The impact of extra long-term storage of logging residues on fuel quality in Estonian conditions – a case study

Estonia has formally ratified the Paris Agreement on Climate Change, aligning with the EU commitment to reduce greenhouse gas (GHG) emissions. In accordance with the ‘Estonia 2035’ development strategy and the Long-term Development Programme for the Estonian Energy Sector up to the year 2035 (ENMAK 2035, 2023), the national objective by 2050 is to achieve an 80% reduction in GHG emissions compared to 1990 levels, aiming to become a climate-neutral country. To attain these ambitious goals, there is a significant emphasis on increasing the utilization of wood as a renewable energy source.

While the annual increment of Estonian forests is estimated at about 16 million cubic meters (Mm3), environmental constraints designate 33.2% of forest areas for protection (Forest, 2023), limiting the further growth of harvesting volume. Consequently, one solution to augment the share of energy wood in Estonia’s energy balance involves focusing on less exploited assortments of raw materials and meticulous logistics planning.

In recent years, milder winters have posed challenges to the extraction of residues and local transport of wood fuels on soft and unfrozen soils. This situation complicates the reliable supply of wood fuels. In instances where logging residues on unfrozen soils are inaccessible, there is a necessity to process older piles of logging residues. The quality of wood chips derived from such piles is lower compared to the norm.

Research on the storage of logging residues (Filbakk et al., 2011; Jirjis, 1995; Jirjis & Lehtikangas, 1993; Nilsson et al., 2013; Nurmi & Hillebrand, 2001; Routa et al., 2015a, 2015b, 2016, 2018) and wood chips (Anerud et al., 2018; Hofmann et al., 2018; Kuptz et al., 2020; Nurmi, 1999; Pari et al., 2017; Whittaker et al., 2018) has consistently shown biomass loss during storage, emphasizing the need to limit storage time. However, existing studies predominantly cover short-term storage, up to 18 months, with little attention given to long-term storage, exceeding two years. Under specific circumstances, there is a demand to utilize older piles of logging residues. Hence, the objective of this study is to investigate the impact of extra long-term storage (up to 8 years) on the properties of logging residues as fuel.

In 2010, sample piles were established in the Järvselja Training and Experimental Forest Centre, specifically in compartments JS177 (subcompartment 8) and JS207 (sub-compartment 3), to monitor the degradation process of logging residues. The piles were created in final felling areas of birch (Betula spp.) and spruce (Picea abies (L.) H. Karst.), employing two storage methods – uncovered and covered piles (Figures 1 and 2).

Figure 1.

Figure 2.

Samples were systematically collected from four types of piles, totalling 17 times over an 8-year period from 2010 to 2018 (Table 1). The purpose was to analyse various properties of woody biomass during storage, including ash content, wood density, moisture content, and calorific value. The analyses adhered to established standards for determining the properties of solid biofuels (EVS-EN ISO 18122, 2015; EVS-EN ISO 18125, 2017; EVS-EN ISO 18134-2, 2017).

Additionally, the diameter drying shrinkage was computed for each sample. This involved measuring the diameter of the branch section in two directions before and after drying. The relative diameter shrinkage (Shr) was calculated by dividing the difference between the wet and dry wood diameter by the wet sample diameter. It was essential to establish a consistent moisture content for this calculation, and the decision was made during the initial sampling on 16 June 2010 (Table 1). The logging residues had been standing and drying in the clear-cutting area, and it was determined to use the base moisture content which was 24.53% for Norway spruces and 25.60% for silver birches.

To characterize the alterations in the properties of logging residues, ANCOVA or regression analysis was employed, utilizing the lm function within the statistical software R (R Core Team, 2019). Subsequently, the ANCOVA equation format was employed to estimate various properties of logging residues, including ash content, moisture content, bulk density, calorific value, shrinkage of diameter, or energy density, represented by the equation: (1) Y=a0+a1⋅t+a2+a3+a4+a5⋅t⋅Sp+a6⋅t⋅C+a7⋅t⋅P, Y = {a_{\it 0}} + {a_{\it 1}} \cdot t + {a_{\it 2}} + {a_{\it 3}} + {a_{\it 4}} + {a_{\it 5}} \cdot t \cdot Sp + {a_{\it 6}} \cdot t \cdot C + {a_{\it 7}} \cdot t \cdot P, where Y – ash content, moisture content, bulk density, calorific value, shrinkage of diameter or energy density; %; t – days from the start of the experiment; a₂ – the coefficient for Norway spruce species, and it is zero for silver birch species; a₃ – the coefficient for an uncovered pile, and it is zero for a covered pile; a₄ – the coefficient for a sample taken from the top of the pile, and it is zero for a sample from the middle of the pile; Sp – species parameter (silver birch – 1, Norway spruce – 2); C – covering parameter (covered – 1, uncovered – 2); P – placement parameter (in the middle of the pile – 1; on top of the pile – 2).

In order to determine the loss of wood mass, it is relevant to establish various relationships within the data. Initially, the correlation between the drying shrinkage of wood diameter (Shr) and the moisture content, as well as the duration of storage, was identified through regression analysis. This involved employing two distinct equations for Norway spruce (Equation 2) and silver birch (Equation 3): (2) Shrspruce=a1⋅Mart+1+a2⋅ln(t⋅Mar+1), Sh{r_{spruce}} = {a_1} \cdot {{{M_{ar}}} \over {t + 1}} + {a_2} \cdot \ln \left( {t \cdot {M_{ar}} + 1} \right), (3) Shrbirch=a0+a1⋅Mar+a1⋅tMar+a3⋅t2, Sh{r_{birch}} = {a_0} + {a_1} \cdot {M_{ar}} + {a_1} \cdot {t^{{M_{ar}}}} + {a_3} \cdot {t^2}, where Shr_species – drying shrinkage of diameter, %; M_ar – moisture content, %; t – days from the start of the experiment; a₀, a₁, a₂, a₃ – equation parameters.

In the regression analysis, an equation (Equation 4) was employed to explain the impact of the season on moisture content: (4) Mar=a0+a1⋅Month+a2⋅Month2, {M_{ar}} = {a_0} + {a_1} \cdot Month + {a_2} \cdot Mont{h^2}, where M_ar – moisture content, %; Month – the sequence number of the month per year (January – 1, February – 2, etc.).

Subsequently, using the base moisture content values (Norway spruce – 24.53% and silver birch – 25.60%) and the number of days since the start of the experiment, the diameter drying shrinkage from base moisture to oven dry was computed for each sample utilizing Equations 2 or 3. Following this, the relative volume of logging residues was calculated at base moisture, considering oven-dry volume as 1: (5) V=(11−Shr100%)2, V = {\left( {{1 \over {1 - {{Shr} \over {100\% }}}}} \right)^2}, where V represents the relative volume at the base moisture compared to the oven-dry volume; Shr indicates the drying shrinkage of the diameter from the base moisture content, %.

The bulk density of the absolute dry material (ρ_od) was assessed for each sample. To calculate the loss of matter during storage, it was necessary to determine the density, which defines the ratio of absolute dry mass to the volume at base moisture content (ρ_bm): (6) ρbm=ρodV, {\rho _{bm}} = {{{\rho _{od}}} \over V}, where ρ_od – the bulk density of the oven-dry material; ρ_bm – the bulk density with oven-dry mass and volume at the base moisture content; V – the relative volume at the base moisture compared to the oven-dry volume.

Regression analysis was subsequently employed to investigate the correlation between storage time and the density of the samples, which was calculated based on dry weight and the initial moisture content volume: (7) ρbm=a0⋅ea1⋅t, {\rho _{bm}} = {a_0} \cdot {e^{{a_1} \cdot t}}, the linear form of this equation is expressed as: (8) ln(ρbm)=ln(a0)+a1⋅t, \ln \left( {{\rho _{bm}}} \right) = \ln \left( {{a_0}} \right) + {a_1} \cdot t, where ρ_bm – the bulk density with oven-dry mass and volume at the base moisture content; t – the number of days from the commencement of the experiment; a₀, a₁ – parameters of the equation.

The assessment of energy loss during the storage of logging residues involved the application of the aforementioned equations. Initially, the net calorific value of the dry matter was computed across various storage durations, ranging from 0 to 3,000 days (Equation 1). Subsequently, a relative net calorific value was derived for distinct storage durations, with the calorific value at the onset of the experiment (0 days) set as the reference point at 100% (R_q).

In the second step, the bulk density of the material at different time intervals was determined using Equation 7, employing oven-dry mass and volume at the base moisture. Next, a relative density was established, considering the material’s density at the commencement (0 days) as 100% (R_ρ).

The relative remaining energy content for diverse storage durations was ascertained by multiplying the relative net calorific value by the relative bulk density using the formula: (9) R=Rq⋅Rρ, R = {R_q} \cdot {R_\rho }, where R – relative remaining energy content, %; R_q – relative net calorific value, %; R_ρ – relative bulk density, %.

ANCOVA was employed to explain the temporal evolution of logging residues properties, as delineated by Equation 1. The outcomes of the analysis are detailed in Table 2, encompassing the statistical parameters associated with the calculation equations for diverse properties. Table 3 displays the parameter estimates derived from Equation 1, along with their corresponding significance probabilities.

The subsequent figures illustrate the influence of storage duration on the characteristics of residues. As depicted in Figure 3, the ash content of all sample types did not surpass 2% within the initial three years of storage. Prolonged storage revealed a noticeable rise in the ash content of birch logging residues, with the highest levels observed in covered birch residues. In contrast, the ash content of spruce residues remained constant throughout the observation period. The lowest ash content was observed in uncovered spruce residues. The extended storage of wood chips derived from Mediterranean poplar plantations revealed a gradual increase in ash content, rising from 2.91% to 3.31% over an 18-month period (Pari et al., 2017). The observed elevation in ash content surpasses that observed in the analysis of logging residues in this study.

Figure 3.

The net calorific value of dry matter in logging residues exhibits a consistent upward trend during storage, as illustrated in Figure 4. Conversely, the net calorific value of moist fuel demonstrates a decline, as depicted in Figure 5, with birch experiencing a more rapid decrease. Covered storage piles maintain a higher net calorific value for moist fuel in comparison to uncovered piles. Additionally, both bulk density and energy density decrease during storage, as evidenced by Figures 6 and 7.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Furthermore, prior research has addressed the issue of dry matter loss. In Norway, the dry matter loss in softwood logging waste was reported to be between 1 and 3% per month (Filbakk et al., 2011), a rate exceeding that observed in our current study. Routa et al. (2015a, 2016) documented a dry matter loss ranging from 0–2.9% per month over an 8-month period, and in a more extended study spanning 10.2–14.5 months, the observed dry matter loss was in the range of 0.07% to 0.95% per month (Routa et al., 2018). These findings are consistent with our own results. A study conducted in Sweden on the storage of common spruce wood chips (including small amounts of pine and birch wood chips) for a period of 7 months revealed a dry matter loss of 5.8% for the covered pile compared to 7.3% for the uncovered pile (Anerud et al., 2018). In a study conducted in Bavaria, where wood chips (produced from spruce logging residues) were stored from May to October (153 days), a dry matter loss of 7.4% was observed (Kuptz et al., 2020). Another study carried out in Bavaria, focusing on the storage of wood chips derived from typical spruce logging residues in stacks, unveiled a monthly dry matter loss ranging from 0.7% to 2.2% (Hofmann et al., 2018). In a study conducted in the United Kingdom, the storage of willow wood chips in stacks for six months resulted in a dry matter loss ranging from 19.8% to 22.6% (Whittaker et al., 2018). According to the findings of these studies, it can be inferred that the loss of dry matter when storing as wood chips is greater than when storing as logging residues.

Figure 8 demonstrates that storage leads to an increase in the drying shrinkage of wood samples over time.

Figure 8.

As observed by Nilsson et al. (2013) in Sweden, the moisture content in wood piles during the summer months typically ranges from 30–40%. Figure 9 depicts the considerable variation in moisture content throughout the year, exceeding 20%. This variability suggests that selecting an opportune moment for chipping residues can significantly enhance the quality of wood chips. Additionally, Figure 9 presents the equations derived from the regression analysis (Equation 4), with parameter estimates detailed in Table 4. During the summer, characterized by lower precipitation and warmer weather, the moisture content of logging residues decreases. Conversely, during the autumn-winter period, the moisture content increases, which is consistent with findings in other studies (Jirjis & Lehtikangas, 1993; Nurmi, 1999; Nurmi & Hillebrand, 2001). A study in Finland indicates that uncovered piles have an approximately 5% higher moisture content than covered piles in winter (Routa et al., 2015b). Our results, as shown in Figure 9, confirm this for Norway spruce logging residues. However, for birch logging residues, the uncovered pile exhibits over 10% more moisture content than the covered pile.

Figure 9.

In Figures 10 and 11, the influence of storage duration and humidity on the shrinkage of diameter and alteration in the moisture content of birch and spruce branches is illustrated. To characterize the correlation between drying shrinkage, storage time, and moisture content, regression analysis was employed. Equation 2 was utilized for the Norway spruce data-set, while Equation 3 was applied to the silver birch data. The parameter estimates resulting from the regression analysis are presented in Table 5.

Figure 10.

Figure 11.

The drying shrinkage percentage of the diameters of silver birch logging residues decreases during storage, with the rate of decrease initially slow and progressively accelerating over time (Figure 10). Conversely, for spruce, an opposite trend is observed, where the drying shrinkage of diameters increases with time (Figure 11). Initially, the shrinkage undergoes a faster change, followed by a slower rate of change.

The disparity in shrinkage between birch and spruce can be attributed to the distinct characteristics of these tree species. Shrinkage is known to be influenced by specimen dimensions, drying speed, and density (Bowyer et al., 2003). In the case of birch, the reduction in drying shrinkage can be explained by the decrease in density. Denser wood typically exhibits higher drying shrinkage, whereas wood with lower density tends to have lower shrinkage (Saarman & Veibri, 2006). However, explaining the increase in drying shrinkage of spruce logging residues is challenging, as the density of spruce logging residues also decreased during storage.

Analysing the changes in the bulk density of residues during storage (see Figures 12 and 13), the decline in the energy content of logging residues over long-term storage was calculated, and the results are illustrated in Figure 14.

Figure 12.

Figure 13.

Figure 14.

The energy loss incurred during the storage of logging residues was determined by applying the net calorific value of the dry matter (Equation 1, Figure 4). Additionally, the bulk density, calculated using the oven-dry mass and volume at the base moisture, was assessed at various time points (Equation 7, Figures 12 and 13). Utilizing Equation 9, the residual energy content for different storage periods was computed and presented in Figure 14 and Table 6.

In their study, Anerudi et al. (2018) observed that the energy loss of wood chips (primarily from Norway spruce and to a lesser extent from Scots pine and birch) of 0.6% when stored in a covered pile and 5.3% when stored in an uncovered pile over a 7-month period.

The data illustrates the decline in energy content of the logging residues during long-term storage, differentiated by species and pile placement over an 8-year period. Notably, birch exhibits a higher loss of dry matter compared to spruce. The 8-year observation indicates approximately a 30% drop for spruce and over 50% for birch. This degradation, attributed to micro-organisms and various processes, also results in a decline in the energy density of the fuel.

Despite the extensive degradation during extra long-term storage, it is noteworthy that the quality of dry fuels remains compliant with standards and remains acceptable for use in boiler houses.

The combustion properties of logging residues as fuel are significantly influenced by their moisture content and bulk density. These two factors have a cascading effect on other properties. Notably, the moisture content of covered piles was found to be lower than in uncovered piles. The ash content, however, exhibited different trends for silver birch and Norway spruce logging residues during the storage period. Silver birch residues showed an increase, while Norway spruce residues demonstrated a decrease. Overall, the ash content did not exhibit a strong dependence on storage time.

Examining the relationship between the net calorific value of dry matter and storage time revealed that silver birch logging residues experienced a greater increase in calorific value compared to Norway spruce. The increase in the net calorific value was attributed to the breakdown of hemicellulose and cellulose, leading to a higher proportion of lignin with a higher calorific value.

The bulk density and energy density of silver birch logging residues decreased at a faster rate than those of Norway spruce. The uncovered logging residues of both tree species had a lower energy density compared to the covered ones. Additionally, the moisture content was lower in summer months than in winter.

Diameter drying shrinkage had a significant correlation with storage time, with silver birch residues experiencing greater reduction than Norway spruce. Organic matter loss during storage was more pronounced in silver birch compared to Norway spruce. The decrease in the density of dry mass per volume unit was significantly higher in silver birch over the storage period, leading to a faster decline in energy density.

In practical terms, it is recommended that the storage time for logging residues should not exceed one or two years due to extensive degradation and dry matter loss. Despite this, the quality of dry fuels remains acceptable and meets standard requirements even after longer storage, making them suitable for use in boiler houses.