Journal of Combustion

Volume 2016 (2016), Article ID 9534063, 10 pages

http://dx.doi.org/10.1155/2016/9534063

## Pyrolysis Kinetic Modelling of Wheat Straw from the Pannonian Region

Faculty of Technical Sciences, University of Novi Sad, Trg Dositeja Obradovića 6, 21125 Novi Sad, Serbia

Received 30 November 2015; Revised 17 February 2016; Accepted 21 February 2016

Academic Editor: Hong G. Im

Copyright © 2016 Ivan Pešenjanski et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

#### Abstract

The pyrolysis/devolatilization is a basic step of thermochemical processes and requires fundamental characterization. In this paper, the kinetic model of pyrolysis is specified as a one-step global reaction. This type of reaction is used to describe the thermal degradation of wheat straw samples by measuring rates of mass loss of solid matter at a linear increase in temperature. The mentioned experiments were carried out using a derivatograph in an open-air environment. The influence of different factors was investigated, such as particle size, humidity levels, and the heating rate in the kinetics of devolatilization. As the measured values of mass loss and temperature functions transform in Arrhenius coordinates, the results are shown in the form of saddle curves. Such characteristics cannot be approximated with one equation in the form of Arrhenius law. For use in numerical applications, transformed functions can be approximated by linear regression for three separate intervals. Analysis of measurement resulting in granulation and moisture content variations shows that these factors have no significant influence. Tests of heating rate variations confirm the significance of this impact, especially in warmer regions. The influence of this factor should be more precisely investigated as a general variable, which should be the topic of further experiments.

#### 1. Introduction

The widespread availability of biomass, which is renewable and has no impact on global warming, has motivated extensive research in the past decade regarding the industrial development of thermochemical conversion plants. Pyrolysis, an important step in all thermochemical processes, is a technology that has the potential to reduce dependence on fossil resources by providing alternative fuels or useful chemicals. Also, pyrolysis will be essential in new technologies such as low temperature carbonization and chemical production from solid fuels. Understanding pyrolysis kinetics is important for the effective design and operation of conversion units.

Extensive reviews based on a number of papers done by Antal Jr. and Varhegyi [1] and Di Blasi [2] show that the pyrolysis of biomass involves a complex series of reactions. Consequently, changes in the experimental heating conditions or sample composition and preparation may affect not only the rate of reaction but also the actual course of the reactions. Most of these observations which are made from reported publications have been mainly focused on woody material and, in a few cases, on agricultural residues. Cereal straws are the predominant biomass in Pannonian agricultural areas, where wood waste is much less available.

Early researches of wheat straw combustion [3–5] have pointed out the fact that gaseous products comprise 70–85% of total pyrolytic reactions products. Temperatures at which they emerge are also notably low (200–450°C) [4, 5]. The most intensive reactions are noticed at around 300°C, so it is very likely that the largest part of gaseous and solid products of reaction under these conditions do not react with considerable acceleration of oxidation at first contact with the oxidant, which could be characterized as ignition. Reference [6] has confirmed the conclusion of Shafizadeh [3] that the thermal decomposition of cellulose over a wide range of mass loss is essentially the same in both air and nitrogen. At the beginning of the decomposition oxygen interacts only with surface cellulose, resulting in only ~3% of mass loss. This gives a reliable basis for the assumption that pyrolysis prevails in regions where chemical kinetics restricts the overall rate of the combustion process.

However, variability of ash composition and ash content in straw (1–15%) seems to have a strong influence on both the pyrolytic characteristics and the product distribution (catalytic effect of the ashes) [2, 7, 8]. This means that results obtained from the analysis of single components cannot be directly applied to wheat straw because of chemical and physical alterations introduced in the separation procedure, the impossibility to reproduce component interactions, and the presence of ash [2, 9]. Ash constituents, especially potassium (K), sodium (Na), and calcium (Ca), act as catalysts for the decomposition process. As chlorine (Cl) and potassium in biomass are water soluble, they can, for the most part, be removed through leaching, thus mitigating their impact on the high-temperature conversion devices. In reality, water or mild acid washing also introduce significant modifications in biomass decomposition characteristics [9, 10].

Quantitative differences between characteristics are caused by several factors which, in addition to the biomass species, include, even for the same sample, the geographical origin, the age, or the specific part of the plant [10, 11]. All these observations justify the need for a detailed investigation of every specified case.

The aims of this research were to investigate the pyrolytic behavior of wheat straw from Vojvodina, to check the influence of experimental conditions on the kinetic parameters, and to provide kinetic information for the evaluation of the processing of these materials.

#### 2. Mathematical Model of Pyrolytic Processes

In reality, thermal decomposition in wheat straw involves numerous physical impacts with complex chemical reactions and ends up with a large number of intermediates and end products. On the other hand, pyrolytic models are created on the basis of visible and measurable reflections. However, our understanding, through experimentation, of the basic processes occurring during the combustion/pyrolytic process is very limited and still questionable [12, 13].

A further general approach is considered in this paper: virgin biomass as the raw material and with gas/volatiles and solid char/coke yield as the end products (one-step global model [13]).

Diffusion effects in TGA tests are particularly important as the pyrolysis process involves the transport of reactant and heat from the external bulk gas phase to the internal particle surface, where the chemical reactions take place. Hence pyrolytic models can be subdivided, depending on intensity of the diffusion effect influence on microparticle models and macroparticle models [14].

Pyrolysis of microparticles includes thermal decomposition of the virgin matter in samples of sufficiently small dimensions, so that the effects of diffusion become negligible, and the intensity of pyrolysis is confined and controlled with kinetics (kinetic region of reaction). This fact enables the phenomenology of kinetics to be investigated on microparticles.

Critical particle dimensions that are set for kinetic control generally range from about 100 to 1000 *μ*m [14] and it has been noticed that they grow with the temperature of pyrolysis [11, 14]. Particles larger than the critical one are characterized by greater diffusional effects which can intensively influence the development of pyrolysis through internal and external temperature gradients, heat inertia, and temperature variations due to the occurrence of exothermic and endothermic reactions, which should be avoided in this work. References [14–16] give a more detailed model that is able to describe the kinetics, the heat transfer, and the mass fluxes that take place during biomass combustion/pyrolysis of samples with macroparticles (more near to real technical objects).

##### 2.1. Kinetic Modelling

Abstracting the concrete content of all processes with the uniform pyrolytic reaction (unified lumped approach, one-step global model), kinetics of the total reaction can be expressed through Arrhenius monomolecular equation of the first order. As the volatiles in this simplified model represent the “subject of motion,” it is natural that the law of change in mass, hypothesis, is set in such a way so that the rate of volatile extraction is proportional to the quantity of remaining volatiles in the fuel. Based on that, the mathematical form of pyrolysis reaction rate can be expressed by the following equation:

In the laboratory conditions measuring the remaining mass of the solid residue (char) () can be done more easily. As needed, (1) may be rearranged to

In the presented form (2) has the dimension of s^{−1}, so that the preexponential multiplier has the dimension of kg/(m^{3}s) (mass flux per volume). The rate of the heterogeneous reaction (sometimes called char reactivity) is expressed per unit of area of reacting material, so for the purpose of comparison with the mass transfer coefficient, as generally is with the use of the complex resistance method, this value is necessary to correct with the multiplier of the ratio of volume and area of combustible matter.

With finding the logarithm of (2) and introduction of Arrhenius variables of temperature and mass loss rate retrospectively, (2) in coordinates of Arrhenius for = const. gets first-order polynomial formwhich is convenient to fit with experimental data. In (4) polynomial coefficients areIt should be mentioned that the differential equation (2), despite its simple form, does not have an analytical solution.

For the reason that organic substances are preexponential multiplier and activation energy dependent on temperature, that is, and , therefore based on that, they have to be calculated in nonlinear form, for example, with degree polynomial:In real coordinates, (6) returns to the following:Both of the models (4) and (6) may be used for the standalone pyrolysis applications as well as a submodel in a CFD code with more complex models [17, 18]. A computer program using C++ is required to perform written calculations.

##### 2.2. Influence of Mass Diffusion through the Gas Phase

In 1924, Nusselt [19] proved through theoretical analysis that smaller particles “burn faster” due to decreased diffusion resistance. The question is how small should the investigated particle size be in which diffusion will occur much faster than the kinetics of pyrolysis?

Knowing that the Nusselt number is a dimensionless temperature gradient integrated over the surface and that the Sherwood number is a dimensionless concentration gradient integrated over the surface, correlations for binary mass transfer coefficients at low mass transfer rates can be obtained the same way and directly from their heat transfer analogs simply by a change of notation (Nu by Sh and Pr by Sc) [20].

Bird et al. (after Froessling, 1938) [20] gave a correlation for predicting the heat transfer to or from a sphere with constant surface temperature by the following empiricism:(which also takes into account current adjustments by the constants in the second summands at 0,60). Equation (8) is also valid for small net heat/mass transfer rates. If there are stagnant conditions (Re = 0), (8) develops to “Newton’s law of cooling”:(this well-known result also provides the limiting value of Nu for heat transfer from spheres at low Reynolds and Grashof numbers). While the Sherwood number isbinary diffusion coefficient can be determined from the Schmidt numberSolving (9), (10), and (11), the mean mass transfer coefficient is obtainedThe Schmidt number for gases does not vary greatly with temperature; it does, however, vary significantly with mixture ratio. The Schmidt number is for a dilute binary mixture of volatile gases (CO_{2}, CO, H_{2}, and CH_{4}) typically present in air [21]; Sc = 0,6–1,0 ≈ 0,8. The viscosity of a volatile mixture is not simply a function of the composition; it varies with temperature. For dilute binary mixtures of typical flue gases with air in a temperature range of 200°C–400°C the viscosity of a volatile-air mixture is , .

When these values are used in (12) we get the hyperbolic dependence of mass transfer coefficient from particle size. For typical wall thickness of the straw mm, the mean mass transfer coefficient is found to be

The requirement that during the experiment mass diffusion does not inhibit the kinetics of pyrolysis is that mass diffusion must take place much faster than the sample mass change occurring through chemical kinetics:where the correction factor is the ratio of particles volume and surface:

#### 3. Methodology: Setup and Procedure

##### 3.1. Samples

Wheat straw samples were collected for research from an agricultural area near Novi Sad (Serbia) in two-year intervals. The wheat type is “Novosadska rana 5” which is commonly sown in Vojvodina (Serbia). The wheat was treated with standard agrotechnical measures. A simple random sampling plan was used in order to get samples immediately after harvest from the collecting warehouse during the dry weather, not exposing the straws to bad climate influences. The collected stalks were air-dried naturally in a dry store room in ambient condition. Its proximate and ultimate analysis are given in Table 1, and its ash melting behavior is given in Table 2.