Pellet Production from Miscanthus: Energy and Environmental Assessment

: The production of wood pellets has grown considerably in the last decades. Besides woody biomass, other feedstocks can be used for pellet production. Among these, miscanthus presents some advantages because, even if speciﬁcally cultivated, it requires low inputs such as fertilisers and pesticides and shows high biomass yield (up to 28 tons of dry matter ha − 1 in Europe). Even if in the last years some studies evaluated the environmental impact of woody pellet production, there is no information about the environmental performances of miscanthus pellet production. In this study, the environmental impact of miscanthus pellet was evaluated using the Life Cycle Assessment approach with a cradle-to plant gate perspective. Primary data were collected in a small-medium size pelletizing plant located in Northern Italy where miscanthus is cultivated to be directly processed. The results highlight how the miscanthus pellet shows lower environmental impact compared to woody pellet, mainly due to the lower energy consumption during pelletizing. The possibility to pelletize the miscanthus biomass without any drying offsets the environmental impact related to the miscanthus cultivation for all the evaluated impact categories (except for Marine eutrophication). In detail, for global warming potential, 1 ton of miscanthus pellet shows an impact of 121.6 kg CO 2 eq. (about 8% lower respect to woody pellet) while for the other evaluated impact categories the impact reduction ranges from 4 to 59%. Harvesting, which unlike the other ﬁeld operations is carried out every year, is by far the main contributor to the impacts of the cultivation phase while electricity is the main contributor to the pelletizing phase.


Introduction
The interest about renewable energy is increasing. The current consumption of energy from fossil sources involves environmental concerns due to its contribution to the GHG emissions [1,2]. In the last decades, thanks to favorable subsidy frameworks, the share of renewable energy has increased [3][4][5]. However, more attention was paid on electricity production [6]. For example, in Europe 29 countries foresee a feed-in-tariff for the electric energy produced from renewable sources [7][8][9]. Regarding the heat, the subsidy framework to promote the production of renewable heat are less widespread even if the production of heat from renewable sources (e.g., biomass) is a well-known solution.
Woody biomass is historically used as energy source to produce heat mainly at household level. Among the woody biomass the pellet presents several advantages. It has higher density compared to wood chips and logs. Furthermore, pellet can be easily handled also in urban and semi urban areas, thanks to the size of the single pellet piece it can be moved with augers and, consequently, it can be efficiently dosed during the feeding of boilers. not usually performed; the crop benefits from the capture of atmospheric nitrogen (about 50 kg·ha −1 ·year −1 [45]).
The miscanthus cultivation area considered in this paper benefits from an average annual rainfall of 805 mm/year. Furthermore, to reduce the risk of failures due to drought, one irrigation is carried out in the first cultivation year during the summer. In addition, as regards the water supply, in the case investigated the crop benefits from a significant contribution from the surface aquifer, being located near the floodplain of the Serio river.
Harvesting is performed yearly, in early spring, when the dry matter content of the biomass is around 90% and after the leaves fall, which add nitrogen to the soil thanks to their decomposition. A self-propelled forage harvester is used, cutting and shredding the miscanthus stalks, then conveying them into trailers coupled to tractors for transporting the biomass to the pelletizing plant (average distance 1.5 km). At the end of the crop cycle, and before the intervention of the stalk chopper, an herbicide treatment is carried out to cancel the resprouting capacity of the miscanthus rhizomes.
In the farm, the biomass is temporarily stored under a canopy, then loaded onto two platforms for transfer via a conveyor belt to a disc separator, which removes larger foreign material. The sieved biomass is then stored in a hopper and later conveyed by means of an auger first to a magnet for the separation of ferrous materials and then to a mill for grinding. The dust produced during grinding is intercepted thanks to a cyclone dust collector and subsequently in a bag filter. A further conveyor transports the milled biomass to the pelletizing press, on which a second bag filter separates the dust produced. After a passage on a bucket elevator with counter-current air for cooling, the pellet is further sifted and finally packaged in big bags. Since the miscanthus has a humidity of 10% upon harvesting, no supplementary drying is required before pelletizing. The produced pellet is directly sold to the citizens and/or local dealers.

Goal and Scope Definition
The goal of this study is assessing the environmental performances of miscanthus pellet produced in a small size pelletizing plant located in Northern Italy (45 • 38 0" N, 9 • 46 0" E). Taking into account that the pellet consumption is increasing in Italy and that the Italian production doesn't satisfy the demand, the local production of pellet from alternative biomass sources is interesting and can be useful for a further development of local production chains.
The outcomes of this LCA can be used to select suitable raw materials for pellet production, to compare pellet produced from different biomass sources and to identify solutions for an effective reduction of the environmental impacts of pellet production.
Even though the geographical scope of this study is Northern Italy, the findings and insights of this study will supply a new reference to future improvement of pellet production from miscanthus or other types of lignocellulosic herbaceous crops (e.g., Arundo donax L.) in areas with similar climate and productive conditions. Moreover, the outcomes of this study can be used to develop the processes and logistics of pellet production for other regions as well.

Functional Unit
According to the ISO 14040 "The functional unit is a key element of LCA which has to be clearly defined. The functional unit is a measure of the function of the studied system and it provides a reference to which the inputs and outputs can be related".
In this study, 1 t of miscanthus pellet packed was selected as functional unit; however, to ease the comparison with pellets produced from different biomass sources, also the calorific value (1 GJ) was taken into account as additional functional unit. The consideration of a mass-based functional unit has also been considered in previous LCA studies available in the literature [22,31,43] together with energy-based functional unit allowing the comparison with alternative production systems independently of the feedstock used [40][41][42].

System Boundary
In this study the LCA approach was applied with a "from cradle to gate" perspective; consequently, all the processes from the raw material extraction to the pellet packaging at the pelletizing plant were included in the system boundary. A "cradle to gate" approach has been widely used in LCA studies focused on renewable energy production from woody biomass [21,22,35].
More in detail, the following steps of the pellet production chain were included: manufacturing of the different production factors consumed (rhizome, fuel, pesticides, maize starch, packaging material), manufacturing maintenance and disposal of capital goods (tractors and operative machines used during miscanthus cultivation, devices and infrastructures of the pelletizing plant) while the steps of distribution, use and end-of-life of pellet were excluded. Regarding the cultivation of miscanthus, the emissions included in the system boundary refer to the combustion of fuel in the tractors and machine engines, the application of pesticides and the nitrogen and phosphorous cycles. No change in soil organic carbon content was considered in accordance with previous studies on perennial energy crops cultivated on arable land [46,47]. Figure 1 shows the considered system boundaries.
a mass-based functional unit has also been considered in previous LCA studies av the literature [22,31,43] together with energy-based functional unit allowing the co with alternative production systems independently of the feedstock used [40][41][42].

System Boundary
In this study the LCA approach was applied with a "from cradle to gate" pe consequently, all the processes from the raw material extraction to the pellet pac the pelletizing plant were included in the system boundary. A "cradle to gate" has been widely used in LCA studies focused on renewable energy product woody biomass [21,22,35].
More in detail, the following steps of the pellet production chain were includ ufacturing of the different production factors consumed (rhizome, fuel, pesticid starch, packaging material), manufacturing maintenance and disposal of capi (tractors and operative machines used during miscanthus cultivation, devices a structures of the pelletizing plant) while the steps of distribution, use and end pellet were excluded. Regarding the cultivation of miscanthus, the emissions in the system boundary refer to the combustion of fuel in the tractors and machine the application of pesticides and the nitrogen and phosphorous cycles. No chan organic carbon content was considered in accordance with previous studies on energy crops cultivated on arable land [46,47]. Figure 1 shows the considered system boundaries.

Inventory Analysis
Primary data about miscanthus cultivation and pellet production were col surveys and interviews with the agronomist of the farm as well as with the m the pelletizing plant. About the cultivation of miscanthus the primary data colle to: the cultivation practice (timing and repetitions of field operations), the mech (characteristics of tractors and operative machines, working time), production fa sumption (fuel, lubricating oil, rhizomes, herbicides, water) and yield.

Inventory Analysis
Primary data about miscanthus cultivation and pellet production were collected by surveys and interviews with the agronomist of the farm as well as with the manager of the pelletizing plant. About the cultivation of miscanthus the primary data collected refer to: the cultivation practice (timing and repetitions of field operations), the mechanization (characteristics of tractors and operative machines, working time), production factors consumption (fuel, lubricating oil, rhizomes, herbicides, water) and yield. Table 1 reports the main inventory data regarding the cultivation of miscanthus.  [2] FC = fuel consumption; [3] Biomass yield 10 t ha −1 the 1st year, 15 t ha −1 the 2nd year then 20 t ha −1 (10% of moisture content); [4] Glyphosate (360 g·L −1 of active ingredient).
Concerning the biomass processing (pelletizing), direct data about the produced pellet were collected, the consumption of energy and other materials (packaging film, maize starch, etc.). Laboratory tests were performed to determine the main characteristics of the produced pellet. More in detail, in according to the standardized technical rules, miscanthus pellet samples were collected and analyzed in order to assess the moisture content, the ash content, the pellet durability indices (PDI) and the heating value (EN 18134-1:2015; EN ISO 18122:2015; EN ISO 18125; UNI EN 15210-1) [48][49][50]. In addition, the percentages of carbon, hydrogen, and nitrogen were determined in accordance with the norm ISO 16948:2015 [51] using a Costech ECS 4010 CHNS-O elemental analyzer. Table 2 reports the main information about the pellet production while Table 3 shows the results regarding the laboratory tests. The miscanthus pellet is characterized by moisture content lower than 10% and by a Heating Value (HHV and LHV) that is lower respect to the woody pellet (18 GJ/ton) [40] but higher than the one produced by orchard pruning residues (16.74 GJ/ton) [52].  Secondary data for the miscanthus cultivation refer to: the emission related to the fuel combustion in the tractor engines modelled according to Nemecek and Kägi [53] and Lovarelli and Bacenetti [54,55]; -the emissions related to the nitrogen and phosphorous compounds in the soil estimated following the IPCC guidelines [56] and Prasuhn [57]; -the emissions of active ingredient of pesticides considered completely released into the soil according to Rivera Schmidt et al. [58].
Background data for the production of diesel fuel, rhizomes, pesticide, electricity, tractors and agricultural machines, maize starch, packaging materials and the pellet producing plant were obtained from the Ecoinvent database ® v.3.6 [59]. The list of the processes retrieved from the Ecoinvent database is detailed in Table 4. Table 4. List of processes retrieved by the databases.

Process Retrieved from Database Note
Tractor, 4-wheel, agricultural {GLO}|market for |APOS, U Used to build soil tillage, planting, soil restoring and transport [1] Diesel {CH}|market for|APOS, U Used to build the different field operations [1] Agricultural machinery, tillage {GLO}|market for|APOS, U Used to build the process soil tillage, planting, and stump chopping [1] Agricultural machinery, unspecified {GLO}|market for|APOS, U Used to build herbicide application [1] Harvester {GLO}|market for|APOS, U Used to build the harvesting operation [1] Agricultural trailer {GLO}|market for|APOS, U Used to build the process the transport operation [1] Lubricating oil {RER}|production|APOS, U Used in the different field operations as well as at the pelletizing plant [ [1] Amount calculated in agreement with Lovarelli and Bacenetti [52].
Regarding the rhizomes production, respect to the process included in the database a higher multiplication factor was considered (55 respect to 50) due to the longer growing seasons and the higher average temperature. Consequently, a production of 550,000 rhizomes per hectare (instead of 500,000 as in the Ecoinvent ® process) was considered (average mass of a rhizome is 70 g).
In agreement with Costantini et al. [61], ionizing radiation was excluded on account of the low prevalence of nuclear power in Italy, while water consumption and land use were excluded due to lack of detailed data about rhizomes and maize starch production. The inventory data were processed using SimaPro ® LCA software v 9.1. (PRé Sustainability, Amersfoort, The Netherlands).

Results and Discussion
The environmental results for miscanthus cultivation are reported in Figure 2 (relative contribution) and in Table 5 (absolute impact values referring to the production of 1 ton of chopped miscanthus). For the different evaluated impact categories Table 5 reports also the main substances and the processes responsible for the total impact.
With a share of the impact ranging from 41 to 72%, the harvesting operation is the main contributor to the impact across all the evaluated impact categories, except Stratospheric Ozone depletion and Marine Eutrophication, where the emission of N and P compounds shows the highest impact. More in detail, dinitrogen monoxide is the substance mostly contributing to stratospheric ozone depletion while nitrate is the one most affecting marine eutrophication. As reported in detail in Table 5, the impact of harvesting is related, for some impact categories, to diesel production (i.e., Fossil resource scarcity) and consumption (i.e., global warming, ozone formation, human health, fine particulate matter formation, ozone formation, terrestrial ecosystems and terrestrial acidification) while for some others, to manufacturing, maintenance and disposal of the forager (harvester) involving mine operations (i.e., terrestrial ecotoxicity and mineral resource scarcity) or the production and management of waste (freshwater eutrophication, freshwater ecotoxicity, marine ecotoxicity, human carcinogenic toxicity and human non-carcinogenic toxicity).
The other operations carried out in the first year of the crop cycle (soil tillage, planting and irrigation) affect the impact by a minimum of 0.03% (marine eutrophication) to a maximum of 21.37% (mineral resource scarcity). transport is responsible for a non-negligible share of the environmental load in the toxicity-related impact categories (from 12% for terrestrial ecotoxicity to 30% for human carcinogenic toxicity). The contribution of the different field operations to human carcinogenic toxicity is mainly due to the emissions related to diesel combustion. The manufacturing of the other production factors (rhizomes and glyphosate) plays a minor role in defining the environmental performance of the miscanthus biomass. Rhizomes production shows a contribution < 5% for all the evaluated impact categories except for freshwater eutrophication (16% due to the emission of N and P compounds due to the fertilizers application) while glyphosate has even a smaller role being responsible of less than 2% of the impact except for freshwater eutrophication (10%, due to the release of phosphorous during its degradation into the soil). The other operations carried out in the first year of the crop cycle (soil tillage, planting and irrigation) affect the impact by a minimum of 0.03% (marine eutrophication) to a maximum of 21.37% (mineral resource scarcity). transport is responsible for a non-negligible share of the environmental load in the toxicity-related impact categories (from 12% for terrestrial ecotoxicity to 30% for human carcinogenic toxicity). The contribution of the different field operations to human carcinogenic toxicity is mainly due to the emissions related to diesel combustion. The manufacturing of the other production factors (rhizomes and glyphosate) plays a minor role in defining the environmental performance of the miscanthus biomass. Rhizomes production shows a contribution < 5% for all the evaluated impact categories except for freshwater eutrophication (16% due to the emission of N and P compounds due to the fertilizers application) while glyphosate has even a smaller role being responsible of less than 2% of the impact except for freshwater eutrophication (10%, due to the release of phosphorous during its degradation into the soil).
Regarding the pelletizing step of the pellet production process, the contribution analysis ( Figure 3) shows how the energy consumption (electricity) is the main responsible of the environmental performances of miscanthus pellet with a share of the impact higher than 40%, except for marine eutrophication, where the main contributor is the miscanthus biomass (due to the emission of N & P compounds during cultivation), and Mineral resource scarcity, where main contributors are the miscanthus biomass (due to the use of tractors and forager and the manufacturing and maintenance of the pelletizing plant). More in detail, manufacturing and maintenance of the pelletizing plant shows a impact lower than 5% for all the evaluated impact categories except than for the toxicity relates ones and for mineral resource scarcity. The impact related to the waste produced during pelletizing (e.g., dust), included in the label "pelletizing plant" has a negligible role (<0.1%) for all the evaluated impact categories.  Regarding the pelletizing step of the pellet production process, the contribution analysis ( Figure 3) shows how the energy consumption (electricity) is the main responsible of the environmental performances of miscanthus pellet with a share of the impact higher than 40%, except for marine eutrophication, where the main contributor is the miscanthus biomass (due to the emission of N & P compounds during cultivation), and Mineral resource scarcity, where main contributors are the miscanthus biomass (due to the use of tractors and forager and the manufacturing and maintenance of the pelletizing plant). More in detail, manufacturing and maintenance of the pelletizing plant shows a impact lower than 5% for all the evaluated impact categories except than for the toxicity relates ones and for mineral resource scarcity. The impact related to the waste produced during pelletizing (e.g., dust), included in the label "pelletizing plant" has a negligible role (<0.1%) for all the evaluated impact categories.
With respect to other LCA studies focused on pellet production [41][42][43][44]52], for the miscanthus pellet there is no heat consumption for biomass drying since the chopped miscanthus has a moisture content (about 10%) that allows the direct pelletizing. Maize starch and packaging materials show a small contribution for all the evaluated impact categories. Table 6 reports the absolute impact for the packed miscanthus pellet considering the 2 selected functional unit: 1 ton and 1 GJ of LHW.
To test the robustness of the achieved results a sensitivity analysis was carried out regarding the biomass yield during miscanthus cultivation as well as about the consumption of electricity and packaging material during pelletizing. For these parameters, a variation of ±20% was considered. The results, reported in Table 7, show how the variation of biomass yield, has an effect higher than 5% for 5 of the 15 evaluated impact categories and, as expected, affects marine eutrophication (the environmental effect where the role of cultivation is higher). Also, the variation of the electricity consumption shows a non-negligible impact for the evaluation impact categories (>10% for 8 for the 15 evaluated impact categories). Finally, differently than for biomass yield and electricity consumption, the variation in the consumption of packaging material has small effect on the environmental performances of miscanthus pellet (lower than 2% for 14 of the 15 impact categories).  [6] Fossil resource scarcity kg oil eq. 3.819 Oil crude Diesel {GLO} market for|APOS [1] Process not included in the database and built by modifying the process Combine harvesting {CH}|processing|APOS, U considering the fuel consumption and the mass of the machine reported in Table 1. Emission due to fuel combustion were modified proportionally to the variation of fuel consumed; [2] Mainly due to the treatment of spoil from hard coal mining; [3] Mainly due to the production of metals; [4] mainly due to treatment of sulfidic tailings, from copper mine operation; [5] Mainly due to the management of waste (landfill and incineration) produced for harvester manufacturing; [6] Mainly due to the metal mine operations. Figure 4 reports the relative comparison between the miscanthus pellet analysed in this study and the woody pellet process included in the Ecoinvent ® database (v3.6): Wood pellet, measured as dry mass {RER}|wood pellet production|APOS, U. The comparison highlights how the pellet from miscanthus shows better environmental performance (lower impact) for all the evaluated impact categories except for Marine eutrophication. The impact reduction for miscanthus pellet ranges from 4% for Fossil resource scarcity to 59% for Fine particulate matter formation and is related to the lower energy consumption (as stated before, heat is need for biomass drying during the woody pellet production). Even if woody pellet is produced using biomass coming from forestry and whose production does not involve soil tillage, planting, and crop management its impact is higher because the produced biomass needs to be dried. Only for marine eutrophication, the impact category mostly affected by phosphate emissions during miscanthus cultivation, the pellet from miscanthus has a higher impact (three times higher) then the woody one. Energies 2021, 14, x FOR PEER REVIEW 9 of 14 With respect to other LCA studies focused on pellet production [41][42][43][44]52], for the miscanthus pellet there is no heat consumption for biomass drying since the chopped miscanthus has a moisture content (about 10%) that allows the direct pelletizing. Maize starch and packaging materials show a small contribution for all the evaluated impact categories. Table 6 reports the absolute impact for the packed miscanthus pellet considering the 2 selected functional unit: 1 ton and 1 GJ of LHW.
To test the robustness of the achieved results a sensitivity analysis was carried out regarding the biomass yield during miscanthus cultivation as well as about the consumption of electricity and packaging material during pelletizing. For these parameters, a variation of ±20% was considered. The results, reported in Table 7, show how the variation of biomass yield, has an effect higher than 5% for 5 of the 15 evaluated impact categories and, as expected, affects marine eutrophication (the environmental effect where the role of cultivation is higher). Also, the variation of the electricity consumption shows a nonnegligible impact for the evaluation impact categories (>10% for 8 for the 15 evaluated impact categories). Finally, differently than for biomass yield and electricity consumption, the variation in the consumption of packaging material has small effect on the environmental performances of miscanthus pellet (lower than 2% for 14 of the 15 impact categories).    59% for Fine particulate matter formation and is related to the lower energy consumption (as stated before, heat is need for biomass drying during the woody pellet production). Even if woody pellet is produced using biomass coming from forestry and whose production does not involve soil tillage, planting, and crop management its impact is higher because the produced biomass needs to be dried. Only for marine eutrophication, the impact category mostly affected by phosphate emissions during miscanthus cultivation, the pellet from miscanthus has a higher impact (three times higher) then the woody one.

Conclusions
The production of wood pellet has grown considerably in the last decades thanks to an increasing demand of this fuel and the interest for renewable energy. However, besides woody biomass, other feedstocks can be used for pellet production. Among these, miscanthus presents some advantages because, even if specifically cultivated, it requires low inputs and shows high biomass yield.
In this study, the environmental impact of miscanthus pellet was evaluated using the LCA approach with a "cradle-to plant gate" perspective. The results highlight how the miscanthus pellet shows lower environmental impact compared to woody pellet mainly due to the lower energy consumption during pelletizing. The possibility to pelletize the miscanthus biomass without any drying offsets the environmental impact related to the miscanthus cultivation for all the evaluated impact categories (except for Marine eutrophication, affected by the emission of P compounds occurring during its run-off and glyphosate degradation). Regarding the cultivation, the harvesting that, differently from the other field operations, is carried out every year, is by far the main contributor to the impact.
The achieved results were mainly derived from a site-specific conditions and local features such as topographic, soil, climatic and agricultural activities. Consequently, additional LCA studies are needed to delve the environmental impact of miscanthus pellet in other contexts. In particular, the aspects that should be carefully considered and investigated furtherly refers the possibility that the cultivation requires additional irrigations or fertilization (due to the cultivation in soils with low nutrient availability) or that drying

Conclusions
The production of wood pellet has grown considerably in the last decades thanks to an increasing demand of this fuel and the interest for renewable energy. However, besides woody biomass, other feedstocks can be used for pellet production. Among these, miscanthus presents some advantages because, even if specifically cultivated, it requires low inputs and shows high biomass yield.
In this study, the environmental impact of miscanthus pellet was evaluated using the LCA approach with a "cradle-to plant gate" perspective. The results highlight how the miscanthus pellet shows lower environmental impact compared to woody pellet mainly due to the lower energy consumption during pelletizing. The possibility to pelletize the miscanthus biomass without any drying offsets the environmental impact related to the miscanthus cultivation for all the evaluated impact categories (except for Marine eutrophication, affected by the emission of P compounds occurring during its run-off and glyphosate degradation). Regarding the cultivation, the harvesting that, differently from the other field operations, is carried out every year, is by far the main contributor to the impact.
The achieved results were mainly derived from a site-specific conditions and local features such as topographic, soil, climatic and agricultural activities. Consequently, additional LCA studies are needed to delve the environmental impact of miscanthus pellet in other contexts. In particular, the aspects that should be carefully considered and investigated furtherly refers the possibility that the cultivation requires additional irrigations or fertilization (due to the cultivation in soils with low nutrient availability) or that drying of the biomass would be needed.