Available online at www.sciencedirect.com ScienceDirect

Energy Procedia 59 ( 2014 ) 134 – 141

European Geosciences Union General Assembly 2014, EGU 2014 yield and feedstock characteristics of energy in Thailand

András Darabanta,b*, Maliwan Haruthaithanasana, Wanida Atklaa, Tepa Phudphonga, Eakpong Thanavata, Kasem Haruthaithanasana

aKasetsart Agricultural and Agro-industrial Product Improvement Institute, Kasetsart University, 50 Ngam Wong Wan, Bangkok, Thailand bBOKU – University of Natural Resources and Life Sciences, Institute of Ecology, Peter Jordan Str. 82, Vienna, Austria

Abstract

Bamboo plantations in two locations in eastern Thailand differed in their biomass yield by an order of magnitude as a result of site productivity and management. The biomass yield of beecheyana and Dendrocalamus membranaceus was comparable, but the moisture content of culms of B. beecheyana was considerably higher as compared to D. membranaceus. With D. membranaceus, internodes had higher moisture content, as compared to nodes. The moisture content decreased with increasing height along culms with both species, but this gradient was stronger with D. membranaceus. The moisture content of culms of B. beecheyana declined with increasing culm age, indicating that older culms are more suitable for energetic utilization. While general feedstock characteristics of the two species were comparable, the calorific content of D. membranaceus was significantly higher than of B. beecheyana. Primarily the upper sections and nodes of older culms of both species are attractive options as bioenergy feedstock, but plantations established on marginal sites without proper plantation management will result in very low yields.

©© 2014 2014 The The Authors. Authors. Published Published by byElsevier Elsevier Ltd. Ltd. This is an open access article under the CC BY-NC-ND license (Peer-reviewhttp://creativecommons.org/licenses/by-nc-nd/3.0/ under responsibility of the Austrian). Academy of Sciences. Peer-review under responsibility of the Austrian Academy of Sciences Keywords: Bamboo; Bambusa beecheyana;biomass;Dendrocalamus membranaceus; feedstock characteristics; plantation productivity; Thailand.

* Corresponding author. Tel.: +43-1-47654-4124; fax: +43-1-47654-4129. E-mail address: [email protected]

1876-6102 © 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/). Peer-review under responsibility of the Austrian Academy of Sciences doi: 10.1016/j.egypro.2014.10.359 András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141 135

Nomenclature

DBH Diameter at Breast Height

LHV Lower Heating Value

HHV Higher Heating Value

1. Introduction

Given wide ecological amplitudes, high growth rates and multipurpose use, bamboo species have always been important resources in those parts of the world they abundantly occur in (e.g. , India and Brazil) [1]. In recent decades, bamboo has turned into a globally important biomass resource [2], partially owing to the rate of economic development in these regions. Besides being a traditional source of energy and other uses, electric valorization of bamboo biomass through gasification has been pioneered recently [3]. Additionally, the acceptance of bamboo in volunteer carbon finance mechanisms amplified its attractiveness as plantation species. These developments have recently led to the expansion of bamboo plantations also in Thailand, even though specific experience on biomass yield, as well as energetic and alternative utilization options are still limited. Allometric functions have been developed for a number of bamboo species in order to estimate above-ground biomass stocks using easy-to-assess proxy variables, such as culm DBH [4, 5]. These allometric functions usually describe culm biomass with very high coefficients of determination. Above-ground biomass stocks of bamboo plantations show a large variation world-wide, with highest stockings around 300 t*ha-1 [6] and the lowest around 25 t*ha-1 [7], respectively. Plantation choices in terms of species, propagation stock, spacing, fertilization, irrigation and harvesting techniques lead to large differences in biomass yield [1]. Intensive plantation management may result in nutrient loss, which can be minimized by restricting biomass removal to culms with nutrients concentration lower than other biomass compartments [8]. K concentrations in bamboo culms, however, are not markedly lower compared to other biomass compartments, and therefore intensively managed bamboo plantations may lead to considerable nutrient loss of especially K [8]. A detailed characterization of biomass feedstock is essential for successful energetic valorization [9], and consists of determination of moisture content, proximate (thermo-chemical reaction) and ultimate (elemental content) analyses, as well as heat of combustion and ash analyses [10]. Primarily the high moisture content, and to a minor extent the K, Si and Cl contents of the ash represent the limiting factors in the utility of bamboo biomass as energy feedstock, while other characteristics are comparable to those of [11]. The moisture content of bamboo culms declines as culms mature and a gradient of moisture content exists between lower and upper culm sections [12], in addition to differences in moisture contents between nodes and internodes. This variation in moisture content along culms, as well as with culm age opens up possibilities to optimize the use of different biomass compartments and age cohorts for different purposes, given adequate knowledge on their properties and on alternative utilization options, such as fiber extraction.

2. Materials and methods

2.1. Sites

Encouraged by an electric utility company, farmers have established small scale bamboo plantations on a few hundred hectares of mostly marginal land in the provinces of Sa Kaew and Nakhon Ratchasima, respectively located in eastern and north-eastern Thailand. One bamboo plantation each was selected for analysis on a low productivity (Sa Kaew) and on a high productivity site (Nakhon Ratchasima). Soil texture in Sa Kaew was sandy loam to sandy clayey loam, while clay dominated in Nakhon Ratchasima. Soil chemical properties differed significantly in terms of contents of organic matter, as well as the most important macronutrients, and the level of this difference reached an order of magnitude with P and K (Table 1). In Sa Kaew, the plantation established in May 2012 consisted of two 136 András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141

adjacent mono-specific blocks of Dendrocalamus membranaceus Munro and of Bambusa beecheyana Munro, both planted on a 6m*2m grid. In Nakhon Ratchasima, the plantation established in December 2011 consisted of two adjacent blocks of different spacings of B. beechayana, planted on a 3m*4m and on a 4m*4m grid, respectively. Both plantations used seedlings raised in nurseries from air layered branches as propagation material. Plantation management in Sa Kaew included irrigation in the initial five months after planting along with mulching, while plantation management in Nakhon Ratchasima included organic fertilization and mulching and regular irrigation once per week using a fixed irrigation system. In the first year, bamboo was intercropped with cassava, taro and eggplant in an agro- scheme.

Table 1: Soil chemical properties at two plantation sites (* indicates significant differences at p”0.05)

Site pH OM [%]* C[%]* N[%]* P[%]* K [mg*kg-1]* Ca [mg*kg-1] Mg [mg*kg-1] Sa Kaew 6.08 1.55 0.89 0.08 14.66 37.74 1271.7 212.08 (±0.34) (±0.18) (±0.09) (±0.02) (±4.84) (±10.64) (±308.2) (±94.51) Nakhon 6.23 2.51 1.41 0.12 160.40 316.40 1511.4 389.18 Ratchasima (±0.43) (±0.52) (±0.23) (±0.02) (±111.95) (±83.80) (±433.3) (±121.44)

2.2. Species

Bambusa beecheyana Munro is a native of southern China and has been introduced to Thailand recently. Culm height ranges from 7-16 meters with internode lengths of 34-41 cm, culm diameter of 6-10 (14) cm and wall thickness of 1.5-2 cm. The sympodial bamboo species with short-necked pachymorph rhizome system forms erect, basally unbranched culms. Shoots emerge in June or July. It is a commercial species in southern China for production with an annual yield of 20-30 Mg*ha-1*a-1 [13]. The fast growing species grows best on heavy, moist soil with good drainage. Dendrocalamus membranaceus Munro is native to northern and parts of central and southern Thailand. Its culms reach 15-18 (25) m in height and 7-10 cm in diameter, with internode lengths of 30-50 cm and branches from the culm base. The culm wall thickness is 0.3-0.6 cm.

2.3. Sampling procedures

Sampling was carried out on three randomly selected plots per block, consisting of three adjacent clumps in each of three adjacent rows, resulting in plots of nine clumps irrespective of spacing. Assessments included clump diameter and vitality, number of fresh shoots, number of live and dead culms by age cohort and the diameter at breast height (DBH) of each culm. The median diameter culm was destructively sampled in each clump and measured for total length, number of internodes, and length of each internode. Subsequently the culm was divided into three (lower, middle, upper) sections of equal length, each measured for green biomass. The central internode along and the node below it was separated from the sections, and measured both for green biomass and after oven drying in the laboratory for dry biomass. Additional destructive sampling of culms outside of research plots was carried out to allow representation of the entire DBH range of culms available in the plantations. All sampling was carried out in February 2014, while in Nakhon Ratchasima the destructive sampling component had to be postponed until December 2014 due to restrictions imposed by the plantation owner. Biomass sub-samples were oven dried at 105±2°C until constant weight. Dry biomass was computed based on the moisture contents of the various sections, weighted by the relative weight contribution of each section to the total culm biomass. The moisture content of culms was determined on the wet basis. Allometric functions of culm biomass based on DBH were developed for both species by evaluating power, logarithmic, polynomial and exponential functions for best fit. Culm biomass per hectare was computed based on data collected in February 2013 for both sites and additionally in December 2013 for Nakhon Ratchasima. The Proximate Analysis of six pooled biomass samples per species was carried out in the laboratory after air drying samples to constancy at approximately 10% moisture content. Higher Heating Values were determined using a bomb calorimeter according to ASTM D 7582-10İ and ASTM D 5865-11a and refer to values at dry basis. András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141 137

2.4. Statistical analyses

Since plantation establishment was farmers’ choice, assessment of various aspects following a fully factorial experimental design was not possible, mandating analyses of various aspects on subsets of the data. Statistical analyses were conducted using SAS 9.2 [14] DQGZHUHFRQVLGHUHGVLJQLILFDQWDWS”$QDO\VHVLQFOXGHGWHVWLQJ for differences in soil chemical properties (PROC ANOVA), as well as the species-specific relationship between culm DBH and culm biomass (PROC NLIN).Total culm biomass per hectare was compared between sites (PROC ANOVA), between species in Sa Kaew only (PROC ANOVA) and between different spacings of B. beecheyana in Nakhon Ratchasima only (PROC MIXED with repeated measures over time). Culm moisture content was tested on the Sa Kaew subset of the data between species, sections and node/internode (PROC GLM), while for Nakhon Ratchasima the comparison was performed considering the factors section, node and age cohort (PROC GLM). Culm biomass constituents resulting from the Proximate Analysis on wet base as well as Higher Heating Values were compared between species (PROC ANOVA).

3. Results and Discussion

3.1. Biomass functions and yield

Allometric functions using a power equation described culm biomass based on DBH best (Fig. 1; r2 = 91% for B. beecheayana, Equation 1;r2 = 95% for D. membranaceus, Equation 2) and resulted in fits comparable to those of allometric functions of culm biomass based on DBH for B. beecheyana in China [15].

Above-ground culm biomass (B. beecheyana) = DBH1.830451265 * 125.0933868 (1)

Above-ground culm biomass (D. membranaceus) = DBH1.968074427 * 123.7175124 (2)

Fig. 1. Allometric functions of culm DBH to culm biomass of two bamboo species. 138 András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141

Fig. 2. (a) Culm biomass of B. beecheyana in two plantations; (b) Culm biomass development of age cohorts of Bambusa beecheyana in Nakhon Ratchasima.

Significant differences in total culms biomass between sites exceeded an order of magnitude in February 2013, with 125.6 t*ha-1 in Nakhon Ratchasima vs. 8.4 t*ha-1 in Sa Kaew (Fig. 2a), even though this difference was confounded by differences in plantation age (14 vs. 9 months, respectively). Culm biomass further increased to 245 t*ha-1 in Nakhon Ratchasima by the time the plantation was harvested at two years of age in December 2013 (data not presented). While no specific references are available on biomass yield of B. beecheyana or D. membranaceus, biomass yield of various bamboo species used in plantations ranges between 4–50 t*ha-1*a-1 , while the total above- ground biomass may reach 25–307 t*ha-1 [6, 7]. Considering that 75–90% of the aboveground biomass is concentrated in culms [6, 16-18], our records for culm biomass of Nakhon Ratchasima are placed among the highest reported in the literature, even without considering the fact that the plantation had not yet reached maximum stocking at two years after planting. As opposed to Nakhon Ratchasima, Sa Kaew placed among the lowest stockings reported for bamboo culm biomass [7, 19]. Spacing showed a significant effect over time on B. beecheyana culm biomass in Nakhon Ratchasima. The tighter spacing of 4m*3m resulted in considerably higher culm biomass until February 2013 (159.2 t*ha-1) than the wider spacing of 4m*4m (92.0 t*ha-1)(b). This possibly indicates that the growing space had not yet been occupied after two years of plantation establishment, even though higher biomass at closer spacing was also reported for Dendrocalamus strictus [20]. No significant differences in culm biomass were observed between species in Sa Kaew (data not presented).

3.2. Moisture content of bamboo culms

The moisture content of culms decreased with increasing height along culms, but differences between species in the gradient of this decline were marginally not significant (Table 2). While moisture content from the bottom to the top sections of B. beecheyana declined from 59.0% to 47.2%, this gradient lay between 55.9% and 38.5% with D. membranaceus (Fig. 3a). A comparable moisture content of 53.7%, assessed irrespective of culm section, was recorded for first-year culms of B. beecheyana in China [15], while a moisture content of 43%, irrespective of culm age was reported from Laos [21]. In Nakhon Ratchasima, the moisture content of B. beecheyana significantly declined from 71.2% with current- year culms to 59.4% with one-year old culms (Fig. 3b), matching a trend reported for the species from China, where moisture content declined from 53.7% with one-year old culms to 39.5% for three-year-old culms and further to András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141 139

38.8% for culms older than three years [15]. The chemical composition and mechanical properties of bamboo culms change with age [22], and in this process the moisture content of culms declines with the age of the bamboo culm, but also with increasing height along the bamboo culm [23].

Table 2. Type III test of fixed effects on arcsin transformed moisture content of culm biomass compartments of two bamboo species.

Source DF Type III SS Mean Square F Value Pr >F species 1 0.14066709 0.14066709 42.98 <.0001 section 2 0.47034952 0.23517476 71.86 <.0001 species*section 2 0.01914449 0.00957224 2.92 0.0575 node 1 0.00810215 0.00810215 2.48 0.1183 species*node 1 0.00778954 0.00778954 2.38 0.1255 section*node 2 0.00122215 0.00061108 0.19 0.8299 species*section*node 2 0.00416763 0.00208381 0.64 0.5308

Lower moisture content of bamboo culms is reported for nodes as compared to internodes [24], but in our case this difference was only significant when analyzed on a partial dataset containing D. membranaceus only (data not presented). Initially high moisture content of bamboo culms affects their utility as energy feedstock, since the energy required to evaporate moisture from the culms will considerably reduce the net calorific yield [3]. An increase in moisture content of 42% on wet base reduces the LHV of biomass by 50% [25], rendering young bamboo culms and lower culm sections with high moisture contents •PRLVWXUHFRQWHQW unattractive for energetic utilization [3].

3.3. Feedstock characteristics of bamboo culms

The Proximate Analysis on wet basis did not reveal any significant differences between the proportions of fractions, however Higher Heating Values of D. membranaceus were significantly higher than of B. beecheyana (Table 3), and also higher than the 18.81 MJ*kg-1 reported for the species from Laos [21], placing our samples in the range of calorific contents of wood [9].

Fig. 3. (a) Moisture contents of culm sections of two bamboo species; (b) Moisture content of two age cohorts of B. beecheyana culms in Nakhon Ratchasima. 140 András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141

Low ash content is a favorable determinant of feedstock quality.Ash contents of herbaceous biomass can be as high as 5–10%, while the ash content of wood is <1%. Bamboo ash contents are variable (0.8–-5% on dry basis [11]), but especially the relatively high K content of ash in lower culm sections is reported to lead to technical problems during gasification [9]. The ash contents of our samples of both species were comparable to those of wood and much lower than reported for D. membranaceus (4.99%) [21], even though comparable to ash contents of Phyllostachys pubescens [22]. The fixed carbon content of our samples was slightly higher than reported elsewhere with 14.9–15.6% [9].

Table 3. Proximate Analysis of bamboo biomass on wet base (MC – moisture content, VM – volatile matter, FC – fixed carbon) and Higher Heating Values on dry base.

Species MC [%] VM [%] Ash [%] FC [%] HHV [MJ*kg-1] B. beecheyana 6.48 74.08 1.92 17.52 19.347 D. membranaceus 6.54 74.17 1.97 17.33 19.513

4. Conclusions

Bamboo plantations in eastern Thailand show vast differences in biomass yield as a result of site productivity and plantation management. Intensive management on high productivity sites can lead to exceptionally high biomass yield, while plantations established on marginal sites without proper plantation management will result in very low yields. Primarily lower culm sections and internodes and young culms show very high moisture contents, rendering their energetic utilization ineffective. While biomass yield does not differ between species, the moisture content of D. membranaceus is significantly lower and the calorific value significantly higher than that of B. beecheyana, indicating that D. membranaceus may be a more attractive option as energy feedstock on suitable sites. Alternative utilization options for lower culm sections and internodes as well as for young culms should be sought.

Acknowledgements

The authors would like to thank Dr. Bunvong Thaiutsa and Dr. Viktor Bruckman for their invaluable support throughout the work, as well as to Dr. Roongreang Poolsiri for conducting soil analyses.

References

[1] V. Kleinhenz, D.J. Midmore, Aspects of bamboo agronomy, Advances in Agronomy, 74 (2001) 99-145. [2] J.M.O. Scurlock, D.C. Dayton, B. Hames, Bamboo: an overlooked biomass resource?, Biomass and Bioenergy, 19 (2000) 229-244. [3] G. Kerlero de Rosbo, J. de Bussy, Open ideas - electrical valorization of bamboo in Africa, in, Enea Consulting, Paris, 2012, pp. 58. [4] A.N. Singh, J.S. Singh, Biomass, net primary production and impact of bamboo plantation on soil redevelopment in a dry tropical region, and Management, 119 (1999) 195-207. [5] T.T. Veblen, F.M. Schlegel, B.E. R, Dry-Matter Production of Two Species of Bamboo (Chusquea Culeou and C. Tenuiflora) in South- Central Chile, Journal of Ecology, 68 (1980) 397-404. [6] P. Shanmughavel, R.S. Peddappaiah, T. Muthukumar, Biomass production in an age series of Bambusa bambos plantations, Biomass and Bioenergy, 20 (2001) 113-117. [7] E. Uchimura, Bamboo cultivation, in: Bamboo Research in Asia. Proceedings of the workshop held in Singapore, 1980, pp. 28-30. [8] B. Kumar, G. Rajesh, K. Sudheesh, Aboveground biomass production and nutrient uptake of thorny bamboo [Bambusa bambos (L.) Voss] in the homegardens of Thrissur, Kerala, J. Trop. Agric, 43 (2005) 51-56. [9] A. Ganesh, Bamboo Characterization For Thermochemical Conversion And Feasibility Study Of Bamboo Based Gasification And Making, in, Energy Systems of Indian Institute of Technology, Mumbai, 2003, pp. 18. [10] R. Stahl, E. Heinrich, H.J. Gehrmann, S. Vodegel, M. Koch, Definition of a standard biomass, in, Forschungszentrum Karlsruhe, Karlsruhe, 2004, pp. 14. [11] C.M.D. Montaño, J.R. Pels, L.E. Fryda, R.W.R. Zwart, Evaluation of torrefied bamboo for sustainable bioenergy production, in: J. Gielis (Ed.) 9th World Bamboo Congress, World Bamboo Organization, Antwerpen, 2012, pp. 809-818. [12] B. Engler, S. Schoenherr, Z. Zhong, G. Becker, Suitability of Bamboo as an Energy Resource: Analysis of Bamboo Combustion Values Dependent on the Culm’s Age, International Journal of Forest Engineering, 23 (2012). [13] W. Hsiung, Bamboo in China: new prospects for an ancient resource, , 39 (1987). [14] SAS Institute Inc., SAS 9.2, in, SAS Institute Inc., Cary, NC, USA., 2002-2008. András Darabant et al. / Energy Procedia 59 ( 2014) 134 – 141 141

[15] S. Wei, Z. Yueguo, C. Senjie, A study of relationship among aboveground factors of Dendrocalamopsis beecheyana var. pubescens (P.F.Li) Keng f., World Bamboo and Rattan, 7 (2009) 8-11. [16] A. Castañeda-Mendoza, J. Vargas-Hernández, A. Gómez-Guerrero, J.I. Valdez-Hernández, H. Vaquera-Huerta, Acumulación de carbono en la biomasa aérea de una plantación de Bambusa oldhamii, Agrociencia, 39 (2005) 107-116. [17] L. Christanty, D. Mailly, J.P. Kimmins, “Without bamboo, the land dies”: Biomass, litterfall, and soil organic matter dynamics of a Javanese bamboo talun-kebun system, Forest Ecology and Management, 87 (1996) 75-88. [18] N. Riaño, X. Londoño, Y. López, J. Gómez, growth and biomass distribution on Guadua angustifolia Kunth in relation to ageing in the Valle del Cauca–Colombia, The Journal of the American Bamboo Society, 1 (2002) 43-51. [19] L. Yiming, L. Peng, W. Wanzhang, Studies on dynamics of carbon and nitrogen elements in Dendrocalamopsis oldhami forest, J Bamboo Res, 17 (1998) 25-30. [20] V. Patil, S. Patil, I. Rao, K. Gnanaharan, C. Sastry, Performance of bamboo under varying spacing and fertility levels, : Current research, (1988) 107-111. [21] B.M.E. Dannenmann, C. Choocharoen, W. Spreer, M. Nagle, H. Leis, A. Neef, J. Mueller, The Potential of Bamboo as a Source of Renewable Energy in Northern Laos, in: Tropentag - Conference on International Agricultural Research for Development, Göttingen, 2007, pp. 5. [22] X.B. Li, T.F. Shupe, G.F. Peter, C.Y. Hse, T.L. Eberhardt, Chemical changes with maturation of the bamboo Phyllostachys pubescens, Journal of Tropical Forest Science, 19 (2007) 6-12. [23] S. Wang, X. Pu, Y. Ding, X. Wan, Aboveground biomass and bamboo shoot nutrients of high altitude bamboos (Fargesia yunnanensis Hsueh et Yi) from different sites in Yunnan province, Journal of Tropical Agriculture, 47 (2009) 48-53. [24] C.N. Krishnankutty, Estimation of moisture content in bamboo for deriving the weight and price conversion factors, in, Kerala Forest Research Institute, Peechi, 2007. [25] V. Francescato, E. Antonini, C. Metschina, C. Schnedl, N. Kranjc, K. Koscik, P. Gradziuk, G. Nocentini, S. Stranieri, Wood fuels handbook, in, Legnaro, 2008, pp. 79.