• Open Access

Environmental impacts of future bioenergy pathways: the case of electricity from wheat straw bales and pellets

Authors


Correspondence: J. Giuntoli, tel + 31 224 565 294, fax + 31 224 565 630, e-mail: Jacopo.giuntoli@ec.europa.eu

Abstract

This study presents the life cycle assessment of electricity generation from straw bales and pellets. Straw is the most abundant biomass residue in Europe and its use for energy purposes is promoted on the premise of high greenhouse gas savings. This assumption has delayed the study of sustainability of straw-fired systems on a broader sense and the literature on the topic is almost absent. This study uses data from specific literature and emissions inventories to model a number of straw pathways. The plant modeled is a medium-scale straw-fired power plant of 50 MWth capacity. The results show that electricity from straw-fired power plants can indeed realize high greenhouse gas savings compared both with existing coal plants and with the European electricity mix. The savings are in the range 70–94%. The influence of the geographical origin of straw is analyzed by using datasets for the cultivation of wheat in five different European countries. The highest emissions are recorded for the case of straw from Spain due to the small yields, whereas cultivation processes in United Kingdom and the Netherlands show high environmental impacts due to the high level of fertilization. Other environmental impacts are evaluated, such as acidification potential, eutrophication, particulate matter emissions, and photochemical ozone formation. The bioenergy system scores worse than the current European electricity mix for all the categories. However, it is important to notice that in Spain and United Kingdom the straw system shows lower impacts compared with the local average coal electricity. Finally, the study investigates the ‘break-even’ distance at which the higher emissions from the pellets production are paid off by the saved emissions in their transport compared with the bales. The results show that no reasonable break-even distance exists for road transport, whereas advantages for pellets are evident in any configuration for transoceanic transport.

Introduction

Agricultural residues are among the cheapest and most widespread biomass resources. More and more utilities around Europe and the world are increasing the use of these biomass fuels for electricity production, mostly in large-scale cogeneration with coal but also in stand-alone biomass plants (Sander & Skøtt, 2007).

Among these residues, wheat straw is not only the most abundant but also the most utilized in the bioenergy sector. Scarlat et al. (Scarlat et al., 2010) calculated that the total availability of crop residues in EU-27 could amount to 258 million dry t yr−1 on average and particularly, wheat straw could account for almost 110 million dry t yr−1. However, with the constant increase in the use of straw, indirect effects might become important: Scarlat et al. (Scarlat et al., 2010) reported several studies underlining the existence of a ‘sustainable removal rate’ of straw above which the quality of the soil (organic carbon, nutrients content, and soil texture) would be negatively affected. This value has been assessed to be in the range 40–50% removal of the harvested straw. In addition, alternative uses of straw such as animal husbandry and mushroom production should also be accounted for: in case straw was diverted to the use of bioenergy, the production and use of alternative materials should be included in the reference system. Adding these numbers up, Scarlat et al. (2010) deduced the final amount of straw (from various types of grains) available to the European energy sector to be in the range of 87.2 million dry t yr−1, corresponding to an energy potential of about 1530 PJ yr−1. This number would cover about 3.2% of the EU27 final energy consumption. Denmark is the pioneer country in the use of straw for energy due to committed policies dating back to the oil crisis of 1973 (Skøtt, 2011), and nowadays uses approximately 1.8 Mt of straw each year for energy.

The logistics of straw utilization for energy are rather complicated: straw is generally used in large square bales which have low energy density, require large spaces for storage and high amounts of energy for handling and for transport. It is even common for utilities to include a specific additional compensation for the storage of the bales on the field in the price paid to the farmers. Such complex logistics are the main reason why nowadays straw is mostly used and traded locally. However, with plans to increase substantially the use of biomass for power and heat and with the traditional sources of bioenergy being already almost fully exploited (e.g., sawdust and shavings) (Obernberger & Thek, 2010), it is important to increase the mobilization of alternative abundant sources like straw.

Large-scale use of straw has been and still is hindered by its specific elemental composition that can cause several issues in industrial boilers such as slagging, fouling, and corrosion (Giuntoli et al., 2009). Nevertheless, progress in this field is steady and efficiencies, as well as trouble-free operation, are steadily improving (Skøtt, 2011; Voytenko & Peck, 2012).

The possibility of using straw in the form of pellets, rather than the more common use in bales form, has been explored by a few power plants (Skøtt, 2011) but it is not yet common practice. The availability of indigenous, abundant resources means that the additional energy required for pellets manufacturing is not paid back by the lower transport costs. But with the development of a large global biomass market, this could become an interesting pathway to better exploit the available resources worldwide.

Scientific literature is rich in terms of characterization of straw as a fuel for thermal conversion (Giuntoli et al., 2009; Verma et al., 2012). Moreover, research has also been carried out to assess the effects of straw removal on soil nutrients and soil carbon content, even though definitive recommendations and emission factors are yet to be defined (Lafond et al., 2009; Lemke et al., 2010; Cherubini & Strømman, 2011).

Abundant literature is also available for the production of liquid biofuels from straw (Gabrielle & Gagnaire, 2008; Cherubini & Ulgiati, 2010), but much less material is to be found regarding the sustainability of power and heat production from straw-fired installations (Searcy & Flynn, 2008). Almost nothing has yet been published with the specific focus on straw pellets traded on international markets, neither on the impact of emissions from the combustion of straw.

The only relevant study, by Sultana and Kumar (2011), presented an analysis of greenhouse gas (GHG) emissions and energy use for the combustion of straw pellets in domestic stoves in British Columbia. They found GHG emissions in the range 6–30 g CO2eq./MJ pellet, with savings of 50% even compared with wood pellets. However, they did not investigate other environmental impact categories, neither the emissions from end use of straw pellets. Nothing else is available to the authors' knowledge regarding straw bales sustainability.

The present study aims at filling this gap in the literature by providing a full life cycle assessment (LCA) of electricity produced by combustion of straw bales and straw pellets.

The system analyzed includes the cultivation of wheat, transportation of bales to the pellet plant, pellet production facility, and combustion in an industrial furnace. To correctly account for the emissions from the cultivation process, economic allocation is applied between wheat grains and straw. However, as this allocation is based on volatile values, a sensitivity analysis has been carried out based on different scenarios.

Geographical location can have an important effect on the environmental impacts of bioenergy due to different yields, fertilizer use, and transport distances. This effect is accounted for in this study by using datasets for the cultivation of wheat in five different European countries: Germany, United Kingdom, Poland, Spain, and the Netherlands.

Finally, the impacts analyzed are not limited to climate change, which is the main focus in current policies, rather several significant categories are included to have a more complete assessment of sustainability.

A comparison between the combustion of bales and pellets is carried out to establish the optimal transport distance above which the production of pellets causes an environmental advantage.

The effects of straw removal from the field are only considered with respect to the additional synthetic fertilizer applied to compensate the nutrients removal. Other effects such as soil carbon depletion, decreased grain productivity, or competition with other sectors using straw are not considered (Cherubini & Strømman, 2011); it is assumed that the straw used in this study is part of the so-called surplus amount and that only 40% of the straw produced on the field is collected.

Materials and methods

This LCA is performed according to the ISO 14040 and 14044 standards (ISO, 2006a,b), using the software GaBi 5, from PE International. The following sections describe the LCA methods and the results obtained according to the scheme provided by the ISO standards.

Goal and scope definition

The LCA is of the attributional type, it analyses the environmental performance of a system producing electricity from straw bales, as it is the current practice, and from straw pellets. The functional unit used is 1 MJ of electricity. The environmental impact categories analyzed in this work are as follows: global warming potential (GWP), eutrophication potential (EP), acidification potential (AP), particulate matter emission (PM), and photochemical oxidant formation potential (POFP).

The scope of the Life Cycle Inventory (LCI) includes the cultivation data for wheat grains and straw in five different European countries, the production of straw pellets and, finally, their utilization in a medium-scale electricity production plant of 50 MWth input.

The physico-chemical properties of straw bales and pellets are summarized in Table 1.

Table 1. Physico-chemical properties of wheat straw considered in this work. Data are on a dry basis
PropertyAmountSource
  1. a

    Fresh matter.

Moisture (% f.m.)a14Deimling & Rehl (2010)
C (%d.m.)45.5Deimling & Rehl (2010)
N (%d.m.)0.65Deimling & Rehl (2010)
S (%d.m.)0.4Giuntoli et al. (2009)
Cl (%d.m.)0.4Giuntoli et al. (2009)
Energy (MJ/kgd.m.)17.5Deimling & Rehl (2010)

System boundaries

The approach of the study is from cradle to grave. The chain of processes covers all the phases from cultivation to the final utilization of straw bales and pellets to produce electricity. Figure 1 summarizes the system boundaries for the straw bales and pellets pathways. The main difference between the two pathways resides in the additional process of pellet manufacture.

Figure 1.

System boundaries for straw bales (a) and straw pellets (b) pathways.

Life cycle inventory

The second step of an LCA involves the construction of the inventory, a systematic inventory of all the energy and material flows, and emissions connected to the straw systems during the entire life cycle.

A number of different data sources are used to generate the compilation and quantification of inputs and outputs for straw production and utilization. Analyses are based on literature data, emissions inventories, and state-of-art technologies for straw conversion systems relevant to Europe. Where necessary and for background processes, the literature data are supplemented with data from commercial databases such as Ecoinvent v2.2. (Ecoinvent, 2010) and Gabi professional.

Cultivation

The cultivation of winter wheat grain and straw was modeled by PE International (Deimling & Rehl, 2010). The dataset uses average and specific data relative to wheat cultivation. The final product of this process is considered to be straw already baled in large square bales (called Hesston bales) that are the typical feedstock for current straw-fed plants (Skøtt, 2011). The dataset does not consider any emission due to direct or indirect land-use change. As mentioned in the introduction, the reference use of straw in agriculture is generally threefold: part of the straw is left on the field to recirculate the nutrients, maintain the soil carbon content, and protect the soil from erosion; part of it is used for animal husbandry and bedding; and another part is generally considered as surplus. In the past, this fraction was burned directly on the field (and the nutrients would be then directly recirculated via the ashes), but this is now forbidden in many European countries (Skøtt, 2011). For this study it is considered that the straw used for energy is part of the ‘surplus’ fraction and thus it does not have an impact on other markets or on the soil productivity. The only land-use change effect that is considered here is that an additional amount of nitrogen has to be applied on the field to compensate for the amount removed with the straw. The effect of straw removal on other nutrients (P and K) is considered irrelevant (Cherubini & Strømman, 2011; Sultana & Kumar, 2011).

Yields, prices, and reference system data (such as the electricity mix generation) are differentiated between the five European countries included in this study: Germany (DE), Spain (ES), the Netherlands (NL), Poland (PL), and the United Kingdom (UK).

Machinery and infrastructures are systematically ignored during the whole study as they are known to have a negligible impact in the analysis of energy systems (UNEP, 2011). The main inputs and yields from the cultivation process are summarized in Table 2.

Table 2. Main inputs and outputs of winter wheat cultivation in five different European countries (Deimling & Rehl, 2010)
ParameterUnitDEUKESPLNLSource
  1. a

    Yields are based on an average over the period 2000–2008.

  2. b

    Yield of straw is based on a proportion grain–straw as 1 : 0.8.

  3. c

    The sources for fertilizer input depend on the specific country; for United Kingdom, data are from DEFRA, 2011. For Poland, data are taken from FAO, 2003. For NL, ES, and DE, data are taken from FAO, 2007.

  4. d

    Active Substance.

Yield grainakg ha−174307890287038008265EUROSTAT (2011)
Yield strawbkg ha−159406310230030406612 
Diesell ha−17575757575KTBL (2008)
Seedskg ha−1140140140140140KTBL (2008)
N-fertilizerkg N ha−11651909565190 c
P-fertilizerkg P ha−1303453109 c
K-fertilizerkg K ha−14042312410 c
Limestonekg ha−1300300300300300 c
Fungicideskg AS ha−1d0.430.520.030.110.57EUROSTAT (2007)
Herbicideskg AS ha−1d1.432.730.570.695.7EUROSTAT (2007)
Insecticideskg AS ha−1d0.040.05n.a.n.a.0.03EUROSTAT (2007)

The cultivation of wheat has two main products, grain and straw. Although it is generally recommended to avoid allocation whenever possible in order not to introduce an artificial variable in the results (ISO, 2006b; Cherubini & Strømman, 2011), in this case the two products cannot be decoupled by simple system expansion. Therefore, the present dataset was built using an economic allocation. Due to the nonexistence of a proper market for straw for bioenergy, the price of straw was assigned based on the cost of straw baling and on the amount of mineral fertilizer which could be substituted with the straw in the case straw was left on the field (see Table 3 for details on the allocation values). The price of fertilizers and grains is taken from Eurostat (EUROSTAT, 2011). The price of straw baling is taken from German sources (KTBL, 2008) and weighted by national price indices.

Table 3. Allocation of emissions by price to straw and grains for the ‘surplus’ base case scenario (Deimling & Rehl, 2010)
ParameterUnitDEUKESPLNL
  1. a

    Fresh matter.

Value of straw€ t−1 fma18.930.126.115.930.6
Value of grain (average 2000–2008)€ t−1 fm126135154129122
Grain-to-straw ratiokg kg−1 1 : 0.8
Allocation%10.715.111.99.016.7

As a part of the sensitivity analysis, allocation based on energy content and on market prices of grain and straw (used for animal husbandry) will be compared for the case of United Kingdom.

Transport

Transport processes are very important in this study as the main advantage of using straw pellets rather than bales is indeed the higher energy density and the subsequent lower transport costs and associated emissions.

Transport of straw bales is assumed to be done by diesel-driven, Euro 4 type, flat-bed trucks with a total gross weight of 20–26 t and 17.3 t of net payload capacity. The large Hesston bales usually weigh about 500 kg per bale and the capacity of the flat-bed truck is limited to a maximum of 24 bales; thus a maximum payload of 12 t can be transported (Skøtt, 2011; Voytenko & Peck, 2012). For the case of straw pellets it is considered that the bales are first transported to the pellet plant via the same type of truck described above; subsequently the pellets are carried to the power plant with a diesel-driven, Euro 4 truck and trailer with a total gross weight of 34–40 t and 27 t of net payload capacity. The payload is considered to be weight limited and thus fully utilized by the pellets load. A utilization factor of 0.85 is used in both cases to account for the return trips of the trucks with partial load.

The inventory of the transport process is taken from Gabi database. The input is diesel fuel. The chosen dataset for the diesel describes a mass-weighted average refinery diesel from crude oil for Europe (EU-27) and is taken from the Gabi database.

The outputs of the process are the combustion emissions. Truck production, end-of-life treatment of the truck, and the infrastructure for the fuel supply chain are not included in the dataset.

The transport distance between the wheat field and the power plant is dependent on the scale of the plant. According to Voytenko and Peck (Voytenko & Peck, 2012), the new straw-fed installations will be mostly large-scale CHP plants with boiler capacities in the range of 50 MWth and a yearly requirement of about 100 kt of straw. Starting from this data, Scarlat et al. (2010) have calculated an optimal supply travel distance of 70 km for Europe. Other authors (Sultana & Kumar, 2011) have indicated longer supply distances up to 94 km for the case of Canada. In this work, we have used a value of 70 km. This will be the basis of comparison for the base cases of bales and pellets. Furthermore, for the latter case it is considered that the pellet plant is on the same site of the power plant and that no additional transport of the pellets is necessary.

For the Scenario 6 analysis, described later, the data for transoceanic transport consider the two most common types of bulk carriers, Panamax and Handysize (Bradley et al., 2009; Sénéchal & Grassi, 2009). The fuel consumption and associated emissions are calculated based on the bulk density of the transported goods and are based on the data published by the International Maritime Organization (Buhaug et al., 2009).

Straw pellets manufacturing

The manufacturing of agri-pellets is not yet a common commercial practice both for technical and economical reasons (Pastre, 2002). However, some experience and data already exist for the case of straw or can alternatively be deduced from the plants producing wood pellets.

The process of pellet manufacturing consists of a few basic subprocesses: drying, size reduction, pelletizing, cooling, and screening (Mani, 2005). Compared with the more common process for the production of pellets from sawdust, the main advantage of straw is the lower moisture content compared with wood. This means that the very energy intensive step of feedstock drying can be avoided. The only energy input needed, then, is the electricity required for various operations, such as: the handling of the straw (including the screening of contaminants), the grinding of raw material to a smaller size and, finally, the pelletization. This electrical power is commonly taken from the grid. In a future perspective process, an internal CHP plant fed with straw bales could provide the process heat and power.

Alternative data in the literature suggest that a drying and a conditioning step might be required to produce straw pellets for domestic use (Sultana & Kumar, 2011), thus this case is considered in the sensitivity analysis.

The values for the inputs of this process, in the base case, are summarized in Table 4. The basis for the data is 1 MJ of dry straw pellets. The value for electricity consumption is an average of the values reported by several sources (Pastre, 2002; GEMIS, 2011; Sultana & Kumar, 2011) and it amounts to 2% of the energy contained in the pellets.

Table 4. Input and output flows of the process of manufacturing of straw pellets in the base case scenario
InputUnitValueSource
Straw (@14% moisture)MJ1.04 
ElectricityMJ0.02GEMIS (2011), Pastre (2002), Sultana & Kumar (2011)
DieselMJ0.002Mani (2005)
OutputUnitValueSource
Straw pellet (@10% moisture)MJ1 

The dataset also takes into account possible losses of dry matter of straw in the pellet factory or even during storage (due to bacterial processes) (Sultana & Kumar, 2011). In this case a 3% of dry mass loss is considered during storage (both for bales and pellets) and an additional 1% loss is considered in the pellet mill. A small amount of diesel is included for internal transportation and logistics within the production plant.

In this work the pellets are considered to be drier than the feedstock even without an explicit drying process based on the heat generated by friction during the milling step. The final moisture content of 10% is considered adequate for most industrial pelleting operations.

Power plant and combustion technology

As indicated in the introduction, nowadays Denmark is the world leader for the use of straw for energy. Cereal straw is mainly used in district heating installations, but the current trend seems to focus on larger-scale CHP plants (Skøtt, 2011; Voytenko & Peck, 2012).

Most of the straw-fired CHP plants produce heat as the main product and electricity as a coproduct, which negatively affects the net electrical efficiency. Moreover, many of the existing plants apply cofiring systems where straw is fed together with municipal solid waste or other types of biomass such as wood chips. Straw is also cofired in coal power plants, but the percentage of straw input in this case is limited by technological problems such as slagging and corrosion.

The vast majority of existing plants use straw bales and moving grate furnaces; in this technology, the straw bales are first shredded into smaller pieces and then fed over a vibrating grate that simultaneously allows the primary air distribution and the bottom ash removal. Other smaller plants even feed to the furnace the entire bale that then burns with the so-called ‘cigar’ behavior (Skøtt, 2011). Straw pellets on the other hand are more suitable for the so-called ‘dust firing’ systems, where they are fed in a burner after being ground to smaller particles. Fluidized beds are also promising and would be able to handle almost any size of pellet. However, the use of straw in fluidized beds is limited by the agglomeration of the sand bed due to the alkali elements present in the feedstock (Bartels et al., 2008). Fluid bed technology could potentially be used to cofire straw with coal (or wood) up to shares which do not create operational problems, but at the moment, this is not yet a commercially exploited option.

The plant modeled in this work is considered to have a capacity of 50 MWth input and in the base case it is considered as an electricity-only power plant. Although this is energetically less efficient than cogeneration, the assumption is for a wider use of straw for energy throughout the whole of Europe where district heating infrastructures are generally much less developed than in Scandinavian countries.

Average electrical efficiencies for existing straw-fired CHP plants are available in the literature (Nikolaisen et al., 1998). In the base case of this study it is assumed that the net efficiency of conversion is the same for straw bales and straw pellets combustion and that is 29%. The net power output is considered to be the useful power sold to the grid minus the internal consumption for auxiliary operations such as feedstock handling and shredding/grinding. Most of the problems in straw combustion derive from the high content of alkali metals and chlorine that at high temperatures can give rise to massive problems with corrosion and slagging (Arvelakis & Koukios, 2002). This is the main reason why the steam temperature in straw boilers needs to be kept lower than in other installations, thus limiting the efficiency of conversion. These problems are not affected by the actual state of the feedstock, whether bales or pellets. Therefore, the assumption of equal efficiency seems acceptable. Given that we consider the internal consumption of the plant within the net electrical efficiency, pellet handling could have some advantages: bales handling requires the use of cranes and very large storage facilities, and the moving grate technology is also responsible for high power consumption. On the contrary, pellets can be automatically fed by pneumatic systems and can be stored in large silos, making their logistics energetically more efficient. Data on pellet grinding or bales shredding are scarce and thus, the energy consumption of these processes is assumed to be the same in the base case. To account for these unknowns, a scenario is introduced in the sensitivity analysis where the efficiency of the pellet plant is increased up to a value of 32%.

Accounting for the combustion emissions is not straightforward as they are technology specific and no data for straw systems yet exist in the traditional LCA databases. Luckily, Denmark, as the main user of straw for energy, provides a good source of empirical data on emissions from straw-fired CHP. For this work, data from the Danish emissions inventory was used (Nielsen et al., 2010). These data are valid for CHP installations with a maximum power output of 25 MW electric, which is comparable with the plant modeled in this work. Moreover, the emission factors used refer to existing plants using straw bales as the input fuel.

Many of the main pollutants are reported in the Danish inventory, making it possible to have detailed values for the emissions of the main greenhouse gases (CH4 and N2O), hydrogen chloride, PAH, and dioxins. Data on particulate matter (PM) are provided as Total Suspended Particles (TSP) which includes all particulates escaping the bag filters and does not differentiate for the size of the particles. To account for the different effects on human health of these emissions, it was decided to use the measured TSP as if it was composed of PM2.5 (i.e., 2.5 μm or smaller). Other pollutants emissions were taken from Ecoinvent 2.2 for a woodchips-fuelled CHP (See Table 5 for the main emissions factors). A separate consideration is required for the emissions of pollutants that could be toxic for humans or for the environment. This will be discussed in more detail in the discussion section.

Table 5. Emission factors for a straw-fired CHP plant. Used for both straw bales and straw pellets pathways
ParameterUnitValueaSource
  1. a

    All values are reported in mass of pollutant per GJ of fuel input to the boiler.

  2. b

    All measurements below detection limit.

Net electrical efficiency%29Sander & Skøtt (2007)
COg/GJin67Nielsen et al. (2010)
NOxg/GJin125Nielsen et al. (2010)
SO2g/GJin49Nielsen et al. (2010)
HClg/GJin56Nielsen et al. (2010)
CH4g/GJin0.47Nielsen et al. (2010)
NMVOCg/GJin0.78Nielsen et al. (2010)
N2Og/GJin1.1Nielsen et al. (2010)
Dust (>PM10)g/GJin112Ecoinvent (2010)
Dust (PM 2.5)g/GJin2.3Nielsen et al. (2010)
Polychlorinated dibenzo-p-dioxins and furans (equivalent) (PCDD/-F)ng/GJin19Nielsen et al. (2010)
Naphthalenemg/GJin12.1Nielsen et al. (2010)
Hexachlorobenzene (HCB)μg/GJin0.11Nielsen et al. (2010)
Benzo[a]pyrene (equivalent)μg/GJin125Nielsen et al. (2010)
Cdbmg/GJin0.32Nielsen et al. (2010)
Hgbmg/GJin0.31Nielsen et al. (2010)
Znmg/GJin0.41Nielsen et al. (2010)

The disposal of the ashes from combustion is an important part of the process: recirculating the bottom ash from straw power plants could balance out or mitigate the nutrients debt caused by the removal of the straw itself. However, although this practice is already partially allowed in Denmark (Skøtt, 2011), this would need to be evaluated from country to country. Therefore, this is not included in this work. Nevertheless, the processes necessary for the bottom ash disposal and their associated emissions are accounted for. Fly ashes are currently not utilized even in Denmark due to their high content of cadmium (Sander & Skøtt, 2007).

Reference system

In the case of electricity from bioenergy, the biomass pathway can generally be compared either with the existing electricity generation mix or with the energy system most likely to be displaced (marginal system). Both methods have positive and negative aspects. By comparing the bioenergy system with the existing mix it is possible to account for the increasing percentage of renewable sources with their low environmental impacts. On the other hand, this penalizes the bioenergy system that would instead most likely displace coal-fired power plants and should thus be evaluated against a ‘dirtier’ reference system.

In this work, emissions savings are estimated comparing the bioenergy system to both the current European electricity generation mix and to the emissions due to coal-fired power plants geographically located in the specific country where the straw is consumed.

The reference systems are taken from the Gabi database of PE International. Infrastructures are not included.

Scenarios description

As mentioned in the previous sections, various parameters may differ from the assumptions used to model the base case. To estimate the influence of such factors on the final results, several scenarios are analyzed in this study. The parameters modified in each scenario are described in Table 6.

Table 6. List of scenarios analyzed and the parameters changed from the base case
ScenarioParameter changed
1Net electrical efficiency of straw pellet plant is increased to 32%
2Cogeneration is considered and an exergy allocation of emissions to electricity and heat is applied
3Drying of straw prior to pelletization is included. A natural gas boiler is used to provide the process heat and steam.
4Different allocation methods are used for the cultivation emissions between wheat grains and straw for the case of United Kingdom. Allocation based on the market price of straw (1), on the energy content (2), and allocation only to the wheat grains (3) (based on the methodology defined by the European Commission for biomass residues).
5aBreak-even distance for road transport: drying of straw pellets with natural gas
5bBreak-even distance for road transport: base case
5cBreak-even distance for road transport: increased efficiency of conversion for straw pellets
5dBreak-even distance for road transport: increased efficiency of conversion for straw pellets + increased payload of pellet trucks to 40 t
6aBreak-even distance for ocean transport: drying of straw with natural gas and transport via a Panamax bulk carrier
6bBreak-even distance for ocean transport: drying of straw with natural gas and transport via a Handysize bulk carrier
6cBreak-even distance for ocean transport: base case and transport via a Panamax bulk carrier
6dBreak-even distance for ocean transport: base case and transport via a Handysize bulk carrier
6eBreak-even distance for ocean transport: increased efficiency of conversion for straw pellets and transport via a Panamax bulk carrier
6fBreak-even distance for ocean transport: increased efficiency of conversion for straw pellets and transport via a Handysize bulk carrier

Scenario 1 assumes that the net efficiency of combusting pellets is higher than combusting straw bales, either because of better combustion efficiency or because of lower internal power consumption. The electrical efficiency of the pellet power plant is then raised to 32%.

In Scenario 2, it is assumed that the plant cogenerates both electricity and heat and therefore all upstream and combustion emissions are allocated to electricity and steam via exergy allocation. The data for exergy allocation are summarized in Table 7.

Table 7. Basic data for Scenario 2: allocation of emissions to electricity and heat based on exergy. Data for temperature of hot water and ambient air taken from COM 2010 (2010); Dones et al. (2007)
PropertyUnitValue
Net electrical efficiency%29
Thermal efficiency%59
Temperature of hot waterK403
Temperature of ambientK273
Exergy electricity1
Exergy heat0.323
Exergy total0.480
Allocation electricity%60.4
Allocation heat%39.6

In Scenario 3, straw is dried and conditioned prior to the milling to have high-quality pellets. The necessary heat and steam is provided by a natural gas boiler with a 90% thermal efficiency (Sultana & Kumar, 2011) (Table 8).

Table 8. Data for Scenario 3: straw pellets manufacturing when drying and conditioning is needed. Heat and steam provided by a natural gas boiler
InputUnitValueSource
  1. a

    This value is increased compared with the base case due to the extra power needed for additional auxiliaries.

  2. b

    From thermodynamics calculations.

Straw (@14% moisture)MJ1.04 
ElectricityMJ0.021aGEMIS (2011); Pastre (2002); Sultana & Kumar (2011)
DieselMJ0.002Mani (2005)
Heat for dryingMJ0.017Sultana & Kumar (2011)
Heat for conditioningMJ0.007[b]
OutputUnitValueSource
Straw pellet (@10% moisture)MJ1 

Scenario 4 analyses the influence of the allocation method for the cultivation emissions between grains and straw. In this scenario, three additional methodologies are compared with the one used in the base case: allocation based on the market price of straw in United Kingdom, allocation based on the energy content of grains and straw, and full allocation to wheat grains [as prescribed in the Directive 2009/28/EC (Directive 2009/28, 2009/28, 2009)]. The allocation factors used are reported in Table 9.

Table 9. Data for Scenario 4: allocation factors between wheat grains and straw
ParameterUnitUK
  1. a

    Average price 2000–2008. Source: (EUROSTAT, 2011).

  2. b

    Average market price 2000–2008. Source: (DEFRA, 2011).

  3. c

    A process for the baling of the straw is added to mimic the recommendations of the Directive 2009/28/EC. The baling process uses 0.01 MJ diesel per MJ of straw baled.

Allocation (‘surplus’)%15.1
Price of wheat grainsGBP84a
Price of strawGBP25.4b
Allocation (‘real price’)%19.5
LHV wheat grainsMJ kg−1 dry17.1
LHV wheat strawMJ kg−1 dry17.5
Allocation (‘energy’)%45.1
Allocation (‘residue’)%0c

Finally, Scenarios 5a-d and 6a-f present the so-called ‘break-even distances’ for which GHG emissions caused by the straw bales pathway are the same as the ones from the pellets pathway. This represents the transport distance above which it is convenient, in terms of GHG emissions, to produce pellets from straw rather than burning directly the bales. Two analyses are carried out to account for the different transportation means: Scenario 5 considers only road transport, whereas Scenario 6 considers transoceanic transport. This distance will vary depending on the manufacturing process (5b and 6a-b) and on the efficiency of conversion (5c and 6e-f).

Results

In a Life Cycle Impact Assessment (LCIA) inventories of emissions and resources consumed are assessed in terms of environmental impacts, to understand and evaluate their magnitude and significance. The impact methods suggested by the ILCD Handbook (ILCD, 2011) are used in this work and the evaluation is done at midpoint. The contribution analysis is also carried out in this section: for each impact, the processes and the chemical compounds with the highest contribution to the impact are identified and underlined. The emissions linked to each category are converted into reference units with characterization factors that are expressed in the method selected.

Global warming potential

Climate change impact represents a well-known environmental problem and a prominent issue in the debate on energy production from biomass sources. It is assessed using the IPCC model characterization factors, also known as GWP, at the 100-year horizon (IPCC, 2007). For simplicity, carbon dioxide absorption by biomass and emission during combustion is not included in the calculations. This is possible because of the annual cycle of the crop considered. The unit for the characterization is kg CO2eq./MJ electricity.

The results for the base case are summarized in Fig. 2 where the contributional analysis is also included. GHG emissions range between 16 and 26 g CO2eq./MJel. for the bales pathways and between 26 and 36 g CO2eq./MJel. for the pellets pathways, as shown in Fig. 2a.

Figure 2.

Global Warming Potential of all the pathways in their base case. (a) Greenhouse gas (GHG) emissions from all the straw pathways, bars are stacked based on the contribution of the most relevant gaseous species. Bars named ‘(Bales)’ represent the pathways where bales are used as feedstock; bars named ‘(Pellets)’ represent the pathways including pelletization. (b) GHG emissions from all the straw pathways and from the reference systems considered. Bars represent the contribution of each gaseous compound, the squared symbols represent GHG savings compared with the coal electricity, and the triangles represent GHG savings compared with the European electricity mix (right y-axis). (c) GHG emissions from all the straw pathways and contribution of the different processes. (d) GHG emissions from all the straw pathways and contribution of the different processes using a transport distance of 1000 km by truck.

It can be seen that, generally, straw to electricity could guarantee substantial GHG savings compared with coal plants, and even compared with the current EU-27 electricity mix (Fig. 2b). The savings against coal systems range between 86% for Dutch straw pellets up to more than 94% for Polish bales. In the case of electricity mix, the savings range between 72.5% for straw pellets produced in Spain, up to 87.5% for straw used as bales in Poland. The main contributor to this impact for the case of bales is nitrous dioxide that is mostly emitted during agricultural operations such as fertilization (De Klein et al., 2006). However, for the pellets pathways the most significant contribution is given by fossil CO2 associated with the electricity consumption in the pellet mill. Methane emissions are mostly associated with field operations, whereas the contribution due to straw combustion is very low. As shown in Fig. 2c, emissions associated with pellet manufacturing are very relevant due to the use of electricity taken from the distribution grid; emissions from the pellet mill account for 27–37% of the total GHG emissions. With a ‘standard’ transport distance of 70 km, the contribution of emissions from transport is relatively small, between 5% and 9% of the total. Increasing the transport distance to 500 km as a representative distance for intra-EU trade, this contribution increases up to 41% of the total emissions for straw bales, but only up to 15–20% for pellets. It is important to notice that even with extremely high transport distances such as 1000 km, the GHG savings compared with the existing electricity mix are still well above 60%, as shown in Fig. 2d.

Significant differences are seen depending on the geographic origin of the feedstock. Despite the small inputs of fertilizers and pesticides, the low yields of grains and straw obtained in Spain are responsible for the highest emissions from cultivation among the considered cases (23 g CO2eq./MJel.). The Netherlands and United Kingdom present the second highest value, mostly due to the high input of N-fertilizer and associated N2O emissions (about 13–14 g CO2eq/MJel. only from N2O).

Eutrophication potential

The impact on eutrophication is divided into two categories depending on the main species responsible. The freshwater eutrophication impact is expressed in terms of kg P eq. whereas the marine eutrophication impact is expressed in terms of kg N eq. Both categories are calculated according to the method ReCiPe (Goedkoop et al., 2012).

The main contributors to the freshwater impact are phosphorous, which is mostly emitted during combustion, and phosphates, released during wheat cultivation (Fig. 3a and b). The main factor responsible for the phosphorous emissions is the combination of high content of P in the straw itself (Vassilev et al., 2010) and the lack of an advanced flue gas cleaning system to capture the fly ash produced. Fertilization is responsible for the phosphate emissions to the environment.

Figure 3.

Eutrophication potential for all the pathways in their base case. (a) Freshwater eutrophication potential (FEP) for all the straw pathways (bales and pellets) and reference systems, bars are stacked based on the contribution of the most relevant species. (b) FEP for all the straw pathways and contribution of the different processes. (c) Marine eutrophication potential (MEP) for all the straw pathways and reference systems, bars are stacked based on the contribution of the most relevant species. (d) MEP for all the straw pathways and contribution of the different processes.

The impact on marine eutrophication is mostly due to field operations during wheat cultivation when the application of N-fertilizers leads to leaching of nitrates from the soil (De Klein et al., 2006). NOx emissions during combustion are responsible for the remaining impact (Fig. 3c and d).

Concerning this impact, the comparison is in favor of the reference systems and this is mostly due to the strict regulations on NOx emissions for larger plants, to the technological advances in abatement techniques of thermal-NOx, and to the absence of emissions from cultivation.

Acidification potential

The AP is expressed in terms of kg SO2-eq and it is calculated according to the method CML 2001 [version November 2010] (Guinée et al., 2002). The results are shown in Fig. 4a and b.

Figure 4.

(a) Acidification potential (AP) for all the straw pathways (bales and pellets) in their base case and reference systems, bars are stacked based on the contribution of the most relevant species. (b) AP for all the straw pathways and contribution of the different processes. (c) Particulate matter emissions (PM) for all the straw pathways in their base case and reference systems and contribution of the most relevant species. (d) PM emissions for all the straw pathways and contribution of the different processes. (e) Photochemical oxidant formation potential (POFP) for all the straw pathways in their base case and reference systems and contribution of the most relevant species. (f) POFP for all the straw pathways and contribution of the different processes.

In this case it is evident how any final assessment of bioenergy sustainability must be accompanied by a careful evaluation of the reference system considered. In fact, all bioenergy pathways are actually potentially more harmful in terms of acidification than the EU electricity mix, with values that are 25–43% higher. The electrical system has no significant emissions of ammonia and Cl due to the lower amounts of N and chlorine in fossil fuels and to the abatement technologies applied for NOx emissions.

Considering straw as a substitute for coal has an important geographical impact: the coal resources used in Spain and United Kingdom are, in fact, responsible for higher sulfur emissions compared with those in Germany and the Netherlands. This implies that using straw in Spain or United Kingdom would indeed have a beneficial effect on air acidification, reducing the impact by 40–70%. In the case of Germany and the Netherlands, instead, a full evaluation reveals that straw would cause 55 up to 81% higher impact on air acidification. With a fully developed biomass market, power generators could include environmental impact among the factors considered for the choice of the fuel and technology. Moreover, it is worth noting, that most of the impact recorded is due to the combustion of straw, where SO2, HCl, and nitrogen oxides are the main substances of concern.

Particulate matter emissions

The particulate matter emission is expressed in terms of kg PM2.5 eq. and it is calculated according to the method Impact 2002+ v.2.1 (Jolliet et al., 2003).

Impacts from the straw pathways are again higher than the electricity reference system by 27–47%. In the case of Spain and United Kingdom, once more, the impact of bioenergy would be favorable compared with the current use of coal decreasing particulate emissions by 50–70%. The main impact is due mostly to the indirect particulate formation caused by the emissions of NOx and SO2 rather than to direct emissions of PM2.5 (Fig. 4c).

Photochemical oxidant formation potential

The impact on photochemical oxidant formation is expressed in terms of kg NMVOC eq. and it is calculated according to the method ReCiPe.

As for the acidification potential and particulate emissions, the main responsible for this impact is the combustion of the straw (Fig. 4f). The main species are NOx and, in minor amounts, SO2 and NMVOCs, as shown in Fig. 4e.

Compared with the reference systems, the straw pathways have an impact that is double the one of the current electricity mix; whereas only in Spain and United Kingdom an advantage is recorded compared with the current use of coal. All other examples have a higher impact of about 50% even compared with 100% coal-fed plants.

Scenario 1: electrical efficiency

As explained in the previous sections, technological issues prevent steam temperatures to be raised above certain critical levels in straw-fired boilers to avoid fouling, slagging, and corrosion. Therefore, the combustion of straw pellets or bales should not present significant differences. However, to account for possible advantages in either combustion or handling of straw pellets, Scenario 1 compares GHG emissions of straw bales and pellets pathways considering an increased electrical efficiency for the latter up to 32%.

Higher conversion efficiency means less straw is needed to produce the same amount of energy; this leads to a decrease in GHG emissions by about 9% compared with the base case due to the lower upstream emissions. Moreover, GHG savings increase by about 2–2.5%.

Scenario 2: exergy allocation to heat and electricity

Calculations presented so far have been based on the hypothesis that the modeled plant would be placed in a location where district heating is not available. However, efforts to increase the use of waste heat from power plants are constantly ongoing and the use of district heating should expand throughout Europe in the next decades.

In the case of production of useful heat together with electricity, emissions would need to be allocated between the two products. The choice in this work was to allocate emissions by exergy. This is also the methodology suggested by the European Commission [COM(2010), 2010]. This allocation method rightfully awards more value to the electricity produced rather than to the low-temperature heat needed for district heating (around 130–150 °C). The allocation by energy, on the other hand, would simply divide emissions between the two products, overlooking the fact that electricity has an intrinsically higher capacity of producing work compared with low-temperature heat.

The allocation factor used in this work is reported in Table 7. In this case, as shown in Fig. 5b, GHG emissions for the electricity produced are reduced by 40% and the achievable savings are then significantly increased to values above 83% even compared with the current electricity mix.

Figure 5.

Results of the scenario analysis. (a) Scenario 1: global warming potential (GWP) for all the straw pathways and the reference system; pathways with straw pellets have improved electrical conversion efficiency. Symbols represent greenhouse gas savings compared with the reference systems (right y-axis). (b) Scenario 2: comparison between GWP for the straw pellets pathways: bars named ‘(Exergy)’ represent the case when emissions are allocated to electricity and heat based on exergy. Bars named ‘(NO all.)’ represent the base case, no emissions allocation to heat. (c) Scenario 3: comparison between GWP for the straw pellets pathways and the reference systems; bars named ‘(Drying)’ represent pathways in which straw needs drying prior to pelletization, and natural gas is the feedstock used to provide process heat. Bars named ‘(NO Drying)’ represent the base case. (d) Scenario 4: comparison between GWP for straw pellets pathways in United Kingdom when different allocations between grains and straw are applied to the emissions from cultivation.

Scenario 3: drying step in pellet manufacturing

Manufacturing of agri-pellets is not yet commercially widespread; therefore, operational practices are also not fully developed. For most sources, straw with moisture in the range 10–15% would not need any drying prior to pelletization because the heat released in the press would bring the moisture content down to levels of about 10% (Pastre, 2002; Sander & Skøtt, 2007). Moreover, industrial plants are able to handle raw materials with moisture up to 20–25% (Nikolaisen et al., 1998), rendering a drying step not necessary in most cases.

Sultana and Kumar (Sultana & Kumar, 2011) stated, instead, that drying is necessary to lower moisture to about 8% and that conditioning with steam is also required to have high-quality straw pellets. However, they seem to focus on producing pellets complying with the existing standards for domestic wood pellets, whereas the straw pellets are not suitable for a domestic market because of the high emissions of pollutants and high production of bottom and fly ash. Anyway, to account for possible different processes that might develop in the markets, Scenario 3 presents the influence of drying and conditioning of the straw pellets using natural gas. The results in Fig. 5c show that GHG emissions are increased by 21–29% when drying is necessary and when it is done by using natural gas. However, GHG savings are still relatively high with values of more than 65% compared with the electricity mix, and above 83% when compared with coal.

Scenario 4: economic vs. energy allocation

An important topic when discussing environmental impacts of biomass residues is whether to assign upstream emissions to the residues or not and which allocation methodology to use. The base case presented in this study used an economic allocation of emissions between wheat grain and straw, as described in previous sections. Figure 5d shows that by changing allocation method the final result could change dramatically.

The allocation by price is complicated in the sense that prices fluctuate with time, which makes predictions and policy decisions very difficult. Therefore, the need for the sensitivity analysis presented below.

In the base case it was chosen to assign a price to straw based only on the costs of baling and on the nutrients content. This was defined as a ‘surplus’ scenario, when the straw used is the one that would have anyway been left on the field. However, if the utilization of straw for bioenergy increases, there will likely be competition between existing markets, such as animal bedding, and bioenergy itself. For this reason, straw prices for energy are likely to increase and eventually reach the current market prices of straw for animal husbandry. For the case of United Kingdom, it was possible to find detailed information on such market prices, and the new allocation is shown in Fig. 5d for the case defined as ‘real’. A rise in price of straw from about 30 euro per tonne to a price of about 37.6 euro per tonne (over a constant grain price) would result in an increased impact on climate change for the straw system of about 18%.

Moreover, with a fully developed market for straw to energy, feedstock prices could rise even further to an extreme case where the price of straw becomes similar or equal to the price of food grains. This is shown in the third case, named ‘energy’, when an energy allocation is applied to the cultivation emissions. In this latest case, the emissions associated to straw energy would more than double.

The scenario called ‘residues’ mimics the approach indicated in the European legislation [COM(2010), 2010] by assigning no upstream emissions to straw except the emissions associated with the baling process. In this case the final emissions are about 60% lower than in the base case.

Scenarios 5 and 6: break-even distances

The production of pellets requires the use of energy and causes additional emissions and environmental impacts compared with the use of simple bales. However, these extra emissions and costs can be paid off by the increased energy density and consequently, the lower emissions for the transportation of pellets. The production of agri-pellets could indeed be a very important step for the creation of a worldwide biomass market beyond wood pellets.

To verify when this trade-off happens, we define the distance at which the GHG emissions from the bales pathways are equal to the ones generated by the pellets pathways as ‘break-even’. For comparison purposes, two simplified scenarios are depicted: in Scenario 5 all the transport distance for pellets is covered by road transport whereas Scenario 6 considers ocean transport via bulk carriers. The transport of bales to the pellet plant is left at 70 km for each combination whereas the other distances are varied. Table 10 summarizes the results for three scenarios.

Table 10. Break-even distances for straw bales and pellets for the different scenarios analyzed
Scenario nr.NotesDistance (based on GHG) [km]
Truck – trailer
5a (Drying)Drying with natural gas prior pelletization1900
5b (Base case)Same efficiency; no drying1100
5c (Extra eff.)Higher efficiency of conversion for pellets; no drying750
5d (Extra payload)Higher efficiency of conversion; no drying; 40 t payload600
Ocean vessels (+70 km bales transport)
6a (Drying)Drying with natural gas prior pelletization; Panamax vessel3050
6b (Drying)Drying with natural gas prior pelletization; Handysize vessel2600
6c (Base case)Same efficiency; no drying; Panamax vessel1750
6d (Base case)Same efficiency; no drying; Handysize vessel1500
6e (Extra eff.)Higher efficiency of conversion for pellets; no drying; Panamax vessel1150
6f (Extra eff.)Higher efficiency of conversion for pellets; no drying; Handysize vessel1000

Regarding road transport, it is possible to see that when natural gas is used for pellet production, the break-even distance is extremely high, basically indicating that no advantage in terms of emissions exists in producing pellets with this technology. Even in the best case considered in this study, the break-even distance (750 km) is still higher than any logistic plan would suggest. In case the pellets could be transported on even larger trucks (Scenario 5d), such as the road-trains available in North America, then the production of pellets would become convenient for distances above 600 km (Table 10).

When sea transport is considered, even in the worst case depicted in Scenario 6a, the break-even distance would be much lower than any transoceanic trade route between Europe and United States. Hence, pelletization is essential to any long-distance trade of straw.

Discussions

The GHG emissions shown in this study are comparable with those quoted in the few available studies in the literature. Sultana and Kumar (Sultana & Kumar, 2011) found higher emissions than the ones presented in this study, reporting a value of about 30.5 g CO2eq./MJ pellet. When the same background conditions used by these authors are used with the dataset of this work (including transport distances and allocation method), values in the range of 24 g CO2eq./MJ pellet are obtained. The difference between the results must be attributed to additional diverging assumptions, such as the inclusion of manufacturing of infrastructure (trucks, plants, etc.). Comparison with the results from other studies is difficult due to the different boundary conditions chosen. Results obtained for wood pellets produced from sawdust range between 10 and 31 g CO2eq./MJ pellet (Magelli et al., 2009; Sikkema et al., 2010; Pa et al., 2012), depending on the methodology chosen for the allocation of upstream emissions to sawdust. Other authors have reported even lower emissions (24 g CO2eq./MJel.) for electricity from forestry pellets (Zhang et al., 2009). No other results were found comparable to the use of straw pellets in industrial applications.

Moreover, this study shows that relatively high GHG savings can be achieved by using straw to substitute fossil sources. All the pathways in the base case result in savings higher than 70% compared with the European electricity mix and even higher than 86% when compared with the current hard coal-fired power plants. These values are higher than the strictest threshold (60%) established by the European Commission for second generation liquid biofuels (Directive 2009/28, 2009) and recommended for heat and power from solid and gaseous biomass [COM(2010), 2010]. Increasing the transport distances up to 500 km by truck does not change this conclusion as savings higher than 69% are still achieved.

The allocation of upstream emissions from the cultivation process, though, could have an important role in the final assessment of the pathways. This work shows that, for the case of United Kingdom, the GHG savings for straw pellets pathways would drop below 50% compared with the electricity mix simply by choosing to allocate emissions based on the energy content (or mass) of the products. If the European Commission had decided to strictly apply this allocation method also to straw, then this pathway would be deemed unsustainable, unjustly penalizing the use of a biomass residue. Applying the choice not to allocate upstream emissions to residues, instead, the straw energy appears extremely favorable, with GHG savings above 90%, as shown in Fig. 5d.

In all the other environmental impacts analyzed, bioenergy systems performed worse than the current European electricity mix. It is important to note that the geographical origin and consumption of the considered biomass has a very important role in assessing its sustainability. In fact, whereas in absolute terms the emissions from wheat cultivation in Spain and United Kingdom might be higher than in the other countries analyzed, when compared with the current coal electricity production in these countries the use of straw would still guarantee advantages in almost all of the impacts studied. In general terms, however, it is clear from these results that the use of biomass for energy should not be deemed sustainable a priori, but should be carefully analyzed case by case and that the analysis should extend beyond the climate change impact.

Another result from this study is that the end use of biomass is the main responsible for many of the environmental impacts analyzed, mostly due to the emission of pollutants such as NOx, SO2, and HCl from the biomass boiler. This phase is too often overlooked in bioenergy LCAs, as well as issues concerning the emissions of heavy metals from biomass combustion. Straw is known to contain rather high amounts of S, N, and Cl (Giuntoli et al., 2009; Wang et al., 2010; Ren et al., 2011) that can cause high release of pollutants at the time of combustion. Moreover, emissions of NOx from fuel-N are more difficult to curb by simple and cheap primary measures compared with the thermal-NOx produced, for example, during natural gas combustion. Research is still ongoing to optimize combustion conditions to reduce these emissions in biomass boilers (Werther et al., 2000; Giuntoli et al., 2010; Houshfar et al., 2012), but for existing installations, this has an important environmental impact in terms of acidification potential, eutrophication, particulate emissions, and photochemical ozone formation. Furthermore, emissions of sulfur and chlorine compounds are also a concern for their impact mostly on acidification and particulate formation. Newer straw-fired plants intend to install a flue gas condensing system to recover additional heat for district heating. This would have the additional benefit of removing HCl and SO2 from the flue gases thus limiting a major source of pollution (Sander & Skøtt, 2007).

Other toxic pollutants, such as PAH and dioxins are responsible for impacts on the health of humans and ecosystems. The results for these toxicity impacts are as important as they are difficult to evaluate. They are in fact highly sensitive to the quality and availability of experimental data. Furthermore, whereas emission limit values are generally provided by national legislations, the actual emissions of these compounds depend strongly on the technology and the type of biomass used. Formaldehyde is one of the known toxic pollutants that are released during the combustion of woody materials [or even during the lifetime of wood products (EPA, 2010) and biogas (Nielsen et al., 2010)], and should thus be included in any environmental evaluation of these pathways. However, experimental data are basically nonexistent regarding formaldehyde emissions from straw combustion. Moreover, heavy metals contained in the fly ashes from biomass combustion could have an important impact on the environment, but data from actual emission inventories for straw are limited to a few elements (Hg, Zn, Cd) and the data are also of poor quality because they are below detection limits (Nielsen et al., 2010). Due to the great uncertainty related to these emission factors, it was decided not to include any toxicity impact in this work. However, the impact of biomass combustion on human and terrestric ecotoxicity should be investigated further as it could potentially be more important than for the reference fossil system.

Finally, this study has shown that the break-even distance for which emissions from pellet manufacturing are paid back by savings due to more efficient transportation is very high when road transport is considered. At best, the additional emissions would be repaid only by transporting the pellets for more than 750 km by truck, which is unreasonable for logistic and cost reasons. Consequently, straw bales seem to be environmentally the best choice when the straw is used within Europe. However, when considering long-distance transoceanic transport, the higher density of pellets makes them more advantageous than bales under any condition and for any oceanic trade route considered.

Ancillary