Isoconversional nonisothermal kinetic analysis of municipal solid waste, refuse‐derived fuel, and coal

The secure supply of energy sources is a key indicator for the sustainable development of societies. Normally, various types of fossil energy reserves (natural gas, coal, and gasoline) are the major sources of conventional energy demand. The use of fossil fuel for energy production is posing many problems including global warming, reduction in fossil energy sources, and other related environmental issues.1,2 On the other hand, the population rise, economics, and rapid industrialization have also posed energy and environmental concerns in various countries.3 It is estimated that about 1.2 billion people globally still do not have even basic energy infrastructure.4 At present, Pakistan is a developing country and facing serious energy crisis with total energy demand and supply of 93.91, 70.94 Mtoe, respectively.5 Considering, the huge energy supply and demand gap, it is highly recommended to explore sustainable energy alternatives to address the issue of current energy crisis. Coal is one of the abundantly and evenly distributed fossil fuel (~65% of the total fossil fuels) around the globe, unlike oil and gas reserves which are only present in a fewer regions of the world.6 At present, coal is the major source of energy and continues to dominate as a main contributor for energy Received: 26 March 2020 | Revised: 23 May 2020 | Accepted: 12 June 2020 DOI: 10.1002/ese3.778


| INTRODUCTION
The secure supply of energy sources is a key indicator for the sustainable development of societies. Normally, various types of fossil energy reserves (natural gas, coal, and gasoline) are the major sources of conventional energy demand. The use of fossil fuel for energy production is posing many problems including global warming, reduction in fossil energy sources, and other related environmental issues. 1,2 On the other hand, the population rise, economics, and rapid industrialization have also posed energy and environmental concerns in various countries. 3 It is estimated that about 1.2 billion people globally still do not have even basic energy infrastructure. 4 At present, Pakistan is a developing country and facing serious energy crisis with total energy demand and supply of 93.91, 70.94 Mtoe, respectively. 5 Considering, the huge energy supply and demand gap, it is highly recommended to explore sustainable energy alternatives to address the issue of current energy crisis.
Coal is one of the abundantly and evenly distributed fossil fuel (~65% of the total fossil fuels) around the globe, unlike oil and gas reserves which are only present in a fewer regions of the world. 6 At present, coal is the major source of energy and continues to dominate as a main contributor for energy generation in developing countries. 7 Pakistan has huge coal reserves that nearly approach to 185 billion tons in its different provinces, especially in Sindh. It is estimated that the recoverable coal reserves in Pakistan are about 1.45 thousand Mtoe with total production capacity of 2. 33 Mtoe per year. 8 The continuous consumption of coal as an energy source can pose severe environmental concerns due to large amount of CO 2 emission in atmosphere. 9 Due to growing concerns of climate change and its associated environmental problems, authorities related to environmental protection are forcing to shift conventional energy sources to renewable energy sources to minimize carbon foot print. 2,10 The generation, disposal, and treatment MSW are major concerns in every country. Currently, Pakistan is one of the developing countries, experiencing severe environmental issues that are mostly linked to mismanagement of MSW disposal. The continuous release of greenhouse gases (GHG) into the atmosphere due to inadequate landfills facilities ultimately results in serious environmental issues. 11 In Pakistan, the estimated release of methane into the atmosphere due to these unusual landfilling practices is about 14.18 Gg per year which contributes almost 22 times greater greenhouse effect than CO2. 12 In light of these concerns, energy from waste sources is getting much attraction worldwide. Currently, the most common and adopted strategies of waste disposals are landfilling, incineration, sea dumping, and composting. 13 In the past few years, waste to energy concept in respective technology has gained great attention. 14,15 Incineration of waste is one of the simplest strategy to convert waste into its corresponding energy. 16 In this regard, a good understanding of the combustion process will be of great help in designing the incinerators with improved efficiency. The combustion process can be understood carefully by knowing the Kinetic parameters. The most popular and simple technique to study the combustion kinetic is thermogravimetric analysis (TGA). Many researchers have utilized the TGA analysis to evaluate kinetics of combustion process in incineration by developing various kinetic models. [17][18][19][20][21][22] Typically, two kinds of kinetic models are employed such as model-free (isoconversional) and model-fitting. In this paper, isoconversional methods were adopted, as these methods are reflected as more reliable and accurate compared to complex data fitting methods that are generally dependent to type of reaction mechanism and lead to imprecise data estimation over wide experimental range. 17,23,24 However, model-free strategy does not depend on a reaction mechanism for fitting of respective thermodynamic data that ultimately results in precise error-free modeling. Another significant issue is variation of activation energy during the combustion process, as combustion process consists of more than one steps with different activation energies. In model-fitting methods, whole combustion process is characterized by a single value of activation energy. Thus, value of activation energy obtained is an average value and cannot depict all the steps involved during combustion. Conversely, isoconversional models address this issue and estimate value of activation energy as function of fractional conversion ( ) that varies continuously till the end of complete combustion. 25,26 In this work, the thermogravimetric analysis is carried out, following by application of model-free methods to propose a set of the kinetic parameters for combustion of coal, MSW, and RDF, to provide fundamental information for optimization of combustion process. The novel objective was to investigate the variance in thermal decomposition and kinetic parameters of locally available sustainable fuels. The thermal behavior and kinetic data of MSW and RDF in comparison with coal with application of isoconversional are rare, it will help researchers and policymakers in great deal regarding planning of combustion and cocombustion of these fuels. The characteristics and thermogravimetric results of these samples used for evaluation of kinetics parameters are presented in our previous work. 27

| MATERIALS AND METHODS
The preparation of experimental samples (MSW and RDF) was based on representative components of MSW collected from city Lahore, Punjab, Pakistan, as described in our previous paper, 28 whereas low-rank coal sample was taken from Chiragh reserves located in province of Punjab, Pakistan. Thermogravimetric analysis was performed with the sample size of 20 mg in Mettler Toledo (TGA/DSC 1 STAR, USA) System. Initially, samples were heated from room temperature to 105°C. Four specific heating rates (10, 20, 30, and 40°C/min) were adopted to reach final temperature of 1000°C. A holding time of 10 minutes and 30 minutes was insured at temperature of 105°C and 1000°C, respectively. Synthetic air with constant flow rate of 80 mL/min was provided for all complete runs of experiments.

| KINETIC MODELING
Generally, for solid particles the rate of reaction is given as follows: Here And α can be ascribed to fractional conversion and denoted as Further, f(α) can be denoted as a function of fractional conversion (α). The equation can be simplified in derivative form using Equation (2) at respective ramping rates for temperature increase.

| Isoconversional methods
Many researchers have pointed that the prediction of reaction kinetics with model-fitting approach always poses inaccurate estimation due to reaction mechanism involvement. [29][30][31] Conversely, isoconversional approach does not follow any preassumption for reaction mechanism that ultimately reduces the possibility of errors for homogenous and heterogeneous combustion kinetic data. This model-free approach can play an effective role to estimate activation energies in wide range of experimental temperatures and conversions. The isoconversional methods applied in this study are Flynn-Wall-Ozawa (FWO), Friedman, Kissinger-Akahira-Sunose (KAS), and Vyazovkin.

| Vyazovkin integral model
This model-free approach is also known as nonlinear integral method that is outcome of revised expression of temperature integral from Equation (6).
As for isoconversional models, g(α) is not dependent on heating rate so, The above equation can be rewritten as follows: The integral value of I(E α , T α,i ) can be evaluated by resolving a complex integration or by using Senum-Yang polynomial 4th-degree approximation 44 as described below.
x can be taken as equivalent to α in the previous equation. The activation energies can be calculated by substitution of experimental values (T α and β) in Equation (10) with condition of changing E α to approach a minimum value. The minimum value was achieved for each value of α to estimate activation energy dependency on conversion. This strategy has also been reported reliable and consistent in various studies reported in the literature. 25,31,32,36,40,[45][46][47]

| Flynn-Wall-Ozawa method
This method was proposed by cumulative contribution of Flynn, Ozawa and Wall by the modification of Doyle contribution as described in Equation (12).
The resultant straight lines by plotting Equation (12) vs 1/T αi will give the estimation of activation energy at different conversion levels. 35,37,38,45,48

| TG and DTG profiles
TG and DTG curves for coal, MSW, and RDF were observed individually, as function of time and temperature range of 25°C to 1000°C. Figures 1 and 2 present TG and DTG trends of the sample fuels at specific heating rates (10, 20, 30, and 40°C/min). It is important to define the extent of conversion through an observable adequate physical property. According to different thermal decomposition studies, among all physical properties, mass loss is the most suitable physical property to describe the dependency of the overall rate constant on the temperature. The is mainly due to fact that the mass loss during the combustion process does not depend upon the heating rate. [53][54][55][56] It is broadly accepted that fuel decomposition process mainly contains three stages over complete range of temperature, which are (a) moisture drying, (b) major loss of weight, due to release of volatile organic matter, and (c) consecutive slow combustion of fixed carbon. 57 Last two steps, certainly, are linked with chemical composition (homogeneous/heterogeneous nature) of these samples, as main constituent of these samples has dissimilar degradation profiles, as shown in Figure 2. The physical composition of simulated solid wastes is presented in Table 1. The proximate and ultimate analysis presented in Table 2 showed that all the fuels were low in content of moisture and fixed carbon, as a result, first and third region showing less prominent weight loss. As expected, apart from the heating rate, the decomposition process of MSW and RDF shows a very high weight loss due to high reactivity, whereas coal TG behavior is slower, which is characterized by a low volatile content and high ash content. The mass percent of ash content of coal is ~32.1% that is much higher than that of MSW (~11.1%) and RDF (~12.9%). The decomposition pattern of solid wastes showed that a number of individual shoulders or peaks appeared at lower heating rate, which disappear or overlap as the heating rate is increased. The thermal decomposition of coal results in one major peak, due to release of carbon-containing volatile matter, 58 whereas existence of different peaks in case of solid wastes is credited to the heterogeneous nature of these wastes fuels. The solid wastes show prominent weight loss between 180°C and 550°C, whereas the coal weight loss is mainly between 400°C and 770°C at heating rate of 20°C/ min. The solid wastes exhibit multicomponents sharp weight loss peaks in the fast devolatilization stage. The weight loss of first peak of MSW (~43.1%) and RDF (~39.8%) is mainly contributed by the highly volatile and easily decomposable components in solid wastes such as biodegradables (cellulose and hemicellulose), textile, and paper with temperature range F I G U R E 1 TG curves of coal, MSW, and RDF samples at different heating rates of 180-390°C and 216-376°C, respectively. The second volatile release peak from solid wastes mainly corresponds to combustion of mixed plastic (LDPE, HDPE, and PVC) with almost same temperature range 388-550°C for MSW and RDF. In the carbonization stage, tiny weight loss in MSW and coal is attributed to thermal decomposition of lignin and char present in samples. This result is in agreement with the findings of researchers using MSW and RDF as fuel samples in TGA. 16,17,[59][60][61] The combustion characteristic parameters such as the initial temperature, burnout temperature, and temperature at maximum weight loss obtained from TG and DTG curves for combustion of the studied samples at different heating rates are summarized in Table 3. Thermal behavior and characterization studies at specific heating rates are an imperative criterion for kinetics studies. It is noticed that the heating rate affects the TG and DTG curves. 62 At specific heating rates, contact of fuel particles varies, which change the curve shape and combustion characteristics considerably, as indicated in Figures 1 and 2. It is evident from analysis, with the increasing heating rates, TGA curves are shifted toward the right and DTG curves are slightly shifted toward higher  peak temperatures without changing the thermal profile. This shows that the overall reaction rate in decomposition process is a function of temperature only. 55 Consequently, all the combustion characteristic parameters were shifted to higher values, showing thermal lag at increasing heating rate. This might be credited to heat and mass transfer limitations due to poor thermal conductivity of materials present in samples. 59 Among the selected fuel samples, coal has highest ignition and burnout temperature, while MSW and RDF have considerably low and slight close range of these temperatures. The coal ignition temperature at four heating rates (10, 20, 30, and 40°C/min) varies from 443°C to 460°C. For MSW and RDF, this thermal lag varies from 219°C to 230°C and 238°C to 255°C, respectively. Similar shift of temperature zone/ thermal lag is observed for burnout temperature values. The higher burnout temperature of coal is attributed to higher ash content, which again is main factor for further characterization of burnout process. It is evident that the heating rate had a significant effect on the weight loss rate during combustion process. In combustion process of these samples, the higher heating rates result in lower weight loss (conversion) and high reactivity. The reactivity during combustion process is proportional to the height of DTG peak. In case of coal, as heating rate increased, the weight loss rate at same temperature is increased. As revealed in Figure 2, as heating rate increased from 10 to 40°C/min, the weight loss rate increased from to 0.1 to 0.19% S −1 . Even though, similar trend was observed in case of MSW and RDF, as greater weight loss rate was recorded in both peaks with increasing heating rate. However, in solid wastes overlapping of peaks was prominent at higher heating rates. This may be explained on the basis of residence time during the combustion process. At low heating rate, more residence time results in efficient and effective heat transfer compared to higher heating rate. Similar results of combustion process are reported by researcher using different heating rate such as MSW, 16,61 RDF, 17 and coal. 20 The reactivity of these fuels was investigated at peak temperatures, resulting an order of RDF > MSW > coal.

| Kinetic analysis
In order to describe the dependence of the activation energy (E ) on the conversion degree, four models from model-free kinetic methods were tested. The isoconversional plots of KAS, FWO, and Friedman for conversion ( ) range of 0.1-0.9 are shown in Figure 3. According to reported studies, the parallelism of these lines is accredited to the same reaction mechanism and kinetic behavior. 63 For considered range of Note: T i : the ignition temperature, T f : burnout temperature, T 1 , T 2 : temperature at maximum weight loss rate of first peak and second peak.  Tables 4-6. The correlation coefficient of obtained E is close to unity, which indicates that these models had capability for better fit of experimental data for estimation of kinetics. The average values of obtained E from isoconversional methods are summarized in Figure 4. Many researchers have presented such findings about isoconversional models. 17,25,29,32,64 Activation energy means the minimum energy requirement that must be acquired to start a reaction. This indicates, the reaction with high activation energy needs a high temperature or an extended reaction time.  Table 4. For interval of 0.2-0.9, the activation energies were close to each other with decreasing trend. This thermal event at temperature ~400°C and conversion degree ~0.1 is characterized by slow reaction rate with highest energy barrier ~155 kJ/mol, which means that in case of coal the energy requirement for the main mass combustion was less after initializing the reaction. It is noticeable that the coal activation energy values obtained from Friedman differential method are slightly lower than integral methods. In case of MSW and RDF, activation energy analysis shows tedious dependency on conversion and specifies the typical behavior of complex parallel reaction, involving multiple and consecutive degradation steps. For solid waste fuel, the changing trend of E value obtained by these isoconversional methods is very consistent. The E values of solid wastes reported in the literature show great variation due to nonuniform composition of such kind of fuels. In case of solid wastes with the conversion rates increased from 0.1 to 0.9, two obvious peaks were observed at : 0.1 and 0.6, respectively. As it can be observed from first peak of MSW, between 0.1 ≤ ≤ 0.5 for temperature range between 180 and 365°C shows a progressive decrease in the activation energy from 239 to 180 kJ/ mol, whereas first peak in case of RDF shows a progressive increase in the activation energy (118-165 kJ/mol) for same conversion degree with shorter temperature range between 226 and 356°C. This behavior is credited to complex multistep reaction during biomass decomposition of MSW. As mentioned earlier, the first peak is attributed to combustion of volatile matter and second peak corresponds to combustion of different organic compounds which offer greater energy barrier. Finally, both MSW and RDF at conversion degree 0.6 observed a maximum increase in activation energy values, ~365 kJ/mol and ~290 kJ/mol at corresponding temperature of 445°C and 405°C, respectively. This interval 0.6 ≤ ≤ 0.9 is attributed to components of solid wastes with lower activity (biochar) and complex decomposition phase of PVC, which dictate the rate of reaction and need more energy under high temperature. 65,66 For solid wastes at conversion range ( = 0.1-0.5), smaller fluctuations in E values with increasing trend were observed, whereas at conversion range ( = 0.6-0.9), higher fluctuations with decreasing trend were observed throughout the process. This fluctuation in the E value is generally attributed to heterogeneous nature of solid wastes, which leads to complex reaction system including parallel, competitive, and complex reaction scheme. 57 It is possible to say that the average activation energy of solid wastes obtained from all considered isoconversional methods is very close, compared to average activation energy of coal, as illustrated in Figure 7. It is important to note that the MSW E by the all isoconversional methods were  67 Conferring to kinetic analysis by all four isoconversional models, the selected fuels could be set in subsequent order of activation energy MSW > RDF > coal.

| CONCLUSION
According to TG and DTG analysis, it was observed that thermal decomposition of coal sample comprised single stage, whereas thermal decomposition of MSW and RDF involved complex set of multiple and simultaneous steps related to decomposition of biowaste (ie, food waste, textile, and paper) and plastic. The rate of thermal decomposition of all solid fuels was found increasing with the increase in heating rates. The dependency of the apparent activation energy (E ) on conversion ( ) was estimated through isoconversional approaches using three integral methods (Vyazovkin, KAS, and FWO) and one differential (Friedman) method. In case of coal, monotonic decrease in activation energy was found for selected conversion, whereas solid wastes observe two prominent peaks with conversion ranges between 0.1 ≤ ≤ 0.5 and 0.6 ≤ ≤ 0.9, with maximum activation energy value of MSW (~365 kJ/mol) and RDF (~290 kJ/ mol) at conversion degree 0.6. The average activation energies were found as 89.