Open Access

An air-drying model for piled logs of Eucalyptus globulus and Eucalyptus nitens in Chile

New Zealand Journal of Forestry Science201545:17

https://doi.org/10.1186/s40490-015-0047-6

Received: 16 June 2015

Accepted: 1 September 2015

Published: 5 October 2015

Abstract

Background

Water accounts for around half of the total mass in living trees, and therefore large savings can be achieved if logs are dried before transporting or burning.

Methods

An air-drying model for piled logs of Eucalyptus globulus Labill. and Eucalyptus nitens H.Deane & Maiden was developed. Daily moisture content (MC) loss was modelled based on meteorological variables and pile characteristics. The trial had a factorial design with two species, two debarking treatments (with or without bark) and two log lengths (244 and 350 cm). Independent trials started in July 2007 (winter), October 2007 (spring) and January 2008 (summer). There were five replicate piles per season and treatment. Wood pile masses were weighted weekly or twice weekly using a crane, a 10,000 kg balance and chains to hold the piles.

Results

The main and interactive effects of seasons and treatments on daily MC loss were highly significant. However, the effect of season (climate) was far greater than the main effects of treatments or the season × treatment interaction. Overall, E. globulus dried 20 % faster than E. nitens, debarked logs dried 8 % faster than barked logs and 244 cm logs dried 3 % faster than 350 cm logs. Daily MC loss for the current day was better explained by a power function of MC at the start of the day, daily air relative humidity, daily air temperature and the number of logs per square metre of pile cross-section (or an equivalent average log diameter).

Conclusions

The air-drying model for piled logs can be used to predict drying times (days) to achieve a given target moisture content, providing a new tool for decision-making in forest transport and industrial planning.

Keywords

Log air-drying model Environmental variables Eucalyptus Management decisions

Background

Chile produces around 41 million cubic metres of timber each year from around 2.4 million ha of planted forests. Pinus radiata D.Don (1.470 million ha), Eucalyptus globulus Labill. (541,000 ha) and Eucalyptus nitens H.Deane & Maiden (232,000 ha) are the most widely planted species (Instituto Forestal 2014). Pinus radiata is the main source for sawn timber and long fibre for the pulp industry, while E. globulus and E. nitens are the main providers of short fibre for pulp production, and marginal providers for veneer and other uses (Lignum 2005).

Chilean forest companies commonly dry large quantities of pulp logs at the roadside before being transported to pulp and paper plants, and recently to industrial boilers for the generation of steam, heat and electricity. The rationale for this strategy is that water accounts for around half of the total mass in living trees, and therefore large savings in transport costs are achieved when logs are dried at the roadside. Also, industrial boilers require logs under a 50 % moisture content on a dry basis (MC) to operate efficiently as the calorific power of wood increases as MC decreases. Moreover, yard handling would become unfeasible when considering drying and storing large log quantities for long periods of time.

Air-drying times vary widely from a couple of weeks to several months depending on species, dimensions, local weather conditions and time of year when the material is stacked (Simpson and Wang 2004). Rainfall, temperature, radiation and relative humidity all contribute to the wetting/drying process of wood piles. Also species exhibit a fivefold range in drying times, and water loss slows down as wood water content decreases (Simpson and Wang 2004). Additionally, different log diameters and lengths, debarking and piling specifications may strongly affect air-drying rates. Therefore, the aim of this study was to predict air-drying times for piled logs of E. globulus and E. nitens, specifically to (i) assess the influence of species, log diameter and length, and debarking on daily MC loss; (ii) correlate daily MC loss with meteorological variables; and (iii) discuss the possibility of applying such a model under different environmental conditions. The development of an air-drying model for piled logs may assist decision-makers in predicting drying times (days) to achieve a given target moisture content, and provide a new tool for decision-making in forest transport and industrial planning in Chile.

Methods

Trial specification

An experiment was set to determine MC loss of air-drying eucalypt wood piles in continental Concepción (Rucamanqui, latitude 37° 15′ S longitude 71° 55′ W) in Chile. The climate is Mediterranean with cold wet winters and hot dry summers. Rainfall for the year 2007 was particularly low (696 mm) compared with the previous year (1407 mm). Annual air temperature for 2007 was on average 10.8 °C, with a monthly minimum of 4.3 °C (June) and a monthly maximum of 19 °C (January). Rainfall in January and February 2008 was 10.1 and 7.3 mm, respectively.

Piles were about 1.4 m wide, 1 m tall and either 2.44 or 3.50 m long. Treatments were made out of the combination of three factors: species (E. globulus or E. nitens), debarking (with bark or without bark) and length (244 or 350 cm). Wood piles were stacked with freshly harvested logs (Fig. 1) at three dates: 14-July-2007 (winter), 6-October-2007 (spring) and 17-January-2008 (summer) and were measured for 167, 146 and 43 days, respectively. The total sample size was 120 piles (3 seasons × 2 species × 2 debarking treatments × 2 lengths × 5 replicates).
Fig. 1

a Layout of log piles of E. globulus and E. nitens set for the air-drying experiment. Piles were about 1.4 m wide, 1 m tall and 244 cm or 350 cm in length. Trials were started in winter, spring and summer with three factors: species (E. globulus, E. nitens), debarking (with bark, without bark) and length (2.44 m, 3.50 m). b Detail of pile-weighing procedure and the dynamometer

For convenience, piles were grouped by length. Within each length, different combinations of species and debarking treatments were alternated. Piles were separated by poles anchored in the soil (Fig. 1a). Wood pile masses were weighed weekly or twice weekly using an electronic dynamometer with a capacity of 10 tonnes, hanging from a crane and towing ropes to hold the piles (Fig. 1b). Meteorological information was obtained from an automatic weather station located at a close distance from the piles.

Data analysis

All analyses were undertaken using The R System for Statistical Computing (R Development Core Team, 2010). Variables were tested for normality and homogeneity of variance, and transformations were made as necessary to meet the underlying statistical assumptions of the models used. Analysis of variance (Proc GLM) was used to examine the main and interactive effects of species, lengths and debarking treatment on log moisture content loss. Analyses were conducted at the pile level. Tukey’s least significant difference test was used to distinguish among individual means where applicable with a confidence level of P ≤ 0.05. The effects of species, debarking treatments and lengths on slopes and intercepts of the linear relationship between average daily MC loss and number of logs per square metre cross-section were tested for significance by analysis of covariance.

Daily MC loss modelling

Following Simpson and Wang (2004), daily moisture content (MC) loss (ΔM) per day (% green basis) was modelled as
$$ \varDelta M\kern0.5em =\kern0.5em a\;{M}^b\kern0.5em {T}^c\kern0.5em {H}^d\kern0.5em {N}^e $$
(1)

where M is the MC at the beginning of the day (%, green basis), T is the daily average air temperature (°C), H is the daily average air relative humidity (%), N is the number of logs per square metre cross-section and a, b, c, d and e are parameters to be determined by non-linear regression. Dummy variables were included in parameter “a” of model (1) in order to account for the effect of species, debarking and log lengths, as: a = α 0 + α 1 D 1 + α 2 D 2 + α 3 D 3, where D 1, D 2, D 3 are dummy variables representing species (0 = E. globulus, 1 = E. nitens), debarking (0 = without bark, 1 = with bark) and lengths (0 = 244 cm, 1 = 350 cm), respectively; and α 0, α 1, α 2 and α 3 are parameters to be determined by non-linear regression.

Results

Treatment effects

There were inherent treatment differences in pile volume and mass when first stacked. When freshly stacked, log piles ranged from 1748 to 4299 kg, 2.7 to 6.2 m3 cord (stacked) volume, 1.6 to 3.9 m3 solid volume and 9 to 35 logs m−2 in the cross-section (or equivalently 14 to 33 cm in average log diameter) (Table 1). The initial wood moisture content ranged from 47 to 60 % (average 54.3 %) on a green basis (89 to 149 % on a dry basis, average 119.6 %). Basic wood density (±1 standard error, n = 30) was significantly greater for E. globulus (514 ± 6 kg m−3) than E. nitens (491 ± 5 kg m−3, P = 0.003).
Table 1

Initial and final condition of air-drying log piles (n = 120) of E. nitens and E. globulus in two lengths (244 cm, 350 cm) with and without bark in Rucamanqui, Los Angeles, Chile

Season

Species

Treatment

n

Days

Green mass start (kg)

Water loss (kg)

Water loss (kg day−1)

Total MC loss (%, green)

Daily MC loss (% day−1, green)

Total MC loss (%, dry)

Daily MC loss (% day−1, dry)

Winter

E. nitens

244 bark

5

167

1950 ± 71e

529 ± 16d

3.2 ± 0.1d

27.2 ± 0.4bc

0.16 ± 0.003bc

62.9 ± 0.7a

0.38 ± 0.004a

  

244 debark

5

167

2172 ± 115de

608 ± 39cd

3.6 ± 0.2cd

27.9 ± 0.6bc

0.17 ± 0.004bc

59.8 ± 1.2a

0.36 ± 0.007a

  

350 bark

5

167

2848 ± 103bc

658 ± 20cd

3.9 ± 0.1cd

23.1 ± 0.5d

0.14 ± 0.003d

50.8 ± 1.4b

0.30 ± 0.008b

  

350 debark

5

167

2534 ± 171cd

693 ± 49c

4.1 ± 0.3c

27.4 ± 0.6bc

0.16 ± 0.003bc

57.8 ± 1.1ab

0.35 ± 0.006ab

 

E. globulus

244 bark

5

167

2387 ± 68cde

641 ± 15cd

3.8 ± 0.1cd

26.9 ± 0.6c

0.16 ± 0.003c

63.0 ± 1.6a

0.38 ± 0.009a

  

244 debark

5

167

2751 ± 42c

877 ± 19b

5.3 ± 0.1b

31.9 ± 1.1a

0.19 ± 0.006a

69.0 ± 2.9a

0.41 ± 0.018a

  

350 bark

5

167

3225 ± 123ab

899 ± 31b

5.4 ± 0.2b

28.0 ± 1.1bc

0.17 ± 0.006bc

64.5 ± 2.8a

0.39 ± 0.017a

  

350 debark

5

167

3435 ± 70a

1051 ± 41a

6.3 ± 0.2a

30.6 ± 0.9ab

0.18 ± 0.005ab

65.6 ± 2.3a

0.39 ± 0.014a

  

Mean

40

 

2663 ± 83

744 ± 28

4.5 ± 0.2

27.9 ± 0.5

0.17 ± 0.003

61.7 ± 1.0

0.37 ± 0.006

Spring

E. nitens

244 bark

5

146

2526 ± 81d

993 ± 30cde

6.8 ± 0.2cde

39.4 ± 0.8ab

0.27 ± 0.005ab

93.7 ± 1.8a

0.64 ± 0.012a

  

244 debark

5

146

2942 ± 100cd

958 ± 36de

6.6 ± 0.2de

32.6 ± 0.5cd

0.22 ± 0.004cd

73.1 ± 1.7c

0.50 ± 0.012c

  

350 bark

5

146

3796 ± 152a

1178 ± 65bc

8.1 ± 0.4bc

31.1 ± 1.3d

0.21 ± 0.009d

73.0 ± 2.8c

0.50 ± 0.019c

  

350 debark

5

146

3761 ± 112a

1506 ± 57a

10.3 ± 0.4a

40.1 ± 1.2a

0.27 ± 0.008a

85.1 ± 2.4ab

0.58 ± 0.016ab

 

E. globulus

244 bark

5

146

2487 ± 102d

912 ± 22e

6.2 ± 0.2e

36.8 ± 0.9ab

0.25 ± 0.006ab

86.2 ± 1.7bc

0.59 ± 0.012bc

  

244 debark

5

146

2668 ± 92cd

992 ± 25cde

6.8 ± 0.2cde

37.2 ± 0.6ab

0.26 ± 0.004ab

80.5 ± 1.4bc

0.55 ± 0.010bc

  

350 bark

5

146

3133 ± 67bc

1119 ± 30cd

7.7 ± 0.2cd

35.7 ± 0.4bc

0.24 ± 0.002bc

84.4 ± 1.7bc

0.58 ± 0.011bc

  

350 debark

5

146

3577 ± 192ab

1330 ± 65ab

9.1 ± 0.4ab

37.2 ± 0.5ab

0.25 ± 0.003ab

77.8 ± 2.7c

0.53 ± 0.018c

  

Mean

40

 

3111 ± 90

1124 ± 34

7.7 ± 0.2

36.3 ± 0.5

0.25 ± 0.004

81.7 ± 1.2

0.56 ± 0.009

Summer

E. nitens

244 bark

5

43

2537 ± 124b

466 ± 31d

10.8 ± 0.7d

18.3 ± 0.5d

0.43 ± 0.012d

41.1 ± 1.4c

0.96 ± 0.032c

  

244 debark

5

43

2468 ± 30b

463 ± 17d

10.8 ± 0.4d

18.8 ± 0.6d

0.44 ± 0.015d

39.1 ± 1.5c

0.91 ± 0.036c

  

350 bark

5

43

3296 ± 85a

593 ± 26c

13.8 ± 0.6c

18.0 ± 0.4d

0.42 ± 0.008d

39.9 ± 0.6c

0.93 ± 0.014c

  

350 debark

5

43

3373 ± 174a

635 ± 24bc

14.8 ± 0.6bc

18.9 ± 0.5d

0.44 ± 0.013d

38.9 ± 1.4c

0.91 ± 0.033c

 

E. globulus

244 bark

5

43

2261 ± 56b

593 ± 5c

13.8 ± 0.1c

26.3 ± 0.5bc

0.61 ± 0.011bc

59.8 ± 1.3a

1.39 ± 0.030a

  

244 debark

5

43

2445 ± 62b

727 ± 19b

16.9 ± 0.5b

29.7 ± 0.4a

0.69 ± 0.009a

58.7 ± 1.0ab

1.37 ± 0.023ab

  

350 bark

5

43

3035 ± 104a

753 ± 28b

17.5 ± 0.6b

24.8 ± 0.6c

0.58 ± 0.013c

53.0 ± 1.5b

1.23 ± 0.035b

  

350 debark

5

43

3233 ± 148a

900 ± 43a

20.9 ± 1.0a

27.8 ± 0.3ab

0.65 ± 0.008ab

53.8 ± 1.0b

1.25 ± 0.023b

  

Mean

40

 

2831 ± 76

641 ± 24

14.9 ± 0.6

22.8 ± 0.7

0.53 ± 0.017

48.1 ± 1.4

1.12 ± 0.033

ANOVA/

Covariate

   

**

**

***

***

***

***

ANCOVA

S

   

***

***

***

***

***

***

***

 

T

   

***

***

***

***

***

***

***

 

S × T

   

***

***

***

***

***

***

***

Main and interactive effects of season (S) and treatment (T) were assessed by analysis of variance and covariance. The covariate was the number of logs per square metre in the pile cross-section. Significant differences are presented as ns non-significant, * significant at P < 0.05, and *** significant at P < 0.001. Values are presented as means (± 1 SE) for seasons and treatments. Separation of means was determined by a Tukey test when applicable. Different letters indicate significant differences at P < 0.05

Log moisture content generally decreased over time in all piles across seasons, species, lengths and debarking treatments. Moisture content loss over time for piles of either E. globulus or E. nitens (350 cm log lengths/with bark) stacked in summer, spring and winter are shown in Fig. 2. Other log lengths and debarking treatments followed the same pattern as the one illustrated in Fig. 2. Log MC decreased for both species at higher rates for logs stacked in summer than spring than winter (summer > spring > winter). Also, log MC decreased at decreasing rates as logs became drier for piles stacked in summer and spring (faster at the beginning and slower as time progressed). In contrast, moisture content of logs stacked in winter decreased slowly while weather was cold and damp to decrease at increasing rates as weather conditions improved (Table 1, Fig. 2). Overall, E. globulus dried faster than E. nitens.
Fig. 2

Changes in wood moisture content (%, green basis) for six randomly selected 350 cm logs piles with bark. Open symbols represent E. nitens while closed symbols represent E. globulus

The main effects of season (F 2,95 = 456–1769) on daily log MC loss were far greater than the main effects of treatment (F 7,95 = 15–62) or their interaction (F 14,95 = 7–29), as shown by their F values and P range (Table 1). Daily log MC loss was on average threefold greater for piles stacked in summer (0.53 % per day, green) than in winter (0.17 % per day, green), while two times greater in summer than in spring (0.25 % per day, green). In contrast, the treatment effects were relatively small compared to season: E. globulus dried 3–47 % (average 20 %) faster than E. nitens, debarking increased the drying rate by 3–12 % (average 8 %) and logs 244 cm long dried 1–4 % (average 3 %) faster than logs 350 cm long (Table 1).

Because log diameter (or the number of logs per square metre in the pile cross-section) had such a strong influence on the wood moisture content loss per day, covariance analysis was carried out to separate the effects of the covariate, log diameter or number, (F 1,100 = 60, P < 0.001) from the effect of the factors (species, lengths, debarking, seasons). Slopes (F 2,100 = 3.8, P = 0.02) but not intercepts (F 2,100 = 2.9, P = 0.06) of the linear relationship between daily MC loss and the number of logs per square metre in the pile cross-section were significantly influenced by season (Fig. 3). Slopes and intercepts of these linear relationships were not influenced by species, lengths or debarking treatments when tested within each season (F 7,100 = 0.9–1.3, P > 0.28). However, it is worth noting that for summer, E. nitens showed an insignificantly yet lower slope than E. globulus (Fig. 3).
Fig. 3

The linear relationship between average log moisture content loss per day (%, green basis) and the number of logs per square metre in the pile cross-section. Analysis of covariance showed that slopes, but not intercepts, of these linear relationships were influenced by the season in which the piles were stacked. Various treatments (species, lengths, debarking) did not have any significant influence on these linear relationships. Logs stacked in summer, y = 0.20140 + 0.01534 x, r 2 = 0.61, P < 0.001; spring, y = 0.17089 + 0.00363 x, r 2 = 0.85, P < 0.001; winter, y = 0.13097 + 0.00155 x, r 2 = 0.16, P < 0.01. Open symbols represent E. nitens while closed symbols represent E. globulus

Air-drying model

Model (1) was fitted to the experimental data of daily MC loss on a green (ΔM green) and dry basis (ΔM dry). The general model for ΔM green was:
$$ \varDelta {M}_{\mathrm{green}}\kern0.5em =\kern0.5em {a}_{\mathrm{green}}\;{M}_{\mathrm{green}}^{2.3979}\kern0.5em {T}^{0.2268}\kern0.5em {H}^{-5.6734}\kern0.5em {N}^{0.7987}\kern0.5em ,\kern0.5em {r}^2\kern0.5em =\kern0.5em 0.67,\kern0.5em P<0.001 $$
with
$$ {a}_{\mathrm{green}}\kern0.5em =\kern0.5em 119020-26671\kern0.5em {D}_1\kern0.5em -\kern0.5em 40969\kern0.5em {D}_2\kern0.5em -\kern0.5em 6814\kern0.5em {D}_3 $$

where M is the MC at the beginning of the day (%, green basis), T is the daily average air temperature (°C), H is the daily average air relative humidity (%) and N is the number of logs per square metre cross-section. Dummy variables were included in parameter “a” of model (1) in order to account for the effect of species, debarking and log lengths, as: a = α 0 + α 1 D 1 + α 2 D 2 + α 3 D 3, where D 1, D 2, D 3 are dummy variables representing species (0 = E. globulus, 1 = E. nitens), debarking (0 = without bark, 1 = with bark) and lengths (0 = 244 cm, 1 = 350 cm), respectively.

The relationship between average log diameter, d, and number of logs per square metre of pile cross-section, N, (d = 14.4182 + 45.1968 e-0.12 N , r 2 = 0.91, P < 0.001) can be replaced into (1) if d rather than N is preferred to be used in the calculations.

Model (1) was also fitted to predict daily MC loss on a dry basis (ΔM dry) as
$$ \varDelta {\mathrm{M}}_{dry}\kern0.5em =\kern0.5em {\mathrm{a}}_{dry}\;{\mathrm{M}}_{dry}^{2.3570}\kern0.5em {\mathrm{T}}^{0.2460}\kern0.5em {\mathrm{H}}^{-5.7448}\kern0.5em {\mathrm{N}}^{0.8422}\kern0.5em ,\kern0.5em {\mathrm{r}}^2\kern0.5em =\kern0.5em 0.63,\kern0.5em \mathrm{P}<0.001 $$
with
$$ {a}_{dry}\kern0.5em =\kern0.5em 55319-10470\kern0.5em {D}_1\kern0.5em -\kern0.5em 22015\kern0.5em {D}_2\kern0.5em -\kern0.5em 3641\kern0.5em {D}_3 $$

There was an observed upper bound of moisture content loss (ΔM green) of 1.97 % per day on a green basis (3.96 % per day on a dry basis) given by the experimental data that should be acknowledged within the model. In practice, this means than any estimate over this value should be set to 1.97 % (3.96 % dry basis), and this upper bound would be required to ensure that realistic values of daily MC loss are estimated.

The model was apparently unbiased for a traditional residual analysis (Fig. 4 a) and also when predicting moisture content for each log pile for the duration of the experiment (Fig. 4 b). Rainfall was uncorrelated with the log moisture content loss.
Fig. 4

a Residuals of model used to estimate daily MC loss: ΔM green = (119,020—26,671 D 1 – 40,969 D 2 – 6814 D 3) M 2.3979T 0.2268H -5.6734N 0.7987, r 2 = 0.67, P < 0.001; where D 1, D 2 and D 3 are dummy variables representing species (E. globulus = 0, E. nitens = 1), debarking (without bark = 0, with bark = 1) and lengths (244 cm = 0, 350 = 1); b the linear relationship between observed and estimated longer-term log moisture content (%, green basis) for E. globulus and E. nitens for the whole experimental data

Air-drying times for E. globulus and E. nitens with and without bark are shown in Fig. 5. The graph shows estimated times (days) to reduce the moisture content on a green basis from 60 to 50, 60 to 40, 60 to 30 and 60 to 20 %, for logs stacked in any day of the year. One important observation is that piles stacked in late summer and early autumn had greatly extended drying times, particularly for low target moisture contents (20–30 %, green basis). In contrast, piles stacked in late spring to early summer required the least drying times compared with other stacking dates.
Fig. 5

Estimated air-drying times to achieve reductions of 10, 20, 30 or 40 % in log moisture content (green basis, percent MCg, right axis) stacked on any day of the year in Rucamanqui, Los Angeles, Chile. Calculations were made for an initial moisture content of 60 % on a green basis, log lengths of 244 cm and 22 logs per square metre (~18 cm diameter logs). Dashed lines represent extrapolated values. Panels (a) and (c) correspond to drying times of E. globulus with and without bark; while (b) and (d) to E. nitens with and without bark, respectively

The influence of log diameter (or the number of logs per square metre of cross-section) on the air-drying time of E. globulus logs (244 cm) with bark is shown in Fig. 6a. For instance, for logs stacked in May, a 30 % reduction in green moisture content would be achieved in 206 days for an average log diameter of 18 cm (22 logs m−2), and this period would increase by 70 days for average log diameters of 30 cm (9 logs m−2). Air-drying time was also strongly influenced by initial log moisture content for E. globulus (Fig. 6b). For instance, for logs stacked in April, a 30 % reduction in green moisture content would be achieved in 218, 248 and 290 days starting with initial wood moisture contents of 60, 55 and 50 % (green basis), respectively. A similar pattern was observed for the air-drying time of E. nitens logs (244 cm) with bark in relation to either log diameter (or the number of logs per square metre of cross-section) (Fig. 6c) or initial log moisture content (Fig. 6d), although with extended drying times compared to E. globulus.
Fig. 6

a Influence of log diameter (or the number of logs per square metre of pile cross-section) on predicted air-drying times to achieve a reduction of either 10 or 30 % in log moisture content (percentage, green basis) of E. globulus stacked on any day of the year in Rucamanqui, Los Angeles, Chile. Calculations were made for an initial moisture content of 60 %. b Influence of initial log moisture content on predicted air-drying times to achieve a reduction of either 10 or 30 % in wood moisture content (percentage, green basis) stacked on any day of the year in Rucamanqui, Los Angeles, Chile. Panels c, d are the equivalents of (a) and (b) for E. nitens. Predictions were made for 244 cm logs with bark

Discussion

Drying times of Eucalyptus logs can vary up to eightfold depending on weather conditions. Such large time variation poses tremendous uncertainty in flow planning and quality control (Simpson and Wang 2004). Additionally, moisture content has an impact on transport costs and bioenergy yields, and therefore accurately determining air-drying times becomes critical for minimising costs and maximising returns. As expected, there is a clear seasonal effect with logs drying faster when stacked in late spring to early summer than in late summer to early autumn, which can be explained by observed values of air relative humidity and temperature being the main environmental drivers of log moisture content loss. Although climate (season) was the main drive of log moisture content loss, there were also important effects of species and debarking treatments on this variable. E. globulus dried on average 20 % faster than E. nitens, and debarking enhanced moisture content loss by 8 %. For instance, in order to reduce 20 %, the moisture content on a green basis, E. globulus logs, with bark, harvested/stacked in January would take 22 days compared to 29 for E. nitens. Similarly, debarking would reduce drying times from 22 to 17 days in E. globulus and from 29 to 20 days in E. nitens to achieve a 20 % reduction in moisture content for piles freshly harvested/stacked in January. Visser et al. (2014) also found faster drying of debarked logs of P. radiata, suggesting that radial water movements are a significant contributor to the log drying process. Log lengths did not have a significant effect on moisture content loss in our study, although 244-cm logs dried on average 3 % faster than 350-cm logs.

Generally, species with low-density wood dry faster than higher density ones (Simpson and Hart 2001). However, the opposite result was found in the current study. E. globulus (which has a basic wood density of 514 kg m−3) dried on average 20 % faster than E. nitens, (which has a 5 % lower basic density (491 kg m−3)). This result may indicate that water movement is more related to fibre conductivity than wood density. Muneri and Raymond (2001) reported that E. globulus exhibits greater fibre length and coarseness than E. nitens which may point in such direction.

The daily MC loss decreased at decreasing rates as the log MC dropped. For instance, logs of E. globulus with bark would take 22, 26 and 35 days in order to reduce 20 % moisture content starting with and initial wood moisture content of 60, 55 and 50 % on a green basis, respectively (provided they are stacked in January). Log diameter also showed to have a strong influence on drying times, e.g. 20 % reductions in moisture content in 18 cm logs would take 22 days if stacked in January, and that time would almost double for 30 cm logs. A similar pattern was observed for E. nitens. This was also observed by Visser et al. (2014), for P. radiata logs, arguing that smaller diameter logs and split logs exhibit a larger surface area to mass ratio and a smaller distance water must move from the inner log to the surface in order to evaporate. Since wood-free water flows primarily along the stem evaporating form the log ends (Walker 2006), there must be a secondary movement across the stem if small diameter logs, large split logs and debarked logs dry faster than large whole logs (Abbot et al. 1997, Visser et al. 2014).

Daily MC loss (ΔM) scaled positively following a power function with the MC at the beginning of the day (M), average air temperature (T) and the number of logs per square metre in the cross-section (N), and scaled negatively with air relative humidity (H). Simpson and Wang (2004) found the same variables to significantly explain daily MC loss, but found that a power model fitted best Ponderosa pine (Pinus ponderosa Douglas ex C.Lawson) while a linear model was most suitable for Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco). Rietz (1970) considered also wind speed and rainfall to explain air-drying rates of lumber. Surprisingly, in our study, rainfall was uncorrelated to air-drying rates, which we assumed was explained because most water intercepted by the piles was evaporated, while only marginal amounts of water infiltrated the logs.

Intensively tended forest plantations of E. globulus and E. nitens in Chile provide biomass for the short- fibre pulp industry and more recently for industrial boilers for the generation of electricity and heat. The companies are looking to reduce the moisture content of the wood in order to diminish the transport cost of pulp logs and to increase the energy power of the biomass. Kiln-drying would be largely uneconomical because of large log diameters (long residence times) and low log prices not justifying high kiln-drying costs. Hence, air-drying becomes the natural alternative to kiln-drying for fuelwood (Filbakk et al. 2011). However, air-drying times are known to vary largely from several weeks to several months depending on species, dimensions, local weather conditions and time of year when the material is stacked (Simpson and Wang 2004). Most studies on air-drying times are based on stacked lumber, and only a few, notably Simpson and Wang (2004), have determined air-drying times for logs, and to the best of our knowledge none has worked with logs of Eucalyptus species.

The model fitted for E. globulus and E. nitens in the current study may be potentially applied to other sites provided that some validation can be obtained. This approach was the same originally developed by Simpson and Hart (2001), who applied a model fitted for only a few locations to many climates across the USA. Once underlying processes describing air-drying rates are explained by environmental variables, then we may correctly assume that the same processes would explain changes for other climates. Reinforcing this idea, Filbakk et al. (2011) found that air-drying rates of whole trees of broadleaved species at three sites in Norway were mainly explained by air temperature and precipitation independent of site. The simplicity of the model developed in this study, based on four variables only, may provide guidance to decision-makers in the forestry industry to minimise transport costs based on currently available meteorological data. This is something that managers may consider since moisture content has an impact on transport costs and energy yields. The model is now routinely used by Forestal Mininco S.A. to predict log air-drying times at the roadside across different environmental conditions.

Conclusions

In summary, the effect of season (climate) was far greater than the effect of species, debarking or log length on log moisture content loss. Overall, E. globulus dried 20 % faster than E. nitens, debarked logs dried 8 % faster than barked logs and 244 cm logs dried 3 % faster than 350 cm logs. Daily MC loss was explained by a power model of MC at the beginning of the day, average daily air temperature, average daily air relative humidity and the number of logs per square metre cross-section of the piles (or equivalently average log diameter). The model needs an upper bound of 1.97 % daily MC loss on a green basis (3.96 % on a dry basis). The model can be used to estimate air-drying times of piles stacked any day of the year, to any final MC, for a wide range of log diameters (in the approximate 14- to 33-cm range), and for different environmental conditions; providing a new tool for decision-making in forest transport and industrial planning.

Declarations

Acknowledgements

This research was sponsored by Forestal Mininco S.A. The experiments and measurements undertaken for this paper comply with the current laws of Chile. None of the authors have any competing interests in the manuscript.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Authors’ Affiliations

(1)
Faculty of Forestry & Conservation, University of Chile
(2)
Forestal Mininco S.A.

References

  1. Abbot, P, Lowore, J, Khofi, C, & Werren, M. (1997). Defining firewood quality: a comparison of quantitative and rapid appraisal techniques to evaluate firewood species from a Southern African savanna. Biomass & Bioenergy, 12, 429–437.View ArticleGoogle Scholar
  2. Filbakk, T, Høibø, O, & Nurmi, J. (2011). Modelling natural drying efficiency in covered and uncovered piles of whole broadleaf trees for energy use. Biomass and Bioenergy, 35, 454–463.View ArticleGoogle Scholar
  3. Instituto Forestal. (2014). Anuario Forestal 2014. [Technical Report]. Boletín Estadístico No 144 (p. 159). Concepción: Instituto Forestal. http://wef.infor.cl/publicaciones/publicaciones.php. Access date: August 23, 2015.Google Scholar
  4. Lignum. (2005, April). Eucalyptus nitens, Nueva Especie Latifoliada Industrial. Revista Lignum. Google Scholar
  5. Muneri, A, & Raymond, CA. (2001). Nondestructive sampling of Eucalyptus globulus and E. nitens for wood properties; II. Fibre length and coarseness. Wood Science and Technology, 35, 41–56.Google Scholar
  6. Rietz, RC. (1970). Air drying lumber in a forklift yard. [USDA forest service research note Fpl-0209]. Madison, WI, USA: US Department of Agriculture, Forest Service Forest Products Laboratory.Google Scholar
  7. Simpson, WT, & Hart, CA. (2001). Method for estimating air-drying times of lumber. Forest Products Journal, 51(11/12), 56–63.Google Scholar
  8. Simpson, WT, & Wang, X. (2004). Estimating air-drying times of small-diameter ponderosa pine and Douglas-fir logs. Forest Products Journal, 54(12), 24–28.Google Scholar
  9. Visser, R, Berkett, H, & Spinelli, R. (2014). Determining the effect of storage conditions on the natural drying of radiata pine logs for energy. New Zealand Journal of Forestry Science, 44: 3.View ArticleGoogle Scholar
  10. Walker, J. (2006). Primary wood processing (2nd ed.). Dordrecht, Netherlands: Springer Verlag.Google Scholar

Copyright

© Bown and Lasserre. 2015