Properties of Pellets from Forest and Agricultural Biomass and Their Mixtures

Abstract

Pellets can be produced not only from forest dendromass but also from agricultural dendromass derived from short rotation coppice (SRC) plantations, as well as surplus straw from cereal and oilseed crops. This study aimed to determine the thermophysical properties and elemental composition of 16 types of pellets produced from four types of forest biomass (Scots pine I, alder, beech, and Scots pine II), four types of agricultural biomass (SRC willow, SRC poplar, wheat straw, and rapeseed straw), and eight types of pellets from mixtures of wood biomass and straw. Another aim of the study was to demonstrate which pellet types met the parameters specified in three standards, categorizing pellets into thirteen different classes. As expected, pellets produced from pure Scots pine sawdust exhibited the best quality. The quality of the pellets obtained from mixtures of dendromass and straw deteriorated with an increase in the proportion of cereal straw or rapeseed straw in relation to pure Scots pine sawdust and SRC dendromass. The bulk density of the pellets ranged from 607.9 to 797.5 kg m⁻³, indicating that all 16 pellet types met the requirements of all six classes of the ISO standard. However, it was determined that four types of pellets (rapeseed, wheat, and two others from biomass mixtures) did not meet the necessary requirements of the Premium and Grade 1 classes. The ash content ranged from 0.44% DM in pellets from pure Scots pine sawdust to 5.00% DM in rapeseed straw pellets. Regarding ash content, only the pellets made from pure Scots pine sawdust met the stringent requirements of the highest classes, A1, Premium, and Grade 1. In contrast, all 16 types of pellets fulfilled the criteria for the lower classes, i.e., Utility and Grade 4. Concerning the nitrogen (N) content, seven types of pellets met the strict standards of classes A1 and Grade 1, while all the pellets satisfied the less rigorous requirements of classes B and Grade 4.

1. Introduction

Solid biofuels, primarily in the form of wood chips, pellets, or briquettes, continue to be the leading source of energy production from renewable energy sources (RESs). This trend is evident both in the average of EU countries, where solid biofuels account for approximately 40% of energy production, and particularly in Poland, where this figure rises to around 60% [1]. The proportion of solid biofuels varies depending on the type of energy generated. In Poland, in 2023, solid biofuels accounted for approximately 14% with regard to electricity generation, whereas their share was significantly higher in heat energy generation, reaching 89% [1]. Thus, district heating (at the regional, local, and individual levels) is a sector in which solid biomass continues to play a very important role. Moreover, due to plans for decarbonizing district heating in Poland, which involve replacing coal-based systems with RES systems, primarily biomass systems, the demand for this solid biofuel is expected to continue growing. The main sources of solid biomass for energy, particularly wood biomass, come from forests and the timber industry [2,3,4]. Coniferous tree species dominate Poland’s forests, with the Scots pine being the most common, covering 58.6% of the total forest area [5]. In agriculture, straw is primarily obtained from cereals and oil crops [4,6,7], which make up approximately 65% and approximately 11% of the overall crop structure, respectively [8]. Additionally, agricultural biomass can be harvested from perennial industrial crop plantations in the form of straw, as well as non-woody herbaceous and woody materials [9,10,11]. Among these plants, willow and poplar are the most popular choices in Poland [4].

Solid biomass, while an energy feedstock in its original form, is a very heterogeneous biofuel with diverse and variable properties. This variability arises from its different types, sources of origin, harvesting conditions, and handling procedures throughout the entire logistics chain for a particular feedstock. Therefore, solid biomass in its original form is often not that appealing due to its high moisture content or an unfavorable weight-to-volume ratio, which subsequently results in higher transport costs and higher costs of its conversion into energy. These limitations can be overcome through the biomass pelletization process [12,13]. This method produces pellets that are more convenient for logistics, utilization, and energy generation [2,14,15], making them more suitable for both larger power plants and small, individual boiler rooms [13,16]. Pellets are a densified form of biomass, characterized by greater homogeneity and higher energy density compared with biomass in its original form. Due to these properties, pellets can be transported over longer distances while remaining economically competitive and environmentally friendly. Pellets are an important part of the bioenergy market, as they enable the introduction of sustainable solutions and replace fossil fuels, thereby contributing to an increased share of RESs. The use of pellets in the EU for sustainable heat production reduces dependence on imported fossil fuels and maintains European dominance in the design and manufacture of equipment for pellet production and energy generation from pellets [17]. In addition, pellets help reduce air emissions through the replacement of old, inefficient boilers with modern, automated pellet boilers [15,17].

Over almost 20 years, from 2005 to 2023, global pellet production increased significantly from several million to 48.8 million Mg [17]. Recent years have also seen an upward trend, with global pellet production reaching 47.6 million Mg in 2022 and with an increase of approximately 2.6% in 2023. It should be emphasized that the EU is the largest pellet producer and consumer. Pellet production in the EU remained stable throughout 2022 and 2023, amounting to 20.7 million Mg. Pellet consumption in the EU during the same period reached 22.7 million Mg and 22.0 million Mg, respectively, indicating that the EU consumed more wood pellets than it produced. Since imports of wood pellets from Russia to the EU have been suspended in recent years due to sanctions, it has been necessary to look for alternative sources of supply and/or increase production within the EU. Therefore, pellets were imported from the USA and Brazil, and EU countries such as Spain, Poland, France, Lithuania and Germany increased the production of this solid biofuel by 12%, 10%, 10%, 10% and 4%, respectively [17].

Considering the above, it is reasonable to expect that the demand for feedstocks used in pellet production will continue rising. This is largely because pellets are currently made from wood processing residues, e.g., sawdust, shavings, or possibly wood chips [18]. The primary tree species used to produce pellets from residues are Scots pine, European larch, and birch [19,20]. However, considering the EU policy aimed at limiting the use of forest biomass for energy purposes, the supply of these feedstocks is not expected to increase. As a result, the quantity of forest-based feedstocks available for pellet production may be limited, while demand for these feedstocks in pellet production plants is likely to increase. This may result in reduced availability of the feedstocks and higher prices. A certain alternative in this situation may be to use other biomass types for pellet production, e.g., agricultural biomass in the form of wood biomass from SRC plantations or straw from annual and perennial plants [21,22]. Pellets can also be produced from various types of residues, including post-extraction biomass [23,24] and mixtures of various biomass types [13,25]. It has been demonstrated that mixing agricultural biomass (straw) with forest biomass (pine sawdust) improves the mechanical properties of pellets. For example, adding 30–50% sawdust to straw increased the bulk density and mechanical strength of pellets while reducing ash content and increasing calorific value. Such mixtures enabled the production of high-quality pellets that met selected quality standards [26,27,28]. However, it should be emphasized that effective pellet production requires, above all, proper pre-grinding of the biomass and the maintenance of optimal moisture content in the raw material, as excessive moisture can lead to low pellet quality, while insufficient moisture can hinder the pelletization process. The pelletization process involves compressing biomass under high pressure through cylindrical holes in a die, which results in the formation of pellets. After leaving the die, the pellets are hot and must be cooled to achieve the right hardness and stability. The cooling process also prevents moisture condensation during storage. In the next step, the pellets are screened to remove dust, and then they are packaged for further distribution [29,30,31]. Therefore, the key parameters affecting pellet quality include die geometry, temperature, pressure, moisture content, and fineness of the raw material, as well as the presence of natural binders. The above-mentioned process variables have a direct impact on pellet quality, including its strength, density, and calorific value. Optimal selection of these parameters allows for increased energy efficiency of the process, reduced energy consumption, and stable parameters of the finished fuel [29,32].

However, it should be noted that, as indicated above, solid biomass is a highly diverse and heterogeneous feedstock in terms of its thermophysical properties and elemental composition. Therefore, pellets produced from it will also possess different characteristics, which subsequently affect their use for energy purposes and contribute to potential environmental emissions. Pellets, like other solid biofuels, are used to generate heat and electricity through various conversion processes, such as combustion, gasification, or pyrolysis [15,33,34,35]. However, the characteristics of the feedstock (biomass) used to produce these pellets can lead to various technological issues and environmental emissions. For example, high ash content in pellets can cause slagging in the burner, furnace, and other boiler components [36,37,38,39]. In addition, high nitrogen content in pellets is correlated with increased NO_x emissions, as nitrogen oxide emissions rise with higher nitrogen content in the fuel [40]. The sulfur content of biomass is also responsible for harmful emissions into the atmosphere [41].

Therefore, it is crucial to evaluate the quality of pellets made from various feedstocks, as this impacts both the end users of the fuel and its producers. This assessment is especially important given the rising consumption of pellets and the potential limitations in the availability of feedstocks for their production [13,19]. Therefore, this study aims to determine the thermophysical properties and elemental composition of pellets produced from four types of forest biomass, four types of agricultural biomass, and eight types of pellets from biomass mixtures.

2. Materials and Methods

2.1. Materials of the Study

The study examined 16 types of pellets, including 4 types of pellets from forest biomass (Scots pine I pellets, alder pellets, beech pellets, and Scots pine II pellets); 4 types of pellets from agricultural biomass (SRC poplar pellets, SRC willow pellets, wheat pellets, and rapeseed pellets); 4 types of pellets made from mixtures of the most common forest biomass in Poland (Scots pine sawdust) and wheat straw or rapeseed straw, in two percentage ratios (Scots pine–wheat 75/25 pellets, Scots pine–wheat 50/50 pellets, Scots pine–rapeseed 75/25 pellets, and Scots pine–rapeseed 50/50 pellets); and 4 pellets made from mixtures of the most common SRC biomass in Poland (willow) and wheat straw or rapeseed straw, in two percentage ratios (willow–wheat 75/25 pellets, willow–wheat 50/50 pellets, willow–rapeseed 75/25 pellets and willow–rapeseed 50/50 pellets) (

Table 1. Types of pellets from forest biomass, agricultural biomass, and their mixtures.

Pellet Type	Biomass Type	Species	Species Share (% w/w, Dry Basis)	Pellet Code
Scots pine I	Forest biomass–sawdust from debarked wood	Scots pine (Pinus sylvestris L.)	100	SPI
Grey alder	Forest biomass–chips from non-debarked wood	Grey alder (Alnus incana (L.) Moench),	100	GA
European beech	Forest biomass–chips from non-debarked sawmill residues	European beech (Fagus sylvatica L.)	100	EB
Scots pine II	Forest biomass–chips from branches from forest logging	Scots pine (Pinus sylvestris L.)	100	SPII
SRC poplar	Agricultural biomass–chips from seasoned four-year-old SRC poplar shoots	Poplar (Populus nigra × P. maximowiczii Henry)	100	P
SRC willow	Agricultural biomass–chips from seasoned four-year-old SRC willow shoots	Willow (Salix viminalis L.)	100	W
Wheat	Agricultural biomass–wheat straw collected after winter wheat harvesting	Wheat (Triticum aestivum L.)	100	Wh
Rapeseed	Agricultural biomass–rapeseed straw collected after winter rapeseed harvesting	Rapeseed (Brassica napus L.)	100	R
Scots pine I/wheat	Forest biomass–sawdust from debarked wood, and agricultural biomass–wheat straw	Scots pine/Wheat	75/25	SPI 75/Wh 25
Scots pine I/wheat		Scots pine/Wheat	50/50	SPI 50/Wh 50
Scots pine I/rapeseed	Forest biomass–sawdust from debarked wood, and agricultural biomass–rapeseed straw	Scots pine/Rapeseed	75/25	SPI 75/R 25
Scots pine I/rapeseed		Scots pine/Rapeseed	50/50	SPI 50/R 50
SRC willow/wheat	Agricultural biomass–chips from seasoned four-year-old SRC willow shoots, and agricultural biomass–wheat straw	Willow/Wheat	75/25	W 75/Wh 25
SRC willow/wheat		Willow/Wheat	50/50	W 50/Wh 50
SRC willow/rapeseed	Agricultural biomass–chips from seasoned four-year-old SRC willow shoots, and agricultural biomass–wheat straw	Willow/Rapeseed	75/25	W 75/R 25
SRC willow/rapeseed		Willow/Rapeseed	50/50	W 50/R 50

, Figure 1).

All feedstocks for pellet production were sourced in the northeastern region of Poland. Forest biomass in the form of dry Scots pine sawdust was obtained from the processing of debarked wood in a sawmill. Alder chips from non-debarked wood and beech chips from non-debarked residues of these species were also sourced from a sawmill. In turn, Scots pine chips were obtained from branches from forest logging. All the forest biomass types were sourced in autumn (September), and the forest chips were additionally dried in a dryer to a moisture content of less than 10%. Agricultural dendromass was obtained from the two SRC species most commonly cultivated in Poland, namely willow and poplar, as well as from the most abundant post-production residues of cereal crops (wheat straw) and oil crops (rapeseed straw). SRC willow and poplar biomass were sourced from an experimental plantation of the University of Warmia and Mazury in Olsztyn (UWM) and harvested in two stages over a 4-year cycle. In the first stage, during the winter (in March), the entire shoots were cut and then stored in natural conditions in piles for 6 months until autumn (September) of the same year. They were then dried in a dryer and cut into chips using a chipping machine. Wheat and rapeseed straw were sourced in the form of bales during the summer (August) at the time of the harvest from fields owned by the UWM. The straws were then cut into the form of chaff using a recycler (the straw was not dried).

Forest chips and SRC chips were ground in two stages using a hammer mill. In the first stage, a 12 mm mesh screen was employed, and in the second stage, the fraction obtained was further ground using a 6 mm mesh screen. Scots pine sawdust, wheat, and rapeseed straw chaff were ground in a single stage using a hammer mill with a 6 mm mesh screen. This way, analogous fractions of all types of forest and agricultural biomass were obtained and then pelletized in various variants. Sixteen types of pellets were produced, including four from forest biomass, four from agricultural biomass, and eight from biomass mixtures (Table 1). Biomass mixtures were formed by mixing Scots pine chips or SRC willow dendromass with wheat straw or rapeseed straw in proportions of 75%/25% or 50%/50% w/w (dry basis).

Production tests for each pellet type were conducted on a pellet mill with a capacity of up to 350 kg h⁻¹. The pellet mill was equipped with a 27 mm long horizontal die with an opening diameter of 6 mm and was powered by a 30 kW electric motor. Pellet production tests were carried out on the heated die, and the pre-prepared feedstock was fed manually into the pellet mill on a continuous basis. Between tests, the pellet mill operated for approximately 5–7 min without being fed the feedstock to self-clean the die from the previous pellets. In addition, after starting the test with the next type of input feedstock, the first batch of new pellets was rejected, which further reduced the possibility of contamination with the previous material. Each pellet type was produced in three replicates, with approximately 15 kg of each pellet type produced for each replicate. From each replicate, a sample weighing approximately 2 kg was collected. Each cooled pellet sample was then packed in a plastic bag and transported to the UWM laboratory for analysis to determine the properties of this solid biofuel.

2.2. Laboratory Analyses and Comparison of Pellet Properties to Quality Standards

The bulk density of the pellets (kg m⁻³) was determined according to standard PN EN ISO 17828:2016; Solid biofuels - Determination of bulk density. Polish Standardization Committee: Warsaw, Poland, 2016. After separating a laboratory sample, the moisture content of the pellets was determined by the gravimetric method with oven drying (PN-EN ISO 18134-1:2015) [42]. To this end, the pellets were dried in an FD BINDER (Tuttlingen, Germany) laboratory dryer at 105 °C. The dried pellet samples were ground with a Retsch SM 200 (Haan, Germany). mill on a 1 mm mesh screen. All ground pellet samples were then stored in metal containers in a laboratory dryer and used for further analysis. The carbon, hydrogen, and sulfur contents were determined using an Eltra CHS-500 analyzer (Neuss, Germany) in accordance with standards PN-EN ISO 16948:2015-07 [43] and PN-EN ISO 16994:2016-10 [44]. The nitrogen content was determined using the Kjeldahl method with a K-435 mineralizer and a BUCHI B-324 distillation apparatus (Flawil, Switzerland). The chlorine content was then determined using Eschka’s mixture in accordance with standard PN-ISO 587:2000 [45]. The ash content was determined at 550 °C, and the fixed carbon and volatile matter contents at 650 °C, using an Eltra TGA-Thermostep thermogravimetric analyzer (Neuss, Germany), in accordance with standards PN-EN ISO 18122:2016-01 [46] and PN-EN ISO 18123:2016-01 [47]. The higher heating value (HHV) was determined using an IKA C2000 calorimeter (Taufen, Germany) by the dynamic method. The lower heating value (LHV) was then calculated based on the HHV, moisture content, and hydrogen content according to standard PN-EN ISO 18125:2017-07 [48,49]. All laboratory analyses were conducted in three replicates for each pellet type.

To classify the examined pellets, their properties were compared with three standards from different continents. For Europe, it was standard ISO 17225-2:2021-10; Solid biofuels—Fuel specifications and classesPart 2: Graded wood pellets. The International Organization for Standardization: Brussel, Belgium, 2021, “Specifications for graded wood pellets for commercial, residential applications and for industrial use”, which distinguished between six pellet classes—A1, A2, B, I1, I2, and I3 (

Table 2. Pellet classes according to ISO standards.

Feature	Classes
	A1	A2	B	I1	I2	I3
Bulk density (kg m−3)	600–750	600–750	600–750	≥600	≥600	≥600
Moisture (%)	≤10.0	≤10.0	≤10.0	≤10.0	≤10.0	≤10.0
Ash content (% DM)	≤0.7	≤1.2	≤2.0	≤1.0	≤1.5	≤3.0
LHV (GJ Mg−1)	≥16.5	≥16.5	≥16.5	≥16.5	≥16.5	≥16.5
N (% DM)	≤0.3	≤0.5	≤1.0	≤0.3	≤0.3	≤0.6
S (% DM)	≤0.04	≤0.04	≤0.05	≤0.05	≤0.05	≤0.05
Cl (% DM)	≤0.02	≤0.02	≤0.03	≤0.03	≤0.05	≤0.1

). For North America, it was the Pellet Fuels Institute (PFI) standard, “Standard specifications for residential/commercial densified fuel”, which recognized three pellet classes—Premium, Standard, and Utility (

Table 3. Pellet classes according to PFI and KFRI standards.

Feature	PFI			KFRI
	Premium	Standard	Utility	Grade 1	Grade 2	Grade 3	Grade 4
Bulk density (kg m−3)	640.74–768.89	608.70–768.89	608.70–768.89	≥640	≥600	≥550	≥500
Moisture (%)	≤8.0	≤10.0	≤10.0	≤10	≤10	≤15	≤15
Ash content (% DM)	≤1.0	≤2.0	≤6.0	≤0.7	≤1.5	≤3.0	≤6.0
LHV (GJ Mg−1)	N/A	N/A	N/A	≥18.0	≥18.0	≥16.9	≥16.9
N (% DM)	N/A	N/A	N/A	<0.3	<0.5	<0.7	<1.0
S (% DM)	N/A	N/A	N/A	<0.05	<0.05	<0.05	<0.05
Cl (% DM)	≤0.03	≤0.03	≤0.03	<0.05	<0.05	<0.05	<0.05

). Finally, for Asia, it was the Korea Forest Research Institute (KFRI) standard, “Specifications and quality standards for wood pellets”, which distinguished between four pellet classes—Grade 1, Grade 2, Grade 3, and Grade 4 (Table 3).

2.3. Statistical Analysis

The statistical analyses of the results obtained were conducted based on the one-factor variance analysis (ANOVA), with the F and p-values being determined. The primary factor in this analysis was the type of pellet, comprising 16 variants made from forest biomass, agricultural biomass, and their mixtures. For the pellets examined, homogeneous groups were determined using Tukey’s (HSD) multiple-comparison test based on the significance p < 0.05. For the 12 pellet features under analysis (including the thermophysical properties and elemental composition), arithmetic means and standard deviations were calculated. Furthermore, Pearson’s correlation coefficient was calculated for the 12 features analyzed, and the statistically significant relationships (p < 0.05) were examined. Moreover, descriptive statistics were determined for all the pellet types examined, including the mean, median, minimum and maximum values, lower and upper quartiles, and the variation coefficient. In addition, a similarity analysis was carried out for the 16 analyzed types of pellets produced from forest biomass, agricultural biomass, and their mixtures, as well as for the determined thermophysical properties and elemental composition. Simple regression analysis was used to model the impact of reducing the share of agricultural biomass and increasing the share of forest biomass on the main parameters of pellets, i.e., LHV, ash content, and sulfur. The following cases of wheat straw and SRC willow pellets were considered in the modeling: wheat straw 100% + SRC willow 0%; wheat straw 50% + SRC willow 50%; wheat straw 25% + SRC willow 75%; wheat straw 0% + SRC willow 100%. The next model included rapeseed straw 100% + SRC willow 0%; rapeseed straw 50% + SRC willow 50%; rapeseed straw 25% + SRC willow 75%; rapeseed straw 0% + SRC willow 100%. The modeling also included: wheat straw 100% + Scots pine I 0%; wheat straw 50% + Scots pine I 50%; wheat straw 25% + Scots pine I 75%; wheat straw 0% + Scots pine I 100% and rapeseed straw 100% + Scots pine I 0%; rapeseed straw 50% + Scots pine I 50%; rapeseed straw 25% + Scots pine I 75%; rapeseed straw 0% + Scots pine I 100%. The modeling results have been included in the manuscript. To this end, multi-dimensional cluster analysis was applied. Agglomeration was carried out using Ward’s method, with Euclidean distance as the measure of dissimilarity. The cut-off significance were determined based on the Sneath criterion, set at levels of 33% and 66%. All statistical analyses were conducted using STATISTICA 13 software (TIBCO Software Inc., Palo Alto, CA, USA).

3. Results and Discussion

3.1. Thermophysical Properties of Examined Pellet Types

All the examined thermophysical properties, including MC, BD, FC, VM, Ash, HHV, and LHV, varied significantly (p < 0.001), depending on the pellet type (

Table 4. Analysis of variance (ANOVA)—F and p values for studied attributes.

Feature	F	p-Value
Moisture content—MC (%)	277.6	<0.001 *
Fixed carbon content—FC (% DM)	124.0	<0.001 *
Volatile matter content—VM (% DM)	303.7	<0.001 *
Ash content—Ash (% DM)	2952.7	<0.001 *
Higher heating value—HHV (MJ kg−1 DM)	474.9	<0.001 *
Lower heating value—LHV (MJ kg−1)	639.8	<0.001 *
Bulk density—BD (kg m−3)	138.4	<0.001 *
Carbon content—C (% DM)	349.7	<0.001 *
Hydrogen content—H (% DM)	39.3	<0.001 *
Nitrogen content—N (% DM)	1151.4	<0.001 *
Sulphur content—S (% DM)	1198.7	<0.001 *
Chlorine content—Cl (% DM)	492.7	<0.001 *

* significant values (p < 0.05).

). The moisture content of the examined pellets ranged from 4.73 to 9.32% for alder pellets and SRC poplar pellets, respectively (

Table 5. Selected thermophysical properties of pellets from forest biomass, agricultural biomass, and their mixtures.

Pellet Type	MC (%)	FC (% DM)	VM (% DM)	Bulk Density (kg m−3)	HHV (GJ Mg−1 DM)
Scots pine I	5.07 ± 0.03 jk	21.59 ± 0.26 bc	77.97 ± 0.25 ab	705.45 ± 6.15 c	20.31 ± 0.05 a
Grey alder	4.73 ± 0.02 k	21.27 ± 0.02 c	77.43 ± 0.08 de	751.25 ± 1.95 b	19.86 ± 0.01 b
European beech	6.73 ± 0.01 e	19.63 ± 0.04 g	78.21 ± 0.08 a	727.00 ± 3.01 bc	19.87 ± 0.03 b
Scots pine II	5.31 ± 0.10 ij	21.65 ± 0.21 b	75.83 ± 0.11 fg	705.70 ± 3.70 c	19.94 ± 0.01 b
SRC poplar	9.32 ± 0.01 a	20.48 ± 0.05 ef	77.81 ± 0.02 bc	778.90 ± 11.30 a	19.71 ± 0.04 c
SRC willow	8.25 ± 0.48 bc	22.23 ± 0.13 a	76.08 ± 0.07 f	797.45 ± 10.15 a	19.61 ± 0.01 d
Wheat	7.58 ± 0.04 d	20.90 ± 0.05 d	74.98 ± 0.10 i	628.85 ± 6.45 efg	19.11 ± 0.04 g
Rapeseed	8.19 ± 0.01 bc	19.65 ± 0.02 g	75.35 ± 0.07 h	607.85 ± 3.55 g	18.91 ± 0.02 h
Scots pine I 75/wheat 25	8.26 ± 0.09 b	20.86 ± 0.19 d	77.63 ± 0.17 cd	644.23 ± 5.77 def	19.90 ± 0.01 b
Scots pine I 50/wheat 50	7.84 ± 0.18 cd	20.58 ± 0.12 def	77.29 ± 0.11 e	622.79 ± 7.98 fg	19.76 ± 0.03 c
Scots pine I 75/rapeseed 25	6.52 ± 0.06 ef	20.61 ± 0.01 def	77.84 ± 0.03 bc	653.08 ± 6.92 de	19.89 ± 0.02 b
Scots pine I 50/rapeseed 50	6.73 ± 0.04 e	20.36 ± 0.09 f	77.28 ± 0.04 e	633.64 ± 3.64 efg	19.74 ± 0.03 c
SRC willow 75/wheat 25	6.15 ± 0.02 fg	21.69 ± 0.12 b	75.67 ± 0.19 gh	663.75 ± 8.75 d	19.45 ± 0.05 e
SRC willow 50/wheat 50	6.55 ± 0.04 ef	21.54 ± 0.01 bc	75.46 ± 0.01 h	646.36 ± 6.45 def	19.19 ± 0.03 fg
SRC willow 75/rapeseed 25	5.89 ± 0.05 gh	20.73 ± 0.01 de	76.12 ± 0.01 f	661.64 ± 1.22 d	19.28 ± 0.02 f
SRC willow 50/rapeseed 50	5.72 ± 0.11 hi	20.57 ± 0.02 def	75.91 ± 0.02 fg	635.15 ± 7.85 ef	19.23 ± 0.02 f
On average	6.80 ± 1.30	20.90 ± 0.72	76.68 ± 1.08	678.94 ± 27.57	19.61 ± 0.37

a, b, c, d, e, f, g, h, i, and j denote homogenous groups identified separately for each attribute; ±standard deviation.

). The moisture content of four pellets, i.e., SRC poplar, SRC willow, rapeseed, and an SPI 75/Wh 25 mixture, amounted to over 8%. It was therefore concluded that, regarding moisture content, these four pellets failed to meet the stringent criterion (≤8.0%) for the Premium class, as per the PFI standard (Table 3). However, the remaining 12 pellet types met the requirements of all classes across the three standards under analysis, i.e., ISO, PFI, and KFRI (Table 2, Table 3, and Table 5).

In a previous study, the moisture content of 23 pellet types fell within a narrow range of 5.7–6.5% and did not vary significantly [13]. Specifically, the moisture content of SRC willow and SRC poplar pellets was less than 6%, while that for pellets made from Scots pine sawdust was 6.4%. Slightly higher moisture content values for the same three pellet types were also noted in other studies [14,22]. As for the poplar pellets, the moisture content was 7.9% [39]. The moisture content of pellets from wheat straw was 6.2%, and the use of additives—such as bentonite clay, coconut shell biochar, corn, glycerol, and sawdust from Pinus radiata D. Don—in different weight proportions caused this value to vary between 3.5 and 7.2% [50]. In contrast, the moisture content of pellets made from rapeseed straw, depending on the plant fertilization variant, was higher and ranged from 10.2 to 12.2%, while it ranged from 10.2 to 11.1% for a mixture with 10% glycerol [51]. In view of the above, it can be concluded that the moisture content of wheat pellets and rapeseed pellets in the present study was approximately one percentage point (pp) higher and approximately two pp lower, respectively, than the above values found in the literature. At the same time, the relationship of wheat pellets, characterized by a lower moisture content than rapeseed pellets, was maintained. It is well known that the final moisture content of pellets is determined by the moisture content of the output material and the entire pelletization process [52]. The moisture content of commercially produced pellets typically ranges from 7 to 10% [53], with potential for extension beyond this range [54].

The bulk density of the examined pellets ranged from 607.9 to 797.5 kg m⁻³, respectively, for alder pellets and SRC poplar pellets (Table 5). Therefore, all 16 pellet types met the requirements of all six classes for the ISO standard (Table 2 and Table 5). However, regarding the Premium (PFI) and Grade 1 (KFRI) classes, it was found that four pellet types (rapeseed, wheat, and mixtures of SPI 50/Wh 50 and SPI 50/R 50) failed to meet these requirements (Table 3 and Table 5). In contrast, the remaining 12 pellet types met the requirements of all the classes for the three standards analyzed, i.e., ISO, PFI, and KFRI (Table 2, Table 3 and Table 5). Furthermore, it was found that pellets produced from wood feedstocks showed a higher bulk density (a range of 705–797 kg m⁻³, homogeneous groups “a”, “b”, “c”) compared to straw pellets (a range of 609–629 kg m⁻³, homogeneous groups “e”, “f”, “g”). Pellets produced from mixtures of wood biomass and straw were characterized by intermediate bulk density values (Table 5). Moreover, SRC willow pellets and SRC poplar pellets exhibited the significantly highest bulk density (homogeneous group “a”) compared to wood pellets from forest biomass. In another study, the bulk density of SRC willow pellets and SRC poplar pellets was also higher than that of Scots pine pellets (648 kg m⁻³), with the values obtained being lower compared to those obtained in the present study [13]. Furthermore, the cited study showed that increasing the proportion of SRC biomass from 25 to 75% in relation to Scots pine sawdust resulted in an increased bulk density of pellets made from these mixtures. Other studies have also reported bulk densities for Scots pine pellets within a similar range (647–671 kg m⁻³) [14,22,55,56]. It should be noted that the bulk density of SRC pellets may also be determined through the crop harvesting cycle. Pellets made from the biomass of one-year-old willow and poplar shoots had bulk densities of 523 and 499 kg m⁻³, respectively [14,22], which are lower compared to pellets obtained from four-year-old shoots (over 660 kg m⁻³) [13]. Overall, lower bulk density values were noted for pellets from straw or hay, at 498 kg m⁻³ and 478 kg m⁻³, respectively [57]. Thus, the cited values were lower than the values obtained in the present study for wheat pellets and rapeseed pellets by approximately 20%. In contrast, other studies observed a low bulk density of pellets from wheat straw alone (245 kg m⁻³) [33,50]. However, these studies showed that the use of various additives resulted in an increase in this parameter—even up to 665 kg m⁻³—for the variant of pellets made from a mixture of wheat straw (60%), sawdust (10%), corn starch (10%), bentonite clay (10%), and biochar (10%). Thus, the bulk density of pellets is determined by the properties of the output feedstock, the moisture content [58], the method of preparation, the additives and/or mixtures used [22,50], the degree of comminution, and the pressure of the pelletization process [59].

Although beech pellets were characterized by the highest volatile matter content of 78.21% DM (homogeneous group “a”), their fixed carbon content was the lowest, i.e., 19.63% DM (Table 2). In turn, wheat pellets were characterized by the lowest volatile matter content of 74.98% DM (homogeneous group “i”), while their fixed carbon content was 20.90% DM, which is considered a high value. However, of all the pellets under study, the highest fixed carbon content of 22.23% DM was noted in SRC willow pellets (homogeneous group “a”). In another study, SRC pellets also exhibited higher fixed carbon content and lower volatile matter content than pellets from forest biomass [13]. The fixed carbon contents of pellets made from one-year-old shoots of SRC willow and SRC poplar were also over 20% DM, while their volatile matter content reached 79% DM [22]. A high volatile matter content (82% DM) was noted in Norway spruce pellets [60], which is considerably higher than values obtained in the present study. In contrast, hay and straw pellets were characterized by lower contents of volatile matter (76.6–77.0% DM) and fixed carbon (16.7–18.1% DM) [61,62]. Therefore, in comparison to the cited studies, the wheat pellets and rapeseed pellets analyzed in the present study exhibited lower volatile matter content and higher fixed carbon values.

In terms of the ash content, 12 homogeneous groups were distinguished among the 16 pellet types under study (Figure 2). The ash content fell within a wide range from 0.44% DM in Scots pine I pellets (i.e., made from pure sawdust) (homogeneous group “l”) to 5.00% DM in rapeseed pellets (homogeneous group “a”). Wheat straw pellets and rapeseed straw pellets generally contained ash levels that were approximately 2–3 pp higher than those found in forest biomass pellets and SRC biomass pellets. Therefore, adding wheat straw or rapeseed straw in amounts of 25% and 50% w/w to Scots pine sawdust or SRC willow biomass increased the ash contents of the pellets made from these mixtures. It should also be explained that the higher ash value of Scots pine II pellets (2.67% DM), compared with Scots pine I pellets, can be attributed to the fact that they were produced from chips from forest logging. The beech pellets also contained over 2% DM of ash, as they were produced from chips from non-debarked sawmill residues. In alder pellets from chips from non-debarked wood, the ash content was 1.3% DM, while in SRC pellets, it was approximately 1.7% DM. Therefore, only Scots pine I pellets (made from pure sawdust) met the stringent requirements of classes A1, Premium, and Grade 1 for the three standards examined, i.e., ISO, PFI, and KFRI, respectively (Figure 2, Table 2 and Table 3). Five pellet types (alder, SRC willow, SRC poplar, SPI 75/Wh 25, and SPI 75/R 25) met the requirements of the classes B (ISO) or Standard (PFI). However, four pellet types (wheat, rapeseed, W 75/R 25, and W 50/R 50) failed to meet the ISO standard classes. All 16 pellet types met the PFI and KFRI standards for Utility and Grade 4 requirements.

In another study, pellets from SRC willow and SRC poplar biomass (1.6–1.8% DM) were also characterized by higher ash contents compared to pellets from Scots pine and birch sawdust (0.2–0.4% DM) [13]. That study concluded that increasing the proportion of SRC willow biomass from 25% to 75%, in relation to Scots pine sawdust, resulted in an increase in the ash content of pellets made from these mixtures within a range from 200 to 489%. This was due to the very low ash content of pure Scots pine sawdust (0.2% DM). In another study [14], the ash content in Scots pine pellets was higher (0.37% DM), which was more similar to the values obtained in the present study. An even higher ash content (0.60–0.81% DM) of Scots pine pellets was noted when they were produced from wood [63]. However, torrefied Scots pine pellets contained 1.30% DM of ash [64]. In turn, the ash content of pellets from agricultural biomass, i.e., SRC willow harvested over a five-year cycle, was 1.6% DM [65], which was very similar to the SRC willow pellets in the current study. Pellets produced from non-woody agricultural biomass in the form of hay (7.4–7.5% DM), annual plant straw (5.7–5.9% DM) [66,67], or straw of the genus Miscanthus (2.8–8.8% DM) [68,69,70] were characterized by higher ash contents compared with wood biomass. The ash content of rapeseed pellets was even higher (8.9–9.6%) [51] compared to the results of the present study (more than 2-fold), which may be attributed to the fact that it was derived from spring rapeseed rather than from winter rapeseed. The ash content of the pellets from wheat straw was also higher (7.1%) [50] compared to the results of the present study. Additionally, the use of various additives to wheat straw resulted in the ash content of pellets from these mixtures increasing by 8.2 to 16.2%.

Scots pine I pellets (made from pure sawdust) were characterized by the significantly highest HHV of 20.31 MJ kg⁻¹ DM, homogeneous group “a” (Table 5). The other three types of pellets from forest wood biomass were classified into homogeneous group “b”, and their HHV was lower, ranging from 1.8 to 2.2%. The HHV of pellets from agricultural SRC wood biomass were classified into homogeneous groups “c” and “d”, and their HHV was lower than that of Scots pine I pellets by 3.0–3.4%. In contrast, the HHV of pellets from wheat straw and rapeseed straw was classified into the last homogeneous groups, “g” and “h”, and their HHVs were lower by 5.9–6.9% compared to that of Scots pine I pellets. Therefore, increasing the proportion of wheat straw or rapeseed straw in proportions of 25% or 50% w/w, in relation to Scots pine sawdust or SRC willow, reduced the HHV of pellets from these mixtures. Regarding the standards, it was found that all the pellets met the requirements for all classes of the KFRI standard (Table 3). It was also found that HHV was significantly positively correlated with the contents of carbon (0.91) and volatile matter (0.80), bulk density (0.45), and hydrogen content (0.30) and significantly negatively correlated with the contents of nitrogen (−0.93), ash (−0.90), sulfur (−0.78), and chlorine (−0.75) (

Table 6. Pearson’s correlation coefficient for examined features of 16 types of pellets made from forest biomass, agricultural biomass, and their mixtures.

Feature	MC (%)	FC (% DM)	VM (% DM)	Ash (% DM)	HHV (MJ kg−1 DM)	LHV (MJ kg−1)	BD (kg m−3)	C (% DM)	H (% DM)	S (% DM)	N (% DM)	Cl (% DM)
MC (%)	1.00	−0.30 *	0.01	0.18	−0.26	−0.73 *	0.03	−0.09	−0.20	0.19	0.10	0.26
FC (% DM)		1.00	−0.27	−0.35 *	0.22	0.31 *	0.38 *	0.19	0.22	−0.50 *	−0.02	−0.39 *
VM (% DM)			1.00	−0.80 *	0.80 *	0.55 *	0.34 *	0.84 *	0.33 *	−0.56 *	−0.85 *	−0.52 *
Ash (% DM)				1.00	−0.90 *	−0.71 *	−0.56 *	−0.94 *	−0.48 *	0.83 *	0.82 *	0.75 *
HHV (MJ kg−1 DM)					1.00	0.84 *	0.45 *	0.91 *	0.30 *	−0.78 *	−0.93 *	−0.75 *
LHV (MJ kg−1)						1.00	0.29 *	0.67 *	0.26	−0.65 *	−0.71 *	−0.67 *
BD (kg m−3)							1.00	0.42 *	0.29 *	−0.65 *	−0.21	−0.83 *
C (% DM)								1.00	0.40 *	−0.71 *	−0.91 *	−0.66 *
H (% DM)									1.00	−0.27	−0.18	−0.31 *
S (% DM)										1.00	0.64 *	0.80 *
N (% DM)											1.00	0.51 *
Cl (% DM)												1.00

* significant values (p < 0.05).

).

In another study [13], pellets from Scots pine sawdust also showed the highest HHV (20.73 MJ kg⁻¹ DM), while for SRC pellets, this value was 4–5% lower. This relationship was consistent with the findings of the present study, with the difference attributed to the higher HHV of Scots pine pellets in the cited study. For this reason, increasing the proportion of SRC biomass from 25% to 75% in relation to Scots pine sawdust resulted in reduced HHV of pellets made from these mixtures. The high HHV of Scots pine pellets (20.55 MJ kg⁻¹ DM) was also confirmed by another study [14]. The HHVs of SRC pellets from one-year-old poplar and willow shoots were also high and amounted to 19.89 and 19.75 MJ kg⁻¹ DM, respectively [22]. In addition, these values were higher than the results of the present study. This could be due to the higher bark content of one-year-old shoots, which, in general, have a higher HHV than that of wood. Pellets from agricultural wood biomass were characterized by a higher HHV compared with pellets from hay (17.9 MJ kg⁻¹ DM) or straw (17.6 MJ kg⁻¹ DM [71] and 17.02 MJ kg⁻¹ DM, respectively) [33]. The HHV of pellets from rapeseed straw (18.65 MJ kg⁻¹ DM) [53] and (18.18–18.45 MJ kg⁻¹ DM) [51], similar to the present study, also did not exceed 19 MJ kg⁻¹ DM.

It was concluded—based on the HHV, MC, and hydrogen content—that Scots pine I pellets were characterized by the significantly highest LHV of 17.88 MJ kg⁻¹, placing them in homogeneous group “a” (Figure 3). The other three types of pellets from forest wood biomass were classified into homogeneous groups “b” and “c”, and their LHV was lower by 1.7 to 3.7%. In contrast, the LHVs of pellets made from the agricultural wood biomass of SRC willow and poplar were classified into homogeneous groups “g” and “h”, and their LHVs were lower than that of Scots pine I pellets by 7.2 to 7.9%. However, the LHVs of pellets from wheat straw and rapeseed straw were classified into homogeneous groups “i” and “j”, and their LHVs were lower than that of Scots pine I pellets by 8.6 to 10.3%. Therefore, increasing the proportion of wheat straw or rapeseed straw in proportions of 25% or 50% w/w in relation to Scots pine sawdust reduced the HHV of pellets from these mixtures by 3.5 to 5.7%. Regarding the standards, it was concluded that only two pellet types—made from rapeseed and wheat—failed to meet any of the classes of the ISO standard (Table 2, Figure 3). It was also observed that LHV was significantly positively correlated with HHV (0.84), carbon content (0.67), volatile matter content (0.55), fixed carbon content (0.31), and bulk density (0.29), and significantly negatively correlated with moisture (−0.73), nitrogen (−0.71), ash (−0.71), chlorine (−0,76), and sulfur (−0.65) contents (Table 6).

In another study, the LHV of pellets made from Scots pine sawdust was 18.0 MJ kg⁻¹; for SRC willow and poplar pellets, this value was 3.6–3.9% lower [13]. Therefore, this relationship was similar to that in the present study, and other percentage differences were mainly due to the differences in the moisture contents of the pellets. The cited study concluded that increasing the proportion of SRC biomass from 50% to 75% in relation to Scots pine sawdust resulted in reduced LHV of the pellets made from these mixtures by 0.9 and 2.6%, respectively. Scots pine pellets were characterized by the highest LHV (17.69 MJ kg⁻¹) of all the seven pellet types examined, including Scots pine pellets reported in another study [14]. In turn, the LHV (calculated without considering the hydrogen content) of SRC willow pellets produced from three-year-old shoots was 16.3 MJ kg⁻¹ [72], which is lower than the value for SRC pellets in the present study. However, the LHV of pellets from three-year-old poplar shoots was higher (17.7 MJ kg⁻¹) [73], although the cited study also did not consider the hydrogen content. Nevertheless, pellets made from forest wood were characterized by a higher average LHV compared with pellets from SRC biomass [72], which was also confirmed in the present study. The LHV of pellets made from spring rapeseed straw fluctuated around 17 MJ kg⁻¹, with this value determined in dry matter [51]. Thus, considering the moisture and hydrogen contents, the LHV would be approximately 16 MJ kg⁻¹ on an as-received basis, which aligns closely with the value obtained in the present study for this pellet type.

3.2. Elemental Composition of Pellets Under Study

The contents of all the elements analyzed, including carbon (C), hydrogen (H), nitrogen (N), sulfur (S), and chlorine (Cl), varied significantly (p < 0.001) depending on the pellet type (Table 4). Scots pine I pellets, produced from pure sawdust, were characterized by the highest C and H contents and the lowest N, S, and Cl contents of all the 16 pellet types examined (

Table 7. Carbon, hydrogen, and sulfur contents of pellets made from forest biomass, agricultural biomass, and their mixtures.

Pellet Type	C (% DM)	H (% DM)	N (% DM)	Cl (% DM)
Scots pine I	56.35 ± 0.21 a	6.57 ± 0.08 a	0.10 ± 0.01 k	0.003 ± 0.001 h
Grey alder	55.11 ± 0.07 bcd	6.32 ± 0.06 cd	0.30 ± 0.01 g	0.005 ± 0.002 fgh
European beech	53.45 ± 0.46 fg	6.13 ± 0.06 f	0.29 ± 0.02 g	0.010 ± 0.001 fg
Scots pine II	53.29 ± 0.12 g	6.09 ± 0.07 f	0.31 ± 0.01 g	0.004 ± 0.001 gh
SRC poplar	54.38 ± 0.17 de	6.43 ± 0.04 abcd	0.40 ± 0.02 f	0.011 ± 0.002 f
SRC willow	54.19 ± 0.36 ef	6.38 ± 0.09 abcd	0.57 ± 0.01 e	0.010 ± 0.001 f
Wheat	47.40 ± 0.22 l	6.02 ± 0.01 f	0.76 ± 0.04 b	0.074 ± 0.005 a
Rapeseed	48.38 ± 0.20 k	6.04 ± 0.02 f	0.84 ± 0.01 a	0.071 ± 0.003 a
Scots pine I 75/wheat 25	55.75 ± 0.14 ab	6.30 ± 0.05 de	0.12 ± 0.02 jk	0.036 ± 0.001 d
Scots pine I 50/wheat 50	55.37 ± 0.35 bc	6.16 ± 0.05 ef	0.15 ± 0.01 ij	0.046 ± 0.001 c
Scots pine I 75/rapeseed 25	55.61 ± 0.20 ab	6.11 ± 0.01 f	0.17 ± 0.01 i	0.040 ± 0.001 cd
Scots pine I 50/rapeseed 50	54.83 ± 0.08 cde	6.52 ± 0.05 ab	0.22 ± 0.02 h	0.056 ± 0.002 b
SRC willow 75/wheat 25	52.19 ± 0.31 h	6.28 ± 0.02 de	0.62 ± 0.02 d	0.025 ± 0.003 e
SRC willow 50/wheat 50	50.51 ± 0.34 ij	6.17 ± 0.05 ef	0.69 ± 0.01 c	0.060 ± 0.001 b
SRC willow 75/rapeseed 25	51.08 ± 0.32 i	6.46 ± 0.03 abc	0.70 ± 0.02 c	0.038 ± 0.002 d
SRC willow 50/rapeseed 50	50.29 ± 0.12 j	6.46 ± 0.01 abc	0.76 ± 0.01 b	0.057 ± 0.002 b
On average	53.01 ± 2.70	6.28 ± 0.18	0.44 ± 0.26	0.034 ± 0.025

a, b, c, d, e, f, g, h, i, j and k denote homogenous groups identified separately for each attribute; ±standard deviation.

, Figure 4). An analysis of the variation in individual elements in the pellets under study showed that Scots pine I pellets contained the highest amount of carbon (56.35% DM). Mixed pellets with the highest proportion of Scots pine sawdust (SPI 75/Wh 25 and SPI 75/R 25) were classified into the intermediate homogeneous group “ab”. The carbon contents of the other three pellets from forest dendromass were lower by 2.2 to 5.4%. In contrast, when compared with Scots pine I pellets, the content of carbon in pellets made from agricultural dendromass, i.e., SRC poplar and SRC willow, was lower by 3.5 to 3.8%. In addition, it was much lower in the pellets made from rapeseed straw and wheat straw, by 14.1 to 15.9%. Therefore, increasing the proportion of wheat straw or rapeseed straw in proportions of 25% or 50% w/w in relation to Scots pine sawdust or SRC willow reduced the C contents of pellets from these mixtures. In another study [13], the C content in Scots pine pellets was even higher (58.40% DM), and in SRC pellets, it was closer to the results of the present study. Therefore, in the cited study, the difference in C content between the Scots pine and SRC pellets was greater by 6.6 to 8.6%. Moreover, it was concluded that increasing the proportion of SRC biomass from 25% to 75% in relation to Scots pine biomass resulted in a reduction in the carbon content of pellets made from these mixtures, within the range from 1.0 to 7.6%. In another study [14], Scots pine pellets contained less carbon (55.5% DM). The C content of pellets from SRC willow and SRC poplar one-year-old shoots (53–54% DM) was similar to that indicated in the present study [22]. A lower content of this element (50.7% DM) was noted in pellets from SRC willow harvested over a five-year cycle [65]. In another study [74], the C content of pellets from straw was lower (47.9% DM) and similar to the results obtained in the present study for pellets made from wheat straw and rapeseed straw. Another study [51] found that the C content of pellets from spring rapeseed straw ranged from 46.1 to 47.5% DM, while for pellets from wheat straw, the value was 44.3% DM [33]. Even lower contents of this element (41.5–46.2% DM) were determined in pellets made from the straw of the genus Miscanthus [68,69].

In all the 16 types of pellets under study, the hydrogen content was above 6% DM, ranging from 6.02% DM to 6.57% DM, respectively, for wheat pellets (homogeneous group “f”) and Scots pine I pellets (homogeneous group “a”) (Table 7). Regarding the content of this element, the last homogeneous group “f” contained rapeseed pellets and Scots pine II pellets, as well as beech pellets and SPI 75/R 25 pellets. In another study, the H content of 22 types of pellets made from forest dendromass and agricultural SRC dendromass, as well as from their mixtures, was also above 6% DM [13]. The situation was similar for most pellets made from forest biomass (except for oak pellets, 5.8% DM) [14], although in the cited study, the H content (6.1% DM) in Scots pine pellets was lower compared with the results of the present study. Similar H contents (approximately 6.2% DM) [74,75] or slightly lower contents (5.9% DM) [76] were noted in other pellets from forest dendromass. However, the lowest hydrogen content (5.5% DM) was noted for pellets made from the straw of annual crops [77], spring rapeseed (5.4–5.7% DM) [51], wheat (4.9% DM) [33], and the genus Miscanthus (4.9–5.4% DM) [68,69,70].

Scots pine I pellets were characterized by the lowest N content (0.10% DM), and the contents of this element in the other three pellets from forest dendromass were 3-fold higher (Table 7). This was because the other three pellets were made from non-debarked feedstocks. In contrast, the N content of pellets made from agricultural dendromass, i.e., SRC poplar and willow, was four to six times higher than Scots pine I pellets. Wheat straw and rapeseed straw pellets performed even worse in this respect, with nitrogen contents of 0.76 and 0.84% DM, respectively, i.e., values being seven to eight times higher than those for Scots pine I pellets. Therefore, increasing the proportion of wheat or rapeseed straw by 25% or 50% w/w in relation to Scots pine sawdust or SRC increased the N content of pellets from these mixtures. Therefore, only seven pellet types (Scots pine I, alder, and beech, and four pellets from mixtures of Scots pine sawdust and straw) met the stringent requirements for the N content of classes A1 and Grade 1 for the standards ISO and KFRI, respectively (Table 2, Table 3 and Table 7). However, all pellets met the least stringent requirements of classes B (ISO) and Grade 4 (KFRI). The N content of pellets was significantly positively correlated with the chlorine content (0.51) and significantly negatively correlated with the LHV (−0.73) (Table 6). In another study [13], the N content of Scots pine pellets was lower (0.09% DM), whereas in SRC pellets, it was five times higher, similar to the results of the present study. Pellets made from SRC willow one-year-old shoots contained 0.56% DM nitrogen, while those made from poplar contained 0.73% DM nitrogen [22]. Therefore, the cited value for willow pellets was similar, and for the poplar, it was higher compared with the results of the present study. This might have been due to the younger age of the shoots and their higher bark content. In contrast, in pellets from five-year-old SRC willow shoots, the N content was lower (0.54% DM) [65]. Other studies have reported significant variation in the N content of pellets from hay (0.18–1.3% DM) [78,79] and straw (0.33–0.72% DM) [77,80]. This was also confirmed by a study conducted in Poland [51], in which the N content of pellets from spring rapeseed straw ranged from 1.05 to 1.23% DM, as well as a study carried out in Australia [33], in which pellets from wheat straw contained 0.56% N. Considerable discrepancies in the N content—0.24% DM [70], 0.40% DM [69], or 0.90% DM [68]—were determined in pellets made from straw of the genus Miscanthus. The above variation in the N content of pellets may be attributed to different environmental conditions, crop production techniques and technologies for a particular plant species, and the date of biomass harvesting. Nevertheless, it should be noted that pellets from dendromass are a more stable solid biofuel in terms of the N content compared with pellets from straw or hay. It is important to note that, in general, NO_x emissions increase with an increase in the N content in the fuel used, which is particularly relevant when utilizing biomass for energy purposes [51]. Increased NO_x emissions may occur when the weight content of N is ≥0.6%, which is of particular significance with regard to straw or grasses [81].

Scots pine I pellets showed the significantly lowest S content (0.010% DM), whereas the contents of this element in the other three pellets from forest dendromass were higher, by 140 to 160% (Figure 4). This was because the other three pellets were made from non-debarked feedstocks. The S contents of pellets from SRC poplar and willow dendromass were also higher than that of Scots pine I pellets by 100% and 220%, respectively. Wheat straw and rapeseed straw pellets performed even worse in this respect, as their S content was higher than that of Scots pine I pellets by 790% and 1780%, amounting to 0.089 and 0.188% DM, respectively. Consequently, increasing the proportion of wheat straw or rapeseed straw in proportions of 25% or 50% w/w in relation to Scots pine sawdust increased the S content of pellets from these mixtures by 240–330% or 500–760%, respectively. However, when increasing the proportion of wheat straw or rapeseed straw in relation to SRC willow, the S content of pellets from these mixtures increased by 47–75% or 153–206%, respectively. In view of the above, only seven pellet types (Scots pine I, alder, beech, Scots pine II, SRC poplar, SRC willow, and from a mixture of Scots pine sawdust and wheat straw SPI 75/Wh 25) met the stringent requirements of classes A1 and A2, according to the ISO standard in terms of the S content (Table 2, Figure 4). Moreover, two pellet types made from mixtures—SPI 50/Wh 50 and W 75/Wh 25—met the requirements of the other pellet classes (B, I1, I2, I3) according to the ISO standard. In contrast, seven pellet types (Wh, R, SPI 75/R 25, SPI 50/R 50, W 50/Wh 50, W 75/R 25, and W 50/R 50) failed to meet any of the classes according to the ISO and KFRI standards with regard to the S content. The S content of the pellets was significantly positively correlated with the contents of ash (0.83), Cl (0.80), and N (0.64) and significantly negatively correlated with HHV, LHV, BD, FC, and VM (from −0.78 to −0.50) (Table 6). In another study [13], the S content in Scots pine pellets was lower (0.007% DM), and in pellets from SRC poplar and SRC willow, the values were 0.024% and 0.026% DM, respectively. Therefore, in the cited study, increasing the proportion of SRC biomass from 25% to 75% in relation to Scots pine sawdust resulted in an increase in the sulfur content of pellets made from these mixtures, within a range from 86% to 271%. Pellets from SRC willow one-year-old shoots contained 0.043% DM sulfur, while for the poplar, it was 0.049% DM [22]. Therefore, the cited values were higher than the results in the present study, due to the younger age of the shoots and their higher bark content. In different willow pellets, an even higher S content was noted (0.067% DM), despite its longer harvesting cycle [65]. It can, therefore, be concluded that pellets made from SRC dendromass contain a greater amount of S than pellets made from forest wood. Pellets made from wheat straw and the genus Miscanthus also showed a higher S content of 0.11% DM [33,68], which was also confirmed in the present study. In another study [51], the S content of spring rapeseed straw was higher (0.73–0.78% DM) compared to the results of the present study.

Scots pine I pellets had the significantly lowest Cl content (0.003% DM), whereas the contents of this element in the other three types of pellets made from forest dendromass were higher by 33–233% (Table 7). The Cl content of pellets made from SRC poplar and willow dendromass was also higher than that of Scots pine I pellets by 267% and 233%, respectively. Wheat straw and rapeseed straw pellets performed even worse in this respect, as their Cl contents were higher than that of Scots pine I pellets by 2367% and 2267%, amounting to 0.074 and 0.071% DM, respectively. Therefore, increasing the proportion of wheat straw or rapeseed straw in proportions of 25% or 50% w/w in relation to Scots pine sawdust increased the Cl content of pellets from these mixtures by 1100–1433% or 1233–1767%, respectively. However, increasing the proportion of wheat straw or rapeseed straw in relation to SRC willow increased the Cl content of pellets from these mixtures by 150–500% or 280–470%, respectively. In view of the above, in terms of the Cl content, only six pellet types (Scots pine I, alder, beech, Scots pine II, SRC poplar, SRC willow) met the stringent requirements of classes A1 and A2, according to the ISO standard (Table 2 and Table 6). Seven pellet types met the requirements of the PFI standard (the same six types as for ISO, and, additionally, the mixture of W 75/Wh 25). In contrast, the requirements of the KFRI standard were met by 11 pellet types. All 16 pellet types met the quality requirements with regard to the Cl content according to the lowest class I3 of the ISO standard. The Cl content of the pellets was significantly positively correlated with the contents of sulfur (0.80), ash (0.75), and nitrogen (0.51) and significantly negatively correlated with most of the other features (a range from −0.83 to −0.31) (Table 6). Higher Cl contents of Scots pine pellets (0.015% DM), as well as of SRC poplar pellets (0.019% DM) and SRC willow pellets (0.020% DM), were determined in another study [13]. In turn, the content of this element in pellets from SRC willow and SRC poplar dendromass from one-year-old shoots was closer to the results obtained in the present study (0.011– 0.013% DM) [22]. However, a higher Cl content (0.42% DM) was noted for pellets from driftwood [82], straw of the genus Miscanthus (0.56% DM) [68], and spring rapeseed straw (0.84% DM) [51]. It is crucial to note that a high chloride content of fuels is unfavorable for energy efficiency, as it can lower the melting point for ash and increase ash deposition on boiler heating components [83].

3.3. General Characteristics of Pellets from Forest Biomass, Agricultural Biomass, and Their Mixtures

The high variability of the pellets under study regarding S and Cl contents was confirmed by high variation coefficients of 75.3% and 71.8%, respectively (

Table 8. Selected descriptive statistics for examined features of 16 types of pellets from forest biomass, agricultural biomass, and their mixtures (N Valid = 48).

Feature	Mean	Median	Minimum Value	Maximum Value	Lower Quartile	Upper Quartile	Coefficient of Variation (%)
MC (%)	6.80	6.64	4.71	9.33	5.83	7.93	19.12
FC (% DM)	20.90	20.73	19.59	22.36	20.50	21.54	3.44
VM (% DM)	76.68	76.67	74.88	78.29	75.78	77.75	1.40
Ash (% DM)	2.47	2.25	0.43	5.04	1.60	3.08	47.97
HHV (MJ kg−1 DM)	19.61	19.72	18.91	20.36	19.26	19.88	1.89
LHV (MJ kg−1)	16.92	16.86	16.03	17.95	16.60	17.24	2.83
BD (kg m−3)	678.94	658.41	604.30	797.60	635.23	717.80	8.48
C (% DM)	53.01	53.87	47.18	56.56	50.81	55.14	5.09
H (% DM)	6.28	6.27	6.01	6.65	6.11	6.46	2.90
N (% DM)	0.44	0.36	0.10	0.85	0.20	0.69	58.54
S (% DM)	0.057	0.045	0.010	0.188	0.025	0.084	75.27
Cl (% DM)	0.034	0.037	0.002	0.079	0.010	0.057	71.79

). As for S, the minimum value was 0.010% DM, and the maximum value was 0.188% DM; for Cl, the minimum and maximum values were 0.002% and 0.079% DM, respectively. High variation was also noted for the N and ash contents, for which the variation coefficients were 58.5% and 48.0%, respectively. The range of values (minimum–maximum) for N was from 0.10% to 0.85% DM, while for ash, it was from 0.43% to 5.04% DM. In contrast, the moisture content of pellets showed a lower variation (19.1%). As for the other examined features of pellets from forest biomass, agricultural biomass, and their mixtures (FC, VM, HHV, LHV, BD, C and H), the variation coefficient was <10%. Therefore, it can be concluded that, depending on the type of biomass from which pellets have been produced, greater variation can be expected, mainly in relation to ash formation and emissions of sulfur and nitrogen compounds. Unfortunately, these unfavorable features are mainly associated with straw pellets, whereas wood biomass pellets, including those from SRC dendromass, had more favorable properties in this respect.

The simple regression models confirmed a linear relationship between LHV growth and the increase in the share of woody biomass in relation to agricultural biomass (Figure 5). In each of the analyzed cases, the simple regression models were statistically significant (p < 0.05), with standard errors of estimation ranging from 0.08 to 0.27 MJ kg⁻¹. The best models were obtained for SPI/R and SPI/W pine pellets, where the determination coefficients were 0.986 and 0.8034. The smallest dynamics of LHV changes were observed in the case of willow pellets, while the increase in the share of pine biomass significantly increased the rate of LHV growth. In the case of SPI/R, the dynamics of LHV growth was the highest and amounted to 0.0179 MJ kg⁻¹ per 1% increase in the share of woody biomass and a decrease in the share of agricultural biomass. In the case of willow pellets, lower determination coefficients were obtained—0.5886 for W/R and 0.4446 for W/Wh.

The simple regression models confirmed a linear relationship between the decrease in ash content and the increase in the share of wood biomass in relation to agricultural biomass (Figure 6). In each of the analyzed cases, the simple regression models were statistically significant (p < 0.05), with standard errors of estimation ranging from 0.15 to 0.27%. All models were characterized by high determination coefficients from 0.955 for W/R to 0.9914 for SPI/R. A greater reduction in ash content was observed in the case of pine pellets, where the increase in the share of pine biomass significantly reduced the ash content. In the case of willow pellets, the reduction in ash content was at a lower level.

The simple regression models confirmed the linear relationship between the decrease in sulfur content and the increase in the share of wood biomass in relation to agricultural biomass (Figure 7). In each of the analyzed cases, the simple regression models were statistically significant (p < 0.05), with standard errors of estimation ranging from 0.00327 in the case of W/Wh pellets to 0.00839% in the case of W/R pellets. All models were characterized by very high coefficients of determination from 0.977 for SPI/Wh to 0.987 in the case of SPI/R. A greater reduction in sulfur content was observed in the case of pine pellets than in the case of willow pellets.

A cluster analysis for the 12 examined pellet features also showed that, at a cut-off of 2/3 D_max, two separate clusters were formed (Figure 8a). One common cluster specifically contained features unfavorable for energy purposes, including the ash, N, Cl, S, MC, FC, and moisture contents. The remaining features (FC, VM, DB, H, C, HHV, and LHV) formed the second cluster. However, when the accuracy of the analysis was increased, three clusters were formed at a cut-off of 1/3 D_max. The first cluster, containing unfavorable energy-related features, remained unchanged. In contrast, FC, DB, and H separated and formed their second cluster, with VM, C, HHV, and LHV remaining in the third cluster. As for the cluster analysis for 16 types of pellets from forest biomass, agricultural biomass, and their mixtures, at a cut-off of 2/3 D_max, two main clusters were also distinguished (Figure 8b). Six pellets, including wheat straw and rapeseed straw pellets, and four pellets made from mixtures of SRC willow and wheat straw or rapeseed straw, formed one common cluster. The remaining 10 types of pellets, including 4 made from forest dendromass, 2 made from SRC dendromass, and 4 made from mixtures of Scots pine sawdust and wheat straw or rapeseed straw, formed the second independent cluster. Increasing the accuracy of the analysis (at a cut-off of 1/3 D_max) allowed five pellet clusters to be distinguished. Wheat and rapeseed pellets formed one common cluster. The second cluster was formed by four pellets from mixtures of SRC willow and wheat straw or rapeseed straw. The third cluster contained SRC willow and SRC poplar pellets. The fourth cluster contained three types of pellets made from forest dendromass (Scots pine I, Scots pine II, and alder). Beech pellets, along with four types of pellets from mixtures of Scots pine sawdust and wheat straw or rapeseed straw, formed the fifth independent cluster. Therefore, a more detailed analysis showed a marked difference between pellets made from wheat straw and rapeseed straw and pellets made from SRC dendromass, as well as three types of pellets made from forest dendromass. Moreover, pellets made from mixtures of SRC dendromass and straw were in a different cluster than pellets made from mixtures of pure Scots pine sawdust and straw.

4. Conclusions

Pellets play a crucial role in the bioenergy market by providing sustainable solutions that can replace fossil fuels, thereby increasing the share of renewable energy sources. However, the main feedstocks nowadays for pellet production are forest wood residues. The present study demonstrated the possibilities of pellet production from mixtures of forest dendromass or agricultural SRC biomass and cereal straw or rapeseed straw. As expected, pellets produced from pure Scots pine sawdust exhibited the best quality. It was demonstrated, however, that not only pellets from forest dendromass (particularly those from pure Scots pine sawdust) were characterized by favorable thermophysical properties and elemental composition, but also that pellets from SRC dendromass met most requirements of the quality standards (ISO, PFI, and KRFI). However, wheat straw and rapeseed straw pellets ranked lowest in terms of quality. One possible solution to this problem would be to produce pellets from a mixture of dendromass and straw. The quality of the pellets obtained from mixtures of dendromass and straw deteriorated with an increase in the proportion of cereal straw or rapeseed straw in relation to pure Scots pine sawdust and SRC dendromass (willow or poplar). Furthermore, the present study showed that SRC willow or poplar dendromass is a valuable feedstock that can supplement and expand the range of feedstocks used for the production of high-quality pellets, especially when compared with pellets produced from forest biomass chips derived from non-debarked beech wood and from Scots pine branches obtained through forest logging. A marked difference was demonstrated between pellets made from wheat straw and rapeseed straw and those from SRC dendromass, as well as three types of pellets produced from forest dendromass. Pellets made from mixtures of SRC dendromass and straw were also different from pellets generated from mixtures of pure Scots pine sawdust and straw. Naturally, significant differences were observed among the pellets in terms of compliance with the quality requirements for the examined features, as defined by the individual pellet classes in the three standards under analysis. For example, in terms of the ash content, only the pellets from pure Scots pine sawdust met the stringent requirements of the classes A1, Premium, and Grade 1. In contrast, all the 16 pellet types met the requirements of the lower classes, i.e., Utility and Grade 4. However, with regard to the N content, only seven types of pellets met the stringent requirements of classes A1 and Grade 1, whereas all the pellets met the least stringent requirements of classes B and Grade 4. The results of the present study may serve as a basis for considering mixtures of Scots pine sawdust and wheat or rapeseed straw, as well as SRC dendromass and its mixtures with straw for pellet production. The determination of these feedstock types and their mixtures is important from the perspective of both producers and final users of pellets as an important source of renewable energy.