ARTICLE IN PRESS

Energy Policy 38 (2010) 168–180

Contents lists available at ScienceDirect

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

Biomass gasification in cost-optimized district heating systems—A regional

modelling analysis
Martin Börjesson , Erik O. Ahlgren
Energy Systems Technology, Division of Energy Technology, Department of Energy and Environment, Chalmers University of Technology, SE-412 96 Göteborg, Sweden

a r t i c l e in f o a b s t r a c t

Article history: Biomass integrated gasification combined cycle (BIGCC) plants could, in combined heat and power
Received 16 June 2008 (CHP) generation, increase the power-to-heat ratio compared to conventional biomass steam turbine
Accepted 2 September 2009 plants. Furthermore, biomass gasification could also be used for the efficient production of biofuels for
Available online 30 September 2009
transport. In this study, different applications of biomass gasification in connection to district heating
Keywords: (DH) are analysed and contrasted to conventional technology options. An application of the cost-
Biomass gasification optimizing energy system model MARKAL with a detailed description of the DH sector in a
MARKAL southwestern region of Sweden was developed within the study and used in the analysis. Policy
District heating measures for CO2 reduction and for promotion of ‘‘green’’ electricity are assumed, and required subsidy
levels for large-scale production of transport biofuels are calculated. The model also operates with
different supplies of biomass: a local supply at a lower cost and an international supply of refined
biomass at a slightly higher cost. The study shows that investments in BIGCC CHP are often cost-
efficient in cases with low ambitions regarding transport biofuels. However, due to limitations in heat
demand and in local, lower cost, supply of biomass, investment in biofuel production means less
investment in BIGCC CHP and, thereby, a smaller electricity production.
& 2009 Elsevier Ltd. All rights reserved.

1. Introduction and purpose
There are mainly two driving forces for modern use of biomass
for energy purposes: climate concerns and energy security. With
increasingly more ambitious CO2 emission reduction targets and a
will to reduce dependence on imported energy carriers, the
demand for biomass is likely to increase. Although biomass
resources are renewable, the potential is limited, and an
increasing pressure on efficient resource utilization is probable.
The purpose of the present study is to examine the cost-
effectiveness of biomass gasification technologies compared to
conventional technology alternatives in the district heating (DH)
sector. The study investigates whether, and under what condi-
tions, combined heat and power (CHP) generation through
biomass gasification in so-called biomass integrated gasification
combined cycle (BIGCC) plants could be part of the production in
cost-optimized DH systems. Production of biofuels for transport
through biomass gasification is also analysed and, in particular,
what impact on cost-optimized DH systems a large-scale
introduction of biofuel production would have and what subsidy
levels that would be required for such an introduction. Further-
more, environmental effects in terms of impacts on CO2 emissions
 Corresponding author. Tel.: + 46 31772 5244; fax: + 46 31772 3592.
E-mail address: martin.borjesson@chalmers.se (M. Börjesson).
are addressed. A region in the southwestern part of Sweden,
defined by the county of Västra Götaland and the conurbation of
Göteborg, with its specific energy demands and energy infra-
structure is used as a case study.
A global problem like the climate issue requires international
agreements and strategies. However, a local or regional perspec-
tive is central for the actual implementation of actions for
emission reduction, and allows in several aspects a higher degree
of detail than what is possible with more aggregated approaches
on higher geographical levels. Local circumstances and geogra-
phical considerations, which often are important factors in choice
of energy technologies, are more easily integrated in the analysis.
In addition, regional energy systems analyses can, besides adding
insights of efficient energy technology choices in general, also
provide understanding regarding factors such as suitable plant
locations. Since biomass markets often are local or regional
(although refining of the biomass to, e.g., pellets or briquettes
opens up wider markets through decreased transport costs) this
perspective is also essential when it comes to analyses of
biomass-derived products such as biofuels and ‘‘green’’ electricity.
The importance of regional and local initiatives for environmental
issues has been addressed in various forums. In the United
Nations Programme Agenda 21, local actions for sustainable
development is encouraged and recognized to play a vital role
(UN, 1993), and the European Commission has in its plan for
energy efficiency improvements recognized that large saving

0301-4215/$ - see front matter & 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enpol.2009.09.001
 ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180
potentials could be realized by greater decentralization of energy
management to local and regional levels (European Commission,
2000). A few energy systems studies have applied the regional
perspective. Two examples are Salvia et al. (2004), in which the
Basilicata region, Italy, was modelled to analyse cost-effective CO2
reduction strategies focusing on the civil and waste management
sectors, and Carlson (2003), in which the heating sector of
the Östergötland region, Sweden, was analysed with the
primary objective of examining the consequences of including
external costs for emissions. A review of studies done on
modelling for decentralized energy planning is given in Hiremath
et al. (2007).
The present study has partly its origin in the so-called
Biokombi Rya project (Nyström et al., 2007). The Biokombi Rya
project had as its main objective, on different system levels, to
investigate the possibility and possible benefits of integrating
biomass gasification with a 600 MWfuel natural gas combined
cycle (NGCC) plant, the Rya CHP, recently taken in operation in
the city of Göteborg in the Västra Götaland region of Sweden.
However, in the present study, the possibility of retrofitting
existing natural gas plants with biomass gasification is not in
focus. Instead, the approach is more general considering invest-
ments in different types of gasification-based production, and the
biomass gasification technologies are treated as alternatives in the
entire studied region and not primarily as an option in Göteborg.
Other studies that in different ways are concerned with economic
performance of biomass gasification utilities are for instance
Dornburg and Faaij (2001), Marbe et al. (2004) and Fahle n and
Ahlgren (in press). The latter focuses on biomass gasification in
connection with district heating in a similar way as the present
study, although there are several methodological differences.
2. Background
Gasification makes it possible to use biomass in gas combined
cycle (CC) plants and, thereby, to reach a significantly higher
electrical efficiency than in conventional biomass steam turbine
(ST) plants. However, regarding CHP production, the heat
efficiency and the total efficiency (electricity and heat) are lower
than in a conventional biomass ST CHP plant with flue gas
condensation (Nyström et al., 2007).
Biomass gasification can also be used to produce biofuels for
the transportation sector. From the product gas received from the
gasification process, a number of biofuels, e.g. dimethyl ether
(DME), methanol, synthetic natural gas (SNG), hydrogen and
Fischer–Tropsch diesel, can be produced through different types
of gas processing and synthesis steps. The energy efficiency is high
compared to, for instance, production of ethanol through
fermentation of wheat or cellulose (e.g. KAM, 2003; Nyström
et al., 2007).
The use of biomass (including waste and peat) accounted, in
2005, for 18%, or 112 TWh, of the total energy use in Sweden (SEA,
2006a). Main users of biomass for energy purposes are the DH and
industry sectors, which, in 2005, used 42 and 51 TWh, respectively
(SEA, 2006a). The major part of the biomass resources are residues
and by-products from forestry and forest products industry (paper
and pulp, sawmills, etc.). About 1% (1.5 TWh) originates from the
agriculture sector. Energy crops are grown on approximately
70 000 ha, or 2.5% of the total arable land in Sweden. About 70% of
this area is used to grow feedstock for production of biofuels,
either wheat for ethanol or rape for rape methyl ester (RME/
biodiesel), and 20%, or 14 000 ha, is used for cultivation of energy
forest, Salix (willow), which primarily is used for heat and
electricity production (SOU, 2007). The Swedish Salix industry
has a target that the area used for Salix cultivation in 2010 should
169
be increased to 25 000 ha, which is equal to 0.9% of the arable land
in the country (Statistics Sweden et al., 2007). Imports of biomass
and biofuels to Sweden have been estimated to about 5–9 TWh,
and are mainly composed of wood pellets for use in heat
production and ethanol for use in the transportation sector
(SEA, 2006a).
DH is the dominating heating technology in Sweden with a
market share of about 48% of the heating demand for residential
and commercial premises (SEA, 2006a). In 2005, the energy
supply for DH production (excluding electricity production in CHP
plants) totalled to 55 TWh, of which biomass (including waste and
peat) accounted for 66%. This is the result of a substantial increase
during the last decades, e.g. in 1970 the total energy supply to DH
production was 15 TWh, of which 98% was oil (SEA, 2006b). In the
DH sector, biomass is primarily used in boilers for heat production,
although in recent years the CHP production has risen somewhat
(SEA, 2006a). The so far comparably low level of CHP production in
the Swedish DH systems and the potential of increasing the
production were addressed in Knutsson et al. (2006b).
Energy products, such as heat, electricity and transport fuels,
are often influenced by different kinds of policy tools. One
example is the EU trading scheme for CO2 emission allowances,
or tradable emission permits (TEPs), which put a price on CO2
emissions from fossil fuels and, thereby, promotes the use of CO2-
free technologies and fuels. The system has been in operation
since January 2005 and includes in its initial stage a limited
number of sectors in energy-intensive industries and electricity
producers (SEA, 2006a). In Sweden, a CO2 tax was introduced in
1991, which, in 2005, had a level of 102 EUR/ton CO2.1 The
application of the tax is different, depending on where the CO2
emissions take place. It has been applied to full extent to the
transportation sector and for production of space heating and to a
lesser degree in the industry and for electricity production. Today,
the tax is partly being phased out to make way for the EU
emission trading system (EU ETS), although it is still in use for
sectors that are not included in the system (SEA, 2006a).
In order to stimulate renewable electricity production, a
market-based scheme, so-called tradable green certificates
(TGCs), has been introduced in Sweden. The TGC system has
been in operation since May 2003 and will be so until 2030
(SEA, 2006a). In the TGC system, a green certificate is given to the
electricity producer for each produced megawatt hour (MWh) of
renewable electricity. The certificate is sold on a market and thus
creates an extra income besides the income from the electricity
sales. All electricity consumers, except for energy-intensive
industries, are obliged to buy certificates in an amount corre-
sponding to a certain share of their total electricity consumption
and, thereby, a demand for the green certificates is created.
Renewable electricity production technologies, excluding large-
scale hydropower and electricity production from combustion of
municipal waste but currently including CHP production from
peat, are entitled to TGCs (SEA, 2006a). The yearly price averages
for the TGCs were between 2003 and 2006 in the interval
21–26 EUR/MWh (Svenska Kraftnät, 2007). Impacts and correla-
tion between TEPs and TGCs for the Nordic countries were
modelled and analysed in Unger and Ahlgren (2005).
Also the transportation sector in Sweden is influenced by
energy policies, for example, through fuel taxes. In 2006 the
fuel taxes, excluding VAT but including the earlier mentioned CO2
tax, were 0.55 EUR/l petrol (61 EUR/MWh) and 0.41 EUR/l
diesel (41 EUR/MWh). To increase the amount of biofuels in
the transportation sector, biofuels are currently exempt from
the taxation applied for petrol and diesel (SEA, 2006a).
1
Exchange rates used throughout this paper are 9 SEK= 1 EUR (= 1.14 USD).
 ARTICLE IN PRESS
170 M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180

An introduction of a system similar to the TGCs has also been
investigated (SOU, 2004). The EU directive 2003/30/EC states that
the amount of biofuels in the transportation sector in 2010 should
constitute 5.75% of the total transport fuel consumption in the EU
(European Commission, 2003). The target level for 2005 on 2%
was not met on an EU level, but in Germany and Sweden it was
met on a national level (European Commission, 2007). More
recently, a target for 2020 of 10% has been decided (European
Council, 2007).
The county of Västra Götaland is a part of Sweden located in
the southwest of the country. Västra Götaland has a population of
about 1.5 million people and the largest city is Göteborg. The
municipality of Göteborg has a population of about 0.5 million
people (Statistics Sweden, 2007). Of the total 49 municipalities in
Västra Götaland, in 2004 there were deliveries of DH in 34,
although in many of these in a very small quantity (Statistics
Sweden, 2006). The conurbation of Göteborg, or ‘‘Greater
Göteborg’’, includes 13 municipalities, all of which belong to
the county of Västra Götaland except for the municipality of
Kungsbacka, which is included in the county of Halland. In the
present study, the studied region is defined by the county of
Västra Götaland and Greater Göteborg. Table 1 summarizes the
production in the 15 largest DH systems, in terms of delivered
heat, in the studied region in 2004. The included systems in
Table 1 accounted, in 2004, for about 95% of the total DH
production in the studied region (Statistics Sweden, 2006). The
production of electricity from CHP plants amounted, in 2004, to
about 0.5 TWh. After 2004 several new CHP plants have been
taken into operation, mostly small-scale biomass ST plants but
also the 600 MWfuel NGCC plant Rya CHP.
In 2004, the biomass use for DH (including CHP) in the studied
region was about 2.5 TWh, excluding municipal waste and peat
(based on Swedish District Heating Association, 2006; Statistics
Sweden, 2006). For Västra Götaland, the possible potential
outtake of forest residues (tops, branches and needles) has been
estimated to about 1570 kton dry substance (DS). However,
considering practical, technical and environmental aspects only
about 970 kton DS, approximately 5 TWh, is reported to be a
reasonable potential available for energy purposes (about 70% of
the tops and branches, 30% of the needles) (Johansson, 2001).
The land used for Salix cultivation was, in 2001, less than 0.2% of
the total arable land in Västra Götaland (about 900 ha), which
resulted in a Salix production of approximately 40 GWh (Johansson,
2001).
3. Method and model
Computer-based models are powerful tools for analysis of
complex energy systems. Modelling offers the possibility of a
comprehensive and structured system representation and allows
technology options and their system suitability to be contrasted
and analysed. Furthermore, economical, environmental and
technical aspects of studied systems can be evaluated and
compared quantitatively under different conditions and scenarios.
A range of different types of models and modelling tools are
available. Except for the above-mentioned qualities, some model-
ling features are of special significance in this study. Since the DH
sector is in focus, the modelling approach regarding DH systems is
central. DH as a technology can be defined as the centralized
production of heat distributed as hot water in a piping system and
used for heating purposes in premises (SEA, 2006a). One of the
major benefits with DH is the possibility of making use of local
resources, such as heat released from incineration of municipal
waste or wood waste and heat from process industries or power
generation that might otherwise be wasted (Knutsson, 2005).
Because of its characteristics, DH is, in another way than, e.g.,
electricity, geographically restricted in small local systems, all
with their own specific prerequisites. In modelling studies,
however, several DH systems are often aggregated, e.g. in Unger
and Ahlgren (2005) a model of the Nordic energy system is used in
which the Swedish DH sector is represented as one single system.
Although this, depending on the study’s objective, can be a
passable way, Knutsson et al. (2006a) concluded that the level of
aggregation of DH systems can have an important impact on the
results of the study. In the present study, a model that enables
representation of a number of individual DH systems is preferable.
Furthermore, a high level of detail and flexibility concerning the
description of energy technologies, current and future, and energy
carriers is necessary. Since the cost-effectiveness of new technol-
ogies and the optimal use of biomass resources are studied, a
dynamic cost-optimizing model is an appropriate choice.
The MARKAL (MARKet ALlocation) model (Lolou et al., 2004),
developed within the International Energy Agency’s energy
technology system analysis programme (ETSAP), has the proper-
ties required in the present study. Applications of the MARKAL
model are generally linear programming (LP), cost-optimizing
bottom-up models with perfect foresight (no insecurity about
future development). An objective function minimizes the total
system cost (the objective value) within the limits outlined by an

Table 1
Delivered DH, net production of electricity in CHP plants and energy sources for DH (incl. CHP) production, for the largest DH systems in the studied region, i.e. Västra
Götaland and Greater Göteborg, in terms of amount delivered DH (data from 2004 except for Kungsbacka: data from 2003).

DH system DH (delivered) Electricity (production) Biomass fuels Peat Municipal solid Heat pump Ind. waste heat Natural gas Oil Other
(GWh) (GWh) (%) (%) waste (%) (%) (%) (%) (%) (%)

Göteborg 3843 320 11 30 12 29 17 1

Borås 591 120 77 9 4 10
Mölndal 275 71 14 15
Trollhättan 269 100
Uddevalla 264 36 52 1 11
Lidköping 252 18 15 80 4 2
Skövde 246 89 1 10 1
Vänersborg 150 10 90
Kungsbacka 113 2 61 14 26
Mariestad 103 25 96 4
Alingsås 101 98 2
Kungälv 97 95 1 3
Falköping 88 98 2
Mark 80 90 10
Skara 60 90 10

In the table, heat pumps are represented by produced DH, while other alternatives are represented by the input energy.
Based on Swedish District Heating Association (2006) and Statistics Sweden (2006); for the Göteborg system also Göteborg Energi (2005) and Renova (2005).
 ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180
often large number of linear constraints. The achieved solution
represents the lowest system cost alternative, among the many
choices that are possible in the reference energy system (RES)
structured network of energy technologies, that fulfils the
exogenously given energy demands. The network of energy
technologies is user specified and can cover the whole chain
from fuel extraction to end-use energy services. In the present
study, the MARKAL model is applied to a database of the DH
sector in Västra Götaland and Greater Göteborg forming a model
here referred to as the MARKAL_WS model (WS= West Sweden).
The MARKAL_WS model has a time horizon ranging from 2004
to 20292 divided into six model periods, with each being
represented by one model year (2004, 2009, 2014, 2019, 2024,
2029). Each model year is furthermore divided into three seasons:
winter, summer and spring/autumn (intermediate). The DH sector
is, in the MARKAL_WS model, represented by the 15 DH systems
that are presented in Table 1. As previously stated, these systems
account for the main part of the DH in the studied region
(approximately 95%). In the model, the supply/production side of
the included DH systems is described in great detail with a large
number of available energy technology options while other parts
of the energy system, like fuel extraction and end-use technolo-
gies, are described in a less-detailed way.
Each DH system in the model is driven by an individual
demand for DH, which is assumed inelastic to price changes. For
DH systems where estimations of future DH demand levels have
been available, these are used.3 None of these estimations cover,
however, the entire studied time period, and for the remaining
time period the demands are assumed constant. For DH systems
where no estimations on future demand levels have been
available, the delivered DH of 2004 is used as the DH demand
throughout the entire studied time period. Electricity production
from CHP plants is, in the model, sold at exogenously given market
prices and creates an income to the system and, thereby, a
lowered system cost. Consequently, there is no electricity demand
that needs to be fulfilled. Biofuels for transport are treated in a
similar way; they can be sold at exogenously given market prices
but no specific transport fuel demand exists in the model.
The total system cost (the objective value) could thus be described
as the cost for the system’s DH production when credited for the
income of electricity and biofuels sales.
In the present study, the so-called ‘‘lumpy’’ investment option
in MARKAL is used. This turns the MARKAL_WS model into a
mixed-integer programming (MIP) model rather than a strict LP
model. With MIP, selected technologies can be restricted to be
available only for implementation at discrete capacity levels as
opposed to be available at any capacity level without economies
or diseconomies of scale (linearly).
3.1. Modelling introduction of biofuels for transport
It has earlier been shown that the cost-effectiveness of
reducing CO2 through use of biofuels in the transportation sector
is low compared to CO2 reductions in the stationary energy sector
(e.g. Azar et al., 2003). This indicates that an equivalent cost of CO2
emissions in all sectors probably does not give enough economic
incentives to create a large-scale introduction of biofuels for
transport in the near-term future. To reach current ambitions
regarding biofuels (e.g. European Council, 2007) it is therefore
likely that some kind of sector-specific policy tool promoting
2 4
All costs and prices related to the model description and model results are
expressed in monetary value of 2004.
3
Estimations are primarily based on data received from DH companies for
projects presented in OPET and SEA (2004) and Nyström et al. (2007).
171
biofuels will be necessary under the time horizon of interest in the
present study.
In the model, a biofuel subsidy defined as an extra income for
biofuel production in addition to the income from the fuel sales is
used. To study what impact different levels of biofuel production
have on a cost-optimized system, an iterative modelling proce-
dure is applied. Each model case is in a first model run tested
without the biofuel subsidy present (subsidy level=0). Thereafter,
the level of the subsidy is gradually raised in a series of model
runs. In the model, the biorefineries are only available in discrete
capacity levels (see also Section 3.2) and, therefore, biofuel
subsidy threshold levels at which new biofuel plants are
introduced will be found. These threshold values, each corre-
sponding to a certain capacity of biofuel production, are outputs
of the study. Within each model run, the subsidy level is held
constant throughout the whole studied time period.
In the study, it is not analysed what kind of policy tool would
be the most suitable option in reality. The subsidy in the model
could be interpreted as a tax reduction or some other kind of
financial support for biofuel production. Except for the biofuel
subsidy and a cost for CO2 emissions, biofuels and conventional
transport fuels (petrol/diesel) are assumed to be subject to equally
repressive policy measures (energy taxes, etc.), and these are,
therefore, disregarded in the model.
3.2. Energy technologies
The 15 different DH systems included in the model are
described by their current production plants as well as a number
of possible investment alternatives in new production capacity.
The investment costs and other fixed costs for the set of existing
plants are treated as sunk costs and, accordingly, only the variable
costs are accounted for in the model. The existing plants are given
an assumed remaining technical lifetime based on when the plant
was taken into operation.
Investments in new production capacity can be done from
model year 2014. The options for investment in new plants differ
between the DH systems due to local conditions, e.g. size of the
DH demand or availability of natural gas, and include a number of
conventional production technologies, see Table 2, as well as
biomass gasification technologies, see Table 3. The biomass
gasification technologies included are BIGCC CHP plants and
biorefineries for production of biofuels for transport. In this study,
biofuels are represented by SNG and DME although, as previously
mentioned, in reality also other fuels can be produced through
gasification.
The conventional production technologies can be implemented
at any capacity level, i.e. linearly. However, since the DH systems
differ widely in size, the typical plant sizes also differ between the
DH systems. To take economies of scale and the fact that different
plant sizes have different properties into account, there are,
for the same type of technology (Biomass ST CHP, NGCC CHP,
etc.), different input data in different DH systems, see Table 2.
The gasification technologies are handled in another way. In the
model, these technologies are available at discrete capacity levels
and can only be implemented in the specific plant sizes presented
in Table 3. The different size options available for biomass
gasification technologies in each DH system are, therefore, not
linked to the size of the DH system.4
Due to the iterative modelling procedure with a large number of model runs
applied in the study, the option to only allow discrete capacity levels also for
conventional technology options was rejected since this substantially increases the
model solving time.
 ARTICLE IN PRESS
172 M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180

Table 2
Conventional energy technologies available as investment options in model.

Technology Efficiencya Investment cost Fixed O&M cost Variable O&M cost Available in DH systemsb
Combined heat and power plantsc Electricity/total (%) (MEUR/MWelectr.) (% of inv. cost/year) (EUR/MWhfuel)

Biomass ST CHP 1 25/105 3.9 2 2.6 S

Biomass ST CHP 2 27/110 2.3 2 2.6 M
Biomass ST CHP 3 30/110 1.7 2 2.6 L
Biomass ST CHP 4 34/110 1.3 2 2.6 XL
Natural gas CC CHP 1 45/89 1.0 2 0.9 S (NG)
Natural gas CC CHP 2 46/89 0.90 2 0.9 M (NG)
Natural gas CC CHP 3 49/90 0.73 2 0.9 XL (NG)
Natural gas engine CHP 40/88 0.67 4.2 All (NG)
Waste ST CHP 1 22/95 6.2 3 13.9d M (W), L (W)
Waste ST CHP 2 27/95 4.5 3 13.9d XL (W)
Coal ST CHP 1 25/89 3.7 2.5 3.3 S
Coal ST CHP 2 27/89 2.2 2.5 3.3 M
Coal ST CHP 3 30/89 1.7 2.5 3.3 L
Coal ST CHP 4 34/89 1.3 2.5 3.3 XL

Heat plantse Heat (%) (MEUR/MWheat) (% of inv. cost/year) (EUR/MWhfuel)

Biomass HOB 1 110 0.40 2 2.0 S

Biomass HOB 2 110 0.36 2 2.0 M
Biomass HOB 3 110 0.33 2 2.0 L, XL
Coal HOB 1 89 0.37 2.5 2.7 S
Coal HOB 2 89 0.33 2.5 2.7 M
Coal HOB 3 89 0.30 2.5 2.7 L, XL
Oil HOB 1 90 0.20 2 0.7 S
Oil HOB 2 90 0.17 2 0.7 M
Oil HOB 3 90 0.16 2 0.7 L, XL
Natural gas HOB 1 90 0.11 2 0.7 S (NG)
Natural gas HOB 2 90 0.10 2 0.7 M (NG)
Natural gas HOB 3 90 0.91 2 0.7 XL (NG)
Waste HOB 1 95 1.1 3 15.6d M (W)
Waste HOB 2 95 1.0 3 15.6d L (W), XL (W)
Electric heat pump 320 (COP) 0.60 0.5 0.7 All

The technologies may be implemented at any capacity level without economies or diseconomies of scale (linearly); however, the same technology type has somewhat
different properties in different DH system since typical plant sizes between the DH systems differ.
a
Efficiencies on LHV (lower heating value) basis.
b
Classification of DH systems (see also Table 1): S=small: Skara, Mark, Falköping, Kungälv (NG), Alingsås (NG), Mariestad, Kungsbacka (NG), Vänersborg; M =medium:
Skövde (W), Lidköping (W), Uddevalla (W), Trollhättan, Mölndal (NG); L= large: Borås (W); XL = extra large: Göteborg (NG; W); All: S, M, L and XL; NG= DH systems with
natural gas grid; W=DH systems with permission for incineration of municipal waste.
c
Plant properties for combined heat and power plants are primarily based on Bärring et al. (2003); complimentarily on Gustavsson (1997) and Danish Energy Authority
et al. (2005). For investment costs, scaling has been used with the general relationship: Cost2=Cost1(Size2/Size1)a, where the scale factor a for each technology type has
been estimated from Bärring et al. (2003).
d
Including income from waste disposal fee, assumed to 22 EUR/MWhwaste.
e
For heat plants, efficiencies are based on Ekström et al. (2002), Danish Energy Authority et al. (2005), Knutsson et al. (2006b). Investment costs are based on Ekström
et al. (2002) and Danish Energy Authority et al. (2005). A scale factor a of 0.87 estimated from Gustavsson et al. (1995) has been used (see c). For fixed O&M the same
percentage of investment cost per year as for CHPs are assumed. Variable O&M is based on Knutsson et al. (2006b).

Except for the technology options summarized in Tables 2
and 3, the model also includes the option to convert existing
biomass heat-only boilers (HOBs) to CHP production. In reality,
several different concepts are available for such rebuilding,
and the costs differ from case to case, e.g. depending on how
the HOB is constructed and which power-to-heat ratio is being
aimed for (OPET and SEA, 2004). In the model, the rebuilding cost
is assumed to be 1000 EUR/kWe for a power-to-heat ratio of 0.15
(based on OPET and SEA, 2004). As for the biomass gasification
technologies, this technology option is only available in discrete
size levels, which are individually set by each existing convertible
biomass HOB. In the model, the technology option is available in 7
of the 15 included DH systems.
A few additional model assumptions regarding energy tech-
nologies could be mentioned. Industrial waste heat and combus-
tion of municipal waste are treated in a somewhat simplified way
in the model since these options are not of central interest in the
study. The capacity and use of industrial waste heat are assumed
constant at the level of 2004 for the entire studied time period.
The availability of municipal waste is, for each DH system during
the entire studied time period, limited corresponding to the
currently valid, by authorities determined, permitted levels for
municipal waste incineration (see Swedish EPA, 2005). Further-
more, the natural gas grid, available in five of the included DH
systems, is in the model not possible to extend. Natural gas-fired
plants and production plants for SNG are only available options in
DH systems connected to the natural gas grid. Since the DH sector
is in focus, power plants without DH production are not at all
included in the model.
3.3. CO2 tax and promotion of ‘‘green’’ electricity
In addition to the biofuel subsidy, described in Section 3.1, the
study operates with two different policy tools: a cost for CO2
emissions and promotion of biomass-based electricity, here
referred to as ‘‘green’’ electricity. Unlike the biofuel subsidy, these
policy tools are given to the model in a completely exogenous way,
without feedback from the modelling results.
The cost laid upon emissions of CO2 is denoted ‘‘CO2 tax’’,
although the levels used in the study to a higher degree resemble
the levels of the EU ETS than the Swedish CO2 tax. In contrast to
 ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180 173

Table 3
Biomass gasification technologies available as investment options in model.

Technology Capacity Efficiencya Investment costb Fixed O&M costc Variable O&M costc Available in DH systemsd
Combined heat (MWelectr./MWheat) Electricity/total (%) (MEUR/MWelectr.) (% of inv. cost/year) (EUR/MWhfuel)
and power plants

BIGCC CHP 1 10/12 40/86 2.6 2.5 3.3 All

BIGCC CHP 2 30/31 42/86 1.9 2.5 3.3 All
BIGCC CHP 3 60/60 43/86 1.5 2.5 3.3 All
BIGCC CHP 4 100/100 43/86 1.3 2.5 3.3 All
BIGCC CHP 5 130/130 43/86 1.2 2.5 3.3 All

Biofuel production plants (MWbiofuel/MWheat) Biofuel/total (%) (MEUR/MWbiofuel) (% of inv. cost/year) (EUR/MWhfuel)

Biorefinery SNG 1 86.5/29 69/92 (63/85) 1.7 3.5 3.3 All (NG)
Biorefinery SNG 2 173/58 69/92 (63/85) 1.4 3.5 3.3 All (NG)
Biorefinery DME 1 79/19 62/77 (55/68) 2.2 3.5 3.3 All
Biorefinery DME 2 158/38 62/77 (55/68) 1.8 3.5 3.3 All

The technologies are in the model only available at the discrete capacity levels presented in the table (not linearly).
a
All efficiencies are on LHV basis. Efficiencies are, for BIGCC CHP plants, based on Gustavsson and Madlener (2003), Marbe et al. (2004), Uddin and Barreto (2007) and,
for biorefineries, on Ingman et al. (2006). The biorefinery efficiencies without brackets are defined as: biofuel produced divided by biomass and electricity consumed; and:
biofuel and heat produced divided by biomass and electricity consumed. For comparison, the corresponding efficiencies if the consumed electricity was produced internally
within the plant with an electrical efficiency of 30% are presented within brackets.
b
Investment costs are for BIGCC CHP plants based on: Gustavsson and Madlener (2003), Marbe et al. (2004); for biorefineries on: Ingman et al. (2006). Scaling has been
used with the general relationship: Cost2= Cost1(Size2/Size1)a. A scale factor a of 0.7 has been used for both BIGCC and biorefineries, see e.g. Marbe et al. (2004).
c
Based on Marbe et al. (2004), Ingman et al. (2006).
d
For DH systems classification, see Table 2.

both the EU ETS and the Swedish CO2 tax (see Section 2), the CO2
tax used in the present study applies equally to all CO2 emissions
from non-renewable sources5 regardless of sector, i.e. emissions
from heat and power production as well as from transportation
are valued the same. Due to the assumed continuously higher
ambitions regarding CO2 emission reductions, the CO2 tax is
assumed to increase during the studied period. As a model base
assumption, the level rises linearly from 20 EUR/ton CO2 in model
year 2009 to a level of 40 EUR/ton CO2 at the end of the model
time horizon, model year 2029. The subsidy level on green
electricity production is, as a base assumption, set to 20 EUR/
MWhe, comparable with the levels in the Swedish TGC system
from past years (see Section 2). The subsidy level is constant
throughout the entire studied time horizon.
3.4. Energy markets and prices
In order to keep the assumptions transparent, the fossil fuel
prices excluding the CO2 tax are constant throughout the entire
studied time period (in real terms). As base assumptions, the
international market prices for crude oil, natural gas and coal are
assumed to be 26, 20 and 7 EUR/MWh (50, 39 and 14 USD/barrel
of oil equivalents), respectively (primarily based on Mantzos and
Capros, 2006). These international prices are recalculated to prices
for different commodities (light fuel oil, heavy fuel oil, petrol, etc.)
and end-users (DH companies and consumers at filling stations)
with assumed price differences based on statistics (SPI, 2007;
World Bank, 2007; Dong Energy, 2007). Since there is no demand
for transport services in the model, petrol and diesel are not
explicitly included as energy carriers. Petrol and diesel are,
however, assumed to be price-setting for biofuels for transport
(SNG and DME), which are included in the model. It is assumed
that biofuels at the filling stations could be sold at the same price
(on energy basis) as petrol and diesel, including CO2 tax. This
implies that possible differences between the fuels that could
5
Fossil fuels and peat are here defined as non-renewable while biomass and
municipal waste are defined as renewable.
justify differentiated pricing, e.g. differences in vehicle price or
vehicle energy efficiency, are neglected.
The Swedish electricity price is mainly determined by the
variable cost of the marginal production technology in the Nordic
electricity system. For some time, the marginal production
technology has regularly been coal condensing power generation,
but NGCC could also be a possible future marginal production
technology in the system (e.g. ECON, 2002). In the model, it is
assumed that the price-setting technology for electricity is either
NGCC or coal condensing power generation. The lowest variable
electricity production cost of these technologies is assumed to
determine the model electricity price. The production costs
depend on the fossil fuel prices, which are constant throughout
the studied time period, and the CO2 tax, which increases. The
electricity price is divided into three periods per year following
the model seasons used. Due to the variation in electricity
demand over the year (low demand in summer, high demand in
winter) and the difference in efficiencies between the coal
condensing power plants in the system, the efficiency of the
possible coal condensing marginal production technology is
assumed to vary between the model seasons. The difference in
efficiencies between NGCC plants is assumed to be small and is
neglected. Hence, in the model, for each season the price-setting
electricity production technology is assumed according to the
following:
 Winter: coal condensing plants with electrical efficiency
(Ze)= 35%, or NGCC plants with Ze =55%.
 Spring/autumn: coal condensing plants with Ze =40%, or NGCC
plants with Ze = 55%.
 Summer: coal condensing plants with Ze = 45%, or NGCC plants
with Ze = 55%.
In the model, different supplies of biomass are assumed: a
supply of wood chips at a lower cost and a supply of refined
biomass like wood briquettes or pellets at a slightly higher cost.
The supply of wood chips is limited by assumptions on the local/
regional potential of forest residues and energy crops in the form
 ARTICLE IN PRESS
174 M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180
of Salix. Regarding forest residues, a higher supply limit of
970 kton DS, equivalent to 5.2 TWh, is assumed from model year
2014 (based on the estimated potential in Västra Götaland, see
Section 2). Regarding energy crops, the share of arable land used
for Salix cultivation is, in model year 2009, assumed to be equal to
the national target level of 0.9% (see Section 2), which, for the
studied region, corresponds to approximately 4400 ha. Further-
more, it is assumed that the cultivated area increases with 80%
every model period (5 years), analogous to the targeted increase
on national level from 2005 to 2010. Consequently, about 9% of
the total arable land in the studied region, or 45 000 ha, is
assumed to be used for cultivation of Salix in the last model year
(2029). It is further assumed that the Salix production is 60 MWh/
ha, which is in the upper range of today’s yields from new Salix
plantations (Statistics Sweden et al., 2007). At the end of the
studied period, this results in an upper constraint on wood
chips from energy crops of about 2.7 TWh. Since refined biomass
like wood briquettes and pellets are less bulky than wood chips
and can be more easily transported, an international supply is
assumed, which, from the perspective of the studied region, is
considered to be unlimited.
In the beginning of the studied period, the biomass prices are
assumed to be approximately equal to the Swedish average prices
of wood chips and wood pellets/briquettes in 2003–2006 (SEA,
2007). As the CO2 tax rises, the demand for biomass is likely to
increase and eventually cause biomass prices to rise. It is assumed
that it will be favourable to substitute biomass for coal in coal
power plants and that the price of biomass under high CO2 taxes
will be connected to the coal price (see also Axelsson et al., 2007).
The price of wood chips is, therefore, in the later part of the
studied time period, assumed to be equal to the price of coal,
including CO2 tax. The price difference between wood chips and
wood pellets is assumed to be constant throughout the whole
studied period and is determined by the approximate price
difference in 2003–2006 (SEA, 2007). Assumptions regarding
biomass prices and availability of the local supply of wood chips
are visualized in Fig. 1.
3.5. Distribution of biofuels for transport
As previously mentioned, biofuels for transport are, in this
study, represented by two options, DME and SNG. Since the
distribution costs for these fuels will differ from conventional
transport fuels (petrol/diesel), distribution costs are included in
the model.
DME have characteristics similar to LPG (liquefied petroleum
gas) and can be transported on trucks. Since DME is in gaseous
28
26
24 Price Wood Pellets
Availability Energy
22
[EUR/MWh]
Crops (Salix)
20
18
Price Wood Chips
16
Availability
14 Forest Residues
12
Price Hard Coal incl. CO2 tax
10
2004 2009 2014 2019 2024
Fig. 1. Base assumptions regarding availability of local, lower cost biomass (wood
chips) as well as price levels for biomass (wood chips and wood pellets) and coal
including CO2 tax.
phase under atmospheric pressure, the handling of the fuel needs
to be done under a certain overpressure for the fuel to be in liquid
phase (5–10 bar). This implies that the distribution cost probably
will be higher than for petrol and diesel. In this study, a DME
distribution cost of 22 EUR/MWh is assumed. This includes both
the costs for transport from the production site to the filling
station and the costs at the filling station, and is approximately
twice the corresponding cost for diesel (based on Brandberg et al.,
1997).
Transportation of SNG can be done in different ways, e.g. in
pipelines or in compressed form on trucks. Since several parts of
the distribution chain for SNG are relatively capital intensive, not
least the filling stations that are significantly costlier than for
liquid fuels, the amount of delivered fuel should be a factor of
significance for the cost-effectiveness of the distribution. In the
model, two different distribution costs for SNG are assumed: a
lower one at 28 EUR/MWh6 and a higher one at 39 EUR/MWh.7
The lower cost level represents distribution in the natural gas grid,
available in parts of the studied region, and distribution at large
filling stations (in terms of amount delivered fuel). The higher cost
level represents truck distribution in combination with smaller
filling stations. The lower cost distribution is, in the model,
limited while the higher cost distribution is not. The limitation of
the lower cost distribution is in model year 2009 assumed to be
0.43 TWh and thereafter is assumed to rise linearly and in model
year 2029 reach a level of 1.73 TWh. This corresponds to an
increase from 2.5% to 10% of the total transport fuel use in the
studied region in 2004, or an increase from 5% to 20% of the total
transport fuel use in the municipalities with access to the natural
gas grid (SEA and Statistics Sweden, 2005).
3.6. Model cases analysed
Except for the reference case (‘‘REF’’), in which the model base
assumptions described in previous sections apply, several addi-
tional cases with alternative assumptions are included in the
analysis. The alternative cases investigate how changes in certain
parameters one at a time affect the model outcome compared to
the reference case. In each alternative case, one parameter value is
altered while all other parameter values are kept the same, with
the exception for those parameters directly affected by the change
according to assumed relationships described in earlier sections.
The chosen cases represent a selection of relevant variations in
parameter values for which future levels are highly uncertain. For
all model cases, the subsidy levels required in order to obtain
different numbers of investments in biorefineries, and thereby to
reach certain amounts of transport biofuel production, are
assessed according to the procedure outlined in Section 3.1.
9 In Table 4, all analysed cases are presented and important
input parameter values given. In case ‘‘NG Lo’’, the international
8
price for natural gas is assumed to be 15 EUR/MWh instead of 20
7 EUR/MWh as in the base assumptions. A lower natural gas price
6 also affects the electricity price (see Section 3.4). Alternative
assumptions for policy tools are made in cases ‘‘GE Lo’’ and
[TWh]
5
4 6
Cost for distribution of SNG in the natural gas grid is assumed to 15 EUR/
3 MWh (based on Ekengren, 2006), investment cost for an SNG filling station is
assumed to 0.4 MEUR, and filling station operation and maintenance cost is
2 assumed to 8 EUR/MWh (based on Petterson et al., 2006). The annual amount of
1 delivered fuel is assumed to 8 GWh per filling station (fuel supply to about 700
cars), and an annuity factor of 0.096 (15 years, 5%) is used.
7
0 Cost for distribution of compressed SNG on truck is assumed to 18 EUR/MWh
2029 (based on Petterson et al., 2006), investment cost for an SNG filling station is
assumed to 0.4 MEUR, and filling station operation and maintenance cost is
assumed to 8 EUR/MWh (based on Petterson et al., 2006). The annual amount of
delivered fuel is assumed to 3 GWh per filling station (fuel supply to about 250
cars), and an annuity factor of 0.096 (15 years, 5%) is used.
 ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180 175

Table 4
Important model parameter values regarding policy tools, energy prices and biomass gasification plant investment costs for the reference case (REF) and for the alternative
cases.

REF NG Lo GE Lo BIO Lo OIL Hi CO2 Hi INV Hi

Policy tools
CO2 tax (EUR/ton CO2) 20/40 30/60
Subsidy Green Electricity (EUR/MWh) 20 10

Energy prices
Natural gas (EUR/MWh) 24.7 19.4
Hard coal (EUR/MWh) 7.2
Fuel oil, light (EUR/MWh) 33.8 46.7
Fuel oil, heavy (EUR/MWh) 27.7 40.6
Petrol/diesel (EUR/MWh) 44.8 57.8
a
Wood chips (EUR/MWh) 15.6/20.4 17.1/26.9
Wood pellets (EUR/MWh) 22.2/27.0 23.7/33.6
Electricity Winter (EUR/MWh) 49.0/61.2 44.3/51.7 57.6/68.6
Electricity Intermediate (EUR/MWh) 42.9/59.3 42.9/51.7 51.1/68.6
Electricity Summer (EUR/MWh) 38.1/52.7 38.1/51.7 45.4/67.3

Investment costs
BIGCC CHP 1–5 See Table 3 100% 125%
Biorefinery SNG 1–2 See Table 3 100% 125%
Biorefinery DME 1–2 See Table 3 100% 125%

For the alternative cases, only the parameter values that differ from the base assumptions (in case REF) are given. The fuel prices refer to prices paid by DH plants except for
the price of petrol/diesel, which refer to the price at filling stations (without taxes). The fossil fuel prices are given excluding the CO2 tax. If the parameter value is not
constant throughout the whole studied time period, the values for model year 2009 and 2029 are given (separated with /).
a
Availability of wood chips is unlimited.

‘‘CO2 Hi’’. In GE Lo, the subsidy promoting green electricity
production is 50% lower than in the base assumptions, and in CO2
Hi, the CO2 tax is 50% higher than in the base assumptions. Except
for having an impact on the fossil fuel prices including the cost of
CO2 emissions, a modification of the CO2 tax also changes the
biomass prices as well as the price of electricity (see Section 3.4).
In case ‘‘BIO Lo’’, the restrictions on the availability of the lower
cost biomass are disregarded. In case ‘‘OIL Hi’’, the price of crude
oil is 75 USD/barrel instead of 50 USD/barrel as in the base
assumptions, i.e. 39 instead of 26 EUR/MWh. In case ‘‘INV Hi’’, the
uncertainty in investment costs for biomass gasification-based
plants is addressed; the investment costs for these technologies
are here assumed to be 25% higher than in the base assumptions.
A discount rate of 5% is used in all cases.
4. Results
In this section, the results of the model calculations are
presented. First, the reference case is analysed and, thereafter, the
impact on the model results from the parameter variation in the
alternative cases are considered. In the last part of the section, CO2
emissions associated with the studied system are examined.
Fig. 2a–d shows the DH production in the reference case,
classified according to technologies for different levels of the
subsidy for transport biofuels. Fig. 2a illustrates the DH
production without biofuel subsidy. As expected, there is no
biofuel production in a cost-optimized system under these
conditions. The DH production is, in the first part of the studied
time period, to a large degree based on HOBs but, with time, CHP
production increases while HOBs are phased out. Heat pumps
(HPs) account, in the beginning of the period, for a relatively large
share of the DH production while in the end of the period the
production from HPs is negligible. It is primarily different types of
biomass CHP production plants that account for a production
increase. Of the existing biomass HOBs, a large share is converted
into CHP units, and there is a significant increase in the
production from conventional biomass ST CHP. CHP generation
through biomass gasification in BIGCC CHP plants is, under the
assumed conditions, a competitive technology alternative and
accounts for a considerable part of the DH production. Until the
last model year (2029), production from BIGCC CHP exclusively
takes place in the biggest DH system, Göteborg, in which two
130 MWe BIGCC CHP plants are invested in. In the last model year,
two smaller BIGCC CHP plants, 10 and 30 MWe, are also included
in the small-/medium-sized DH systems Kungsbacka and
Trollhättan, respectively. Also production from municipal waste
CHP is to some extent increased during the studied time period,
but the assumed permitted level of municipal waste incineration
constitutes a constraint for further increase.
With the base assumptions, which apply in the reference case,
the subsidy threshold level at which production of biofuels begins
to take place (in 2014) is calculated to 36 EUR/MWh. Fig. 2b
illustrates the DH production in the reference case with the
biofuel subsidy at this level. The introduced biofuel production
capacity is an SNG plant, placed in Göteborg. Due to economies of
scale, the bigger of the two discrete plant sizes available for this
technology (see Table 3) is chosen (173 MWSNG). The primary
impact on the system because of the introduction of biofuel
production is that less investment in the BIGCC CHP technology
takes place compared to situations with lower biofuel subsidies.
This time, a single 130 MWe BIGCC CHP plant is included in the DH
system of Göteborg, and in the end of the studied period, also in
this case, two smaller BIGCC CHP plants are included in the
previously mentioned two small-/middle-sized DH systems. The
DH production from BIGCC CHP is primarily replaced, except by
the waste heat from the SNG plant, by a larger production from
conventional biomass ST CHP plants. The main reason for the
change from BIGCC CHP to biomass ST CHP with an increasing
biofuel subsidy is the limitations in the local, lower cost, supply of
biomass (forest residues and energy crops). When a large part of
the available lower cost biomass is allocated to biofuel production,
the heat efficiency of biomass CHP technologies becomes of
greater significance since a low use of the costlier biomass (wood
pellets) is strived for. Conventional biomass ST CHP is, therefore,
relatively more competitive compared to BIGCC CHP in a situation
with biofuel production (Fig. 2b) than in a situation without
biofuel production (Fig. 2a).
 ARTICLE IN PRESS
176 M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180

10 Other 10 Other
Peat HOB Peat HOB
9 HP NG HOB 9 HP NG HOB SNG Biorefinery

8 8
7 7
NG CHP BIGCC CHP NG CHP BIGCC CHP
6 6
[TWh]

[TWh]
5 Bio ST CHP 5 Bio ST CHP
4 Bio HOB Conv. Bio HOB to CHP 4 Bio HOB Conv. Bio HOB to CHP

3 3
Ind. Waste Heat Ind. Waste Heat
2 2
Waste ST CHP 1 Waste ST CHP
1
Waste HOB Waste HOB
0 0
2004 2009 2014 2019 2024 2029 2004 2009 2014 2019 2024 2029
No biofuel subsidy Biofuel subsidy = 36 EUR/MWh
10 10
Other Peat HOB DME Biorefinery Other Peat HOB DME Biorefinery
9 HP NG HOB SNG Biorefinery 9 HP NG HOB SNG Biorefinery
8 8
7 7
NG CHP NG CHP
6 BIGCC CHP 6

[TWh]
[TWh]

5 Bio ST CHP 5 Bio ST CHP

4 Bio HOB Conv. Bio HOB to CHP 4 Bio HOB Conv. Bio HOB to CHP

3 3
Ind. Waste Heat Ind. Waste Heat
2 2
Waste ST CHP 1 Waste ST CHP
1
Waste HOB Waste HOB
0 0
2004 2009 2014 2019 2024 2029 2004 2009 2014 2019 2024 2029
Biofuel subsidy = 38 EUR/MWh Biofuel subsidy = 39 EUR/MWh

Fig. 2. Production of DH in the reference case for different levels of the biofuel subsidy and, thereby, different numbers of biorefineries included in the system: (a) no
biorefineries, (b) one biorefinery, (c) two biorefineries and (d) three biorefineries

The same general features can be noted when the biofuel
subsidy is raised further. In Fig. 2c, the DH production in the
reference case with a biofuel subsidy of 38 EUR/MWh is
illustrated. At this subsidy level, in addition to an SNG plant also
a DME plant is invested in. The DME plant has a capacity of
158 MWDME and is placed in Borås, the second largest DH system
in the studied region. The lower cost quota of the SNG distribution
(see Section 3.5) is at this point filled, and this is the main reason
for a DME plant instead of a second SNG plant. There is, under
these conditions, place only for one BIGCC CHP plant in the
system, a 130 MWe plant placed in Göteborg. Heat production
from the biorefineries as well as from HPs constitutes a larger
share of the DH production compared to cases with a lower
biofuel subsidy.
At a subsidy level of 39 EUR/MWh, three biorefineries for
production of biofuels for transport are invested in, one SNG plant
and two DME plants. The second DME plant (again with a capacity
of 158 MWDME) is placed in the DH system of Mölndal. Fig. 2d
illustrates the cost-optimized DH production under these condi-
tions. With this amount of biofuel production in the system, no
investments in BIGCC CHP take place. Instead, conventional
biomass ST CHP gets a major share of the DH production.
Fig. 3a and b illustrates the electricity production, electricity
use, production of biofuels and the corresponding required biofuel
subsidy level, for all analysed cases for model year 2019. For each
case, four different threshold levels for the biofuel subsidy and,
thereby, four different amounts of produced biofuel are calculated.
In all tested cases, the same biorefineries as in the reference case
are invested in and utilized to full extent. Consequently, all cases
have the same levels of biofuel production. In the figures, the
different biofuel production levels within each case are labelled
a–d; ‘‘a’’ denotes a situation without biofuel production (0 plants),
‘‘b’’ denotes a situation with a biofuel production of about 1.4 TWh
(one SNG plant), ‘‘c’’ denotes a situation with a biofuel production
of about 2.6 TWh (one SNG and one DME plant), and ‘‘d’’ denotes a
situation with a biofuel production of about 3.9 TWh (one SNG
and two DME plants). For comparison, it can be mentioned that
the biofuel production level in ‘‘d’’ corresponds to about 22% of the
total transport fuel use in the studied region in 2004 (SEA and
Statistics Sweden, 2005). The electricity use refers only to the
electricity use in plants connected to DH and the produced
electricity refers only to the production of electricity in CHP plants
connected to DH (other electricity use in the region and electricity
production in power plants not connected to DH is not included in
the model). Note that Fig. 3a and b illustrates the same results but
arranged differently: by biofuel production level (a–d) in Fig. 3a
and by model case (REF, NG Lo, GE Lo, BIO Lo, OIL Hi, CO2 Hi,
INV Hi) in Fig. 3b.
The reference case and the alternative cases show the same
kind of results regarding the trade-off between biofuels and
electricity. All cases have in common that as the production of
biofuels increases due to higher subsidies, the production of
electricity decreases and at the same time the use of electricity
increases. As previously stated for the reference case, it is
particularly the conditions for investments in BIGCC CHP that
deteriorate with increasing biofuel subsidy, mainly due to the
thereby following increased competition for the local, lower cost,
biomass. The increased use of electricity is primarily because of
the electricity use in biorefineries and in HPs. In most cases, no
BIGCC CHP plants are included in the system by a subsidy level
that allows three biorefineries.
Lower natural gas prices could lead to lower electricity prices,
and lower electricity prices decrease the incentives for BIGCC CHP.
This situation is observed in case NG Lo. Already with one
 ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180 177

Use El. Biofuel Fossil El. Bio El. BIGCC El. Biofuel Sub.
8000 45

7000 40

6000

Subsidy Biofuel [EUR/MWh]

35
Production [GWh]

5000 30

4000 25

3000 20

2000 15

1000 10

0 5

-1000 0
REF
NGLo
GELo
BIOLo
OILHi
CO2Hi
INVHi
REF
NGLo
GELo
BIOLo
OILHi
CO2Hi
INVHi
REF
NGLo
GELo
BIOLo
OILHi
CO2Hi
INVHi
REF
NGLo
GELo
BIOLo
OILHi
CO2Hi
INVHi
a b c d

Use El. Bio El. BIGCC El. Fossil El. Biofuel Biofuel Sub.
8000 45

7000 40

6000 35

Subsidy Biofuel [EUR/MWh]

5000 30
Production [GWh]

4000 25

3000 20

2000 15

1000 10

0 5

-1000 0
a b c d a b c d a b c d a b c d a b c d a b c d a b c d

REF NG Lo GE Lo BIO Lo OIL Hi CO2 Hi INV Hi

Fig. 3. The charts show electricity production, use of electricity (Use El.), production of biofuels for transport (Biofuel) and the corresponding required biofuel subsidy level
(Biofuel Sub.) for all analysed cases in model year 2019. The electricity production is divided in production from: fossil fuel CHP (Fossil El.), conventional biomass CHP
including converted biomass HOBs (Bio El.) and BIGCC CHP (BIGCC El.). For each model case, four biofuel subsidy threshold levels (indicated by filled circles) for when a new
biorefinery is included in the system have been calculated: (a) no biorefineries, (b) one biorefinery, (c) two biorefineries and (d) three biorefineries. In (a), the results are
arranged by number of biorefineries (or equally; amount of biofuel produced) and in (b), by model case.

biorefinery in the system, the conditions for BIGCC CHP are not
favourable enough for investment. Poorer conditions for BIGCC
CHP reduce the competition for biomass and, consequently,
biofuel production becomes profitable at a somewhat lower
subsidy level for the corresponding amount of biofuel. A redu-
ced subsidy for green electricity, as in case GE Lo, has a similar
impact. Again, BIGCC CHP production becomes a less attr-
active investment alternative than in the reference case. The
biofuel production pushes the BIGCC CHP out of the system at
lower production level, and the required level of the biofuel
subsidy, for the same amount of biofuel, is lower than in the
reference case.
An unlimited supply of the lower cost biomass (wood chips), as
in case BIO Lo, makes the issue of competition for biomass
irrelevant. The result is that BIGCC CHP is used to a substantially
higher degree, and conventional biomass ST CHP to a substantially
lower degree, than in the reference case. However, due to the fact
that the demand for DH is limited, there is still a competition
between biofuel production and BIGCC CHP production. As
mentioned earlier, the model requires a demand for all heat
produced; cooling of excess heat is not allowed.
Since oil just in a very marginal way is used for DH production
in the reference case, a higher oil price, as in case OIL Hi, is
essentially limited to influence the threshold levels of the biofuel
subsidy through a higher petrol/diesel price. A higher petrol/
diesel price means that biofuels can be sold at a higher price, and
the biofuel subsidy level needed to make biofuel production
profitable is, therefore, lower.

178
4000
3500
3000
Negative CO2 Emissions [kton]
2500
2000
1500
1000
500
0
-500
a b c d
REF
Fig. 4. The chart shows the studied system’s negative CO2 emissions and total use of biomass for all analysed cases in model year 2019. The electricity production is
assumed to replace (fossil) electricity production at the margin and the biofuels are assumed to replace petrol or diesel. Within each model case, different levels of the
biofuel subsidy give rise to different numbers of biorefineries in the system: (a) no biorefineries, (b) one biorefinery, (c) two biorefineries and (d) three biorefineries. For the
same number of biorefineries, the same amount of biofuel is produced.
A higher tax on CO2 emissions, as in case CO2 Hi, has somewhat
counteracting effects. On the one hand, the electricity price is
higher, which is favourable for the BIGCC CHP technology, but on
the other hand, the biomass prices are also higher, due to the
assumed connection to the coal price (see Section 3.4). In total,
the conditions for BIGCC CHP are more advantageous than in the
reference case, i.e. the higher electricity price has a greater impact
than the higher biomass prices. Due to the higher carbon content
in coal compared to oil, an increase of the CO2 tax increases
the biomass prices relatively more than the petrol/diesel price.
This, in combination with the higher electricity price and the more
advantageous conditions for BIGCC CHP, makes production of
biofuels less favourable than in the reference case, and higher
biofuel subsidy levels are therefore required for the same amount
of biofuel production.
Higher investment costs for biomass gasification-based plants
obviously decreases the competitiveness for these technologies. The
impact on the production and on the required biofuel subsidy for a
25% increase of the investment costs compared to the reference case
is seen in case INV Hi. The resulting production is very similar to
what earlier have been described for NG Lo and GE Lo. However, the
required biofuel subsidy levels are, in this case, higher.
The studied system, i.e. the DH sector in Västra Götaland and in
Greater Göteborg, gives rise to CO2 emissions due to its use of
fossil fuels8 and electricity. However, the production of biofuels
and electricity within the studied system will replace conven-
tional transport fuels (petrol and diesel) and electricity production
outside the system and, thereby, contribute to less CO2 emissions.
The studied system’s net impact on CO2 emissions is, therefore,
of relevance. Fig. 4 illustrates the negative CO2 emissions, or the
CO2 emission savings, together with the total biomass use of the
system for all analysed cases for model year 2019. The CO2
emissions from fossil fuel use are represented as negative
8
Biomass is considered CO2 neutral.
ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180
CO2 Electricity CO2 Biofuels CO2 Fossil Fuels Biomass Use
12
10
8
Biomass Use [TWh]
6
4
2
0
a b c d a b c d a b c d a b c d a b c d a b c d
NG Lo GE Lo BIO Lo OIL Hi CO2 Hi INV Hi
bars while the ‘‘saved’’ CO2 emissions outside the system,
due to the net production of electricity9 and production of
biofuels10 within the system, are represented as positive bars. For
all tested cases, it is clear that the CO2 emission savings outside
the system by far exceed the CO2 emissions from the system’s use
of fossil fuels. The CO2 emissions due to the fossil fuel use is
relatively constant on a comparably low level for all model cases
and biofuel production levels (a–d), while the CO2 emission
savings differ in a wider range. It is notable that although biofuels
for transport are regarded as completely CO2 neutral and to full
extent replace petrol and diesel, the CO2 benefit of biofuels are
in many cases small and in some cases even negative, i.e. as the
production of biofuels increases in each model case (from ‘‘a’’ to
‘‘d’’) the CO2 emission savings increase very little or in some cases
decrease slightly. This is explained by the decrease in electricity
production and the increase in electricity use that follow from the
biofuel production. As previously stated, the marginal electricity
production that is affected by the studied system’s activities
is assumed to be based on fossil fuels, either coal or natural gas
(see Section 3.4). As opposed to the modest CO2 emission savings,
a distinct increase in the total use of biomass can, for each model
case, be seen from an increased biofuel production.
9
The net electricity production (electricity produced subtracted with
electricity used) is assumed to replace electricity production on the margin. As
previously mentioned, see Section 3.4, the marginal electricity production is
different for different model cases, model seasons and model years but is either
NGCC or coal condensing power generation (with varying efficiencies). For all cases
except for NG Lo and CO2 Hi, the marginal electricity production is in model year
2019 based on coal in summer and in spring/autumn while based on natural gas in
winter. For NG Lo and CO2 Hi, also the marginal production in spring/autumn is
based on natural gas (the resulting electricity CO2 emission factors that give rise to
Fig. 4 are in the range 367–820 kg CO2/MWhe).
10
The biofuels are assumed to replace petrol or diesel. A CO2 emission factor
for petrol/diesel of 259 kg CO2/MWhf is used.
 ARTICLE IN PRESS
M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180
5. Discussion and conclusions
In the present study, the cost-effectiveness of different
applications of biomass gasification is analysed. The study
investigates whether, and under what conditions, combined heat
and power (CHP) generation in biomass integrated gasification
combined cycle (BIGCC) plants, as well as production of biofuels
for transport in biomass gasification biorefineries, could be
competitive alternatives to conventional technology options in
district heating (DH) systems.
In the model-generated results, CHP generation in BIGCC plants is
to a high degree included in cost-optimized DH systems and is
thereby, under the assumed conditions, a cost-competitive technol-
ogy alternative. Most favourable conditions for the BIGCC CHP
technology is found in the biggest DH system in the studied region,
Göteborg, since the heat demand in this system is high enough to
allow a BIGCC CHP plant of larger size, which is advantageous due to
economies of scale. In the present study, the influence of plant size
on costs for distribution of biomass and heat is, however, not taken
into account. These are factors that work against positive plant size
scale effects (see e.g. Dornburg and Faaij, 2001). The income from
electricity production is a central parameter regarding the competi-
tiveness of the BIGCC technology since the main difference in output
between a capital-intensive BIGCC CHP plant and a conventional
biomass steam turbine (ST) CHP plant is the power-to-heat ratio,
which is higher for the BIGCC option. The income from the electricity
production depends on the electricity price and on the level of the
green electricity subsidy. The electricity price depends in turn on
several factors, such as fuel prices and the CO2 tax level. While a high
CO2 tax level is favourable for the BIGCC technology, the study
indicates that the opposite could be the case for biorefineries
producing biofuels for transport.
In the present study, synthetic natural gas (SNG) and dimethyl
ether (DME) are included transport biofuel options. Also for
biorefineries, the plant size is a factor of importance for the
economic performance. Since the heat production from biorefi-
neries often is small compared to the output of biofuels, there are,
however, several more possible DH systems that allow a larger
scale biorefinery than a larger scale BIGCC CHP plant. Biorefineries
constitute in such way a possibility to make use of economies of
scale also in small- and mid-sized DH systems. With the oil price
assumptions used in the study (50–75 USD/barrel), biofuels
cannot economically compete with petrol and diesel without
promotion by policy measures. For the introduction of large-scale
biofuel production in cost-optimized DH systems, a required
biofuel subsidy level of 30–40 EUR/MWh was calculated under
assumptions of an oil price of 50 USD/barrel, a CO2 tax level
increasing from 20 to 40 EUR/ton CO2 from 2009 to 2029 and with
base assumptions regarding plant investment costs.11 A higher oil
price would imply a lower subsidy level requirement, while a
higher CO2 tax would lead to a higher required subsidy level since
this would involve an increased competition for biomass and
higher biomass prices. In a situation where biomass prices are
connected to the coal price including the cost of CO2, as suggested
in the present study, an increase in CO2 tax implies a larger
increase in biomass prices than in the petrol/diesel price, due to
the higher carbon content in coal, and this would lower the
profitability of biofuels for transport.
11
If adding the estimated required subsidy levels to the assumed CO2 tax we
receive figures that can be compared to the currently used taxation system in
Sweden. For example, in model year 2029 the CO2 tax is 40 EUR/ton CO2 or about
10 EUR/MWh for petrol/diesel. Required subsidy levels of 30–40 EUR/MWh plus
the CO2 tax then add up to 40–50 EUR/MWh. Today’s fuel taxes in Sweden on
petrol and diesel, which biofuels currently are exempt from, are 61 and 41 EUR/
MWh, respectively.
179
With a limited amount of resources, prioritizing between
available options is essential. In the context of this study, in
particular the constraint on the availability of the local, lower cost
biomass and the restriction of the DH demand are of importance,
and prioritizing between whether to produce green electricity or
transport biofuels is necessary. Even though imported biomass is
available to an unlimited extent, just slightly more expensive than
the local biomass, the study clearly shows a trade-off situation
between investments in BIGCC CHP and investments in biorefi-
neries and, thereby, between production of electricity and
production of transport biofuels. What is observed in the model
results is that when a large part of the available lower cost
biomass resources is allocated to production of biofuels through
high biofuel subsidy levels, conventional biomass ST CHP with
high heat efficiency satisfies the heat demand to a lower total
system cost than the BIGCC CHP technology. The best conditions
for BIGCC CHP plants are, therefore, found in situations with low
ambitions regarding biofuel production, i.e. with low biofuel
subsidy levels. With a marginal electricity production based on
fossil fuels, the negative impact regarding CO2 emissions of less
production of electricity in CHP plants is significant. Since a high
biofuel production level means a lower production level of
electricity in the system than what otherwise is possible, the
net CO2 benefit of biofuels for transport is small. Goals of
increasing renewable electricity production and at the same time
increasing production of biofuels could therefore, to some extent,
be counteractive. If cooling of waste heat, or biorefineries and
biomass power plants not connected to DH, were included options
in the analysis, the competition induced by the limited DH
demand would to large extent be removed. However, from the
perspective of efficient resource utilization it is desirable to make
use of waste heat from production of electricity and biofuels.
In the present study, a tax for regulation of CO2 emissions is
assumed. If instead, a well-functioning CO2 emission cap system is
assumed, the difference in CO2 emissions due to technical changes
would always be zero (an emission reduction would be replaced
by a corresponding emission increase elsewhere in the system).
The cost to reach the emission cap level will, however, be different
depending on which technical solutions that are applied. The
interpretation of the results of the present study (Fig. 4) in the
light of a CO2 emission cap system would therefore be that
solutions that cause high CO2 emissions would imply high costs to
reach emission cap levels, and vice versa.
Quantitative results, such as cost-effective, technology-specific
production levels or required biofuel subsidy levels, are by their
character to a large extent related to the assumptions made and
the region studied, in this case the Västra Götaland region and
Greater Göteborg. However, similar regions with, for instance,
a large DH system surrounded by smaller DH systems of different
sizes or with a comparable blend of existing production
technologies can be found in many locations, and the results
ought thus to be relevant also for other regions. The study
illustrates that regional energy planning could have an important
role to play in order to achieve energy-efficient and cost-efficient
energy systems. The approach could not only generate knowledge
of appropriate technology choices but also give insights of suitable
plant locations to a higher degree than what is possible with more
aggregated approaches regarding geography and system design.
The system that is here studied and optimized is, however, not
controlled by a single actor who simply could implement suitable
actions on the regional level. Instead, the system consists of a
number of subsystems, each controlled by different actors with
probably a bigger interest of optimizing their own subsystem
than the overall system as a whole. The likely outcome would be a
sub-optimized overall system and not an optimized system, which
is the case in the present study. The results are, therefore, not so
 ARTICLE IN PRESS
180 M. Börjesson, E.O. Ahlgren / Energy Policy 38 (2010) 168–180
much a prediction of a future development but rather an
illustration of what could be, and the interpretation of the results
must be done in the light of these prerequisites.
Acknowledgments
The authors are grateful for the funding of this study provided
through the Biokombi Rya project by Göteborg Energi AB, Göteborg
Energi Research Foundation and the Swedish Energy Agency, as
well as through a research grant from the Adlerbertska Research
Foundation. Thanks to Bengt Johansson, FOI (Swedish Defence
Research Agency)/Lund University, for valuable comments.
References
Axelsson, E., Harvey, S., Berntsson, T., 2007. A tool for creating energy market
scenarios for evaluation of investments in energy intensive industry. In:
Proceedings of the 20th International Conference ECOS, Padova, June 25–28,
2007, pp. 1587–1594.
Azar, C., Lindgren, K., Andersson, B.A., 2003. Global energy scenarios meeting
stringent CO2 constraints—cost-effective fuel choices in the transportation
sector. Energy Policy 31, 961–976.
Brandberg, Å., Sävmark, B., Landälv, I., Lindblom, M., 1997. DME—drivmedel för
dieselmotorer (DME—fuel for diesel engines). Swedish Transport and Com-
munication Research Board, KFB-meddelande 1997, vol. 30, Stockholm
(in Swedish).
Bärring, M., Nyström, O., Nilsson, P.-A., Olsson, F., Egard, M., Jonsson, P., 2003. El
från nya anläggningar—2003 (Power from new plants—2003). Elforsk, Report
03:14, Stockholm (in Swedish).
Carlson, A., 2003. Energy systems and the climate dilemma: reflecting the impact
on CO2 emissions by reconstructing regional energy systems. Energy Policy 31,
951–959.
Danish Energy Authority, Elkraft System, Eltra, 2005. Technology Data for
Electricity and Heat Generating Plants. Danish Energy Authority, Copenhagen.
Dong Energy, 2007. Listpris på naturgas (list price on natural gas). Website
information. Available at /www.dongenergy.seS (2007-02-18) (in Swedish).
Dornburg, V., Faaij, A.P.C., 2001. Efficiency and economy of wood-fired biomass
energy systems in relation to scale regarding heat and power generation using
combustion and gasification technologies. Biomass and Bioenergy 21, 91–108.
ECON, 2002. Marginal elproduktion och CO2-utsläpp i Sverige (Marginal electricity
production and CO2 emissions in Sweden). Swedish Energy Agency, ER
14:2002, Eskilstuna (in Swedish).
Ekengren, T., 2006. Fordonsgas Sverige AB personal communication.
Ekström, C., Amnell, G., Anheden, M., Eidenstam, L., Kirkegaard, G., 2002.
Biobränsle från skogen (biofuel from the forest). Swedish Energy Agency, ER
9:2002, Eskilstuna (in Swedish).
European Commission, 2000. Action plan to improve energy efficiency in the
European Community. Communication from the Commission to the Council,
the European Parliament, the Economic and Social Committee and the
Committee of the Regions, COM (2000), 247, Brussels.
European Commission, 2003. Directive 2003/30/EC of the European Parliament
and of the Council of 8 May 2003 on the promotion of the use of biofuels and
other renewable fuels for transport. Official Journal of the European Union
L123/42.
European Commission, 2007. Promoting biofuels as credible alternative to oil in
transport. Memo/07/5, Date 10/01/2007.
European Council, 2007. Brussels European Council 8/9 March 2007. Presidency
Conclusions, Nr: 7224/1/07 REV1.
Fahle n, E., Ahlgren, E.O., in press. Assessment of integration of different biomass
gasification alternatives in a district-heating system. Energy, corrected proof.
Gustavsson, L., Börjesson, P., Johansson, B., Svenningsson, P., 1995. Reducing CO2
emissions by substituting biomass for fossil fuels. Energy 20, 1097–1113.
Gustavsson, L., 1997. Energy efficiency and competitiveness of biomass based
energy systems. Energy 22, 959–967.
Gustavsson, L., Madlener, R., 2003. CO2 mitigation costs of large-scale bioenergy
technologies in competitive electricity markets. Energy 28, 1405–1425.
Göteborg Energi, 2005. Miljörapporter 2004 (Environmental Reports 2004).
Göteborg Energi AB, Göteborg (in Swedish).
Hiremath, R.B., Shikha, S., Ravindranath, N.H., 2007. Decentralized energy
planning; modeling and application—a review. Renewable and Sustainable
Energy Reviews 11, 729–752.
Ingman, D., Larsson, E., Waldheim, L., 2006. Biobränsleförgasning för kraftvärme
respektive energikombinat—integration med Rya kraftvärmeverk i
Göteborg (biomass gasification for combined heat and power and energy
combine—integration with Rya CHP in Göteborg). In: Ahlgren, E., Andersson,
E., Axelsson, E., Börjesson, M., Fahle n, E., Harvey, S., Ingman, D., Karlsson, S.,
Larsson, E., Löfgren, Å., Waldheim, L. Ådahl, A., 2007. Biokombi Rya—Slu-
trapporter från ingående delprojekt (Biokombi Rya—final reports from
part projects). Chalmers Energy Centre—CEC, Report CEC 2007:3, Göteborg
(in Swedish).
Johansson, T., 2001. Biobränslen i Västra Götaland—en förstudie (Biomass in Västra
Götaland—a prestudy). The County Administration of Västra Götaland,
Publication 2001: 26, Vänersborg (in Swedish).
KAM, 2003. Final report from the ecocyclic pulp mill programme, 1996–2002.
KAM-Report A100, STFI Packforsk, Stockholm.
Knutsson, D., 2005. Simulating conditions for combined heat and power in the
Swedish District Heating Sector. Diss. Chalmers University of Technology,
Göteborg, New Series No. 2356.
Knutsson, D., Sahlin, J., Werner, S., Ekvall, T., Ahlgren, E.O., 2006a. HEATSPOT—a
simulation tool for national district heating analyses. Energy 31, 278–293.
Knutsson, D., Werner, S., Ahlgren., E.O., 2006b. Combined heat and power in the
Swedish district heating sector—impact of green certificates and CO2 trading
on new investments. Energy Policy 34, 3942–3952.
Lolou, R., Goldstein, G., Noble, K., 2004. Documentation for the MARKAL family of
models. Energy Technology Systems Analysis Programme. Available at /http://
www.etsap.orgS (2004-11-16).
Mantzos, L., Capros, P., 2006. European Energy and Transport Trends to
2030—Update 2005. Directorate-General for Energy and Transport, European
Commission, Luxembourg.
Marbe, Å., Harvey, S., Berntsson, T., 2004. Biofuel gasification combined heat and
power—new implementation opportunities resulting from combined supply of
process steam and district heating. Energy 29, 1117–1137.
Nyström, I., Ahlgren, E., Andersson, E., Börjesson, M., Fahle n, E., Harvey, S., Ingman,
D., Karlsson, S., Larsson, E., Löfgren, Å., 2007. Biokombi Rya—Biobränsleför-
gasning satt i system (Biokombi Rya—Biomass Gasification in a
System). Chalmers Energy Centre—CEC, Report CEC 2007:2, Göteborg
(in Swedish).
OPET, SEA (Swedish Energy Agency), 2004. OPET RES-e: El från förnybara
energikällor—Förutsättningar för småskalig kraftvärmeproduktion från
biobränsle i Västra Götaland (Electricity from Renewable Energy Sources—
Conditions for Small Scale Combined Heat and Power Production from Biomass
in Västra Götaland). Available at /www.opet.seS (2006-03-22) (in Swedish).
Petterson, A., Losciale, M., Liljemark, S., 2006. LCNG-studie—Möjligheter med LNG
i fordonsgasförsörjningen i Sverige (LCNG-study—possibilities with LNG in the
vehicle gas supply in Sweden). Report SGC 167, Malmö (in Swedish).
Renova A.B., 2005. Miljörapport 2004 (Environmental Report 2004). Renova AB,
Göteborg (in Swedish).
Salvia, M., Pietrapertosa, F., Cosmi, C., Cuomo, V., Macchiato, M., 2004. Approaching
the Kyoto targets: a case study for Basilicata region (Italy). Renewable and
Sustainable Energy Reviews 8, 73–90.
SEA (Swedish Energy Agency) and Statistics Sweden, 2005. Deliveries of gasoline
and fuel oil divided into municipalities in 2004. Statistics Sweden, EN 13 SM
0501, Örebro.
SEA (Swedish Energy Agency), 2006a. Energy in Sweden 2006. Swedish Energy
Agency, ET 2006:45. Eskilstuna.
SEA (Swedish Energy Agency), 2006b. Energy in Sweden 2006—Facts and Figures.
Swedish Energy Agency, ET 2006:44, Eskilstuna.
SEA (Swedish Energy Agency), 2007. Prisblad för biobränslen, torv m.m. Nr 3/2007
(Price sheet for biofuel, peat etc. No. 3/2007). Swedish Energy Agency,
Eskilstuna (in Swedish).
SOU (The Swedish Government Official Reports), 2004. Introduktion av förnybara
fordonsbränslen (Introduction of renewable fuels for vehicles). SOU 2004: 133,
Ministry of the Environment, Stockholm (in Swedish).
SOU (The Swedish Government Official Reports), 2007. Bioenergi från jordbruket—
en växande resurs (Bio energy from the agriculture—a growing resource). SOU
2007: 36, Ministry of Agriculture, Stockholm (in Swedish).
Statistics Sweden, 2006. Kommunala energibalanser (Municipal energy balances).
Website information. Available at /www.scb.se/EN0203S (2006-08-15) (in
Swedish).
Statistics Sweden, 2007. Population statistics. Website information. Available at
/www.scb.se/BE0101-ENS (2007-08-15).
Statistics Sweden, Swedish Board of Agriculture, Swedish Environmental Protec-
tion Agency, The Federation of Swedish Farmers, 2007. Hållbarhet i svenskt
jordbruk 2007 (Sustainability in Swedish Agriculture 2007). Statistics Sweden,
Örebro (in Swedish).
Svenska Kraftnät, 2007. Marknadsstatistik för elcertifikatsystemet (market statis-
tics for the electricity certificate system). Website information. Available at:
/https://elcertifikat.svk.se/S (2007-09-25) (in Swedish).
Swedish District Heating Association, 2006. Statistik 2004 (Statistics 2004).
Swedish District Heating Association, Report February 2006, Stockholm (in
Swedish).
Swedish EPA (Environmental Protection Agency), 2005. Strategi för hållbar
avfallshantering (Strategy for Sustainable Waste Disposal). Swedish Environ-
mental Protection Agency, Stockholm (in Swedish).
Swedish Petroleum Institute (SPI), 2007. Statistics Sweden—Prices. Website
information. Available at /www.spi.se/eng/S (2007-03-15).
Uddin, S.N., Barreto, L., 2007. Biomass-fired cogeneration systems with CO2 capture
and storage. Renewable Energy 32, 1006–1019.
UN (United Nations), 1993. Agenda 21: Earth Summit—The United Nations
Programme of Action from Rio. United Nations Department of Public
Information, New York.
Unger, T., Ahlgren, E.O., 2005. Impacts of a common green certificate market on
electricity and CO2-emission markets in the Nordic countries. Energy Policy 33,
2152–2163.
World Bank, 2007. Commodity Price Data (Pink Sheet). January 4, 2007, World
Bank, Washington, DC. Available at /www.worldbank.orgS (2007-01-19).