Insight into the Effects of Inorganic Element Catalysis and Basic Fuel Properties on the Self-Sustained Smoldering Process of Sewage Sludge

Abstract

The objective of this study is to investigate the effects of inorganic element catalysis and basic fuel properties of sewage sludge on pyrolysis kinetics and self-sustained smoldering characteristics. The sludge pyrolysis process was explored by thermogravimetric and iso-conversion methods, and it was found that the pyrolysis process can be divided into two stages, which are mainly determined by the organic and inorganic components of the fuel. The inorganic components (e.g., Na, Fe and Mn) have a significant catalytic effect on the release of volatiles and the decomposition of macromolecules. The smoldering experiment revealed that the smoldering front and the evaporation front propagated at stable but different speeds. Among the five fuels, SS4 has the highest smoldering temperature (1070 °C) and the lowest propagation velocity (0.7 cm/min of smoldering velocity and 0.3 cm/min of evaporation velocity), while the carbon density mainly determines the heat release in the oxidation process, and the inorganic elements play a significant catalytic role at different temperatures. The obtained thermodynamic and smoldering characteristics facilitate the development and optimization of the disposal of sewage sludge, emphasizing the importance of considering feedstock composition.

1. Introduction

Sewage sludge (SS) is the main by-product of the wastewater treatment process, containing massive organics, pathogens and other harmful substances, with an output of around 40 million tons annually in China [1]. Inappropriate disposal of SS can lead to serious contamination of groundwater and soil, as well as threats to human health [2]. Both high energy consumption and low profitability are the major hurdles that hinder the wide application of SS disposal technologies (e.g., incineration [3,4], pyrolysis [5,6], gasification [7,8] and hydrothermal treatment [9], etc.). Smoldering technology is emerging as a promising method to solve the above two problems very well [10,11]. The smoldering combustion process consists primarily of pyrolysis and oxidation. When organic components are ignited successfully at high temperatures, the heat released by oxidation is transferred to the pyrolysis region, where char and volatile products are produced. Self-sustaining smoldering can be achieved when the heat generated by the reaction is sufficient to overcome heat loss and maintain the reaction’s propagation, so less external energy input is required and the operating costs are lower [12].

In fact, the basic fuel properties, which mainly consist of organic matter and inherent inorganic elements, dominate the progress of sludge smoldering and attributes of the derived products [13,14]. Generally, the organic components in the sludge, i.e., protein, polysaccharide and humic acid, account for 30–70% of the total solid phase of the sludge [1], while the content of inorganic components are more complex. Inorganic matter in sludge mainly includes mineral and metallic compounds introduced during flocculation or sludge-conditioning processes. The different inorganic elements have little effect on the structure, but have essential catalytic effects on the pyrolysis conversion of the feedstock [15]. Many studies have investigated the effect of inherent inorganic elements on the pyrolysis of solid fuels, including biomass, coal and sewage sludge, by means of demineralization and/or impregnation [16,17,18]. Those studies revealed that inherent inorganic elements showed different effects on the pyrolysis of solid fuels, depending on the type and content of the metal element and pyrolysis temperature. Thermogravimetric (TG) analysis is important to understand the thermochemical conversion of solid waste. The study [19] mixed biomass with coal and then analyzed its burning process using a TG analyzer; the result is that polysaccharides can increase the rate of devolatilization and accelerate charcoal combustion. The TG analysis of additional catalysts (alkali/alkaline earth metal compounds) in burning of municipal solid waste found the flame temperature reduced [20]. Similar studies were done by [21,22] and found that the catalyst can increase char formation and the rate of the volatility, thus greatly affecting product distribution.

One potential problem is that most studies in the field of the effects of inorganic element catalysis and basic fuel properties are mostly limited to one feedstock. As an example, Tian et al. reported that inorganic minerals in sludge catalyzed the conversion of N-containing compounds to N₂, thus reducing the production of NO_x [23], whereas Tang et al. came to discrepant results [24]. Apparently, the inconsistency of these results may be due to the failure to consider the effect of the differences in components of different SS. Therefore, it is necessary to reveal the potential relationship between the fuel composition and smoldering properties, which could provide essential information for the selection of sludge sources and the optimization of the smoldering conditions.

This work investigates the effects of inherent inorganic element catalysis and basic fuel properties of SS on pyrolysis kinetics and self-sustained smoldering characteristics. Based on the fact that pyrolysis can determine the following smoldering combustion process, the pyrolysis behaviors are investigated firstly and the kinetic parameters are obtained by the TG analysis. Employing a multi-device coupled online system (thermogravimetry–Fourier-transform infrared spectroscopy–mass spectrometry (TG–FTIR–MS) for real-time qualitative and semi-quantitative analysis of pyrolysis products can be useful and necessary to provide more accurate and detailed information for revealing the genuine pyrolysis process. The effects of fuel properties of SS on smoldering combustion are also discussed. Altogether, this study provides new guidance for considering smoldering combustion as an alternative SS disposal option.

2. Results and Discussion

2.1. Basic Fuel Properties of SS

Table 1. Basic fuel properties of the SS.

	Proximate Analysis			Ultimate Analysis					HHV e MJ/kg	LHV f MJ/kg
	V a %	A b %	FC c %	C %	H %	N %	S %	O d %
SS1	37.0	57.7	5.4	20.4	2.9	3.6	0.6	72.5	9.4	9.2
SS2	43.0	49.9	7.1	23.9	3.6	4.0	0.8	67.7	10.3	9.6
SS3	42.2	50.1	7.7	23.9	3.6	4.1	1.0	67.4	10.2	9.5
SS4	30.1	55.0	14.9	35.5	3.5	0.2	0.3	60.6	17.4	16.7
SS5	38.7	60.6	0.6	15.9	3.0	2.4	4.9	73.8	6.8	6.1

^a Volatile matter content, dry basis. ^b Ash content, dry basis. ^c Fixed carbon content, dry basis. ^d By difference. ^e Higher heating value on dry basis, measured using a bomb calorimeter. ^f Lower heating value on dry basis, measured using a bomb calorimeter.

shows the results of proximate, ultimate and heat value analysis of the SS. The samples were dried at 105 °C for 24 h in a drying oven to remove the moisture content before the test. The comparison of the basic fuel properties for five SSs shows significant differences in their various sources, especially between municipal sludges and industrial sludges. The ash content of around 50–60% observed in all SS is noticeably higher than that of other common fuels (e.g., coal and biomass), thus resulting in low calorific values. The low content of organic components indicates that part of the organics of activated sludge was broken down during the process of sewage digestion. The fuel properties of municipal sludges are similar, in which high content of nitrogen suggests a large amount of protein [25]. The characteristics of industrial sludges vary greatly, with SS4 having the highest carbon content and HHV (LHV), which are 35.5% and 17.4 MJ/kg (16.7 MJ/kg), respectively. Meanwhile, SS5 has the lowest carbon content and HHV (LHV), and the highest sulfur content of 4.9%.

2.2. Characteristics of Inorganic Elements

Contents of the main inorganic elements—alkali metal elements (K, Na), alkali earth metal elements (Ca, Mg, Al) and transition metal elements (Fe, Cu, Cr, Mn)—are shown in Figure 1. In addition to the content, the occurrence forms of metal elements in SS are diverse.

Group 1 alkali metal elements (K, Na)—occur predominately in water-soluble/ion-exchangeable forms (i.e., organically bound); these elements are generally highly reactive during combustion and contribute to the formation of inorganic phases with lower melting points [26]. A detailed analysis has been introduced in our previous study [27]. Therefore, alkali metal elements can promote the escape of volatile components and improve the internal heat and mass transfer during sludge pyrolysis [28]. This catalytic effect is further explained in the subsequent analysis.

Group 2 alkali earth metal elements (Ca, Mg, Al)—Ca and Mg present predominately in water-soluble/ion-exchangeable forms. Al is predominantly found in the insoluble portion of the fuel [26,27].

Group 3 transition metal elements (Fe, Cu, Cr, Mn)—present predominately in the water-soluble, ion-exchangeable and acid-soluble phases. Transition metal elements do not always exist in fuel, and for different types of materials, there are different transition metal elements in different forms [29]. These elements tend to act as catalysts [30,31].

Actually, many studies have investigated the catalytic mechanisms of different metal elements on thermochemical conversion [32,33,34,35]. During the pyrolysis process, alkali and alkaline earth metal (AAEM) cations act as virtual cross-linking points, which are most likely to bond with oxygen-containing groups, especially carboxyl and hydroxide. It has been speculated that, during this process, in the absence of oxygen, some AAEM species associated with the –COO⁻ groups at the outer particle surface may leave the particle, together with the –COO group, if the bond between the –COO⁻ group and the fuel matrix was broken first. The –COONa or –COOCa groups in fuel are isolated from each other, where Na represents alkali metals and Ca represents alkaline earth metals. In the presence of oxygen, these AAEM species become the active centers for the chemisorption of O₂, leading to the formation of surface complexes. These surface complexes bond the AAEM species tightly to the fuel/char matrix. With the release of CO₂, Na or Ca originally associated with a –COO group in the coal matrix may still bond to the fuel/char matrix (–CM),

$$(-\mathrm{COO}-\mathrm{Na})+(-\mathrm{CM})=(\mathrm{CM}-\mathrm{Na})+{\mathrm{CO}}_2$$

(1)

or for Ca,

$$(-\mathrm{COO}-\mathrm{Ca}-\mathrm{OOC}-)+(-\mathrm{CM})=(-\mathrm{COO}-\mathrm{Ca}-\mathrm{CM})+{\mathrm{CO}}_2$$

(2)

$$(-\mathrm{COO}-\mathrm{Ca}-\mathrm{CM})+(-\mathrm{CM})=(\mathrm{CM}-\mathrm{Ca}-\mathrm{CM})+{\mathrm{CO}}_2$$

(3)

continuously serving as a virtual cross-linking point. For those AAEM species retained in the fuel/char matrix, which still contains significant amounts of substitutional groups, particularly oxygen-containing groups (e.g., phenolic groups), the AAEM species are likely to be bonded to these oxygen-containing groups. As the temperature is further increased, the newly formed Ca–CM or Na–CM bonds are not very stable and are broken again to generate free radical sites together with the release of oxygen-containing species or aliphatic materials. Some of the AAEM species may leave the particle.

$$(\mathrm{CM}-\mathrm{Na})=(-\mathrm{CM})+\mathrm{Na}$$

(4)

$$(\mathrm{CM}-\mathrm{Ca}-\mathrm{CM})=(-\mathrm{CM})+(-\mathrm{Ca}-\mathrm{CM})$$

(5)

$$(-\mathrm{Ca}-\mathrm{CM})=(-\mathrm{CM})+\mathrm{Ca}$$

(6)

$$(-\mathrm{CM})=(-{\mathrm{CM}}^{\ast })+\mathrm{gas}$$

(7)

The AAEM species are also likely to be highly reactive and recombine with the free radicals generated to form more stable Ca–CM or Na–CM bonds:

$$(-{\mathrm{CM}}^{\ast })+(-\mathrm{Ca}-\mathrm{CM})=({\mathrm{CM}}^{\ast }-\mathrm{Ca}-\mathrm{CM})$$

(8)

$$(-{\mathrm{CM}}^{\ast })+\mathrm{Na}=({\mathrm{CM}}^{\ast }-\mathrm{Na})$$

(9)

For the release of organic volatiles, the net effect of the repeated Ca–CM or Na–CM bond-forming and bond-breaking processes described above is that tar precursor fragments are repeatedly linked to the char matrix through reactions involving free radicals and are thermally cracked. During this thermal-cracking process, the more aliphatic parts in a tar fragment, including the smaller aromatic ring systems, are released as gas. Some more aromatic units, especially the larger aromatic ring systems, may eventually become a part of the char.

The catalytic mechanisms of transition metals and alkali/alkaline earth metals are very different. Manganese oxide mainly undergoes the redox process through the change of its own valence to achieve the catalytic effect, and this process can reduce the activation energy of the overall combustion reaction and accelerate the conversion rate of intermediate products to final products, so as to achieve the purpose of obtaining higher conversion efficiency at the same temperature [34]. Actually, the change of valence of manganese oxide is a transfer process of oxygen atoms, where oxygen is transferred from the inner part of the catalyst to the surface step by step, and then the surface oxygen and organic matter combine to produce carbon dioxide or water, and the catalyst then stores oxygen in the reverse direction through the oxidation of air. Moreover, the activity of the catalyst is also shown by the adsorption effect of active sites on reactants, and some studies found that manganese oxide has a certain adsorption effect on both methyl and hydroxyl groups, and its catalytic mechanism is shown in Figure 2 [35].

As shown in Figure 1, all three types of municipal sludge have a low content of inorganic elements. The contents of AAEM are similar and almost no other transition metal elements are present except Fe. The content of inorganic elements in the two industrial sludges differed greatly. The high content of transition elements, especially Mn, in SS4 indicates that the ash component in sludge has excellent positive catalytic effects on sludge pyrolysis and smoldering combustion. The high content of Na and Fe in SS5 indicates the high activity of the sludge.

2.3. Thermogravimetric and Kinetic Analysis

The TG–DTG curves of pyrolysis of SS at a heating rate of 5, 7.5 and 10 °C/min are shown in Figure 2. The different amounts of solid residue remaining after thermal decomposition of all samples can be explained by the diverse organic and inorganic compositions [36]. Therefore, the pyrolysis process of the SS can be divided into two stages: stage 1 at medium-low temperature represents the decomposition of organic components, and stage 2 at high temperature represents the decomposition of inorganic components. The dehydration stage is neglected here because the samples were dried in a drying oven at 105 °C for 24 h during pretreatment. Although the main decomposition temperature regions are similar, their DTG peak heights and positions at each decomposition region are different. In particular, the main decomposition temperature region (between 350 and 600 °C) in stage 1 of SS4 is higher than that of other sludges. The main decomposition temperature region (between 125 and 500 °C) in stage 1 of SS5 is lower, while the DTG peak heights in stage 2 are higher than those of other sludges. From

Table 2. Characteristic parameters for pyrolysis obtained at various heating rates.

Sample	$\overline{{\mathit{T}}_{\mathit{i}}^{\mathit{a}}}$ °C	$\overline{{\mathit{T}}_{\mathit{m}}^{\mathit{b}}}$ °C	$\overline{-\mathit{D}\mathit{T}{\mathit{G}}_{\mathit{m}}^{\mathit{c}}}$ %/°C	${\mathit{M}}_{\mathit{f}}^{\mathit{d}}$ %
SS1	207.88	272.04	1.11	65.01
SS2	223.68	279.25	1.39	58.43
SS3	224.02	275.92	1.44	53.75
SS4	319.02	545.64	1.10	65.47
SS5	160.01	355.83	0.78	48.35

^a Average ignition temperature at different heating rates, by TG tangent method. ^b Average maximum decomposition peak temperature at different heating rates. ^c Average maximum decomposition rate at different heating rates. ^d Residue mass at the set temperature.

, it could be observed that the initial decomposition temperature of SS4 (319.02 °C) is much higher than the three municipal sludges, while SS5 (160.01 °C) is significantly lower than them. The maximum mass loss rate of SS5 is 0.78% at 355.83 °C, which is also lower than other sludges. This can be explained by the low volatile contents and high ash content in SS5, as expected from the proximate analysis results (see Table 1). In addition, the high content of alkali metal elements, especially Na, in the ash of SS5 greatly promotes the release of volatile matter. Similar results were obtained by Li and Kim et al. [37,38], where AAEM were effective in reducing the peak temperature of the maximum decomposition rate. In fact, the high Fe content in dyeing sludge greatly catalyzes the thermal degradation of organic components during pyrolysis [39].

Stage 1 is the main decomposition stage which results in a weight loss of 65–85% of the initial weight. The different DTG peaks in stage 1 indicate the presence of organic fractions such as humus, carbohydrates, protein and lipid decomposition. Bach et al. [40] reported decomposition of proteins and lipids to take place at 150–550 °C and 210–310 °C, respectively. Furthermore, the cellulosic materials were found present in the sewage sludge samples decomposing at a temperature between 240 and 400 °C [41]. The decomposition of these components may overlap over different temperature regions. Stage 2 denotes the decomposition of inorganic elements, and only a clear degradation peak of SS5 is visible in Figure 3, indicating the decomposition of carbonates at a high temperature [42]. This is also illustrated by the minimum residue mass of SS5 (48.35%) in Table 2.

Iso-conversional methods were used to analyze the sludge pyrolysis from the kinetic model equations for Flynn–Wall–Ozawa (FWO), Kissinger–Akahira–Sunose (KAS) and Starink methods, respectively. Conversions between 0.1 and 0.8 were considered for this study, and the corresponding E_a values are presented in Figure 4. The curves exhibit a similar shape corresponding to the considered iso-conversional method, and the activation energy values are nearly the same. The E_a values on the pyrolysis of all SS firstly increase to a peak and then gradually decrease. By analyzing the E_a of different components in the sludge, Lee et al. [43] found that the content of organic and inorganic components in the sludge had a great impact on it at all conversions. Similar results are obtained here, where the activation energy dependence on conversion suggests complex reaction schemes. The occurrence time of an E_a peak in Figure 4 is similar to the TG–DTG curve. For municipal sludges, the E_a peak at α = 0.3–0.65 corresponds to the peak DTG temperature range of stage 1 at 200–400 °C. The E_a value of SS4 decreases and remains stable after α = 0.25, which is partly because of the catalytic effect of ash in sludge. In addition, the E_a value of SS5 increased until α = 0.65, which is due to the thermal decomposition of inorganic components. Lin [41] believed the reason for similar phenomena as the decomposition of inorganic matter required higher energy to start and generated undesirable products. Actually, Tang [24] investigated the activation energy of sludge before and after pickling and ash removal and discovered that the inherent inorganic elements in sludge have a catalytic effect on the decomposition of organic fractions of sludge at low pyrolysis temperatures, lowering the reaction activation energy. However, because the iso-conversion method does not account for the multi-step reactions in sludge pyrolysis, the apparent activation energy does not allow for a detailed discussion of the mixing effect of inorganic elements.

2.4. TG–FTIR–MS for the Gaseous Products

Figure 5 and Figure 6 show the FTIR and MS detection of gaseous products and main functional groups from SS by 10 °C/min. Figure 5a and Figure 6a show that the absorption peak of CH₄ (3015 cm⁻¹, m/z = 16) reaches the maximum at different temperatures, which corresponds to the pyrolysis of their slight hydrocarbon content, among which SS4 produces the most. CH₄ is a typical high calorific gas fuel, usually produced by random breakage of polymer chains, and it has also been suggested that CH₄ comes from secondary cracking and methanation of bio-oil [44]. The absorption peak at 3600–4000 cm⁻¹ is attributed to the hydroxy functional group, indicating the formation of H₂O. In addition to the free water and bound water release at low temperatures, the characteristic absorption peak of water can also be observed at high temperatures, which may be from the water bound with inorganic substances or produced by the organic decomposition reaction. Figure 5c and Figure 6c show that the production of CO increases at a high temperature, which is due to the Boudouard reaction of CO₂ and char (C + CO₂ → 2CO). As shown in Figure 5d, the IR spectra have high absorbance in the regions of 2400–2240 cm⁻¹ and 600–680 cm⁻¹, corresponding to CO₂, which is attributed to saccharide and protein decomposition [45]. The trends of the absorption peak with temperature have coincided with the results of the TG–DTG. As the content of inorganic components in SS5 is high, there are two distinct peaks in the CO₂ absorbance curve, which correspond to the decarboxylation of organic matters at a low temperature and the thermal decomposition of inorganic components at a high temperature, respectively. The MS intensity of CO₂ (Figure 6d) is much higher than that of the other gaseous products, indicating that the release of CO₂ is the most important reason for the mass loss. Among them, SS5 has the highest CO₂ production, which may be due to the catalytic effect of AAEM [37]. The gaseous products also include esters, aldehydes and acids, represented by the C=O and −COOH functional groups in Figure 5e,f. Their formation at a low temperature is mainly due to the thermal decomposition of fiber, saturated aliphatic chains and proteins. As mentioned earlier, the content of organic fractions in municipal sludges is higher than that in industrial sludges, resulting in more C=O and −COOH. The gradual increase in −COOH in SS5 in Figure 5f may be due to the good catalytic effect of Fe₂O₃ on the cracking of oil tar, allowing oxygenated organics to be directly cracked into smaller organics and H₂O through deoxygenation and cracking processes [45].

2.5. Smoldering Combustion

Smoldering combustion can be thought of as proceeding in two steps. The solid fuel is decomposed through endothermic pyrolysis, producing gases and solid char. The char is then oxidized, usually in a highly exothermic process in which it is converted to gaseous products and ash [46,47]. The fuel properties have a great effect on the two steps of the reaction, and the oxidation is controlled by pyrolysis, as the pyrolysis product is used as the oxidation reactant.

Temperature history is the most widely used data for smoldering process analysis. The traces of a set of thermocouples placed along the center of the reactor are shown in Figure 7. Taking SS1 as an example to analyze the advancement of the smoldering process, during the external heating period, the temperature rises slowly and there is an evaporation plateau of 100 °C. Subsequently, when the temperature of TC1 rises to the ignition temperature of SS1 (250 °C, exceeding the ignition temperature obtained from TG analysis, as shown in Table 2), the heater was turned off and fresh air was introduced. Generally, the ignition process dominates the initial temperature rise for thermocouples 1 to 4. The different slopes in the temperature curves indicate that the smoldering process is a multi-stage process dominated by different mechanisms [48]. The initial reaction of the volatiles produced steam that condensed downstream. The water then evaporated as the reaction front approached. The other evaporation plateau occurring between 60–80 °C evidences the moving process of the evaporation front. As the smoldering front approached the thermocouple, the temperature began to rise sharply after the water had evaporated. The high heating rate from the evaporation temperature to the peak temperature in the temperature curve indicates that the pyrolysis process occurs extremely fast in the actual smoldering combustion, and the smoldering velocity is primarily controlled by the oxidation process.

The data of peak temperature extracted from Figure 7 are presented in Figure 8. In the absence of external energy being applied, the successive temperature peaks from TC4 to TC14 indicate the advancement of the smoldering front in a self-sustaining manner, while the data from the first 3 cm and last 6 cm were affected by ignition and surface heat loss, respectively. Figure 9 shows the average peak temperatures in the self-sustaining smoldering stage, and it is found that the smoldering temperature of SS4 is much higher than that of the other SS. Combined with the previous analysis, it can be considered that this is due to the high content of organic components in SS4, and the inorganic components in the sample (e.g., Mn) play a key catalytic role in the heterogeneous reaction that occurred in the reactor. The above two superiorities of fuel properties are also reasons for the high LHV of SS4.

Typically, the smoldering reaction front is considered particularly thin according to the large-activation-energy asymptotics [49,50]. Based on temperature histories, the smoldering velocity is calculated from the time lapse of the front arrival at two consecutive thermocouples and the known separation distance, as is the evaporation velocity. Figure 10 shows that the smoldering front and the evaporation front propagated at stable but different speeds. This is because the smoldering propagates in a completely dried material, and the moisture only dictates if self-propagation is possible or not [51]. Furthermore, Figure 10b highlights the complicated effects of fuel properties on the smoldering reaction, which is hypothesized to have resulted from the organic and inorganic compositions in samples. It is known that the C content and LHV of SS4 are the highest, which is equivalent to the highest content of fuel per unit volume, leading to its potential to release more heat once it smolders in a self-sustaining manner. This explains the peak temperature results in Figure 9. On the other hand, it is known from the previous analysis that the temperature required for thermal decomposition of SS4 is the highest; therefore, it took longer to destroy when the smoldering front arrived, which accounts for the lowest smoldering propagation velocity.

3. Materials and Methods

3.1. Materials

Five different SS feedstocks were collected from different wastewater treatment plants. Municipal sludges, defined as SS1, SS2 and SS3, were delivered from Wuhan and Guangzhou, China. Additionally, industrial sludges, defined as SS4 and SS5, were delivered from Zhuhai, China. Enough samples were taken at one time to meet all test needs. The moisture contents of municipal sludges are more than 80%, and they were dried outdoors for five days at an ambient temperature of 27 °C, while the moisture contents of industrial sludges are around 5%. All samples were mechanically crushed to obtain uniform samples with a particle size of less than 180 μm and kept in the cabinet.

3.2. Kinetics Theory

The kinetics of heterogeneous solid-state thermal decomposition is dominated by the fundamental equation [52]:

$$\frac{d\alpha }{dT}=k\left(T\right)f\left(\alpha \right)=A\exp \left(\frac{-E_a}{RT}\right)f\left(\alpha \right)$$

(10)

where A, E_a, R and T are the pre-exponential coefficient, the activation energy, the universal gas constant and the sample absolute temperature, respectively. f (α) represents the kinetic model that describes the rate of conversion dependence on the conversion α.

$$\alpha =\frac{W_0-W_t}{W_0-W_{\infty }}$$

(11)

where W₀, W_t and W_∞ are the sample weights at the beginning, at time t and at the end of the pyrolysis stage, respectively. Under a constant heating rate (β = dT/dt), the time dependence of the conversion rate can be transformed into a temperature dependence which can be used to rewrite the differential form (Equation (3)) or the integral form (Equation (4)) of the decomposition kinetic expression.

$$\frac{d\alpha }{dT}=\frac{A}{\beta }\exp \left(\frac{-E_a}{RT}\right)f\left(\alpha \right)$$

(12)

$$g\left(\alpha \right)={\displaystyle {\int }_0^T\frac{A}{\beta }\exp \left(\frac{-E_a}{RT}\right)}dT$$

(13)

The pyrolysis of SS is a particularly complicated process that is considered the presence of several complex components and their parallel and/or sequential reactions [41]. Therefore, the iso-conversional method was used here to evaluate the apparent activation energy to ensure its reliability and objectivity. In this paper, FWO, KAS and Starink methods were employed to evaluate the apparent activation energy of feedstock at conversions because of their good adaptability and validity for the model-free approach.

Starink examined two iso-conversional techniques (FWO and KAS) and found out that both conform to the expression given below [53,54,55]:

$$\ln \left(\frac{\beta }{T_{\alpha }^x}\right)=C_x-\frac{y{E}_{\alpha }}{R{T}_{\alpha }}$$

(14)

where for KAS x = 2, y = 1, for FWO x = 0, y = 1.052 and for Starink x = 1.8, y = 1.0037. Their linearized form can be expressed as:

$$\ln \left(\frac{\beta }{T_{\alpha }^2}\right)=C_x-\frac{E_{\alpha }}{R{T}_{\alpha }}$$

(15)

$$\ln \left(\beta \right)=C_x-1.052\frac{E_{\alpha }}{R{T}_{\alpha }}$$

(16)

$$\ln \left(\frac{\beta }{T_{\alpha }^{1.8}}\right)=C_x-1.0037\frac{E_{\alpha }}{R{T}_{\alpha }}$$

(17)

At a constant value of conversion rate α, the thermogravimetric data recorded at different heating rates β help to derive a straight line whose slope allows evaluation of the apparent activation energy.

3.3. Apparatus and Methods

3.3.1. Proximate, Ultimate and Heating Value Analyses

Standard methods were used to determine the volatile matter and ash content, GB/T 212-2008, while the fixed carbon (%) was determined by difference. Complete elemental analyses of representative samples were performed in a specialized diagnostic laboratory (Elementar, Vario EL cube, Langenselbold, Germany). The heating values were obtained by bomb calorimeter (Bente, WZR-1TCⅡ, Changsha, China).

3.3.2. Inorganic Element Analysis

Inductively coupled plasma atomic emission spectroscopy (ICP–AES) (PerkinElmer, Optima 8000DV, Waltham, MA, USA) was used to detect the contents of inorganic elements. Referring to our previous study [27], the acids selected for use in sample digestion were 65% HNO₃, 36% HCl, 30% H₂O₂ and 40% HF. These acids were chosen as each helps in the decomposition of a specific part of SS. The biological material is destroyed by HNO₃ and HF acts against mineral aluminosilicate matrixes. HCl is a strong acid and is used for the decomposition of organic samples in combination with HNO₃. H₂O₂ can significantly improve the oxidation capacity of the digestion system, and the decomposed substances are water and oxygen, which do not affect the subsequent detection. For each test, 0.1 g of sludge sample was weighted to the microwave digestion tank, mixed with acid and placed in the microwave digestion system (Anton Paar, Multiwave PRO, Buchs, Switzerland). Then, the acid was removed and moved to a 20 mL of volumetric flask. A blank sample was set up in the experiment to exclude the influence of acid.

3.3.3. Thermogravimetric Analysis

Thermogravimetric analysis of pyrolysis was investigated by TGA apparatus (SDT650, TA Instruments, New Castle, DE, USA). All five SSs were non-isothermally heated at different heating rates: 5, 7.5 and 10 °C/min. The samples were heated from room temperature to 105 °C, kept at 105 °C for 20 min to remove the moisture content, and then heated from 105 °C to 900 °C. Argon carrier gas at 100 mL/min and 5.0 ± 0.1 mg of sample were applied to make an inert atmosphere and minimize the heat and mass transfer effects. The TGA and derivative TG (DTG) curves were drawn based on the change in residual mass of SS during non-isothermal heating at different heating rates.

3.3.4. TG–FTIR–MS Analysis

The TG–FTIR–MS system (STA449 F3, Germany; Bruker Tensor 27 FTIR, Bremen, Germany; OmniStar MS, Beijing, China) was used to analyze the composition of gaseous products in the process of sludge pyrolysis. For this, 50 mg of SS were pyrolyzed in STA at 10 °C/min under 20 mL/min of an argon atmosphere. Then, the gaseous products travel through the heated transfer line to MS and FTIR. FTIR spectra were recorded from 4000 to 600 cm⁻¹. By the multi-equipment coupled online system, the release of different gas products and functional groups and their changing trends during the sludge pyrolysis process can be qualitatively understood.

3.3.5. Smoldering Combustion Setup

Figure 11 illustrates the experimental setup. It consists of the smoldering reactor, the airflow control system, the igniting control system and the data acquisition system. Smoldering combustion tests of SS were carried out in a cylindrical quartz reactor with a diameter of 15 cm and a height of 40 cm. A mass flow meter controlled the flow of air, which is supplied by an air compressor. A centrally symmetric air distribution plate was set up to ensure that the air flow was distributed evenly. An electrical heater (600 W) in the shape of a disc was used by the temperature controller to ensure uniform heating. The temperature was monitored by twenty thermocouples placed on the reactor’s central axis at 1 cm intervals each; then, the data were sent to a data logger. Before the experiment, the bottom of the quartz reactor was filled with clean sand with different particle sizes to have a rectifying effect. A porous matrix was created with the necessary heat retention and air permeability properties for smoldering combustion by mixing the sludges with sand and water. The sand was used here because it is a low-cost commodity and it has been identified as an effective agent to increase the porosity of fuels for application to smoldering treatments [12]. The mixed fuel was then loaded into the reactor with a filling height of about 20 cm and covered the clean sand to ensure the complete reaction of the top fuel. During the experiment, the heater was turned on and the temperature was set to 600 °C. When the temperature of TC 1 reached the ignition temperature of the SS, the heater was turned off, and the air was introduced with a flow rate of 31.8 L/min until the mixed fuel was completely destroyed.

4. Conclusions and Outlook

Applied smoldering has been shown to be a viable waste-to-energy strategy for high-moisture-content wastes like SS, while few studies have considered the effects of components in the different feedstock. This study addressed this knowledge gap by considering the effects of inorganic element catalysis and basic properties on the SS self-sustained smoldering process. The following conclusions were obtained.

(1)

The basic fuel properties and inorganic element composition of various sludges are quite different, especially between industrial sludges. The pyrolysis process of the SS can be divided into two stages, which are basically determined by the organic and inorganic compositions of the fuel.

(2)

The iso-conversional methods (FWO, KAS and Starink) are appropriate for evaluating the apparent activation energy of the complex mixtures of sludge, as indicated by the dependence of the apparent activation energy on the conversions. The apparent activation energy of municipal sludges shows a tendency of increasing and then decreasing with the conversion rate, ranging from 53 to 182 kJ/mol overall, whereas the industrial sludges do not exhibit a uniform trend. It can be considered that the organic components include proteins, lipids, polysaccharides and humic acid, while the inorganic components mainly play a catalytic role in the process of decomposition.

(3)

Employing a multi-device coupled online system (TG–FTIR–MS) for real-time qualitative and semi-quantitative analysis of pyrolysis products can be useful and necessary to provide more accurate and detailed information for revealing the genuine pyrolysis process. It reveals that the main gaseous products released during the heating process are CO₂, H₂O, CH₄ and CO, which are produced by the pyrolysis of organic macromolecules and inorganic substances.

(4)

In the smoldering experiment, it is found that the smoldering front and the drying front propagated at stable but different speeds. Furthermore, the propagation velocity of the char oxidation front was significantly affected by the carbon density and inorganic element catalysis of sewage sludge, in which the carbon density mainly determines the heat release in the oxidation process and the inorganic elements play a significant catalytic role at different temperatures.

However, it is acknowledged that this is an initial study and the correlation between sludge components and products needs to be explored. The catalytic effect of inorganic elements on the smoldering process is too intricate; future research will focus on how to quantitatively determine the catalytic behavior of inorganic elements and obtain the mechanism of catalytic interaction of inorganic elements.