto increase its production of heat. This will, in turn, decrease theamount of by-products available for export versus increasing thequantity of logging residues necessary to import. However,regarding logging residues the other Swedish industries will

ging residuesa,b Roundwooda,b

0 m3 (GWh) V m3 solid�1 (V GWh�1) 0000 m3 V tonne�1

3 (5266) 51 (24) 12,786 75(1767) 45 (20) 4290 69

6 (5883) 8 (4) 14,285 48(334) 38 (18) 810 71

3 (2265) 15 (7) 5500 428 (19,404) 24 (11) 47,116 681 (12,969) 30 (14) 31,490 556 (20,965) 18 (9) 50,905 59(1155) 25 (12) 2804 53(1107) 49 (23) 2687 367

9 (4898) 15 (7) 11,893 37(338) 23 (11) 820 41

1 (11,808) 16 (8) 28,672 636 (4511) 23 (11) 10,953 546 (3709) 9 (4) 9005 45(737) 9 (4c) 1789 53

4 (5499) 21 (10) 13,352 56882 (37,766) 28 (13) 91,700 442 (3342) 34 (16) 8115 43

050 (143,721) 23.3 (12) 348,972 70.7

Swedish Forest Agency [45].ET II [20] and FAOSTAT [24].

s price was assumed to be identical to the price in Slovakia. Forest resource potentialsel market.

Table 4Benchmark input consumption by industry, input and country.

Country By-products Logging residues Roundwood

Heata Powerb RWFc Pulpd Heata Powerb RWFc Pulpd Sawmille0000 m3 (GWh) 0000 m3 (GWh) 0000 m3 (GWh) 0000 m3 (GWh) 0000 m3 (GWh) 0000 m3 (GWh) 0000 m3 (GWh) 0000 m3 0000 m3

Austria 1026 (2166) 1822 (3848) 406 (858) 5120 (10,813) 1026 (2166) 1822 (3848) 102 (214) 5120 9892Belgium 257 (542) 535 (1130) 72 (151) 1407 (2972) 257 (542) 535 (1130) 18 (38) 1407 2690Czech Republic 132 (279) 619 (1307) 27 (58) 1995 (4214) 132 (279) 619 (1307) 7 (14) 1995 8153Denmark 667 (1408) 1914 (4042) 429 (905) 3 (6) 667 (1408) 1914 (4042) 107 (226) 3 470Estonia 425 (897) 24 (51) 18 (38) 183 (386) 425 (897) 24 (51) 5 (10) 183 3300Finland 1123 (2372) 10,624 (22,438) 374 (791) 29,505 (62,315) 1123 (2372) 10,624 (22,438) 94 (198) 29,505 22,444France 422 (891) 1431 (3021) 45 (95) 6636 (14,014) 422 (891) 1431 (3021) 11 (24) 6636 20,000Germany 8735 (18,448) 4069 (8594) 232 (491) 7629 (16,113) 8735 (18,448) 4069 (8594) 58 (123) 7629 34,432Hungary 37 (79) 707 (1494) 12 (25) 3 (6) 37 (79) 707 (1494) 3 (6) 3 1248Italy 44 (92) 1996 (4216) 356 (753) 1365 (2882) 44 (92) 1996 (4216) 89 (188) 1365 1147Latvia 1997 (4217) 6 (13) 60 (127) 3 (6) 1997 (4217) 6 (13) 15 (32) 3 7950Netherlands 918 (1939) 1916 (4046) 45 (95) 310 (655) 918 (1939) 1916 (4046) 11 (24) 310 448Poland 563 (1188) 801 (1692) 271 (572) 2761 (5832) 563 (1188) 801 (1692) 68 (143) 2761 12,856Portugal 632 (1335) 1319 (2785) 45 (95) 5165 (10,908) 632 (1335) 1319 (2785) 11 (24) 5165 1890Slovakia 149 (314) 3 (7) 9 (19) 1614 (3408) 149 (314) 3 (7) 2 (5) 1614 4845Slovenia 41 (86) 94 (198) 0.3 (1) 405 (856) 41 (86) 94 (198) 0.1 (0.2) 405 1403Spain 677 (1429) 1412 (2982) 52 (109) 5228 (11,042) 677 (1429) 1412 (2982) 13 (27) 5228 7343Sweden 4647 (9814) 6901 (14,575) 1960 (4140) 32,086 (67,766) 4647 (9814) 6901 (14,575) 490 (1035) 32,086 58,200United Kingdom 2 (4) 1948 (4114) 3 (5) 904 (1909) 2 (4) 1948 (4114) 1 (2) 904 5059EU19 22,491 (47,500) 38,140 (80,553) 4416 (9329) 102,316 (216,104) 22,491 (47,500) 38,140 (80,553) 1105 (2333) 102,316 192,199

a Own calculations based on Svensk Fjarrvarme [44]; Wibe [49].b Own calculations based on Franco and Giannini [26].c Own calculations based on Energirådgivningen [17]; Bjerg et al. [6].d Own calculations based on the Finnish Forest Association [25].e FAOSTAT [24].

Table 5National indicative targets for share of renewable power.

Country RES-E TWh 1997 RES-E % 1997 RES-E % 2010

Austria 39.05 70 78.1Belgium 0.86 1.1 6Czech Republic 2.36 3.8 8Denmark 3.21 8.7 29Estonia 0.02 0.2 5.1Finland 19.03 24.7 31.5France 66.00 15 21Germany 24.91 4.5 12.5Hungary 0.22 0.7 3.6Italy 46.46 16 25Latvia 2.76 42.4 49.3Netherlands 3.45 3.5 9Poland 2.35 1.6 7.5Portugal 14.30 38.5 39Slovakia 5.09 17.9 31Slovenia 3.66 29.9 33.6Spain 37.15 19.9 29.4Sweden 72.03 49.1 60United Kingdom 7.04 1.7 10EU19 349.95 13.9 22

Sources: Ref. [16] and Hinsch [30].

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26 S23

demand less of this input, thus resulting in a decrease in the netimport quantity of logging residues as Fig. 6 indicates. Similararguments can be applied to all countries when analysing theresults of scenario 1.

The implementation of the RES-E Directive (scenario 2) affectsthe power sector. Hence, for all other industries the production willnot change. The RES-E Directive is applied on a national basis,implying that the power sector in each country has to increase itsproduction by a certain amount. With the RES-E implemented thelargest net exporter of by-products will be Finland, Austria andLatvia. Latvia is one of the countries using less by-products andmore logging residues in order to reach its goal. Hence, Latviawill be a net importer of logging residues and a net exporter ofby-products. In contrast, Sweden will use more by-products and

less logging residues, compared to the baseline scenario, resultingin a large increase in the net import of by-products and a high netexport level of logging residues. Thus, the Swedish power sectorwill use less logging residues in its production and can increase itsexport to other countries where the marginal product of loggingresidues is higher.

As for roundwood, only small volumes are being used by theenergy sector. The model simulations show that the roundwoodmarket will be left unaffected by an increased demand for by-products and logging residues. Hence, the increased demand forforest fuels by the energy sector will not have any explicit conse-quences for the roundwood market.

Implementing the EU energy policies will result in an increase inthe total trade volume for both by-products and logging residues.This is illustrated in Table 6 in which the change in the traderesulting from scenario 1 and scenario 2 is compared to the base-line scenario.

With the implementation of the White Paper the total trade ofby-products will increase with ten percent. In the baseline scenariothe total EU net export is 40 TWh. This amount increased to43.6 TWh in scenario 1. In addition, the total trade of loggingresidues increased by 14 percent. With the implementation of theRES-E Directive the trade increases even more. The total tradeincreases by 50 and 67 percent for by-products and logging resi-dues to 59.9 and 86.9 TWh, respectively.

Since the elasticities of substitution between the input factorsare important parameters in the EFTM, a sensitivity analysis hasbeen conducted to check how sensitive the results are to changes inthese parameters. In the analysis, the elasticities of substitution arechanged by �1 for the heat and power industries, �0.1 for thewoodfuel refining industry and �0.5 for the pulp industry. Thesensitivity results are reported in Table 7.

Decreasing the elasticities of substitution will decrease the netimport of by-products by 1 percent and logging residues by 2.Increasing the elasticities of substitution will increase the netimport of logging residues by 2 percent and decrease the net importof by-products by 2 percent. Overall, changing the elasticity of

-15000

-10000

-5000

05000

1000015000

Swed

en

Finl

and

Est

onia

Spai

n

Pola

nd

Aus

tria

Ger

man

y

Fran

ce

Slov

akia

Slov

enia

Port

ugal

Cze

chR

epub

lic

Den

mar

k

Bel

gium

Hun

gary

The

Net

herl

ands

Lat

via

Uni

ted

Kin

gdom Ital

y

By-products baseline By-products S1 By-products S2

(’00

0 m

3 solid

)

-30000-20000-10000

010000200003000040000

(GW

h)N

et im

port

of

indu

stri

al b

y-pr

oduc

ts

Fig. 5. Net import of industrial by-products by country under different scenarios.

-15,000

-10,000

-5,000

0

5,000

10,000

15,000

20,000

-30,000-20,000-10,000

010,00020,00030,00040,000 Sw

eden

Finl

and

Est

onia

Spai

n

Pola

nd

Aus

tria

Ger

man

y

Fran

ce

Slov

akia

Slov

enia

Port

ugal

Cze

chR

epub

lic

Den

mar

k

Bel

gium

Hun

gary

The

Net

herl

ands

Lat

via

Uni

ted

Kin

gdom Ital

y

By-products baseline By-products S1 By-products S2

Net

impo

rt o

f lo

ggin

g re

sidu

es

(GW

h)(’

000

m3 so

lid)

Fig. 6. Net import of logging residues by country under different scenarios.

Table 6Percentage change in total trade of by-products and logging residues compared tothe baseline scenario.

Input Scenario 1 Scenario 2

By-products þ10 þ50Logging residues þ14 þ67

Table 7Percentage change in net import with different elasticities of substitution.

Variable By-products Logging residues

Net import, lower substitution elasticity �1 �2Net import, higher substitution elasticity �2 þ2

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26S24

substitution only produces relatively small changes suggestingreasonably robust results.

5. Discussion and conclusions

The development of an EU forest-based biomass market isa wide and complex issue. There are several institutional obstacles,in addition to the more economic issues, that currently act asa restriction to the trade. However, the trade development has been

accelerated by regional imbalances of resources and demand. Thisstudy is an attempt to assess how EU trade levels of specific forestfuels are affected by different EU energy policies. Two specific EUpolicies are assessed: the full implementation of the White Paperand RES-E. Both the White Paper and RES-E Directive implicitlysupport an increasing use of forest-based biomass for energypurposes.

Although partial equilibrium models do not account for asmany linkages between product groups as computable generalequilibrium (CGE) models do, they can provide a transparent andfocused trade analysis of how a limited number of forest prod-ucts are affected by various policies. For instance, a limit of theEuropean forest trade model outlined in this study does notaccount for linkages between the forest product markets and thecapital and labour markets. The model explicitly assumes that theprice of both labour and capital is fixed and unaffected by anychange in the endogenous variables. However, bearing this inmind, the results can still be interpreted with some degree ofconfidence.

A general conclusion it that an increasing level of trade inforest-based biomass might mitigate the transitional problem ofthe EU energy system from fossil fuel based to renewable energysources. Generally, the implementation of the RES-E Directive hasa larger affect on trade volumes than the implementation theWhite Paper. The implementation of the RES-E Directive resultsin an increase of total trade in logging residues and by-products

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26 S25

by 50 and 67 percent, respectively. The corresponding increase intrade volumes from the implementation of the White Paper is 10and 14 percent for by-products and logging residues, respectively.

Depending on implemented policy the results differ. A countrythat was net importing by-products given the White Paper imple-mentation can instead be net exporting by-products when imple-menting the RES-E Directive. Hence, a country can transform froma net exporter to a net importers (and vice versa) between thescenarios.

The scenarios reinforce the picture of the EU forest fuel marketsas a complex and multi-layer subject. There exist many differentand credible alternative future scenarios that the markets candevelop towards. The results also open up new discussions andprovide information for the purposes of policy makers. For firms,the scenarios provide knowledge that can be utilised in strategicdecision making. Thus, studying European trade in forest fuels isimportant for understanding how different industry sectors will beaffected by the EU policies aimed at promoting the use of renew-able energy sources.

Acknowledgements

The authors would like to thank the Kempe Foundations and theSwedish Energy Agency for their generous financial support.

Appendix

The calibrated CES cost functions for the heat, power and refinewoodfuel industries is expressed as:

Ci;r ¼yi;r

yi;rCi;r

"qi;r

�Pn

r

Pnr

�1�si;r

þð1� qi;rÞ�

Pmr

Pmr

�1�si;r# 1

1�si;r

(A1)

where Ci,r is the total cost for industry i in country r, y is theproduction level, P is the input price for input n and m, s is theelasticity of substitution and q is the input value share. Therespective variables benchmark values are denoted as C, y and P.The input value share can be expressed as:

qi;r ¼Pn

r Dni;r

Pnr Dn

i;r þ Pmr Dm

i;r

(A2)

where D is the input demand and D is benchmark value. The cali-brated demand functions can be expressed as:

Dni;r ¼ Dn

i;ryi;r

yi;r

Pn

r ci;r

Pnr ci;r

!si;r

(A3)

Where c is the unit cost, c its benchmark value and they areexpressed respectively as:

ci;r ¼ ci;r

"qi;r

�Pn

r

Pnr

�1�si;r

þð1� qi;rÞ�

Pmr

Pmr

�1�si;r# 1

1�si;r

(A4)

ci;r ¼

Pnr Dn

i;r þ Pmr Dm

i;r

yi;r

!(A5)

Finally, the production level is expressed as:

yi;r ¼ yi;r

"qi;r

Dn

i;r

Dni;r

!ri;r

þð1� qi;rÞ

Dmi;r

Dmi;r

!ri;r# 1

ri;r

(A6)

where ri;r ¼ ðsi;r � 1Þ=si;r . The optimisation problem can then beformulated as:

Min CTotEU ¼

Xi

Xr

Ci;ryi;r (A7)

with the restrictions that domestically available inputs are usedbefore import startsX

i

Dnr;i ¼ Sn

r þ NMnr (A8)

and that no trade outside the EU is allowedXr

NMnr ¼ 0 (A9)

where S is the benchmark value of input supply and NM is the netimport. Since no trade is allowed outside the EU, the total EU supplyof inputs has to equal total EU demand. Expressed as equations, the‘‘union security-of-supply’’ condition can be written as:X

i

Xr

Dnr;i ¼

Xr

Snr (A10)

In the baseline scenario each industry sector must produceexactly the benchmark production level, hence yi;r ¼ yi;r , indi-cating that the ratio yi;r=yi;r will equal unity. In contrast, yi;r isallowed to vary in scenario 1 under the condition thatP

iyi;r ¼ 1:32

Pi

yi;r for i¼ heat, power and RWF. This suggests thateach country must increase its total utilisation of forest-basedbiomass by 32 percent. In scenario 2, all industry sectors ina country, besides the power industry, must produce their respec-tive benchmark production levels. For the power industry theforest-based biomass power generation must increase witha certain percentage. Hence, the yi;r=yi;r ratios are fixed, for the heatand refined woodfuel industries it will equal unity whilst thenational targets stated in the RES-E Directive will determine theratio for the power industry. In all scenarios the pulp industry mustproduce exactly the benchmark production level, hence yr ¼ yr ,indicating that the ratio yi;r=yi;r equal unity.

References

[1] Adams DM, Haynes RW. The 1980 softwood timber assessment market model:structure, projections, and policy simulations. Forest Science Monograph 22;1980.

[2] Alakangas E, Vesterinen P. Export–import possibilities and fuel prices – task 2.ALTENER AFB-net V. Jyvaskyla, Finland: VTT Energy; 2001.

[3] Alakangas E, Vesterinen P. EUBIONET biomass survey in Europe – summaryreport. Jyvaskyla, Finland: VTT Processes; 2003.

[4] Alavalapati JRR, Adomowicz WL, White WA. Random variables in forest policy:a systematic sensitivity analysis using CGE models. Journal of ForestEconomics 1999;5(2):143–69.

[5] Binkley CS, Percy M, Thompson WA, Vertinsky IB. A general equilibriumanalysis of a reduction in harvest levels in British Columbia. The ForestryChronicle 1994;70(4):449–54.

[6] Bjerg J, Fjallstrom T, Grassi G, Janssen R, Karapanagiotis S, Nilsson N, et al.Pellets in southern Europe. The state of the art of pellets utilisation insouthern Europe. New perspecitves of pellets from agri-residues. In: 2ndworld conference on biomass for energy, industry and climate protection,Rome, Italy; 2004.

[7] Bohringer C, Boeters S, Feil M. Taxation and unemployment: an appliedgeneral equilibrium approach. Economic Modelling 2005;22:81–108.

[8] Bohringer C, Rutherford TF, Wiegard W. Computable general equilibriumanalysis: opening a black box; 2003. ZEW discussion paper 03-56, Mannheim.

[9] Bonnefoi B, Buongiorno J. Comparative advantage of countries in forest-products trade. Forest Ecology and Management 1990;36:1–17.

[10] Boyd R, Krutilla K. The impact of the free trade agreement on the U.S. forestrysector: a general equilibrium analysis. In: Nemetz P, editor. Emerging issues inforest policy. Vancouver: University of British Columbia Press; 1992.

[11] Boyd R, Doroodian K, Abdul-Latif S. The effects of tariff removals on theNorth American lumber trade. Canadian Journal of Agricultural Economics1993;41:311–28.

R. Lundmark, A. Mansikkasalo / Journal of Cleaner Production 17 (2009) S18–S26S26

[12] Bredstrom D, Carlsson D, Lundgren J, Mason A, Ronnqvist M. Supply chainoptimization in the pulp mill industry – IP models, column generation andnovel constraint branches. European Journal of Operational Research2004;156:2–22.

[13] Bruce IA. Forestry and wood-based industries in the Australian economy:a computable general equilibrium analysis. Australian Forestry1988;51(4):238–45.

[14] Buongiorno J, Gilless JK. A model of international trade of forest products, withan application to newsprint. Journal of World Forest Resource Management1984;1:65–80.

[15] Buongiorno J, Tomberlin D, Turner J, Zhang D, Zhu S. The global forest productsmodel. USA: Academic press; 2003.

[16] Directive 2001/77/EC of the European Parliament and of the Council of 27September 2001 on the promotion of electricity produced from renewableenergy sources in the internal electricity market. Official Journal of theEuropean Communities, L 283/39, 2001.

[17] Energirådgivningen. Pellets, <www.energiradgivningen.se>; 2006.[19] Ericsson K, Nilsson L. International biofuel trade – a study of the Swedish

import. Biomass and Bioenergy 2004;26(3):205–20.[20] EUBIONET II. WP1: current situation and future trends in biomass fuel trade in

Europe – country reports, <www.eubionet.net>; 2006.[21] European Commission. Communication from the Commission. Energy for the

future: renewable sources of energy. White Paper for a Community Strategyand Action Plan. European Commission COM(97)599 final; 1997.

[22] European Pellet Centre, <www.pelletcentre.info>; 2006.[23] EUROSTAT database, <www.europa.eu.int/comm/eurostat/>; 2006.[24] FAO – Food and agriculture organization of the United Nation. FAOSTAT

forestry, <www.fao.org>; 2006.[25] Finnish Forest Association. Produkter av tra, <www.smy.fi>; 1999.[26] Franco A, Giannini N. Perspectives for the use of biomass as fuel in

combined cycle power plants. International Journal of Thermal Sciences2005;44:163–77.

[27] Frankel JA. The regionalization of the world economy. In: National Bureau ofEconomic Research Project Report Series. Chicago: University of ChicagoPress; 1998.

[28] Gilless JK, Buongiorno J. PAPYRUS: a model of the North American pulp andpaper industry. Forest Science Monographs, 28; 1987.

[29] Hillring B. World trade in forest products and wood fuel. Biomass and Bio-energy 2006;30:815–25.

[30] Hinsch C. Going forward. New Energy 2003:36–8.[31] Hughes E. Biomass cofiring: economics, policy and opportunities. Biomass and

Bioenergy 2000;19(6):457–65.[32] Jianbang G, Smith CT. Availability of logging residues and potential for elec-

tricity production and carbon displacement in the USA. Biomass and Bio-energy 2006;30(12):1011–20.

[33] Kallio M, Dykstra DP, Binkley CS. The global forest sector: an analyticalperspective. New York, USA: Wiley; 1987.

[34] Karlsson M, Klugman S, Moshfegh B. A Scandinavian chemical wood-pulp mill.Part 2. International and model mills comparison. Applied Energy2007;84(3):340–50.

[35] Leamer EE. Sources of international comparative advantage. Boston, MA: MITPress; 1984.

[36] Lin CI. Processing and exporting raw materials: a computable general equi-librium analysis of the Indonesian log and plywood industries. CanadianJournal of Development Studies 1996;17(1):7–30.

[37] Lundmark R. Dependencies between forest products sectors: a partial equi-librium analysis. Forest Products Journal 2007;57(9):79–86.

[38] Lundmark R, Soderholm P. Brannhett om svensk skog – en studie om råvar-ukonkurrensens ekonomi. Stockholm, Sweden: SNS forlag; 2004.

[39] Passalacqua F, Janssen R, Grassi G, Sandovar A, Tsoutsos T, Fjallstrom T, et al.Pellets in southern Europe. The state of the art of pellets utilisation insouthern Europe. New perspectives of pellets from agri-residues. In: 2ndworld conference on biomass for energy, industry and climate protection, 10–14 May, Rome, Italy; 2004.

[40] Prestemon JP, Buongiorno J. Comparative advantage in U.S. interstate forestproduct trade. Journal of Forest Economics 1997;3(3):207–28.

[41] Rehn M. Technology in the pulp and paper industry – empirical studies ofscale economies, productivity growth and substitution possibilities. Disserta-tion 20, Swedish University of Agricultural Sciences, Department of ForestEconomics; 1995.

[42] Samuelson PA. Spatial price equilibrium and linear programming. AmericanEconomic Review 1952;42(3):283–303.

[43] Sisask A. En studie av fjarrvarmeverkens produktionsteknologi. Umeå,Sweden: Department of Forest Economics, Swedish University of AgriculturalSciences; 2001. Report 125.

[44] Svensk Fjarrvarme. Fjarrvarme – och kraftvarmestatistik 2004. Stockholm,Sweden: Svensk Fjarrvarme; 2004.

[45] Swedish Forest Agency. Swedish statistical yearbook of forestry 2006. Jon-koping, Sweden: Sweden’s official statistics; 2006.

[46] Takayama T, Judge GG. Spatial and temporal price and allocation models.Amsterdam, The Netherlands: North-Holland Publishing; 1971.

[47] Vanek J. The natural resource content of foreign trade, 1870–1955, and therelative abundance of natural resources in the United States. Review ofEconomics and Statistics 1959;41:146–53.

[48] Wahlund B, Yan J, Westermark M. A total energy system of fuel upgrading bydrying biomass feedstock for cogeneration: a case study of Skellefteå bio-energy combine. Biomass and Bioenergy 2002;23:271–81.

[49] Wibe S. Substitution och teknologisk forandring inom den svenska fjarrvar-mesektorn. Umeå, Sweden: Department of Forest Economics, SwedishUniversity of Agricultural Sciences; 2001. Report 124.