Cost-effective policy instruments for greenhouse gas emission reduction and fossil fuel substitution...

Energy Policy 39 (2011) 3261–3280

Contents lists available at ScienceDirect

Energy Policy

0301-42

doi:10.1

n Corr

sity of N

Austria.

E-m

journal homepage: www.elsevier.com/locate/enpol

Cost-effective policy instruments for greenhouse gas emission reduction andfossil fuel substitution through bioenergy production in Austria

Johannes Schmidt a,n, Sylvain Leduc b, Erik Dotzauer c, Erwin Schmid a

a Institute for Sustainable Economic Development, University of Natural Resources and Life Sciences, Peter Jordan Straße 82, A-1190 Vienna, Austriab International Institute for Applied Systems Analysis, Schlossplatz 1, A-2361 Laxenburg, Austriac Malardalen University, P.O. Box 883, SE-72123 Vasteras, Sweden

a r t i c l e i n f o

Article history:

Received 9 December 2009

Accepted 3 March 2011Available online 7 April 2011

Keywords:

Bioenergy policy

Bioenergy with carbon capture and storage

Spatially explicit modeling

15/$ - see front matter & 2011 Elsevier Ltd. A

016/j.enpol.2011.03.018

espondence to: Institute for Sustainable Econ

atural Ressources and Life Sciences, Feistma

ail address: [email protected] (J. S

a b s t r a c t

Climate change mitigation and security of energy supply are important targets of Austrian energy

policy. Bioenergy production based on resources from agriculture and forestry is an important option

for attaining these targets. To increase the share of bioenergy in the energy supply, supporting policy

instruments are necessary. The cost-effectiveness of these instruments in attaining policy targets

depends on the availability of bioenergy technologies. Advanced technologies such as second-

generation biofuels, biomass gasification for power production, and bioenergy with carbon capture

and storage (BECCS) will likely change the performance of policy instruments. This article assesses the

cost-effectiveness of energy policy instruments, considering new bioenergy technologies for the year

2030, with respect to greenhouse gas emission (GHG) reduction and fossil fuel substitution. Instru-

ments that directly subsidize bioenergy are compared with instruments that aim at reducing GHG

emissions. A spatially explicit modeling approach is used to account for biomass supply and energy

distribution costs in Austria. Results indicate that a carbon tax performs cost-effectively with respect to

both policy targets if BECCS is not available. However, the availability of BECCS creates a trade-off

between GHG emission reduction and fossil fuel substitution. Biofuel blending obligations are costly in

terms of attaining the policy targets.

& 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Climate change mitigation and security of energy supply areamong the main drivers of current European energy policies (COM,2010). Austria committed to a 13% reduction in greenhouse gas(GHG) emissions with respect to the 1990 reference year for theKyoto commitment period of 2008–2012 (Council, 2002) and to anincrease in renewable energy production by 11 percentage pointsuntil 2020 under EU directive 2009/28/EC. Currently, Austria is farfrom reaching the Kyoto target (Umweltbundesamt Austria, 2010),and significant efforts will be necessary to meet the 2020 energytargets (Nakicenovic et al., 2008). More forestry biomass utilization(Schadauer, 2009) and agricultural biomass production are stillpossible and necessary if Austria is to reach the ambitious climatechange mitigation and renewable energy targets. Particularly after2020, global GHG emission reductions will need to be considerablyhigher than the Kyoto targets to limit the maximum warming of theglobal mean temperature to 2 1C (Roeckner et al., 2010).

ll rights reserved.

omic Development, Univer-

ntelstraße 4, A-1180 Vienna,

chmidt).

1.1. Policy instruments

Several policy instruments are in place or under discussionto facilitate the achievement of these policy targets in Austria.This paper focuses on bioenergy as a climate change mitigationoption and discusses policies that directly or indirectly supportbioenergy production in Austria. Policies that are introduced toreduce GHG emissions and thus indirectly promote bioenergy,including a CO2 tax and the EU Emission Trading Scheme(ETS), are compared to policies that directly support bioenergyproduction, including biofuel blending obligations, feed-intariffs for biomass power production, and subsidies for biomassfurnaces.

Policy instruments that are designed to directly reduce GHGemissions are expected to be cost-effective in reducing GHGemissions if they allow an efficient allocation of reduction effortsamong market participants and technologies. Even in that case,cost-effectiveness with respect to fossil fuel substitution may below because GHG emission reduction does not necessarily corre-spond with fossil fuel substitution (Schmidt et al., 2010a). More-over, long-term technological development may not be triggeredby these policies because in the main, technologies are chosenthat are close to being competitive on the market (Sanden and

www.elsevier.com/locate/enpol

dx.doi.org/10.1016/j.enpol.2011.03.018

mailto:[email protected]

dx.doi.org/10.1016/j.enpol.2010.03.030

Nomenclature

A Biomass supply costs (h)aagrar

i,sc,pl Decision variable for separable programming foragricultural biomass (dimensionless)

bi,o,u Decision variable for separable programming forforestry biomass (dimensionless)

bdirecti,k,b Biomass transported directly to consumers (e.g., fuel-

wood) (MWhbiomass)bplant

i,j,b,l,t Biomass transported to plants (MWhbiomass)emtaxed CO2 emissions that may be taxed by policy (all

emissions without agriculture) (tCO2)p Price of wood supply (h MWh�1)q Quantity of wood supply (MWh)qbio

j,t Heat production in the plant (MWh)

qdhj,h,ns,t Heat transportation from plant to district heating

network (MWh)

qdhfh,ns,t Fossil district heat production (MWh)totem

qgash,ns Amount of gas used for heating (MWh)

totem Total GHG emissions in tCO2 equivalents (tCO2e)ugnet

h,ns Binary variable for investment in gas network(dimensionless)

udneth,ns Binary variable for investment in district heating

network (dimensionless)upipe

j,h,ns Binary variable for investment in transportation pipe-line (dimensionless)

uplantj,l Binary variable for plant investment (dimensionless)

zbioj,c Amount of energy commodity produced in a

plant (MWh)zbio

j,k,c,t Amount of energy commodity transported to consu-mers in each period (MWh)

zfossilk,f Amount of fossil fuel used to satisfy demand (MWh)

Parameters

e,eo Elasticity of biomass supply (dimensionless)Zbio

c,d Efficiency of converting a bioenergy commodity touseful energy (dimensionless)

Zbiodb,d Efficiency of converting biomass directly to useful

energy (e.g., fuelwood) (dimensionless)Zconv

j,l,c Conversion efficiency in bioenergy plants (dimensionless)

Zdhh,t Efficiency of distributing heat in district heating net-

work (dimensionless)Zdhf

h,t Efficiency of fossil district heating boiler (dimensionless)

Zfossilf ,d Efficiency of converting fossil fuel to useful energy

(dimensionless)Zgas Efficiency of converting natural gas to heat

(dimensionless)Zheat

j,l Heat efficiency in bioenergy plants (dimensionless)

Ztransj,h,ns,t Transportation efficiency of heat pipeline

(dimensionless)bi,o Maximum sustainable yield (MSY) in supply cell

(MWh y�1)bj,l,t Production capacity of plant (MWh)bfs,o MSY in federal state (MWh y�1)C Constant in wood supply function (dimensionless)cccs Costs of carbon capture and storage (h tCO2

�1)cdhf

t Costs of fossil districting heat fuel (h MWhfuel�1 )

cdneth,ns Annualized costs of investment in district heating

network (h y�1)cem CO2 price (h tCO2

�1)cfossil

f Price of fossil fuel (h MWhfuel�1 )

cfossilinvf Investment costs necessary at consumer for fossil

technology (h MWhfuel�1 )

cgast Costs of natural gas (h MWhgas

�1)

cgneth,ns Costs of natural gas (h MWhgas

�1)

cinvc Investment cost necessary at consumer for bioenergy

technology (h MWhbiomass�1 )

cinvdk,b Investment cost necessary at consumer for direct

biomass consumption (e.g., fuelwood) (h MWhbiomass�1 )

cpipej,h,ns Annualized costs of investment in transportation

pipeline (h y�1)cplant

j,l Annualized costs of plant investment (h y�1)

cprodj,l Variable costs of bioenergy production (h MWhbiomass

�1 )

csupi Costs of biomass supply (h MWhbiomass

�1 )

ctransbi,j,b Costs of biomass transportation from i to j

(h MWhbiomass�1 )

ctranscj,k,c Costs of transporting commodity from j to k

(h MWh�1)ctransd

i,k,b Costs of biomass transportation from i to k

(h MWhbiomass�1 )

dk,d Energy demand (MWh y�1)

eagrarsc,b,pl GHG emissions in tCO2 equivalents from fertilizer

application in agricultural production(tCO2e MWhbiomass

�1 )eccs

l,b Carbon capture rate in plant (tCO2 MWhbiomass�1 )

edhf CO2 emission factor of fossil district heating fuel(tCO2 MWhfuel

�1 )egas CO2 emission factor of natural gas (tCO2 MWhgas

�1)efossil

f CO2 emission factor of fossil fuels (tCO2 MWh�1)

etransi,j,b CO2 emission factor of biomass transportation from i

to j (tCO2 MWhbiomass�1 )

etransi,k,b CO2 emission factor of commodity transportation

from i to k (tCO2 MWhbiomass�1 )

etransj,k,c CO2 emission factor of commodity transportation

(tCO2 MWhcommodity�1 )

f blendc,d Mandatory share of bioenergy commodity in final

energy demand (dimensionless)p Price of forest wood (h MWh�1)pagrar

i,sc,b,pl Price of agricultural products (h MWh�1)

qagrari,sc,b,pl Amount of agricultural products (MWh)

q� Supply of wood harvests (MWh)q�u Quantities for separable programming (dimensionless)qqi,o,qfs,o Observed quantities of wood harvests total, in

supply cell, in federal state (MWh y�1)

qDh,t Heat demand in settlement (MWh season�1)

qDh,ns,t Heat demand in district heating network (MWh

season�1)qpipe

ns,t Capacity of heat transportation pipeline (MWhseason�1)

txs,txff Binary parameters controlling which fossil fuels are

priced (dimensionless)

Subscripts

b Feedstock (dimensionless)c Energy commodity (dimensionless)d Energy demand for useful energy (dimensionless)f Fossil fuels (dimensionless)fs Federal state (dimensionless)h Settlements supplied by heat networks (dimensionless)i Biomass supply site (dimensionless)j Plant location (dimensionless)

J. Schmidt et al. / Energy Policy 39 (2011) 3261–32803262

k Demand region (dimensionless)l Technology and size of bioenergy plant (dimensionless)ns District heating network size (dimensionless)o Forest ownership (dimensionless)

pl Price level of agricultural biomass (dimensionless)sc Price scenario for agricultural biomass (dimensionless)t Season (dimensionless)u Index for separable programming (dimensionless)

J. Schmidt et al. / Energy Policy 39 (2011) 3261–3280 3263

Azar, 2005). The following GHG emission reduction policies areconsidered in our analysis:

�
A CO2 tax is currently under discussion in the European Unionand is proposed as one measure of financing the EU budget. It isapplied to all fossil fuels before combustion according to thecarbon content of the respective fuel. Denmark, Finland, Italy,Ireland, and Sweden are EU members that have already imple-mented a CO2 tax. However, none of these countries apply auniform tax on all fossil fuel consumers and exempt some sectorsfrom the tax. Tax rates vary from 10 to 150 h tCO2
�1 (Pope andOwen, 2009). As marginal abatement costs are not known withcertainty, the total amount of GHG emission abatement cannot bedetermined a priori when a tax is being introduced. Bioenergywith carbon capture and storage (BECCS) producers are notdirectly affected by a CO2 tax because they do not consume fossilfuels. To give an incentive for installing CCS in bioenergy plants, abonus equal to the level of the CO2 tax would have to beintroduced to pay for negative CO2 emissions from BECCS (Ricci,2010).
� The EU ETS was introduced in 2005. About 250 facilities that
emit around one-third of total GHG emissions were affected inAustria. CO2 emission permits were allocated for free in thestart-up phase by applying a grandfathering principle. The CO2

emission allowances can be traded on a market to attainefficient allocations. The second phase of the EU ETS is from2008 to 2012. In Austria, the main differences from the firstphase (from 2005 to 2008) are a slightly lower cap (reductionfrom 33.19 to 32.8 MtCO2) and the auctioning of 1.2% of theallowances in the initial phase of the scheme (Federal Ministryof Agriculture, Forestry, Environment and Water Management,2007; Paoletta and Taschini, 2006). As GHG emissions frombiomass are assumed to be carbon-neutral, no ETS allowanceshave to be acquired for biomass combustion.

Direct promotion of specific bioenergy technologies isexpected to have low cost-effectiveness because only a subsetof available technologies is usually subsidized. However, techno-logical development may be triggered for these technologies,which will make them competitive in the long term (Sandenand Azar, 2005). We include the following policies in our analysis:

�
Feed-in tariffs for different forms of renewable electricity pro-duction such as bioenergy power plants, small water powerplants, and photovoltaic power are defined in the Austrianrenewable energy law. Facilities that are included in the supportscheme receive feed-in tariffs for 12 years. Feed-in tariffs arechosen to be close to the production costs of specific technolo-gies and therefore depend on the respective technology. � The EU directive 2003/30/EC sets an indicative target for all
member states to blend a share of 5.75% of biofuels withgasoline until 31st December 2010. Austria ratified this law in2003. In 2009 the EU issued directive 2009/28/EC, whichdemands, among other targets, a share of 10% of renewableenergies in the transportation sector. In contrast to directive2003/30/EC, this regulation does not dictate the utilization ofbiofuels. Electric cars or public transportation relying on renew-ably produced electricity also come within the ambit of thedirective. However, biofuels are one major technological option

for attaining the targets in the short term (Sandy Thomas,2009).
� Subsidies are granted for wood heating systems by all federal
states in Austria. These subsidies are granted if a heating systemis installed in a new building or if an old heating system isreplaced. Technological development is not the main aim of thispolicy because heating furnaces are technically mature andsignificant technological improvements cannot be expected.

While each of the policy instruments presented work differently,their overall targets are the same: they are implemented to reduceGHG emissions and shift energy production from fossil fuels torenewable energies. The cost-effectiveness (i.e., the costs of attainingthe policy targets) is likely to vary between the policy instruments.The cost-effectiveness of policies that aim at increasing the produc-tion of certain bioenergy commodities such as biomass power (feed-in tariffs), biofuels (biofuel targets), or biomass heat (pellet sub-sidies) depends directly on the performance and costs of bioenergyproduction technologies. While the cost-effectiveness of non-tech-nology specific policies, i.e., the carbon tax and the ETS, depends onall available low carbon technologies, bioenergy technologies willhave a major influence because bioenergy production will mostlikely have to contribute a significant share of GHG emissionreduction and renewable energy production in Austria (Kalt et al.,2010; Nakicenovic et al., 2008; Steininger and Voraberger, 2003).We restrict our analysis to bioenergy technologies therefore. Highlyefficient power production with Biomass Integrated Combined CyclePlants (BIGCC) in the power sector and second-generation biofuelproduction based on woody biomass in the transportation sector aretechnologies that are expected to increase bioenergy productivitysignificantly (Havlık et al., 2010; Lange, 2007; Marbe et al., 2004;Steenhof and McInnis, 2008). In addition, BECCS is considered to bea possible major contributor to future climate mitigation efforts(Azar et al., 2010; Luckow et al., 2010; Katofsky et al., 2010).However, it is not yet clear how much these technologies cancontribute to carbon mitigation and renewable energy targets andhow the respective policy instruments will perform in terms ofattaining these policy targets at least cost, if the technologies inquestion become available.

1.2. Challenges in bioenergy modeling

The relative performance of the policy instruments discussednot only relies on technological development but also on thespatial variation of demand and supply. We have identified thefollowing important factors: (i) energy distribution costs, parti-cularly for district heating, depend on heat demand densities andon the distance from plant to demand (Grohnheit and Mortensen,2003; Ivezic et al., 2008; Konstantin, 2007); (ii) bioenergytechnologies rely on different types of feedstock, for example,starchy energy crops for first-generation biofuels versus shortrotation lignocellulose or forest wood for second-generationbiofuels. The availability and the costs of different types ofbiomass depend on the biophysical characteristics of the landand on current agricultural production systems; (iii) transporta-tion costs have a significant impact on the final cost of thebioenergy supply chain and vary according to technology, plantsize, and the distance from plant to production sites (Eriksson andBjorheden, 1989; Luckow et al., 2010; Richard, 2010). Spatial

ply

mo

de

lin

gD

em

an

dm

od

eli

ng

Te

chn

olo

gie

s/p

oli

cie

sse

lect

ed

for

cost

-e

ffe

ctiv

eG

HG

em

issi

on

red

uct

ion

tio

ns

on

tota

la

va

ila

bil

ity

of

bio

ma

ssS

cen

ari

os

Bio

ma

ssb

ase

dh

ea

ta

nd

po

we

rp

rod

uct

ion

,

bio

fue

lsp

lay

min

or

role

tio

ns

on

tota

la

va

ila

bil

ity

of

bio

ma

ssS

cen

ari

os

Bio

fue

lp

oli

cyco

stly

;ca

rbo

nta

xp

refe

rsh

ea

ta

nd

po

we

rp

rod

uct

ion

tio

ns

on

tota

la

va

ila

bil

ity

of

bio

ma

ssS

cen

ari

os

Ma

inly

he

at

an

dfu

el

pro

du

ctio

nse

lect

ed

;n

o

po

we

rp

rod

uct

ion

s,a

nd

sho

ws

tra

de

-off

be

twe

en

oo

dp

rod

uct

ion

De

ma

nd

curv

es

Wit

ho

ut

CC

S:

fue

l,p

ow

er

an

dh

ea

tp

rod

uct

ion

;

wit

hC

CS

:m

ain

lyp

ow

er

pro

du

ctio

n

s,b

ut

do

es

no

tsh

ow

de

clin

ein

foo

dD

em

an

dcu

rve

sN

od

eta

ile

da

sse

ssm

en

to

fb

ioe

ne

rgy

tech

no

log

ies

tech

nic

al

bio

ma

ssp

ote

nti

al

Bo

tto

m-u

pm

od

el

Po

we

r

lan

dp

rod

uct

ivit

y,

an

dsh

ow

str

ad

e-o

ff

rgy

an

dfo

od

pro

du

ctio

n

De

ma

nd

no

tco

nsi

de

red

On

lyb

iofu

els

ass

ess

ed

tio

ns

on

cost

sa

nd

av

ail

ab

ilit

yo

fS

cen

ari

os—

Bo

tto

m-u

pH

ea

ta

nd

po

we

r,b

iofu

el

po

licy

cost

ly


variation of biomass supply and heat demand is high in Austria(Schmidt et al., 2010b). The location of plants in relation tofeedstock supply and energy demand is therefore relevant.Modeling efforts that spatially and explicitly integrate biophysicaland logistical constraints in bioenergy supply chains, as well asfacility and infrastructure constraints in energy demand, mayincrease the reliability of modeling results. Additionally, feedstockproduction on agricultural land competes with food and feedproduction. The trade-offs between food and energy crop produc-tion should therefore be made explicit in bioenergy models(Bryan et al., 2010).

Previous research on this topic made a comparative assessmentof bioenergy policies. Global, continental, and regional studies havebeen conducted using different methodologies. They producedifferent results and come to different conclusions regardingoptimal application of technologies and policies (see Table 1).However, a regional spatially explicit bioenergy model that includesfeedstock production considers competition for land use, modelsenergy production technologies, and distribution systems in detail,and is applied to analyze the cost-effectiveness of policy instru-ments, which has not yet been published in the scientific literature.This article presents such a model to analyze the cost-effectivenessof two classes of energy policy instruments with respect to GHGemission reduction and fossil fuel substitution in Austria. Instru-ments that are designed to reduce CO2 emissions, including a CO2

tax and the ETS, and technology-specific instruments for thepromotion of bioenergy, including feed-in tariffs for biomass powerplants, biofuel policies, and pellet subsidies, are assessed. The trade-offs between the two policy targets, assuming the availability ofnew bioenergy technologies, are analyzed. Bioenergy technologiesare modeled explicitly while other low carbon and renewableenergy technologies are implicitly considered by adjusting demand.

The article is structured as follows: Section 2 introduces themodel and data, as well as the bioenergy technologies included inthe assessment, and describes the policy scenarios. Section 3presents the results including costs, GHG emission reductions,and fossil fuel substitution under the policy scenarios. Section 4discusses the article with respect to the results and derives majorconclusions. Model details can be found in the Appendix.

Ta

ble

1S

tud

ies

on

cost

so

fb

ioe

ne

rgy

tech

no

log

ies

an

dp

oli

cyin

stru

me

nts

.

Art

icle

Mo

de

lin

ga

pp

roa

chB

iom

ass

sup

Glo

ba

la

nd

con

tin

en

tal

mo

de

lsA

zar

et

al.

(20

03

)O

pti

miz

ati

on

mo

de

lFi

xe

da

ssu

mp

Be

rnd

es

an

dH

an

sso

n(2

00

7)

Op

tim

iza

tio

nm

od

el

Fix

ed

ass

um

p

Gie

len

et

al.

(20

03

)O

pti

miz

ati

on

mo

de

lFi

xe

da

ssu

mp

Luck

ow

et

al.

(20

10

)P

art

ial

eq

uil

ibri

um

mo

de

lS

up

ply

curv

e

en

erg

ya

nd

f

Ma

gn

ee

ta

l.(2

01

0)

Co

mb

ine

db

ott

om

-up

an

dto

p

do

wn

mo

de

l

Su

pp

lycu

rve

pro

du

ctio

n

Re

gio

na

lm

od

els

Alf

on

soe

ta

l.(2

00

9)

Op

tim

iza

tio

nm

od

el

Mo

de

lin

go

f

Bry

an

et

al.

(20

10

)C

ost

calc

ula

tio

ns

Mo

de

lin

go

f

be

twe

en

en

e

Ko

nig

(20

11

)O

pti

miz

ati

on

mo

de

lFi

xe

da

ssu

mp

bio

ma

ss

2. Data and methodology

2.1. A techno-economic spatially explicit model

A spatially explicit, techno-economic mixed integer program(MIP) is developed and applied to assess the cost-effectiveness ofdifferent policy instruments in attaining the policy targets ofreducing GHG emissions and substituting fossil fuels in Austria.The model minimizes the costs of supplying Austria with transpor-tation fuels, heat, and electricity from either bioenergy or fossil fuels.It is static and simulates 1 year of operation. The year is split intotwo heating seasons to consider the differences in heat demandbetween winter and summer. The current model version considersdomestic biomass supply and energy demand and does not allowimports and exports of biomass or bioenergy commodities. Themodel determines which bioenergy plants (i.e., pellets, first-genera-tion ethanol or biodiesel, second-generation methanol, BIGCC orBECCS, and heating) of a specific size and specific location will bebuilt and which demand regions will be supplied with bioenergyand/or fossil fuels. Direct delivery of fuelwood from forest produc-tion sites to demand regions is possible. Each plant produces variousenergy commodities (Fig. 1). They replace fossil fuels in heating,power generation, and transportation. It is assumed that pellets andfuelwood are burnt in boilers of households or community heatingnetworks, power is transmitted to the national grid, surplus heat is

Forest

Biomass

Transportation

fuel demand

Single building

heating demand

Gasoline

Power production

(Steam, BIGCC with

and without CCS)

Power demand

Network bound

heat demand

Fossil power

(Fuel mix gas

and coal)

Gas in

single-

dwellings,

gas fired

district

heating

Heating

oil

Methanol plant

(Second generation,

with and without CCS)

Agricultural

Biomass

Pellets plant

Ethanol plant

(First generation, with

and without CCS)

Heating plant

Biodiesel plant

(First generation)

Community

district heating

demand

Fig. 1. Diagram of the mixed integer programming model.


delivered to district heating networks, and biofuels replace gasolinefor transportation purposes. The objective function is minimized andconsists of the costs of biomass supply from forestry and agriculture,biomass transportation (i.e., transportation of energy crops andforest biomass from biomass supply sites to plants and of fuelwoodfrom biomass supply sites to consumers), plant investment annu-ities, district heating infrastructure annuities, investment annuitiesfor heating furnaces, CCS costs, commodity transportation (i.e.,transportation of pellets and transportation fuels from plants toconsumers), and the costs of the fossil reference technologies.Biomass supply curves endogenously determine the price of feed-stock from forestry and agriculture, while prices of fossil fuels aregiven exogenously. Energy demand is defined exogenously byscenario assumptions. The modeling approach focuses on bioenergytechnologies and does not explicitly consider other low-carbon orrenewable technologies. Instead, their development is implicitlyconsidered in the demand scenarios (see Section 2.4). Energyefficiency measures, particularly in the heating sector, as well asan expansion of non-bioenergy renewable power production, aremodeled by reducing the demand for final energy. The model is ableto assess the relative cost-effectiveness of policy instruments withrespect to attaining two policy objectives (i.e., reducing GHGemissions and substituting fossil fuels in the light of availablebioenergy technologies). We assess energy supply system costsassociated with a shift from fossil fuels to bioenergy; however,transaction costs caused by the policy instruments are not consid-ered in the analysis. Taxes currently applied to both fossil andbioenergy fuels are not included in the model. A detailed descriptionof the mixed integer program can be found in the Appendix.

2.2. Biomass resources

We consider biomass from forestry and agriculture as possiblefeedstock for bioenergy production. Biomass supply curves areused to model the amount of biomass that can be provided atdifferent prices. However, data are separately available for for-estry and agriculture. The sectors are therefore implemented withdifferent approaches in the MIP.

2.2.1. Forestry biomass

An inverse biomass supply curve using a constant elasticityfunction is applied:

p¼ Cq1=e ð1Þ

where p is the price, q is the quantity, C is a constant, and e is thesupply elasticity. The integral:

A¼

Z q�

0p q�q� �1=e

dq¼ q�q� �1=e

q�pe

1þeð2Þ

yields the area A under the supply curve and thus the biomasssupply costs when biomass amount q* is produced. Parametersp and q represent an observed price and quantity, respectively.Schwarzbauer (1997) estimates different supply elasticities eo fordifferent types of forest ownership (index o). Supply elasticities ofprivately owned small forests are estimated to be lower (around0.4) than of large private forests (around 0.5). The supplyelasticities of state-managed forests are even higher (around0.7). Data on observed quantities qfs,o by ownership are availableon the level of federal states (index fs), while price p is a singlenational value because local price variations are not reported(Federal Ministry of Agriculture, Forestry, Environment and WaterManagement, 2009). We expect price variations to be of minormagnitude, and we account for transportation costs explicitly inthe model. To further refine the spatial resolution of the biomasssupply, we use spatially explicit estimates on the maximumsustainable yield (MSY). Details of the modeling of MSY can befound in Schmidt et al. (2010b). Denoting the MSY of an owner-ship (index o) in one supply cell (index i) with bi,o, the observedquantity qi,o for that cell is determined by

qi,o ¼ qfs,obi,o

bfs,o

ð3Þ

The total observed value qs,o is multiplied by the proportion ofestimated MSY bi,o and total MSY of that state and ownership bfs,o.The aggregated forest wood supply curve, with indications of theMSY, and observed prices and quantities in the year 2009 can befound in Fig. 2. Eq. (4) shows a transformation of Eq. (2), in which

0 10 20 30 40

0

50

100

150

Forestry: Quantity of Biomass (TWh)

Mar

gina

l Cos

t (€

MW

h-1)

Small ForestownersLarge ForestownersState Owned ForestsJoint Supply CurveObserved Price and QuantityMaximum Sustainable Yield

Agriculture: Quantity of Biomass (TWh)

Pric

e In

crea

se E

nerg

y C

rops

(%

)

0

50

100

150

200

250

300

0 2 4 6 8 10

GrainsForage CropsOilseedsSrc

Fig. 2. Forest wood supply curve (left) and agricultural supply curve (right). Note: The right picture shows the increase in short rotation cellulose (src) production and the

respective decrease in the production of the most important food and feed crops if prices for short rotation lignocellulose are increased. Both supply curves are an

agreggate for Austria.


supply by ownership and grid cell is included. The function isconvex and can therefore be linearized by separable programming(Jensen and Bard, 2002), which is also shown in the followingequation:

q�i,oqi,o

� �1=eo

q�i,opeo

1þeo�

eo

1þeoqi,op

Xu

q�u� �1=eo q�ubi,o,u ð4Þ

where parameter q�u denotes the share of the observed amount ofbiomass that is produced (e.g., a value of 0.1 means that 10% ofthe amount of biomass observed in the reference period issupplied). Variable bi,o,u is a decision variable involved in theseparable programming and u is the index of the separable steps.The amount of forestry biomass in a supply cell available forbioenergy production (see model details in the Appendix) islimited byXj,b,l

bplanti,j,b,l,tþ

Xi,k,b

bdirecti,k,b r

Xo,u

qi,o q�u� �1=eo q�ubi,o,u ð5Þ

The termP

j,b,lbplanti,j,b,l,tþ

Pi,k,bbdirect

i,k,b represents biomass trans-

ported from supply regions (i) to bioenergy plants (j) of differenttypes (l) and to consumers (k) in season (t). Index b indicates thetype of biomass, in this case, forest wood. The following convexitycondition is necessary in separable programming:X

u

bi,o,u ¼ 1 ð6Þ

2.2.2. Agricultural biomass

Biomass growth on agricultural land for different crops underdifferent management options and crop rotations is simulatedusing the biophysical process model Environmental Policy Inte-grated Climate (EPIC) (Izaurralde et al., 2006; Williams, 1995).Some of the outputs of EPIC are gross margin calculations, whichare input into the spatially explicit land use optimization modelPositive Agricultural Sector Model Austria (PASMA) (Schmid andSinabell, 2007) to find optimal land use management choices bymaximizing total gross margin subject to resource endowmentsat municipal level. Prices of crops are taken from the year 2006(Statistik Austria, 2010) and linearly extrapolated to 2030, basedon the OECD agricultural outlook (OECD, 2009). About 40 agri-cultural crops with three intensification levels are represented inPASMA. They have been grouped into seven categories for theMIP, including grains, oil seeds, forage crops, ethanol crops, oilcrops, short rotation cellulose, and others. Ethanol crops(i.e. starchy and sugar crops such as wheat and sugar beet) canbe used for first-generation ethanol production, while oil crops(e.g., sunflower and rapeseed) are used for biodiesel production.

Heat, power, and second generation biofuels may be producedfrom short rotation cellulose (e.g., short rotation poplar). Spatiallyexplicit biomass supply curves are generated with PASMA bysteadily increasing prices for energy crops from 0% to 300% whileleaving the prices of all other crops constant. Thus, points on thesupply curve are generated accounting for intensification andland use changes (i.e., bioenergy crops substitute food and feed).Combinations of single crop price changes (e.g., ethanol cropsonly) and multi-crop price changes (e.g., ethanol and oil crops)will depict a wide set of supply responses. An aggregated supplycurve for short rotation lignocellulose can be found in Fig. 2. Thefigure shows that price increases of 100% are necessary to triggersubstantial amounts of short rotation cellulose production inAustria. The increases in production cause significant decreases

in food and feed production. Parameters pagrari,sc,b,pl (price of crops)

and qagrari,sc,b,pl (quantity of crop) represent points on the supply

curve. Index i denotes the supply region, index sc indicates theprice scenario, b is the index for the crop type, and pl is the pricelevel of the energy crop. Costs of agricultural production aredetermined byXi,sc,f ,pl

pagrari,sc,b,plq

agrari,sc,b,pla

agrari,sc,pl ð7Þ

where aagrari,sc,pl is the decision variable for separable programming as

outlined in the previous section. Transportation of energy crops toplants

Pj,b,lb

planti,j,b,l is limited byX

j,b,l

bplanti,j,b,l,t r

Xcr,pl

qagrari,sc,b,pla

agrari,sc,pl ð8Þ

where b is any type of biomass except for forest wood. Thereare two dimensions (price scenario sc and price level pl) overwhich the convexity of the curve has to be guaranteed in eachsupply cell:Xsc,pl

aagrari,sc,pl ¼ 1 ð9Þ

2.3. Biomass conversion technologies

There are numerous bioenergy conversion technologies, eithercommercially developed or under research. We select a subset ofthese technologies based on the current deployment and avail-ability of the technologies, as well as on a literature review fortechnologies that are still under research. Input costs for thetechnologies are based on Kalt et al. (2010), who estimated thecosts of bioenergy technologies up to 2030, applying a learningcurve approach for modeling decreases in costs of technologies.BECCS is not modeled by Kalt et al. (2010). Several other sourcesare taken into account to estimate these costs. Costs and


technological data for bioenergy production are listed in Table 2.Solid biomass technologies are available for heating of individualbuildings or small settlements (i.e., community heating net-works). Fuelwood is a cheap option with respect to investmentand fuel costs (Kalt et al., 2010) and is widely used in Austria.Pellets systems are more efficient, comfortable to handle, andneed less labor on the part of the users (Ammann and AlbisserVogeli, 2009; Gustavsson et al., 2005). A large increase in installedcapacity has been observed in the last years (Statistik Austria,2009). The model considers both technologies. However, theopportunity costs of fuelwood users caused by additional timerequirements for handling fuelwood boilers are not considered.This may create a bias toward fuelwood boilers. This affects the

Table 2Investment costs, conversion efficiencies, and carbon capture and storage characteristi

Parameter name

Investments costs plants (capacity 100 MWbiomass)a

Steam power

Steam power with CCS

BIGCC

BIGCC with CCS

Ethanol first generation

Ethanol first generation with CCS

Biodiesel first generation

Methanol second generation

Methanol second generation with CCS

Pellets

Heating plant

Conversion efficiencyCommodity Plant Type

Power Steam

Steam with CCS

BIGCC

BIGCC with CCS

Heat Steam

Steam with CCS

BIGCC

BIGCC with CCS



Fuels Ethanol first generation (without and with CCS)

Biodiesel first generation



Pellets Plantj Pellets

CCS: CO2 emissions captured in productionBIGCC with CCS

Ethanol first generation with CCS


Investment costs at consumersk

Fuelwood boiler 15 kW

Fuelwood boiler 100 kW

Pellets boiler 15 kW

Pellets boiler 100 kW

Conversion efficiencies at consumersFuelwood boiler: 15 and 100 kW

Pellets boiler: 15 and 100 kW

a Costs are given in the biomass capacity of the plants. Cost differences between p

plants is significantly smaller (see efficiencies).b Kalt et al. (2010).c Azar et al. (2006); Rhodes and Ketih (2005); and Uddin and Barreto (2007).d Dornburg and Faaij (2001); Marbe et al. (2004); and Uddin and Barreto (2007).e Bonijoly et al. (2009); and Lindfeldt and Westermark (2009).f Azar et al. (2003); Grahn et al. (2007); and Hamelinck and Faaij (2001).g Hedenus et al. (2010); Lindfeldt and Westermark (2009); and Luckow et al. (201h Leduc et al. (2008); Leduc et al. (2009).i Polagye et al. (2007); Thek and Obernberger (2004).j Electricity consumption of pellets plants induces costs and CO2 emissions, whichk Investment costs per MWh are calculated by determining the annuity of total in

competition between the two bioenergy technologies, but doesnot change the overall model results.

Power is currently produced from biomass mainly with steamengines in combined heat and power (CHP) plants (Marbe et al.,2004). BIGCC plants produce power more efficiently (Dornburgand Faaij, 2001; Marbe et al., 2004), but they are more costly.Currently no commercial facilities are installed worldwide. Weinclude both power producing technologies to explicitly addressthe trade-offs between them. Power production from biogasbased on agricultural products is not considered. High feedstockand transportation costs make biogas power production morecostly than power production in BIGCC or steam engine plantsbased on woody biomass feedstock (Konig, 2011). Both steam

cs of bioenergy technologies.

Unit Value References

Million h 48 b

Million h 50 c

Million h 78 d

Million h 79.5 c

Million h 35 b

Million h 40 e

Million h 9 b

Million h 87 f

Million h 89 g

Million h 5 b, i

Million h 17 b

% 29 b

% 20 c

% 42 d

% 30 c

% 52 b

% 52 c

% 43 d

% 43 c

% 8 h

% 8 g

% 47 b

% 62 b

% 59 f

% 57 g

% 100 b, i

tCO2 MWhbiomass�1 0.36 c

tCO2 MWhbiomass�1 0.05 e

tCO2 MWhbiomass�1 0.22 g

h MWhbiomass�1 76 b




% 80 b

% 89 b

lants with and without CCS are relatively low because the electric capacity of CCS

0).

are considered in the variable costs and in the emissions of the plant.

vestment costs and assuming 1800 load hours per year.


engines and BIGCC can be integrated with CCS (Azar et al., 2006;Uddin and Barreto, 2007).

A wide range of technologies is available and under develop-ment for the use of biomass in the transportation sector. These arecommonly differentiated in terms of being first- or second-genera-tion production technologies. While first-generation fuels rely onstarchy, sugar, and oil crops as feedstock, second-generation bio-fuels allow the conversion of lignocellulosic feedstock, includingforest products and short rotation lignocellulose from agriculture,into fuels. First-generation biofuels are commercially available, andethanol and biodiesel facilities are operating around the world.Second-generation biofuels are currently only produced in a veryfew commercial facilities worldwide (Bacovsky et al., 2010). Themost important first-generation biofuel production technologies inAustria are ethanol and biodiesel production (Kranzl and Haas,2008), which are both included in this analysis. Because second-generation biofuels are still at the early development phase, variousproduction technologies are still under research. Hydrolysis andsubsequent fermentation competes with gasification of biomass.From what is known today, gasification, which allows the produc-tion of various fuels such as Fischer Tropsch (FT) diesel, DME, ormethanol (Demirbas, 2006), is estimated to be more competitivethan hydrolysis (Bram et al., 2009; Lange, 2007). The uncertaintiesin production costs and GHG emission reduction potentials of DMEand methanol are high. Comparative studies can be foundin Wahlund et al. (2004) and Semelsberger et al. (2006). Becauseof the uncertainties, modeling of the two methodologies will notproduce significant advantages for one technology over the other.We therefore chose gasification of biomass and subsequent metha-nol production as the representative second-generation productiontechnology. Because of lower conversion efficiencies, FT dieselproduction costs are higher and GHG emission reductions are lowerthan those of DME and methanol fuels (Sues et al., 2010). FT dieselis therefore not considered in the model. Carbon capture andstorage can be applied to first-generation ethanol and to second-generation methanol production (Azar et al., 2010; Faaij, 2006;Hedenus et al., 2010; Lindfeldt and Westermark, 2009). Relativelyclean CO2 streams are generated in the production process that canbe compressed, transported, and stored directly without the needfor cleaning processes, as in power plants. However, only a limitedamount of the carbon stored in the biomass can be captured, asmost of the CO2 is released when the biofuels are combusted in thevehicles (Hedenus et al., 2010).

Bioenergy with carbon capture and storage (BECCS) is anemerging technology and one of the few options that allowsnegative CO2 emissions to be achieved (Kraxner et al., 2003), whichmay be necessary to manage climate risks effectively (Obersteineret al., 2001). The carbon that is captured at the bioenergy produc-tion sites is transported by pipeline or ship to the final disposalplace, which may be oil, gas or coal fields, deep saline aquifers, oroceans. Costs for CCS after capture at the plant arise mainly fromtransportation and injection of the CO2 into the reservoir. Accordingto Hendriks and Grais (2004)costs for transportation over distancesfrom 500 to 2000 km are estimated to be 10 h tCO2

�1. Major oil andgas fields, as well as saline aquifers, can be found within thisdistance from Austria (Hendriks and Grais, 2004). The total storagecapacity in Europe is estimated to be 86.8 GtCO2, which amounts to18 years of total GHG emissions in the European Union in 2008.Injection costs range from 1.1 to 11.4 h tCO2

�1, depending on thetype and depth of storage. Others estimate transportation andinjection costs to be in the range of 13–42 h tCO2

�1 (Azar et al.,2006; Rhodes and Ketih, 2005; Uddin and Barreto, 2007). Weassume those costs to be at 25 h tCO2

�1. However, the technologicaldevelopment, storage capacities, storage security, and the ecologicaleffects of carbon storage remain uncertain (Holloway, 1997; Thistleet al., 2006; van der Zwaan and Gerlagh, 2009). We therefore run all

scenarios with and without the option for CCS to account for theseuncertainties. For all plants, a scaling factor of 0.7 (Marbe et al.,2004; Hamelinck and Faaij, 2001) with respect to investment costsis assumed (i.e., increasing the size of plants decreases the specificinvestment costs).

Better integration of bioenergy and CCS in industrial processesmay create significant CO2 emission reductions (Mollersten et al.,2006; Mollersten et al., 2004; Mollersten et al., 2003). However, adetailed modeling of the technologies in the relevant industrieswould be necessary to assess the GHG emission reductionpotentials. This is out of the context of this analysis. A roughcalculation shows that the introduction of carbon capture andstorage to the pulp and paper industry could currently savearound 4 MtCO2 if the technology is applied to all Austrian pulpand paper industries, assuming a carbon recovery rate of 90%. Thecosts are estimated to be between 30 and 40h (Mollersten et al.,2006). The industry had a capacity of around 6.63 Mt of biomassin 2005 (Schwarzbauer and Stern, 2010).

2.4. Energy demand

Heating demand is estimated in a spatially explicit way with abottom-up model that combines average consumption values withprivate dwelling areas by age cohort and with the number ofemployees for commercial buildings and industrial low-tempera-ture applications. High temperature heat (industrial heat) is notconsidered in the model. The bottom-up model is validated withnational consumption values for heating fuels (Statistik Austria,2009b). Heating demand is estimated in cells of 1 km2 size forAustria. Demand is split into three demand classes depending onthe heating demand density. The first class (network bound heat)includes all cells that have a heating demand density of above10,000 MWh km�2 a�1. Gas and district heating networks competein these cells. Between 5000 and 10,000 MWh km�2 a�1, supply isbased on community heating networks with a maximum boilercapacity of 100 kW (community district heat). Demand of up to5000 MWh km�2 a�1 is exclusively supplied by boilers in singlebuildings with a capacity of up to 15 kW (single building heat). Thedemand classes are chosen so that district heating and gas supply isreproduced for the base year 2008. A detailed description of theheating demand model can be found in Schmidt et al. (2010b).Future heat demand is assumed to change drastically because of thedemoliton of old buildings, construction of new buildings, andretrofitting of existing buildings (Kalt et al., 2010). Heating demandin the year 2030 is therefore estimated assuming demolition andretrofitting rates for the various age cohorts based on linear trendsfrom 1990 to 2009 (Statistik Austria, 2009a). The spatially explicitdistribution of demolished and newly constructed buildings isbased on spatially explicit population growth estimates (OROK,2009). Also, a growth in square meter of housing area per person isassumed, following a linear trend from 1991 to 2009 (StatistikAustria, 2009a). Combining these factors, a 27% reduction in heatingdemand on the part of private dwellings (from 65 to 48 TWh a�1)from the base year level 2008 is estimated, and found to be in linewith other scenarios (Kranzl and Haas, 2008). However, demand isshifted from low density regions to higher density regions, as citiesare expected to grow significantly, while population in rural areas isexpected to decline (OROK, 2009). Demand for commercial and lowtemperature process heat is assumed to stay constant, as efficiencygains and production increases are assumed to level out. Transpor-tation fuel demand is assumed to grow by 12% (Kranzl and Haas,2008) from the 2008 level of 82 TWh a�1 (Statistik Austria, 2009b).Spatially explicit data on the distribution of the population is usedto allocate fuel demand to demand regions. Power demand isexpected to grow by 25% (Kranzl and Haas, 2008). We assume that


the share of thermal power production remains unchanged (i.e.,renewable energy production and fossil power production haveequal growth rates), which will lead to an increase in thermalpower production from 19 TWhpower a�1 in 2008 (E-Control, 2009)to 24 TWhpower a�1 in 2030. Table 4 shows the resulting demandfor power, heat, and transportation.

2.5. GHG emissions and reference technologies

We account for GHG emissions that are produced within theregion modeled. Leakage effects caused by national carbon policiesare not considered. With respect to bioenergy, the most importantleakage effect concerns GHG emissions that are caused by directand indirect land use change in other world regions (Havlık et al.,2010; Lapola et al., 2010; Searchinger et al., 2008). Althoughpossible GHG emissions from indirect land use changes are notquantitatively assessed, the methodology applied allows thedecline in domestic production of food and feed crops caused byadditional production of energy crops to be shown. The followingdirect GHG emissions are considered: N2O emissions from fertili-zer application in agriculture, CO2 emissions from biomass andcommodity transportation, CO2 emissions in the bioenergy pro-duction process (e.g., from power consumption), CO2 emissions offossil fuel combustion, and negative CO2 emissions from BECCS.Changes in the carbon sequestration of forests due to increasedfuelwood consumption are not included. The carbon released inbioenergy production and consumption is assumed to be neutra-lized by the plant growth. This assumption is consistent with thecurrent version of the UNFCCC reporting guidelines for the KyotoProtocol, which assumes that woody biomass use in energyapplications is GHG emission neutral (UNFCCC, 2006). GHG emis-sion reductions achieved by bioenergy technologies, as well as theeconomic competitiveness of these technologies, depend on thereference technologies chosen. Wood pellet and fuelwood boilers

Table 3Assumptions for demand scenario in 2030.

Sector Demand 2030(TWh y�1)

Change from2008a (%)

Transportation 89 12

Thermal power 24 25

Single building boiler heat and

community district heat demand

31 �37

Network bound heat 30 5

a The demand in 2008 is taken from Statistik Austria (2009b) and E-Control

(2009).

Table 4CO2 emissions and costs of fossil fuel reference technologies in the baseline scenario.

Bioenergytechnology

Fossil reference technology CO2 emissions ofreference technolog

Pellets/fuelwoodboilers

Heating oil boilers 0.28 tCO2 MWhoil�1

Power—no CCS Fuel mix 80% gas and 20% coal 0.39 tCO2 MWhpower�1

Power—with CCS Fuel mix of 80% gas and 20% coal 0.05 tCO2 MWhpower�1

District heating Gas fired heating in single building

boilers, gas fired district heating

0.2 tCO2 MWhgas�1

Ethanol, methanol,and biodiesel

Gasoline 0.26 tCO2 MWhgasolin�1

a Emissions are based on Emission Factors in the Austrian Emissions Inventory (Um

will most likely replace heating oil boilers in single buildings andcommunity district heating. They are similar in operation and alsoneed fuel storage space (Gustavsson et al., 2005). We assume thatsurplus heat from bioenergy plants competes with gas-fireddistrict heating and gas-fired single-building boilers in the net-work-bound heating cells. The costs of the district heating infra-structure are determined by applying an exponential cost functionthat depends on heat demand density (Konstantin, 2007). Thecosts of competing gas networks are assumed to be half the costsof district heating networks (Torekov et al., 2007). BIGCC andBIGCC with CCS are thermal power plants and are therefore able toproduce constant base load power unlike other renewable energytechnologies. In the model, power production therefore competeswith thermal fossil power production. A mix of fossil fuelsconsisting of gas (80%) and coal (20%) is modeled, assuming thatcoal is slowly faded out from the current production share of 40%and that oil with a current share of 8% further loses relevance inpower production (E-Control, 2009). A sensitivity analysis is usedto determine the effects of different power generation fuels. In theCCS scenario, fossil power production applies CCS, which leads todecreased GHG emission factors and increased power prices.Methanol is blended with gasoline for utilization in the transpor-tation sector (Leduc et al., 2008) and is therefore regarded as adirect substitute for gasoline. Surplus heat from bioenergy plantsreplaces gas, which is either used locally in boilers or as fuel indistrict heating networks. CO2 emissions from the different bioe-nergy technologies with regard to the reference technologies andcosts of the reference technologies can be found in Table 3.

2.6. Policy scenarios

We assess the effect of energy policies in a scenario for theyear 2030 with a focus on GHG emission reductions and fossil fuelsubstitution in energy production and transportation. GHG emis-sions in these sectors currently account for around 49% of totalAustrian GHG emissions in 2008 (Umweltbundesamt Austria,2010) (Table 4). The scenario for the oil price is based on theIEA world energy outlook (International Energy Agency, 2009). Itis assumed that the spread between gas and oil remains compar-able to today and that no significant spread between heating oiland gasoline develops. The power price is modeled by looking atfossil fuel prices and the investment costs of thermal powerplants. The power price and CO2 emissions of fossil powerproduction are varied in a sensitivity analysis. The assumptionsfor the sensitivity analysis are stated in Table 5. We assess thecost-effectiveness of five energy policy instruments with respect

ya

Costs of reference technology

Investment costs for oil boilers: 47 h MWhfuel�1 for boilers of 15 kW and

32 h MWhfuel�1 for boilers of 100 kW (Kalt et al., 2010).

Heating oil price: 65 h MWhfuel

75 h MWhpower�1

Investment costs and efficiencies of fossil power production from Tzimas

and Georgakaki (2010)

100 h MWh�1

Investment costs and efficiencies of fossil power production with CCS

from Tzimas and Georgakaki (2010)

Investment costs for gas boilers 32 h MWh�1 (Kalt et al., 2010)

Gas price: 42 h MWhfuel

e Gasoline price: 65 h MWhfuel

weltbundesamt Austria, 2010).

Table 5Energy prices and CO2 emissions of fossil reference technologies used in the scenarios and the sensitivity analysis.

Scenario Power costs(h MWhpower

�1 )CO2 emissions of powerproduction (tCO2 MWhpower

�1 )Costs of heating technology(single buildings) including fuel (h MWhheat

�1 )CO2 emissions of heatingtechnology (tCO2 MWhheat

�1 )

Without CCSBase scenario 75 0.38 124 0.27

Coal scenario 60 0.67 124 0.27

Gas scenario 80 0.32 124 0.27

Heat pump scenario 75 0.38 119 0.13

With CCSBase scenario 100 0.05 124 0.27

Coal scenario 75 0.08 124 0.27

Gas scenario 106 0.04 124 0.27

Heat pump scenario 100 0.05 119 0.13

Table 6Levels of policy instruments considered in the scenario analysis.

Policy instrument Description Range Increment

TX Tax on CO2 emissions of all fossil fuels 2–150 h tCO2�1 5 h tCO2

�1

TS CO2 emission trading scheme 2–150 h tCO2�1 5 h tCO2

�1

FI Feed-in tariffs for biomass power production 80–120 h MWh�1 5 h MWh�1

BF Biofuel shares imposed 0.20–12.00% 0.40%

PE Subsidies to pellet and fuelwood boilers 50–1000 h kW�1 45 h kW�1


to reducing GHG emissions and substituting fossil fuels. The cost-effectiveness of policies is defined such that the attainment ofenergy policy targets (i.e., GHG emission reduction and fossil fuelsubstitution) is cost-minimal in the energy supply system. Wemodel policies that focus on reducing GHG emissions and thatdirectly promote bioenergy technologies. The base scenario (BA)does not contain any policy intervention. The first class of policyscenarios includes the following two scenarios:

�
The TX policy scenario taxes CO2 emissions of all fossil fuels,including private heating and transportation fuels. It isassumed that BECCS gains direct payments for negative CO2
emissions.
� The EU ETS is modeled in the TS scenario where the application
of a carbon price is limited to power and district heat produc-tion only. Fossil fuel consumption in single building boilers,community district heating, and transportation is not includedin the trading scheme. Trading of allowances in the EU ETS isnot modeled explicitly. Constant allowance prices are assumedinstead. BECCS is assumed to receive credits for negative CO2

emissions.

The second class includes the following three scenarios:

�
Feed-in tariffs that guarantee fixed levels of power prices aremodeled in the FI scenario. � In the BF scenario, fixed shares of biofuels in transportation
fuel consumption are imposed.
� Investment subsidies are guaranteed for pellets and fuelwood
furnaces in the PE scenario.

Table 6 reports the levels of the policy instruments considered inthe analysis. All scenarios are assessed with and without BECCSbecause of the uncertainty of technological development. In theBECCS scenarios it is assumed that fossil fuel-based power produc-tion also relies on CCS, which decreases emissions but increasesproduction costs. Costs, GHG emission reductions, and fossil fuel

substitutions in the scenarios are compared to the base scenario(i.e., relative changes in GHG emissions, fossil fuel substitution, andcosts in relation to the base scenario are used to measure the effectof the policy instruments).

3. Results

3.1. GHG emission reduction, fossil fuel substitution, costs, and

technologies

3.1.1. Without CCS

In the base scenario without any policy intervention, GHGemissions are slightly higher than in 2008 (see Table 7). Thedecrease in heating demand and lower CO2 emissions from powergeneration are more than compensated for by increases in powerand transportation fuel demand. A shift from heat to power andtransportation fuel production can be observed, which is due tolower heat demand levels, higher power prices, and more efficientBIGCC production. Biomass consumption increases slightly incomparison to 2008. Forestry provides the whole bioenergy feed-stock. The results of the policy scenarios are shown in Fig. 3. Acarbon tax on all modeled sectors (TX) leads mainly to an increaseof biomass utilization for heating. Up to a price of 50 h tCO2

�1,forestry provides the entire biomass. For higher prices, agricultureproduces additional short rotation lignocellulose. At a price of75 h tCO2

�1, some limited amount of first-generation ethanol pro-duction is triggered because marginal prices of lignocellulose fromagriculture are high in comparison to marginal prices of ethanolcrops. However, the total amount of ethanol production remainssmall. About 2.5 TWh of ethanol is produced at a CO2 price of150 h tCO2

�1. The scenarios TS and FI show similar results. Powerproduction is the main mitigation option in TS, with GHG emissionsreduced to a minor extent by district heating. FI solely subsidizespower production. GHG emission savings and fossil fuel substitu-tion are low in comparison to TX because the private heat sectoris not included in the mitigation efforts. In the BF scenario,

Table 7Model results of the base scenario with and without CCS.

Base 2030 without CCS Base 2030 with CCS

Amount in

scenario

Change

from

2008a

Amount in

scenario

Change

from

2008a

GHG emissions (MtCO2e) 44 þ2 39 �3

Biomass heat production (TWhheat)Single- and multi-dwelling

buildings

4 �13 2 �15

District heatingb 7 þ6 8 þ7

Fossil fuel heat production (TWhheat)Heating oil, single- and

multi-dwelling buildingsc

30 þ7 32 þ9

Gas, district heating 8 þ1 7 0

Gas, heat production in

buildings

18 þ4 18 þ4

Biomass powerproduction (TWhpower)

5 þ3 12 þ10

Fossil fuel powerproduction (TWhpower)

19 0 12 �7

Biofuel production(TWhfuel)

4 þ0.3 0 �3.7

Gasoline consumption(TWhfuel)

85 þ3 89 þ7

Total biomassconsumption (TWhbiomass)

25 þ2 29 þ6

a 2008 values taken from Statistik Austria (2009b), E-Control (2009),

and Umweltbundesamt (2010).b Guesstimate for 2008.c 2008 values include coal and electricity.


second-generation methanol is chosen as production technology forbiofuels, which relies mainly on woody biomass from forestry. GHGemissions are comparable to the baseline scenario, while fossil fuelsubstitution is lower. The main feedstock is forest wood, if biomassheat is directly subsidized (PE). Only a small fraction of shortrotation lignocellulose is used to produce pellets for heat produc-tion. The relative performance of the policies is shown in Fig. 5. TXis superior in attaining both targets cost-effectively (i.e., GHGemission reduction and fossil fuel substitution). In TX, a mix ofbioenergy technologies is chosen. No single technology can there-fore be regarded as superior with respect to cost-effective GHGemission reduction and fossil fuel substitution. There is less cost-effectiveness in all the other policy scenarios, as some technologiesare excluded from the mitigation efforts. In these scenarios,biomass utilization in highly cost-effective applications (such asheating) may be reduced in comparison to the baseline scenariobecause increasing biomass prices make the application in non-subsidized sectors non-competitive. There are, however, differencesin the performance of these instruments. TS and FI performcomparably, while subsidies for heating (PE) are more costly inreducing GHG emissions and increasing fossil fuel substitution. Theinvestment costs for fossil- and biomass-based heating boilers aresignificantly higher than for oil heating boilers, which explains thehigh costs of PE. The biofuel policy (BF) performs worse than anyother policy instrument. Fossil fuel substitution is actually negativein comparison to the base scenario.

3.1.2. With CCS

The use of CCS in fossil power production leads to a significantdecrease in GHG emissions in this sector in the base scenario(see Table 7). Overall, a decrease of 3 Mt CO2e in comparison to2008 is modeled. A big shift from heat to power production frombiomass occurs because the utilization of CCS in the powerproduction sector increases power prices substantially and

therefore makes power production based on biomass more profit-able. Short rotation cellulose from agriculture provides 1 TWh asbioenergy feedstock in the base scenario. Fig. 4 shows detailedresults of the policy scenarios. At low CO2 prices, biomass powerproduction without CCS and, to a lesser extent, heat production,increase in the TX scenario. GHG emission reductions are largerfor biomass heating than for biomass power production becauseof the low CO2 emissions of fossil CCS plants. At CO2 prices above70 h tCO2

�1, BECCS becomes competitive. Methanol and powerproduction with CCS have a comparable share of total production,while heat production declines. Methanol CCS is chosen althoughthe CO2 capture rate is lower than for power production. How-ever, the reference technology for methanol is fossil gasoline thathas higher CO2 emissions in comparison to the reference technol-ogy for power (i.e., fossil power with CCS). A large amount of shortrotation lignocellulose comes from agriculture. Forestry produc-tion is also increased in Austria. GHG emission savings from thetransportation sector and the power sector decrease significantlybecause of the negative CO2 emissions caused by BECCS. Morethan half of the GHG emissions of the sectors observed can bereduced at a CO2 price of 150 h tCO2

�1. However, fossil fuelsubstitution is low because additional fossil power productionhas to be in place when biomass power with CCS is substituted forbiomass power without CCS. In the TS scenario, there is norelevant change in the supply structure up to a price of70 h tCO2

�1, at which biomass based power production with CCSbecomes competitive. Biomass power with CCS gradually replacesbiomass power production without CCS. GHG emission savingsare significant and reach 40% of total GHG emissions. However,fossil fuel substitution is low.

Agricultural production of lignocellulose increases signifi-cantly. The promotion of bioenergy by feed-in tariffs (FI) causesan increase in biomass power production without CCS. Thus whilefossil fuel substitution is substantial, almost no effects on GHGemission reduction can be observed. Biofuel production in the BFscenario mainly replaces existing power and heat production.GHG emission savings are low and fossil fuel substitution doesnot occur at all. Small amounts of agricultural feedstock are usedfor biofuel production. Heat production in PE gradually replacesother bioenergy applications. Short rotation lignocellulose is usedfor the production of pellets to supply heating demand. GHGemission savings are not comparable to TX and TS because noBECCS is deployed.

There is a clear trade-off between fossil fuel substitution andGHG emission reduction when CCS is available, which is shown inFig. 5. The TX and TS scenarios allow a very significant reductionin GHG emissions; however, effects on fossil fuel substitution arelow. FI performs very well with respect to this policy target, butGHG emission reductions are comparably expensive and limited.Subsidies for heating systems and the biofuel policy lead toreductions in GHG emissions but fossil fuel substitution is low.

3.1.3. Technologies

BIGCC dominates steam engines in TS and FI. When CCS isavailable, only BIGCC is deployed. Second-generation methanolproduction is the main biofuel production technology deployed inthe biofuel scenarios. In the TX scenario without CCS, ethanol ischosen as the feedstock availability for methanol is constrained dueto the high utilization of short rotation lignocellulose by the othersectors. Biodiesel is only deployed at a very small scale (o0.3 TWhin all scenarios). If CCS is available, second-generation methanolwith CCS clearly dominates first-generation ethanol with CCS. Withthe exception of pellets plants, plant sizes are always chosen at themaximum capacity (i.e., 300 MWbiomass for all technologies).Although logistic costs are high for such plants, lower investmentcosts for the plants make up for the additional transportation costs.

Fuel

Sub

stitu

tatio

n (T

Wh)

Bio

mas

s U

tiliz

atio

n (T

Wh)

0

GH

G E

mis

sion

s(M

tCO

2e)

Ene

rgy

Prod

uctio

n(T

Wh)

Feed-in Tariff (€ MWh−1)

70

Biofuel Target (% of Total Consumption)

0

Subsidy to fuel wood and pellets boilers (€ kW−1)

0

Methanol

Ethanol

PowerDistrict Heat

Private Heat

Power

District HeatPrivate Heat

Methanol

Power


Methanol

PowerDistrict Heat

Private Heat

Methanol

Private Heat

District HeatPower

Methanol

Fossil Fuel SubstitutionGHG Emissions Agricultural Feedstock

Forest Biomass

0

20

40

60

Fuel

Sub

stitu

tatio

n (T

Wh)

Bio

mas

s U

tiliz

atio

n (T

Wh)

0

20

40

60

Fuel

Sub

stitu

tatio

n (T

Wh)

Bio

mas

s U

tiliz

atio

n (T

Wh)

0

20

40

60

Fuel

Sub

stitu

tatio

n (T

Wh)

Bio

mas

s U

tiliz

atio

n (T

Wh)

0

20

40

60

Fuel

Sub

stitu

tatio

n (T

Wh)

Bio

mas

s U

tiliz

atio

n (T

Wh)

0

20

40

60

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60E

nerg

y Pr

oduc

tion

(TW

h)

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

Carbon Price (€ tCO2−1)

50 100 150 0


50 100 150

0


50 100 150 0


50 100 150

80 90 100 110 120

Feed-in Tariff (€ MWh−1)

70 80 90 100 110 120

2 4 6 8 10 12


0 2 4 6 8 10 12

200 400 600 800 1000


0 200 400 600 800 1000

Fig. 3. GHG emission savings, fossil fuel substitution, and biomass utilization from forestry and agriculture (left) and technological mix (right) in scenarios without CCS:

(a) TX, (b) TS, (c) FI, (d) BF, and (e) PE.


3.2. Alternative reference technologies

Previous sensitivity analyses of the model (Schmidt et al.,2010a) have shown that fossil fuel prices and the characteristicsof fossil reference technologies have the most significant influ-ence on model results. Uncertainties in the development of thecosts and efficiencies of technologies are less important. Insteadof conducting a full sensitivity analysis on all model parameters,we analyze the performance of policy instruments when the costsand CO2 emissions of the fossil reference technologies arechanged. We test for the sensitivity of results of the scenarioswith and without CCS to changes in fossil power production

technology and the availability of an additional heating technol-ogy. There are two alternative fossil power production cases: inthe first case, only gas is used as fuel, while in the other one, onlycoal. We consider heat pumps as alternative reference technologyfor the heating of single buildings. The parameters applied in thesensitivity analysis are reported in Table 5.

3.2.1. Without CCS

Fig. 8 shows the relative performance of the policies in thethree scenarios without CCS. Generally, the ordering of the

Fossil Fuel Substitution

GHG Emissions Agricultural Feedstock


Fuel

Sub

stitu

tion

(TW

h) B

iom

ass

Util

izat

ion

(TW

h)

0

Ene

rgy

Prod

uctio

n(T

Wh)

Feed-in Tariff (€ MWh-1)100


0


0

Power w/o CCSMethanol w CCS

Power w CCSDistrict HeatPrivate Heat


Power w/o CCS

Power w CCS


Power w/o CCS

District Heat

Private Heat

Power w/o CCS Methanol w/o CCS

Private Heat

District HeatPower w/o CCS

Forest Biomass

Ethanol w/o CCS

0

20

40

60

Fuel

Sub

stitu

tion

(TW

h) B

iom

ass

Util

izat

ion

(TW

h)

0

20

40

60

Fuel

Sub

stitu

tion

(TW

h) B

iom

ass

Util

izat

ion

(TW

h)

0

20

40

60

Fuel

Sub

stitu

tion

(TW

h) B

iom

ass

Util

izat

ion

(TW

h)

0

20

40

60

Fuel

Sub

stitu

tion

(TW

h) B

iom

ass

Util

izat

ion

(TW

h)

0

20

40

60

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60

Ene

rgy

Prod

uctio

n(T

Wh)

0

20

40

60

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

GH

G E

mis

sion

s(M

tCO

2e)

10

30

50

50 100 150


0 50 100 150


0 50 100 150


0 50 100 150

110 120 130 140 150 160Feed-in Tariff (€ MWh-1)

100 110 120 130 140 150 160

2 4 6 8 10 12


0 2 4 6 8 10 12

200 400 600 800 1000


0 200 400 600 800 1000

Fig. 4. GHG emission savings, fossil fuel substitution, and biomass utilization from forestry and agriculture (left) and technological mix (right) in scenarios with CCS:

(a) TX, (b) TS, (c) FI, (d) BF and (e) PE.


policies with respect to cost-effectiveness is stable. The onlychange in this order occurs if a low carbon/low cost heatingtechnology (i.e., heat pumps) is available. In that case, BF per-forms better than PE. However, quantitative differences betweenpolicy instruments do change. The performance of PE with regardto fossil fuel substitution is close to the performance of TS and FIin the coal power scenario. The costs of substituting fossil fuel forbiomass fuel in heating are similar to those for biomass powergeneration because of low fossil power prices. In the gas powerscenario, fossil fuel substitution by FI and TS is comparable to theTX scenario because of higher fossil power costs. GHG emissionreductions with TS and FI are costly because gas-fueled power

generation has lower CO2 emissions. When heat pumps areavailable, TX, TS, and FI have a similar performance becausemitigation efforts are mainly concentrated in the power sector.The PE policy instrument performs less well than BF in that case.

3.2.2. With CCS

Fig. 8 shows the relative performance of the policy instru-ments for alternative reference technologies with CCS. Again, theordering of the policy instruments is stable, with the exception ofthe third scenario where the availability of heat pumps decreasesthe cost-effectiveness of PE in relation to BF and FI.

0GHG Emission Reductions

(MtCO2e a-1)

Cos

ts (

M€

)TXTSFI

BFPE

0

TXTSFI

BFPE

Fossil Fuel Substitution(TWh a-1)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

1 2 3 4 5 6 5 10 15 20

0GHG Emission Reductions

(MtCO2e a-1)

0Fossil Fuel Substitution

(TWh a-1)

1 2 3 4 5 6 5 10 15 20

Fig. 5. GHG emission reductions (left) and fossil fuels substituted (right) in relation to costs in the scenario without CCS (upper) and with CCS (lower).

0

Cos

ts (

M€

)

0

Fossil Fuel Substitution (TWh a−1)

0 −5


−2

TXTSFI

BFPE

TXTSFI

BFPE

TXTSFI

BFPE

Emission Reductions (MtCO2e a−1)

Emission Reductions (MtCO2e a−1)

Emission Reductions (MtCO2 e a−1)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

1 2 3 4 5 6 5 10 15 20

0 5 10 15 20

−5Fossil Fuel Substitution (TWh a−1)

0 5 10 15 20

1 2 3 4 5 6

0 2 4 6

Fig. 6. GHG emission reductions (left) and fossil fuel substitution (right) in relation to costs for alternative reference technologies without CCS, (upper) coal power,

(middle) gas power, and (lower) heat pumps.


3.3. Feedstock

The TX, TS, and FI policy scenarios lead to an increase inagricultural biomass, mainly short rotation lignocellulose. Fig. 6

shows that in the TX, TS, and FI scenarios without CCS (compar-able to those with CCS), the production of grains, forage crops,and oil seeds decreases significantly. Grain production is reducedto 50% of the production level of the baseline scenario, oilseeds to

0

Emission Reductions (MtCO2 e a−1)C

osts

(M

€)

TXTSFI

BFPE

0


0


TXTSFI

BFPE

−5


−1


TXTSFI

BFPE

−5


−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

Cos

ts (

M€

)

−100

100

300

500

1 2 3 4 5 6 5 10 15 20

1 2 3 4 5 6 0 5 10 15 20

0 1 2 3 4 5 6 0 5 10 15 20

Fig. 7. GHG emission reductions (left) and fossil fuels substituted (right) in relation to costs for alternative reference technologies with CCS, (upper) coal power, (middle)

gas power, and (lower) heat pumps.

GrainsOilseeds

Forage CropsEthanol crops

Oil cropsSrc

Carbon Price (€ tCO2)-1

0

1000

2000

0 50 100 150

Carbon Price (€ tCO2)-1 Feed -in Tariff (€ MWh-1)

0 50 100 150

Agr

icul

tura

l Pro

duct

ion

(100

0 t)

0

1000

2000

3000

4000

Agr

icul

tura

l Pro

duct

ion

(100

0 t)

0

1000

2000

3000

4000

80 90 100 120

Agr

icul

tura

l Pro

duct

ion

(100

0 t) 4000

3000

Fig. 8. Food and energy crop production in the TX (left), TS (middle), and FI (right) scenario without CCS.


60%, and forage crops to 72% in the TX scenario of 150 h tCO2�1.

Production of forest wood is increased from around 25 TWh in thebaseline scenario to 38 TWh in the TX scenario. TS and FI showcomparable production increases (Figs. 7 and 8).

4. Discussion and conclusions

4.1. Discussion

The results of our model analysis confirm to some degree theresults of other studies. Berndes and Hansson (2007) and Konig(2011) find that the European biofuel policy is costly in

comparison to other measures, while Azar et al. (2003) concludethat biofuels are not competitive with other mitigation options.Gielen et al. (2003) find that biofuels have to play a major role in astringent GHG reduction scenario if no other mitigation optionsare available in the transportation sector. The potentials foradditional supply of biomass to the transportation sector are,however, low in Austria if power and heat production is based onbiomass. Luckow et al. (2010) conclude that mainly biomass-based power production with CCS is a major mitigation optionwith very stringent GHG emission targets. We conclude incontrast that methanol CCS can also contribute to GHG emissionreductions because very significant reductions in the transporta-tion sector can be achieved, which are otherwise not possible, at


least as long as electric cars are not introduced at large scale. Allresults with respect to CCS depend on the development of thattechnology, which remains uncertain.

The large substitution of food crops for energy crops, especiallyin the TX scenario (i.e., 50% reduction of grain production), willincrease agricultural imports to or decrease exports from Austria.Additional agricultural imports or fewer exports may triggerdirect and indirect land use changes in other regions that willlikely offset GHG emission reductions from bioenergy productionin Austria (Havlık et al., 2010; Lapola et al., 2010; Searchingeret al., 2008). An intensive utilization of domestic agriculturalresources for bioenergy will thus likely have negative effects onthe total carbon balance of the policy instruments. Only BECCSallows very significant GHG emission reductions that may offsetsuch indirect effects. Increasing the supply of domestic forestwood can deliver some additional bioenergy resources withoutreplacing existing crop production; however, the total potentialfrom sustainable forest management is limited to around38 TWhbiomass a�1 (Schadauer, 2009). A strategy that relies onimporting biofuels instead of producing them domestically couldbe more effective. For instance, production of Brazilian ethanol issignificantly less expensive than European biofuels (CerqueiraLeite et al., 2009; Nass et al., 2007; de Vries et al., 2010). However,serious social and ecological concerns are connected with biofuelproduction in tropical countries (Delzeit and Holm-Muller, 2009;Martinelli and Filoso, 2008).

While bioenergy technologies are explicitly modeled, otherlow carbon technologies are implicitly considered by reducingenergy demand. This guarantees that the deployment of bioe-nergy is eventually restricted by the demand for renewableenergy in the model. However, the cost-effectiveness of non-technology specific policies, i.e., the carbon tax and the ETS,depends on all available low carbon technologies, includingnon-bioenergy technologies. Explicitly including additional lowcarbon technologies in the model would change the total amountof GHG emission reduction and fossil fuel substitution estimatedat different policy levels. However, the relative ordering withrespect to cost-effective GHG emission reduction of the carbontax and the ETS in comparison to the other policies would not bechanged. Results of our analysis show that the carbon tax,followed by the ETS, is the most cost-effective solution. Addingadditional, non-bioenergy low carbon production technologieswill not cause the policies to be less cost-effective (i.e., morecostly) in reducing GHG emission reduction, i.e., cost-effective-ness is either increased or not changed at all because the twopolicies cause the deployment of cost-minimal technologies forGHG emission reduction. The case for fossil fuel substitution isdifferent: the carbon tax and the ETS are not always the mostcost-effective policies with respect to fossil fuel substitution.Introducing BECCS reduces cost-effectiveness of these policies.Additional low-carbon technologies may therefore change therelative ordering of the policies. Even if the relative ordering ofthe policies is changed, the trade-off between fossil fuel substitu-tion and GHG emission reduction from bioenergy still remains incase BECCS becomes available. The deployment of cost-minimaltechnologies for fossil fuel substitution is not guaranteed by thecarbon tax and the ETS. However, the link between GHG emissionreduction and fossil fuel substitution is strong for most low-carbon technologies with the exception of CCS—CCS is explicitlyregarded. The results of our analysis can therefore also beconsidered a lower bound on cost-effectiveness with respect tofossil fuel substitution.

We have assumed exogenous development of different newbioenergy conversion technologies. These assumptions rely on avast literature survey, but cost estimates remain uncertain. Theinfluence on model results is however limited, as previous

sensitivity studies have shown (Schmidt et al., 2010a). Theinfluence of prices and GHG emissions of reference technologiesmainly affect the competition between CHP and heat productionin single building boilers and community district heating boilers.

Policy instruments that support the development of newtechnologies such as feed-in tariffs and biofuel policies may beless cost-effective than technology-neutral instruments like car-bon taxes. However, the former may be dynamically moreefficient in terms of promoting emerging technologies that allowsubstantial technological advances but still need R&D beforebeing competitive with well-established technologies. The meth-odology applied allows the costs that have to be expected fromthese policy instruments to be estimated and is able to identify asubset of technologies that may be worth subsidizing.

4.2. Conclusions

We presented a modeling approach to cost-effectively assessenergy policy instruments and technologies, as well as theirimpacts on the structure of bioenergy supply in Austria. Thespatially explicit integration of the whole bioenergy supply chain,from biomass production to energy distribution, allows theintegration of various important cost factors that are not coveredby generalized energy system models. For instance, biophysicalconstraints are integrated in an economic biomass productionoptimization model to construct spatially explicit supply curvesfor different types of agricultural energy crops. The explicitassessment of the trade-offs between food, feed, and energy cropproduction is therefore possible and has been applied to theAustrian case. Spatially explicit estimation of energy demandallows energy distribution to be modeled in detail, including thecompetition of district heating and natural gas networks.

The model results indicate that a carbon tax on all fossil fuels iscost-effective with regard to both policy targets (i.e., GHG emis-sion reduction and fossil fuel substitution), if CCS is not available.A trade-off between the two targets exists if CCS is deployed on alarge scale. The fact that some sectors, particularly the privateheating sector, are missing from the ETS reduces its cost-effec-tiveness in comparison to a carbon tax on all fossil fuels. The cost-effectiveness of technology-specific instruments can be ordered inthe following way if CCS is not available: feed-in tariffs beforesubsidies to heating boilers before biofuel blending obligations.This ordering is robust to a change in fossil reference technologieswith the exception of the availability of a low cost heatingtechnology, which makes the biofuel policy perform better thansubsidies on heating boilers. The ordering of policy instrumentswith respect to fossil fuel substitution changes significantly if CCSis available. In this case, a carbon tax and ETS lead to largereductions in GHG emissions, but fossil fuel substitution is low.Feed-in tariffs show contrary results such that low GHG emissionreductions and high amounts of fossil fuel substitution areachieved. Subsidies on heating boilers are in between the carbonpolicies and the feed-in tariffs, while the biofuel policy performsless well than all other policies with regard to both CO2 emissionreductions and fossil fuel substitution.

BECCS can deliver far higher GHG emission reductions thantechnologies without CCS. GHG emission reductions of 20% can beachieved without CCS at the maximum CO2 emission price of150 h tCO2

�1, while 67% of GHG emissions are reduced with CCS.Attaining very stringent GHG emission targets is therefore mucheasier when CCS is available. However, 20% of the fossil fuels aresubstituted at the maximum CO2 price without CCS, while only11% are substituted when CCS is available. Biofuel technologies donot play a significant role in policy scenarios without CCS, butmethanol with CCS is a cost-effective mitigation option, because


the fossil reference technology (i.e., fossil gasoline) does not allowCCS. The biofuel policy is costly and ineffective with regard toGHG emission reduction and fossil fuel substitution. Biofuelpolicies are often designed not only to attain energy policytargets, but also foster rural economic development (Berndesand Hansson, 2007; Lehrer, 2009). Nevertheless, rural develop-ment goals can be more efficiently combined with energy policygoals if biomass resources are directed to other conversionchains than transportation fuel production (i.e., heat or powerproduction).

Forestry biomass is the preferred feedstock. In agriculture,mainly short rotation cellulose is chosen as feedstock, whilethe use of other crops in ethanol and biodiesel production islimited. Results depend on the availability of BIGCC as a powerproduction technology. If BIGCC becomes commercially available,current feed-in tariffs should be directed specifically to largeBIGCC projects instead of promoting CHP steam technology.Model results indicate that economics of scale of big productionplants are higher than increasing transportation costs (i.e., bigplant projects are less costly than small ones). This should beacknowledged in the design of future energy policies, andour spatially explicit mixed integer program is able to accountfor it.

Acknowledgments

This article has been supported by the Austrian ClimateResearch Program project ‘‘Greenhouse Gas Reduction throughSecond Generation Biofuels in Austria (GHG-SEBA)’’, as well as bythe project ‘‘Energieversorgung aus Land- und Forstwirtschaft inOsterreich unter Berucksichtigung des Klima- und GlobalenWandels in 2020 und 2040 (Energ.Clim)’’. We are grateful to veryvaluable comments of two anonymous reviewers and want tothank the team of the Vienna Scientific Cluster (VSC) computergrid for the support when running the model scenarios onthe VSC.

Appendix. Optimization model

The MIP model builds on previous work (Leduc et al., 2009,2010; Schmidt et al., 2010a, 2010b) and minimizes the costs forsupplying demand regions (index k) with different forms of energyproducts (index d) from either biomass plants or fossil fuels.Woody biomass and agricultural feedstock (index b) are trans-ported from supply regions (index i) to possible plant locations(index j) where different conversion technologies (index l) may beemployed to produce different commodities (index c). Plants ofdifferent size and type (index l) are allowed to assess the trade-offbetween increasing costs due to increasing transportation dis-tances and decreasing investment costs due to economies of scalewith growing plant sizes. Biomass can also be transported todemand regions for direct use (e.g., burning of log wood in localfurnaces).

Ethanol, methanol, and pellets are transported to the demandregions by truck. Power is directly distributed to the power grid,while district heat is delivered to settlements (index h) usingpipelines. District heating or gas networks of different sizes (indexns) have to be built in the settlements to allow heat or gasdistribution. Bioenergy competes with fossil fuels (index f). Themodel is static and models 1 year of operation. All investmentcosts are annualized assuming an interest rate of 10% and 25years of economic lifetime. Heating seasons (index t) are used todifferentiate between seasonal heating demands.

The total costs in the objective function f(b,z,q,u) are minimized:

f ðb,z,q,uÞ ¼Xi,o,u

eo

1þeoqi,op q�u

� �1=eo q�ubi,o,uþX

i,sc,b,pl

pagrari,sc,b,plq

agrari,sc,b,pla

agrari,sc,pl

þXi,j,b,l

ctransbi,j,b þcprod

j,l,b þeccsl,b cccs

� �bplant

i,j,b,lþXi,k

ctransdi,k,b þcinvd

k,b

� �bdirect

i,k,b

þX

j,l

cplantj,l uplant

j,l þXj,k,c,t

ctranscj,k,c þcinv

c

� �zbio

j,k,c,tþXk,f

cfossilf þcfossilinv

f

� �zfossil

k,f

þXj,h,ns

cpipej,h,nsu

pipej,h,nsþ

Xh,ns

cdneth,ns udnet

h,ns þXh,ns,t

cdhft qdhf

h,ns,tþXh,ns

cgneth,ns ugnet

h,ns

þXh,ns,t

cgast qgas

h,nsþcememtaxed ðA1Þ

where

emtaxed¼ txsXi,j,b,l

etransi,j,b bplant

i,j,b,lþXi,k,b

etransi,k,b bdirect

i,k,b þXj,k,c,t

etransj,k,c zj,k,c,t

0@

1A

þ txff

Xk,f

efossilf zfossil

k,f þXh,ns,t

egasqgash,ns,tþ

Xh,ns,t

edhf qdhfh,ns,t�

Xi,j,b,l

eccsl bplant

i,j,b,l

0@

1AðA2Þ

totem¼ emtaxedþX

i,sc,b,pl

eagrarsc,b,plq

agrari,sc,b,pla

agrari,sc,pl ðA3Þ

The different summands in the objective function represent

(1)
Biomass supply costs from forests, as described in Section 2.2. (2) Agricultural supply costs, as described in Section 2.2. (3) Biomass transportation costs (parameter ctransb
i,j,b ) from supply

to plant, variable bioenergy production costs (parameter

cprodj,l,b ) and carbon capture and storage costs (parameter cccs

times the amount of CO2 emissions captured by unit of

biomass eccsl,b ) times the amount of feedstock (variable bplant

i,j,b,l ).

(4)
Direct biomass transportation costs (ctransdi,k,b ) from supply to
demand (e.g., fuelwood), investment costs at demand (cinvdk,b ,

e.g., fuelwood boiler) times the amount of biomass trans-

ported directly to demand (bdirecti,k,b ).

(5)
Annualized costs of plant investments (parameter cplantj,l )
times the binary variable for plant selection (uplantj,l ).

(6)
Costs for transporting energy commodities to demand regions
(parameter ctranscj,k,c ) plus investment costs (parameter cinv

c )

times the amount of commodities produced in each period

(variable zbioj,k,c,t). Costs for power transportation and invest-

ment are zero, as it is assumed that the power can be solddirectly to the power grid. For ethanol, methanol, and pellets,transportation costs from plants to demand regions by truck

are considered. Investment costs cinvc are zero for biofuels

because no additional investments are necessary to operatecars with ethanol or methanol. For pellets, the investmentcosts represent the investment costs for pellet boilers.

(7)
Costs for fossil fuels (parameter cfossilf ) plus investment costs
necessary at demand (parameter cfossilinvf e.g., oil boiler) times the

amount of fossil fuels used in a demand region (variable zfossilk,f ).

(8)
Annualized costs of building a pipeline from the plant to the
settlement (parameter cpipej,h,ps) times the binary variable for

pipeline selection (upipej,h,ps).

(9)
Annualized costs for installing a district heating network in
the settlement (parameter cdneth,ns ) times the binary variable

for district heating network selection (udneth,ns ).


(10)
Costs for producing fossil district heat (parameter cdhft )
times the amount of fossil district heat (variable qdhfh,t )

produced.
(11) Annualized costs for installing a gas network in the settle-
ment (parameter cgash,ns) times the binary variable for gas

network selection (ugash,ns).

(12)
Costs for natural gas (cgast ) times the amount of natural gas
consumed (qgash,ns,t).

(13)
CO2 emissions (emtaxed) thta consist of: (i) CO2 emissions of
biomass transportation (CO2 emission factor etransi,j,b ) from

supply to plant, (ii) CO2 emissions of biomass transportation

(CO2 emission factor etransi,j,b ) from supply to demand, (iii) CO2

emissions of commodity transportation (CO2 emission fac-

tor etransj,k,c ), (iv) CO2 emissions of fossil energy production

(CO2 emission factor efossilf ), (v) CO2 emissions of gas heat

production (CO2 emission factor egas), (vi) CO2 emissions of

fossil district heat production (CO2 emission factor edhf ), and(vii) CO2 emission savings by BECCS in bioenergy production(CO2 emission factor eccs

l ). Those CO2 emissions (empriced)

are multiplied by the CO2 price (parameter cem). Binary

parameters txs and txff control if all fossil fuels or only part of

the fossil fuels are taxed by a specific policy instrument.
(14) Total GHG emissions (empriced) are calculated as sum of
CO2 emissions and as sum of N2O emissions from fertilizerapplication in agricultural production. The emission factor

eagrarsc,b,pl describes the CO2 equivalent of N2O emissions.

The objective function in Eq. (A1) is minimized subject to thefollowing constraints. Forest biomass utilization is restricted byXj,b,l,t

bplanti,j,b,l,tþ

Xi,k,b

bdirecti,k,b r

Xo,u

qi,o q�u� �1=eo q�ubi,o,u,b¼ forestwood ðA4Þ

as described in Section 2.2. The convexity condition necessary forthe linearization of the supply curve is guaranteed byX

u

bi,o,u ¼ 1 ðA5Þ

Agricultural production is restricted byXj,b,l,t

bplanti,j,b,l,t r

Xcr,pl

qagrari,sc,b,pla

agrari,sc,pl,ba forestwood ðA6Þ

as described in detail in Section 2.2.The convexity of the supply curves is guaranteed byX

sc,pl

aagrari,sc,pl ¼ 1 ðA7Þ

The plant size constraints production is byXi,b

bplanti,j,b,l,t rbj,l,tu

plantj,l ðA8Þ

where parameter bj,l,t is the production capacity of plant j inperiod t. Index l indicates the size and technology of the plant(e.g., 50 MW steam CHP and 100 MW BIGCC). The commodityproduction in each period (variable zbio

j,c,t) is determined by thebiomass input and conversion efficiency (parameter Zconv

j,b,l,c) shownin the following equation:X

i,l

Zconvj,b,l,cbplant

i,j,b,l,t ¼ zbioj,c,t ðA9Þ

District heat production is modeled with variable qbioj because

it is distributed differently from the other commodities:Xi,b,l

Zheatj,b,l bplant

i,j,b,l,t ¼ qbioj,t ðA10Þ

where Zheatj,b,l is the conversion efficiency for heat. Distribution of

commodities to demand regions k is restricted byX

k

zbioj,k,c r

Xt

zbioj,c,t ðA11Þ

where variable zbioj,k,c denotes the amount of a commodity c

transported from plant location j to demand region k.Energy demands (parameter dk,d) are satisfied by direct bio-

mass utilization, different commodities from bioenergy produc-tion (variable zbio

j,k,c), and fossil fuels (variable zfossilk,f ):

Xi,b

Zbiodb,d bdirect

i,k,b þXj,c

Zbioc,dzbio

j,k,cþX

f

Zfossilf ,d zfossil

k,f ¼ dk,d ðA12Þ

In the equation, parameters Zbiodb,d , Zbio

c,d , and Zfossilf ,d describe the

efficiency of converting biomass, bioenergy commodities, andfossil fuels to forms of useful energy.

Heat production limits the amount of heat available for districtheating. Seasonal supply of heat in the plants is restricted byXh,ps

qdhj,h,ps,t rqbio

j,t ðA13Þ

The production of heat by bioenergy plants qdhj,h,ns,t plus the

production of fossil district heating qdhfh,ns,t and the production of

heat in single-household gas-fired boilers qgash,ns,t have to meet the

demand (parameter qDh,t) in each period, which is guaranteed by

Zdhh,t

Xj,ps

Ztransj,h,ns,tq

dhj,h,ns,t

0@

1AþX

ns

Zdhf qdhfh,ns,t

0@

1AþZgasqgas

h,ns,t ¼ qDh,t ðA14Þ

where parameter Ztransj,h,ns,tdenotes the heat losses in the pipe system

from the plant to the settlement. Losses in the heat distribution

network within the settlement are modeled by parameter Zdhh,t .

Parameters Zdhf and Zgas are introduced to describe conversionefficiencies of fossil district heating and gas-fired boilers.

The sum of heat produced by the bioenergy plant and by thefossil district heating boiler has to match the district heatingdemand (parameter qD

h,ns,t) in settlement h. This is modeled by

Zdhh,t

Xj,ps

Ztransj,h,ps,tq

dhj,h,ps,t

0@

1AþX

ns

Zdhf qdhfh,ns,t

0@

1A¼X

ns

qDh,ns,tu

dneth,ns ðA15Þ

The same is the case for gas-supplied settlements:

Zgasqgash,ns,t ¼ qD

h,ns,tugash,ns ðA16Þ

The existence of a transportation pipeline, in the event that asettlement is supplied by a bioenergy plant, is ensured by

qdhj,h,ns,t rqpipe

ns,t upipej,h,ns ðA17Þ

where parameter qpipeps,t denotes the capacity of the pipeline.

The policy instruments are implemented in the following way:the price of carbon emissions is controlled by the value of cem. Thebinary parameters txs and txf

f control which CO2 emissions aretaxed by the particular policy instrument. Feed-in tariffs aremodeled by setting the fossil power price cfossil

f to the level ofthe tariff. The investment costs for pellet furnaces cinv

c aredecreased in the pellet subsidy scenario. The compliance withbiofuel blending obligations is guaranteed byX

j

zbioj,c Z f blend

c,d

Xk

dk,d ðA18Þ

where f blendc,d is the mandatory share of a bioenergy commodity in

total useful energy demand.


The MIP is finally defined as

min f b,z,q,uð Þ�

s:t:ðA4Þ�ðA18Þ

0raagrari,sc,pl,bi,o,u,bdirect

i,k,b ,bplanti,j,b,l ,emtaxed,qbio

j ,qdhj,h,ps,t ,q

dhfh,ns,t ,q

gash,ns,

totem,zbioj,c ,zbio

j,k,c,t ,zfossilk,f

ugneth,ns ,udnet

h,ns ,upipej,h,ns,u

plantj,l A 0,1f g: ðA19Þ

References

Alfonso, D., Perpina, C., Perez-Navarro, A., Penalvo, E., Vargas, C., Cardenas, R.,2009. Methodology for optimization of distributed biomass resourcesevaluation, management and final energy use. Biomass and Bioenergy 33,1070–1079.

Ammann, H.I.B., Albisser Vogeli, G., 2009. Comparison of wood-fired heatingsystems. Agrarforschung 16, 64–69.

Azar, C., Lindgren, K., Andersson, B.A., 2003. Global energy scenarios meetingstringent CO2 constraints—cost-effective fuel choices in the transportationsector. Energy Policy 31, 961–976.

Azar, C., Lindgren, K., Larson, E., Mollersten, K., 2006. Carbon capture and storagefrom fossil fuels and biomass—costs and potential role in stabilizing theatmosphere. Climate Change 74, 47–79.

Azar, C., Lindgren, K., Obersteiner, M., Riahi, K., van Vuuren, D., den Elzen, K.,Mollersten, K., Larson, E., 2010. The feasibility of low CO2 concentration targetsand the role of bio-energy with carbon capture and storage (BECCS). ClimaticChange 100, 195–202.

Bacovsky, D., Dallos, M., Worgetter, M., 2010. Status of 2nd Generation BiofuelsDemonstration Facilities in June 2010, A Report to IEA Bioenergy Task 39.

Berndes, G., Hansson, J., 2007. Bioenergy expansion in the EU: cost-effectiveclimate change mitigation, employment creation and reduced dependency onimported fuels. Energy Policy 35, 5965–5979.

Bonijoly, D., Fabbri, A., Chapuis, F., Laude, A., Ricci, O., Bauer, H., Grataloup, S.,Gali�egue, X., 2009. Technical and economic feasibility of the capture andgeological storage of CO2 from a bio-fuel distillery: CPER artenay project.Energy Procedia 1, 3927–3934.

Bram, S., De Ruyck, J., Lavric, D., 2009. Using biomass: a system perturbationanalysis. Applied Energy 86, 194–201.

Bryan, B.A., King, D., Wang, E., 2010. Biofuels agriculture: landscape-scale trade-offs between fuel, economics, carbon, energy, food, and fiber. GCB Bioenergy 2,330–345.

Cerqueira Leite, R.C.D., Verde Leal, M.R.L., Barbosa Cortez, L.A., Griffin, W.M., GayaScandiffio, M.I., 2009. Can Brazil replace 5% of the 2025 gasoline worlddemand with ethanol? Energy 34, 655–661.

COM, 2010. Energy 2020. A Strategy for Competitive, Sustainable and SecureEnergy. European Commission.

Council, E., 2002. Council Decision of 25 April 2002 concerning the approval, onbehalf of the European Community of the Kyoto Protocol to the United NationsFramework Convention on Climate Change and the joint fulfilment ofcommitments thereunder, Official Journal.

Delzeit, R., Holm-Muller, K., 2009. Steps to discern sustainability criteria for acertification scheme of bioethanol in Brazil: approach and difficulties. Energy34, 662–668.

Demirbas, A., 2006. Present and Future Transportation Fuels. Energy sources, PartA: Recovery, Utilization, and Environmental Effects 30, 1473–1483.

Dornburg, V., Faaij, A., 2001. Efficiency and economy of wood-fired biomassenergy systems in relation to scale regarding heat and power generationusing combustion and gasification technologies. Biomass and Bioenergy 21,91–108.

de Vries, S.C., van de Ven, G.W., van Ittersum, M.K., Giller, K.E., 2010. Resource useefficiency and environmental performance of nine major biofuel crops,processed by first-generation conversion techniques. Biomass and Bioenergy34, 588–601.

E-Control, 2009. Time series of power production (Jahresreihen: Gesamte Elek-trizitatsversorgung Aufbringung), E-Control.

Eriksson, L.O., Bjorheden, R.W., 1989. Optimal storing, transport and processing fora forest-fuel supplier. European Journal of Operational Research 43, 26–33.

Faaij, A., 2006. Modern biomass conversion technologies. Mitigation and Adapta-tion Strategies for Global Change 11, 335–367.

Federal Ministry of Agriculture, Forestry, Environment and Water Management,2009. Logging Report 2008 (Holzeinschlagsmeldung 2008), Federal Ministry ofAgriculture, Forestry, Environment and Water Management.

Federal Ministry of Agriculture, Forestry, Environment and Water Management,2007. National Allocation Plan for Austria for the period 2008–2012 (Natio-naler Zuteilungsplan fur Osterreich gemaß y 11 Emissionszertifikategesetz furdie Periode 2008–2012). Federal Ministry of Agriculture, Forestry, Environ-ment and Water Management.

Gielen, D.J., Fujino, J., Hasimoto, S., Moriguchi, Y., 2003. Modeling of global biomasspolicies. Biomass and Bioenergy 25, 177–195.

Grahn, M., Azar, C., Lindgren, K., Berndes, G., Gielen, D., 2007. Biomass for heat oras transportation fuel? A comparison between two model-based studies.Biomass and Bioenergy 31, 747–758.

Grohnheit, P.E., Mortensen, B.O.G., 2003. Competition in the market for spaceheating. District heating as the infrastructure for competition among fuels andtechnologies. Energy Policy 31, 817–826.

Gustavsson, L., Mahapatra, K., Madlener, R., 2005. Energy systems in transition:perspectives for the diffusion of small-scale wood pellet heating technology.International Journal of Technology Management 29, 327–347.

Hamelinck, C.N., Faaij, A., 2001. Future prospects for production of methanol andhydrogen from biomass. Journal of Power sources 111, 1–22.

Havlık, P., Schneider, U.A., Schmid, E., Bottcher, H., Fritz, S., Skalsky, R., Aoki, K., Cara,S.D., Kindermann, G., Kraxner, F., Leduc, S., McCallum, I., Mosnier, A., Sauer, T.,Obersteiner, M., 2010. Global land-use implications of first and second genera-tion biofuel targets. Energy Policy. doi:10.1016/j.enpol.2010.03.030.

Hedenus, F., Karlsson, S., Azar, C., Sprei, F., 2010. Cost-effective energy carriers fortransport—the role of the energy supply system in a carbon-constrainedworld. International Journal of Hydrogen Energy 35, 4638–4651.

Hendriks, C., Grais, W., 2004. Global Carbon Dioxide Storage Potential and Costs. Ecofys.Holloway, S., 1997. Safety of the underground disposal of carbon dioxide. Energy

Conversion and Management 38, 241–245.International Energy Agency, 2009. World Energy Outlook. International Energy

Agency.Ivezic, D., Zivkovic, M., Tanskovic, T., Dajic, N., 2008. An economic model for the

appraisal and selection of energy supply system. Applied Thermal Engineering29, 1440–1446.

Izaurralde, R., Williams, J., McGill, W., Rosenberg, N., Jakas, M.Q., 2006. Simulatingsoil C dynamics with EPIC: model description and testing against long-termdata. Ecological Modelling 192, 362–384.

Jensen, P.A., Bard, J.F., 2002. Operations Research Models and Methods. Wiley.Kalt, G., Kranzl, L., Haas, R., 2010. Long-term strategies for an efficient use of

domestic biomass resources in Austria. Biomass and Bioenergy 34, 449–466.Katofsky, R., Stanberry, M., Frantzis, L., 2010. Achieving climate stabilization in an

insecure world: does renewable energy hold the key? Report. InternationalEnergy Agency—Renewable Energy Technology Development.

Konig, A., 2011. Cost efficient utilisation of biomass in the German energy systemin the context of energy and environmental policies. Energy Policy 39,628–636.

Konstantin, P., 2007. Handbook Energy Industry (Praxisbuch Energiewirtschaft).Springer, Berlin Heidelberg.

Kranzl, L., Haas, R., 2008. Strategies for the Optimal Development of BiomassPotentials in Austria Until the Year 2050 Aiming at a Maximal Reduction ofGreen House Gas Emissions (Strategien zur optimalen Erschließung derBiomassepotenziale in Osterreich bis zum Jahr 2050 mit dem Ziel einermaximalen Reduktion an Treibhausgasemissionen). Energy Economics Group,University of Technology Vienna.

Kraxner, F., Nilsson, S., Obersteiner, M., 2003. Negative emissions from BioEnergyuse, carbon capture and sequestration (BECS)—the case of biomass productionby sustainable forest management from semi-natural temperate forests.Biomass and Bioenergy 24, 285–296.

Lange, J., 2007. Lignocellulose conversion: an introduction to chemistry, processand economics. Biofuels, Bioproducts and Biorefining 1, 39–48.

Lapola, D.M., Schaldach, R., Alcamo, J., Bondeau, A., Koch, J., Koelking, C., Priess, J.A.,2010. Indirect land-use changes can overcome carbon savings from biofuels inBrazil. Proceedings of the National Academy of Sciences 107, 3388–3393.

Leduc, S., Schmid, E., Obersteiner, M., Riahi, K., 2009. Methanol production bygasification using a geographically explicit model. Biomass and Bioenergy 33,745–751.

Leduc, S., Schwab, D., Dotzauer, E., Schmid, E., Obersteiner, M., 2008. Optimallocation of wood gasification plants for methanol production with heatrecovery. International Journal of Energy Research 32, 1080–1091.

Leduc, S., Starfelt, F., Dotzauer, E., Kindermann, G., McCallum, I., Obersteiner, M.,Lundgren, J., 2010. Optimal location of lignocellulosic ethanol refineries withpolygeneration in Sweden. Energy 35, 2709–2716.

Lehrer, N., 2009. Bio)fueling farm policy: the biofuels boom and the 2008 farm bill.Agriculture and Human Values 27, 427–444.

Lindfeldt, E.G., Westermark, M.O., 2009. Biofuel production with CCS as a strategyfor creating a CO2-neutral road transport sector. Energy Procedia 1,4111–4118.

Luckow, P., Wise, M., Dooley, J., Kim, S., 2010. Large-scale utilization of biomassenergy and carbon dioxide capture and storage in the transport and electricitysectors under stringent CO2 concentration limit scenarios. InternationalJournal of Greenhouse Gas Control 4, 865–877.

Magne, B., Kypreos, S., Turton, H., 2010. Technology Options for low stabilizationpathways with MERGE. The Energy Journal 31, 83–108.

Marbe, G., Harvey, S., Berntsson, T., 2004. Biofuel gasification combined heat andpower—new implementation opportunities resulting from combined supplyof process steam and district heating. Energy 29, 1117–1137.

Martinelli, L.A., Filoso, S., 2008. Expansion of sugarcane ethanol production in Brazil:environmental and social challenges. Ecological Applications 18, 885–898.

Mollersten, K., Gao, L., Yan, J., 2006. CO2 Capture in pulp and paper mills: CO2

balances and preliminary cost assessment. Mitigation and Adaptation Strate-gies for Global Change 11, 1129–1150.

Mollersten, K., Gao, L., Yan, J., Obersteiner, M., 2004. Efficient energy systems withCO2 capture and storage from renewable biomass in pulp and paper mills.Renewable Energy 29, 1583–1598.

dx.doi.org/10.1016/j.enpol.2010.03.030


Mollersten, K., Yan, J., Moreira, J., R., 2003. Potential market niches for biomassenergy with CO2 capture and storage—opportunities for energy supply withnegative CO2 emissions. Biomass and Bioenergy 25, 273–285.

Nakicenovic, N., Haas, R., Schleicher, S.P., Klettner, C., Kletzan, D., Koppl, A., Turk,A., 2008. Assessment of Austrian Contribution Toward EU 2020 Target Sharing.Austrian Institute of Economic Research WIFO.

Nass, L.L., Pereira, P.A.A., Ellis, D., 2007. Biofuels in Brazil: an overview. CropScience 47, 2228–2237.

Obersteiner, M., Azar, C., Kauppi, P., Mollersten, K., Moreira, J., Nilsson, S., Read, P.,Riahi, K., Schlamadinger, B., Yamagata, Y., 2001. Managing climate risk. Science294, 786–787.

OECD, 2009. OECD-FAO Agricultural Outlook 2009. OECD Publishing.OROK, 2009. Scenarios for The Spatial and Regional Development of Austria in the

European Context (Szenarien der Raumentwicklung Osterreichs 2030. Regio-nale Herausforderungen und Handlungsstrategien, Wien 2009). Osterrei-chische Raumordnungskonferenz (OROK).

Paoletta, M.S., Taschini, L., 2006. An Econometric Analysis of Emission TradingAllowances. Swiss Finance Institute.

Polagye, B.L., Hodgson, K.T., Malte, P.C., 2007. An economic analysis of bio-energyoptions using thinnings from overstocked forests. Biomass & Bioenergy 31,105–125.

Pope, J., Owen, A.D., 2009. Emission trading schemes: potential revenue effects,compliance costs and overall tax policy issues. Energy Policy 37, 4595–4603.

Rhodes, J.S., Ketih, D.W., 2005. Engineering economic analysis of biomass IGCCwith carbon capture and storage. Biomass & Bioenergy 29, 440–450.

Ricci, O., 2010. Providing adequate incentives for biomass carbon capture andgeological storage. In: First International Workshop on Biomass & CarbonCapture and Storage, Universite d’Orleans.

Richard, T.L., 2010. Challenges in scaling up biofuels infrastructure. Science 329,793–796.

Roeckner, E., Giorgetta, M., Crueger, T., Esch, M., Pongratz, J., 2010. Historical andfuture anthropogenic emission pathways derived from coupled climate–carbon cycle simulations. Climatic Change, 1–18.

Sanden, B.A., Azar, C., 2005. Near-term technology policies for long-term climatetargets–economy wide versus technology specific approaches. Energy Policy33, 1557–1576.

Sandy Thomas, C., 2009. Transportation options in a carbon-constrained world:hybrids, plug-in hybrids, biofuels, fuel cell electric vehicles, and batteryelectric vehicles. International Journal of Hydrogen Energy 34, 9279–9296.

Schadauer, K., 2009. Wood and biomass study (holz- und biomassenstudie). BFWPraxis Information, 18.

Schmid, E., Sinabell, F., 2007. On the choice of farm management practices afterthe reform of the Common Agricultural Policy in 2003. Journal of Environ-mental Management 82, 332–340.

Schmidt, J., Leduc, S., Dotzauer, E., Kindermann, G., Schmid, E., 2010a. Cost-effective CO2 emission reduction through heat, power and biofuel productionfrom woody biomass: a spatially explicit comparison of conversion technol-ogies. Applied Energy 87, 2128–2141.

Schmidt, J., Leduc, S., Dotzauer, E., Kindermann, G., Schmid, E., 2010b. Potential ofbiomass-fired combined heat and power plants considering the spatialdistribution of biomass supply and heat demand. International Journal ofEnergy Research 34, 970–985.

Schwarzbauer, P., 1997. Austria. In: B. Solberg, (Ed). Demand and Supply Analysesof Roundwood and Forest Products Markets in Europe. Proceedings of the 1stWorkshop of the Concerted Action Project AIR3—CT942288, Helsinki.

Schwarzbauer, P., Stern, T., 2010. Energy vs. material: economic impacts of a

‘‘wood-for-energy scenario’’ on the forest-based sector in Austria—a simula-

tion approach. Forest Policy and Economics 12, 31–38.Searchinger, T., Heimlich, R., Houghton, R.A., Dong, F., Elobeid, A., Fabiosa, J.,

Tokgoz, S., Hayes, D., Yu, T., 2008. Use of U.S. croplands for biofuels increases

greenhouse gases through emissions from land use change. Science 319,

1238–1240.Semelsberger, T.A., Borup, R.L., Greene, H.L., 2006. Dimethyl ether (DME) as an

alternative fuel. Journal of Power Sources 156, 497–511.Statistik Austria, 2009a. Dwellings 2009, Results of the Dwellings Survey in the

Microcensus, Yearly Mean 2009 (Wohnen, 2009; Ergebnisse der Wohnung-

serhebung im Mikrozensus Jahresdurchschnitt, 2009).Statistik Austria, 2009b. Energy Balances 1970–2008 (Energiebilanzen 1970–

2008). Statistik Austria.Statistik Austria, 2010. Price of Products in Agriculture and Forestry 2003–2009

(Land und Forstwirtschaftliche Erzeugerpreise 2003–2009). Statistik Austria.Statistik Austria, 2009. Total Use of Energy Sources (Gesamteinsatz Aller Ener-

gietrager 2003/2004–2007/2008). Statistik Austria.Steenhof, P.A., McInnis, B.C., 2008. A comparison of alternative technologies to de-

carbonize Canada’s passenger transportation sector. Technological Forecasting

and Social Change 75, 1260–1278.Steininger, K.W., Voraberger, H., 2003. Exploiting the medium term biomass

energy potentials in Austria. Environmental and Resource Economics 24,

359–377.Sues, A., Jurascık, M., Ptasinski, K., 2010. Exergetic evaluation of 5 biowastes-to-

biofuels routes via gasification. Energy 35, 996–1007.Thek, G., Obernberger, I., 2004. Wood pellet production costs under Austrian and

in comparison to Swedish framework conditions. Biomass & Bioenergy 27,

671–693.Thistle, D., Sedlacek, L., Carman, K.R., Fleeger, J.W., Bewer, P.G., Barry, J.P., 2006.

Simulated sequestration of industrial carbon dioxide at a deep-sea site: effects

on species of harpacticoid copepods. Journal of Experimental Marine Biology

and Ecology 330, 151–158.Torekov, M.S., Bahnsen, N., Qvale, B., 2007. The relative competitive positions of

the alternative means for domestic heating. Energy 32, 627–633.Tzimas, E., Georgakaki, A., 2010. A long-term view of fossil-fuelled power

generation in Europe. Energy Policy 38, 4252–4264.Uddin, N.S., Barreto, L., 2007. Biomass-fired cogeneration systems with CO2

capture and storage. Biomass & Bioenergy 32, 1006–1019.Umweltbundesamt Austria, 2010. Austria’s National Inventory Report 2010.

Umweltbundesamt Austria.UNFCCC, 2006. Updated UNFCCC Reporting Guidelines on Annual Inventories

Following Incorporation of the Provisions of Decision 14/CP.11. United

Nations.Wahlund, B., Yan, J., Westermark, M.O., 2004. Increasing biomass utilization in

energy systems: a comparative study of CO2 reduction and cost for different

bioenergy processing options. Biomass and Bioenergy 26, 531–544.Williams, J.R., 1995. The EPIC model. In: Singh, V.P. (Ed.), Computer Models of

Watershed Hydrology. Water Resources Publications, Highlands Ranch, Color-

ado, pp. 909–1000.van der Zwaan, B., Gerlagh, R., 2009. Economics of geological CO2 storage and

leakage. Climate Change 93, 285–309.

Cost-effective policy instruments for greenhouse gas emission reduction and fossil fuel substitution...

Documents

Transcript of Cost-effective policy instruments for greenhouse gas emission reduction and fossil fuel substitution...