Characterization of residues from plant biomass . . . CHARACTERIZATION OF RESIDUES FROM PLANT BIOMASS FOR USE IN ENERGY GENERATION

The use of plant residues for energy purposes is already a reality, yet in order to ensure suitability and recommend a given material as being a good energy generator, it is necessary to characterize the material through chemical analysis and determine its calorifi c value. This research aimed to analyze different residues from plant biomass, characterizing them as potential sources for energy production. For the accomplishment of this study, the following residues were used: wood processing residue (sawdust and planer shavings); coffee bean parchment and coffee plant stem; bean stem and pod; soybean stem and pod; rice husk; corn leaf, stem, straw and cob; and sugar cane straw and bagasse. For residue characterization the following analyses were done: chemical analysis, immediate chemical analysis, calorifi c value and elemental analysis. All procedures were conducted at the Laboratory of Forest Biomass Energy of the Federal University of Lavras. In general, all residues showed potential for energetic use. Rice husk was found to have higher lignin content, which is an interesting attribute as far as energy production is concerned. Its high ash content, however, led to a reduction in calorifi c value and fi xed carbon. The remaining residues were found to have similar energetic characteristics, with corn cob showing greater calorifi c value, followed by coffee plant stem, both also containing higher levels of carbon and fi xed carbon. A high correlation was found of higher calorifi c value with volatile materials, carbon and hydrogen contents.


INTRODUCTION
Brazil is a leading forest and agricultural producer for a variety of reasons, including area availability for cultivation, possibility of introducing assorted cultures, geographic position favoring intense solar radiation throughout the year, tropical climate, besides its extremely rich biodiversity and advanced technology, all of which allow Brazil to hold a privileged position in the fi eld of agrarian sciences.

Ramos e Paula , L. E. de et al.
As a result, Brazil has become one of the world's largest producers of wood and agricultural products and is known by many as 'the barn of the world'.In spite of that, intensive production generates large amounts of residue and that can cause serious environmental problems.
According to Vale and Gentil (2008), residue can be defi ned as any discarded material from activities relating to a production process, being a potential hazard to the environment and, consequently, to society.By contrast, residues can go from being a hazardous risk to becoming generators of profi t if they are turned into raw materials for other processes, reducing both the price and the demand for the main product.
Wood is an important input whose value has increased in past decades, yet it is not being used fully (RIBEIRO; ANDRADE, 2005).According to Lima and Silva (2005), every process involving wood transformation generates residues but only 40% to 60% of the total volume of logs harvested is actually used.
The processing of agricultural crops also generates large amounts of residue.According to Instituto Brasileiro de Geografia e Estatística -IBGE (2006), in 2006 alone Brazil produced around 11 million tons of bulk, unprocessed rice which in turn generated 2.6 million tons of rice husk.Rocha et al. (2006) argue that the ratio of processed coffee beans to residue can reach 50%.According to Macedo (2001), each ton of sugar cane (culm) produces 140 kg (dry matter) of bagasse, 150 kg of sugar and 140 kg (dry matter) of stubble, lost from fi eld burns.According to Chagas et al. (2007), residue from bean crops includes straw and stalk, making up 60% (in weight) of the crop.Soybean is produced on a larger scale in Brazil and thus generates an equally larger amount of residue.According to Bose and Martins Filho (1984), soybean stubble corresponds to as much as 120% to 150% of soybean weight.
Lignocellulosic residue can be reused as raw material in a different process from the original process, for instance, it can be used energetically for generation of heat or electricity in generator groups or thermoelectric plants (QUIRINO, 2003).
According to Vale and Gentil (2008), biomass residues are solid fuels that can be used directly as they are, under moisture controlled conditions, and be transformed by mechanical processes into small particles such as wood chips or sawdust, or else pressed into briquettes.
The use of residues from plant biomass for energy purposes is already a reality.However, in order to ensure suitability and recommend a given material as being a good energy generator it is necessary to characterize the material through chemical analysis and determine its calorifi c value.
Thus the objective of this work is to analyze different lignocellulosic residues, characterizing them as potential sources for energy production.

MATERIAL AND METHODS
In order to conduct this study, lignocellulosic residues were obtained in the municipality of Lavras/MG, including wood processing residues (planer shavings and sawdust), coffee bean parchment and coffee plant stem, bean stem and pod, soybean stem and pod, rice husk, corn leaf, stem, straw and cob, plus sugar cane straw and bagasse.The residues were submitted to analysis at the Laboratory of Forest Biomass Energy of the Federal University of Lavras.
The material was ground by a Wiley blade apparatus, using the fraction that passed through a 40mesh screen but was retained in a 60-mesh screen.The residues were stored under controlled temperature (20 ± 3)°C and relative humidity (65 ± 2)% for moisture homogenization.
An elemental analysis was performed to start with for determination of carbon, hydrogen, nitrogen, sulfur and, by difference, oxygen contents of the material.For that, approximately 2 mg of residue was weighed in a tin sample holder.The aggregate (residue + sample holder) was then placed in the carousel of an Elementar® elemental analyzer.The analysis was performed on one sample at a time.The gases necessary for the operation included helium, used as carrier gas, and oxygen, used as ignition gas.The temperature of the combustion tube, which is located inside the equipment, at the moment the sample was fed into the carousel was 1,150 °C.After combustion, gases were stripped to the reduction tube and then to the detection column.Elements were determined by a thermal conductivity detector whereby each element interacted and had its specifi c peak.The computer coupled to the equipment made the relevant calculations and derived values expressed as percentage.
The determination of silicon content in rice husk was done using the molybdenum colorimetric method, according to Furlani and Gallo (1978).Characterization of residues from plant biomass ... The chemical analysis followed Standards M3/89, M11/77 and M70/71 of Associação Brasileira Técnica de Celulose e Papel -ABTCP (1974), for determination of extractives content, ash content and lignin content respectively.The holocellulose content (H) was done on the basis of Equation 1, using percentage in relation to dry matter of components.
The immediate chemical analysis was based on Standard 8112 of Associação Brasileira de Normas Técnicas - ABNT (1983), to determine volatile materials and ash content.Fixed carbon content (CF) was derived from the difference (Equation 2).
The analysis for determination of higher calorifi c value was performed in a Parr® calorimeter, following standard 8633, ABNT (1984).Determination of lower calorifi c value (PCI) was based on Equation 3.
An analysis of correlation was performed to verify associations between the characteristics being analyzed.

RESULTS AND DISCUSSION
Table 1 provides elemental analysis results for residues.
In analyzing Table 1 it is noted that the sulfur content was low for all residues, ranging from 0.1% to 0.4%.The presence of sulfur in fuels is undesirable due to erosion problems and release of SO 2 gas after combustion.
Table 2 provides reference values for the elemental analysis of some residues.Rice husk had higher oxygen content and lower carbon and hydrogen contents, being in agreement with Jenkins (1990).
Elemental analysis results for coffee plant stem and coffee bean parchment were similar to results for wood residues, except nitrogen content, found higher in coffee residues.Lower nitrogen content in wood residues resulted in a high C/N ratio.Analysis values for coffee bean parchment are in agreement with Brum et al. (2008).
Sugar cane bagasse and straw, as well as wood residues, showed similar results to those by Seye et al. (2003).
The elemental analysis of bean straw (stem + pod), as conducted by Oliveira (2009), provided similar results to results found in this study.
The percentages of corn stem and leaf elements are in agreement with Raveendran et al. (1995), who also studied corn plant (stem + leaf).Results for corn cob and straw were similar to the literature result in Table 2.
A higher C/N ratio was observed for sawdust, as shown in Table 1, followed by planer shavings, both of which being wood residues.According to Munalula and Meincken (2009) the higher nitrogen content is related to environmental impacts and air pollution as a function of the formation of toxic nitrogen oxides and nitric acid.
Table 3 provides average values of total extractives, lignin, ash and holocellulose contents found in the residues being assessed.Rice husk was found to have lower extractives content and higher ash content, in comparison with other residues.The high ash content is associated with a large amount of silica (SiO 2 ); the analysis detected 8.32% of silica.The lignin content obtained by Reyes et al. (1997) for rice husk was similar to fi ndings in this research, unlike ash content.Additionally, rice husk showed a low holocellulose content, in association with its high ash content.
The reference values for these analyses are provided in Table 4.The result found for ash content in coffee plant stem is close to the result found by Pereira (2008), and bean parchment values were found similar to those in literature.
The extractives content found in cane bagasse differed from the fi nding of Pitarelo (2007), possibly because in his experiment the sugar cane originating the bagasse was burned and washed prior to being ground.The lignin content was similar to the value found by Marabezi et al. (2009).
The chemical analysis of bean residues (Table 3) provided differing component percentages for stem and pod; the broadest difference found was in extractives and lignin contents.From Table 4 it is noted that prior studies consider the aggregate straw (stem + pod) for analysis.
As for the chemical analysis of planer shavings and sawdust, no signifi cant difference was found, and values are in agreement with results found by Mori et al. (2002) and Trugilho et al. (2003).
The chemical analysis of corn stem and leaf provided similar results, the same occurring for straw and cob, except the lignin content for the latter two.
The chemical analysis of soybean residues, stem and pod, provided differing values.The ash content of pod is in agreement with Silva et al. (2008).The values found for soybean were not very dissimilar to bean values, though that was expected as both belong to the pulse family.Characterization of residues from plant biomass ...  Extr., Lign., Ash, Hol.= Extractives, lignin, ash and holocellulose contents.

Ramos e Paula , L. E. de et al.
Results of immediate analysis and higher calorifi c value of residues are provided in Table 5.Table 6 contains reference values for these analyses.The immediate analysis and calorifi c value of rice husk are in agreement with Diniz et al. (2004), Jenkins (1990) and Souza et al. (2005).Characterization of residues from plant biomass ... The immediate analysis of coffee plant stem and coffee bean parchment provided similar results to wood results.
The values determined in this study for sugar cane bagasse are in agreement with Jenkins (1990) and Seye et al. (2003), except volatile materials content.
Sugar cane straw was found to have similar fi xed carbon value and calorifi c value to bagasse.Seye et al. (2003) observed lower fi xed carbon contents than this study.
The immediate analysis results for bean residues, stem and pod, provided similar values of volatile materials and fi xed carbon.The calorifi c value of the stem was slightly higher, possibly due to its higher lignin content.
The calorifi c value of wood residues is within the range determined by Brito (1993) and Brito and Barrichello (1978), as provided in Table 6.
The immediate analysis of corn stem and leaf provided similar results of volatile materials to values found by Raveendran et al. (1995), who studied corn plant as an aggregate, in other words, stem plus leaf.The volatile materials content in corn cob was similar to literature results, unlike ash content and calorifi c value.The immediate analysis and calorifi c value of corn cob are in agreement with Jenkins (1990).Quirino et al. (2005) observed an inferior calorifi c value for cob straw.
Results for soybean residues proved similar to results for bean residues.Silva et al. ( 2008) observed a higher fi xed carbon value for soybean pod than this study did.
The lower calorifi c value is within the expected range for lignocellulosic materials.Corn cob provided the highest value and rice husk, the lowest.
In observing results of residue analyses, dissimilarity is observed, in most cases, in relation to literature results.According to Brum (2007), it should be taken into account that the chemical constitution of materials is conditional on several factors, including soil composition, climate, harvest season, disease and weed presence, planting method, among others, all of which can cause even plants of the same species to differ in composition.Differences may also occur in the way the material is sampled and in the analysis methodology.Additionally, by their very nature, residues are remainders of other processes and thus may well be contaminated or stored unsuitably.
The most signifi cant coeffi cients of correlation for the characteristics being assessed in residues are provided in Table 7.A high positive correlation was found of higher calorifi c value with volatile materials, carbon and hydrogen contents, and a high negative correlation was found of higher calorifi c value with oxygen content.
A high correlation was expected of lignin, extractives and fi xed carbon contents with higher calorifi c value, despite results.This is possibly due to the great variability in residues used in this study.Holocellulose content, however, was found to have a moderate, positive correlation with higher calorifi c value.

CONCLUSIONS
In general, all residues showed good potential for use in energy generation.In the elemental analysis, the carbon content varied within the range expected for lignocellulosic materials, except for rice husk.This latter residue showed higher lignin content, but also high ash and oxygen contents and low carbon content, Ramos e Paula , L. E. de et al.
which probably led to a reduction in calorifi c value and fi xed carbon content.The remaining residues showed similar characteristics, with special mention of corn cob which showed greater calorifi c value, followed by coffee plant stem.All residues showed high levels of volatile materials and low levels of fi xed carbon.Higher calorifi c value had a high positive correlation with volatile materials, carbon and hydrogen contents, and a negative correlation with oxygen content.An alternative way of improving the energetic properties of residues, reducing the level of volatile materials and increasing the calorifi c value, is applying torrefaction to residues or blending them with fi ne charcoal particles.

Table 1 -
Elemental analysis of residues.
Ramos e Paula , L. E. de et al.

Table 2 -
Reference values for elemental analysis.Valores de referência para a análise elementar.

Table 4 -
Reference values for chemical analysis.

Table 6 -
Reference values for immediate analysis and calorifi c value.Valores de referência para análise imediata e poder calorífi co.

Table 7 -
Main coeffi cients of correlation between assessed characteristics.