Research on Thermal Behaviors and NO_x Release Properties during Combustion of Sewage Sludge, Sawdust, and Their Blends

Abstract

To investigate the thermal behaviors and NO_x emission properties during combustion of sewage sludge (SS), sawdust (SD), and their blends (SS5SD5, SS3SD7, and SS1SD9 with SD proportions of 50, 70, and 90 wt %, respectively), tests were conducted using thermogravimetry–mass spectrometry (TG-MS), Fourier transform infrared spectroscopy (FTIR), and a tube furnace in this study. Results indicated that hydrogen in the fuel was mainly released during volatile combustion, and carbon conversion proceeded during the whole combustion process. With the SD proportion increasing, samples exhibited better combustion characteristics. Compared to SD, SS emitted more NO_x due to its higher nitrogen content but showed lower conversion ratios from fuel nitrogen to NO_x, and the NO_x yields decreased significantly with the increase in SD proportion. NO_x emissions of higher volatile samples were more sensitive to temperature, and NO_x yields of SD and SS1SD9 continued to decrease from 800 to 1000 °C, whereas NO_x yields of SS, SS5SD5, and SS3SD7 changed slightly from 800 to 900 °C and decreased significantly from 900 to 1000 °C. Synergistic effects of cocombustion on NO_x emission varied with the blending ratio and temperature. SS5SD5 and SS3SD7 always presented a positive NO_x reduction performance, and SS1SD9 exhibited opposite NO_x reduction effects at different temperatures. Synthetically considering the SS disposal capacity, combustion characteristic, and NO_x yield, an SS proportion of around 30% in blends is more recommended in practical applications.

1. Introduction

Sewage sludge (SS) is an inevitable byproduct of wastewater treatment. The amount of SS has increased steadily worldwide with the rapid expansion of urban cities and wastewater treatment capacities over the past decades.^1,2 Lots of undesirable substances in SS, such as viruses, bacteria, parasites, and heavy metals, pose a constant threat to the environment and human health.^3,4 The common disposal methods of SS include landfill, agricultural application, composting, and incineration, but the practical application of each method varies all over the world.⁵⁻⁷ Nonstandard or irregular treatments do much harm to the environment and human health. Incineration is an efficient and convenient method to achieve sludge reduction and reutilization, and the abundant numbers of studies on the combustion of solid fuels provide rich research methods and utilization experiences for sewage sludge combustion.⁸⁻¹¹ Considering the poor fuel properties of SS due to its high ash content and low calorific value, a practical way to achieve efficient utilization of SS is cocombustion with some other high-quality fuels such as most agricultural and forestry wastes, which can also make remarkable contributions to the reduction of CO₂ emission.

Considering the high nitrogen content of most SS, the investigation of both thermal behaviors and NO_x release properties is of great significance for the application of cocombustion. Thermogravimetric analysis (TG) has been widely used to assess thermal behaviors and kinetics of cocombustion of SS and various fuels. The release of typical gas species during pyrolysis of SS and biomass was also investigated by thermogravimetry–mass spectrometry (TG-MS) and thermogravimetric Fourier transform infrared spectroscopy (TG-FTIR),^12,13 which contributed to the better NO_x emission control. The results of TG-MS analysis during cocombustion of straw and SS suggested an SS proportion of less than 30% in blend fuels.¹⁴ The synergistic effect on NO_x and SO₂ reduction was observed during the cocombustion of SS and three kinds of biomasses.¹⁵ Meanwhile, the study on volatile-N and char-N during pulverized coal combustion in the O₂/CO₂ atmosphere at a high temperature¹⁶ helped obtain the fuel nitrogen conversion mechanism. Previous research on NO_x emission mainly focused on the total yield of NO_x during combustion, whereas the relevance between NO_x release and combustion process has not been discussed in detail. The NO_x release properties versus time make sense, but it is insufficient to assess the NO_x emission during actual combustion due to the uncertainty of burnout time, fuel amount, and airflow velocity. However, the carbon conversion proceeds throughout the entire combustion process and represents the combustion status to a great extent. The carbon conversion ratio always ranges from 0 to 100%, which makes the carbon conversion ratio become an excellent medium for connecting actual combustion with bench-scale tests. Therefore, the NO_x emission versus time and carbon conversion process is indispensable to a better understanding of NO_x emission performance during combustion, which also contributes to more effective and accurate measures for initial NO_x control during combustion.

The mass loss, residue evolution, and NO_x emission properties during combustion were of great importance to the study on combustion of SS and SD and have been discussed in this paper. The mass loss and gas release properties of SS, SD, and their blends in different proportions were investigated by TG-MS analysis under nonisothermal conditions. FTIR analysis was used to investigate the residue evolution characteristics of SS and SD during the heating process in an air atmosphere. Isothermal experiments in a tube furnace were also performed to study NO_x emission properties with the process of time and carbon conversion during combustion. Therefore, a clear understanding of the cocombustion process was obtained, which contributed to evaluating the feasibility of cocombustion of SS and SD, achieving initial NO_x control and eventually finding a practical, secure, and optimized utilization method for these two resources. This study will also provide an important reference for industrial applications of SS and biomass.

2. Materials and Methods

2.1. Materials

SS and SD were obtained from a municipal wastewater treatment plant in Yantai (Shandong Province, China) and a poplar wood plant in Hanzhong (Shaanxi Province, China), respectively. Both SS and SD were dried at 105 °C for 24 h in an oven. The dried SS and SD were milled and sieved into 75–147 μm in diameter, and samples for each test were well ground evenly in a mortar by definite mass proportions. Three types of mixtures with SD mass proportions of 50, 70, and 90 wt % were named as SS5SD5, SS3SD7, and SS1SD9, respectively.

The proximate (GB/T 28731-2012) and ultimate (GB/T 28728-2012, GB/T 28732-2012, and GB/T 28734-2012) analysis results and lower heating values (LHVs, GB/T 30727-2014) of samples are presented in Table 1. As shown in Table 1, there were distinct differences between the compositions of SS and SD. With the increase in SD proportion, the average combustible material content increased, the average ash content and nitrogen content decreased, and the fuel properties were remarkably affected. The XRF results of ash at 550 °C are shown in Table 2. SS ash showed remarkable Ca and Fe contents, and the K and Ca were the significant components in SD ash. All three blends presented similar ash compositions to SS due to the much higher ash content of SS than SD.

Table 1. Proximate and Ultimate Analyses and Lower Heating Values (LHVs)^a.

	proximate analysis (wt %)				ultimate analysis (wt %)
samples	Mad	Aad	Vad	FCad	Cad	Had	Oad	Nad	Sad	LHV (MJ/kg)
SS	7.2	47.1	40.5	5.2	22.9	3.5	14.1	4.2	0.97	9.8
SS5SD5	5.2	24.6	59.6	10.6	34.3	5.1	27.6	2.5	0.64	13.5
SS3SD7	4.4	15.7	67.2	12.7	38.9	5.7	33.0	1.8	0.50	15.0
SS1SD9	3.6	6.8	74.8	14.8	43.4	6.3	38.4	1.1	0.36	16.5
SD	3.2	2.3	78.6	15.9	45.7	6.7	41.1	0.8	0.30	17.2

^aNotes: ad is air-dried basis; M is moisture; A is ash; V is volatile matter; FC is fixed carbon.

Table 2. Ash Compositions of Samples at 550 °C (wt %).

samples	Ca	K	Na	Mg	Fe	Al	Zn	Si	P	S	C	O
SS	16.21	1.54	1.21	5.11	6.60	5.72	0.72	5.84	3.49	1.89	3.10	48.3
SS5SD5	16.09	2.61	1.20	4.97	6.31	5.47	0.69	5.61	3.39	1.83	3.05	48.5
SS3SD7	15.95	3.90	1.19	4.81	5.97	5.17	0.65	5.34	3.27	1.76	2.99	48.7
SS1SD9	15.44	8.58	1.16	4.22	4.71	4.08	0.53	4.34	2.82	1.50	2.76	49.5
SD	13.72	24.61	1.03	2.23	0.42	0.34	0.11	0.93	1.31	0.63	1.99	52.3

2.2. Experimental Apparatus and Procedure

In this study, TG-MS tests, FTIR analysis, and tube furnace experiments were conducted, and the schematic diagram of the experimental system is shown in Figure 1.

Figure 1.

Schematic diagram of the experimental system.

TG-MS tests of SS, SD, and their blends were conducted by the Thermo Mass Photo instrument. For each test, 5 mg of sample was put into an Al₂O₃ crucible and heated in the TG-MS apparatus at a heating rate of 10 °C/min from room temperature to 1000 °C with a 21 vol % O₂/79 vol % He flow rate of 100 mL/min. The mass loss characteristics were obtained by the TG results, and the release properties of gas products were obtained by the corresponding ion current intensity signals from MS.

To obtain SS and SD residues at specified temperatures, SS and SD were put into a tube furnace (GSL 1200X) with an airflow rate of 1 L/min, and then, the temperature increased from room temperature to preset temperatures (200, 300, 400, 500, 600, 700, 800, and 900 °C) at a heating rate of 10 °C/min. Residues were taken out and cooled with liquid nitrogen immediately when the tube furnace reached the target temperatures. SS, SD, and their residues were analyzed using a Fourier transform infrared spectrometer (Nicolet iS50 FTIR). In addition, to study the NO_x emission properties of SS, SD, and their blends at various combustion temperatures, 0.1 g of samples was put into the tube furnace, which was preheated to 800, 900, and 1000 °C, respectively. Air with a flow rate of 1 L/min was injected into the furnace to provide the necessary oxygen for combustion. A flue gas analyzer (GA-21plus) was used to monitor the O₂, CO₂, CO, NO, and NO₂ concentrations during the whole combustion process. This study was focused on the NO_x yields and the correlation between NO_x release and carbon conversion.

To investigate potential interactions during cocombustion of SS and SD, the experimental and theoretical values for specific parameters, such as TG, NO_x yield, and the ratio of fuel nitrogen to NO_x, were also compared in this study.

3. Results and Discussion

3.1. Combustion Characteristics of SS and SD

3.1.1. TG-MS Analysis

The TG, DTG, and DSC curves of SS and SD at a heating rate of 10 °C/min under a 21 vol % O₂/79 vol % He atmosphere are presented in Figure 2. According to the MS results, strong ion intensity signals of m/z of 18 (H₂O⁺), 28 (CO⁺), and 44 (CO₂⁺) were detected during the heating process. The DTG curves and main gas release curves of SS and SD during the heating process are demonstrated in Figure 3.

Figure 2.

TG, DTG, and DSC curves of (a) SS and (b) SD.

Figure 3.

DTG and gas release curves of (a) SS and (b) SD.

As exhibited in Figures 2a and 3a, after a slight weight loss of 7.35% between 40 and 180 °C by the removal of moisture, SS mainly experienced three weight-loss stages during the heating process. A dramatic weight loss of 31.40% between 180 and 400 °C (Stage I) was visible along with the distinct evolution of H₂O, CO₂, and CO corresponding to the devolatilization and combustion of most volatiles. During Stage I, the release of H₂O was mainly related to thermal decomposition reactions (mainly dehydration) and the oxidation of hydrogen in fuel; the CO evolution was mainly caused by the decomposition of labile compositions like alkyl aryl ethers and the oxidization of light hydrocarbyl groups.¹⁷ CO₂ was mainly released from the decomposition of carboxyl and carboxylate groups,¹⁸ as well as the oxidization of carbon-containing compounds. Then, there was a weight loss of 12.81% between 400 and 650 °C (Stage II) with remarkable CO₂ evolution, which mainly resulted from the combustion of residual organics with low reactivity and fixed carbon. Meanwhile, the slight CO yield during Stage II mainly resulted from the incomplete oxidization of carbon. The high ash content of SS decelerated the devolatilization, resulting in a much higher weight loss during Stage II (12.81%) than the fixed carbon content (5.2%) in Table 1. Finally, calcium carbonate decomposition resulted in a weight loss of 9.42% between 650 and 750 °C (Stage III) with abundant CO₂ yield. SS presented a lower final residual mass fraction (m_b) of 38.33% than the ash content of 47.1% (Table 1) mainly due to carbonate decomposition of Stage III.

As shown in Figures 2b and 3b, SD experienced a slight weight loss of 2.33% between 40 and 160 °C with a slight H₂O release peak by the removal of moisture; a sharp weight loss of 74.62% between 220 and 400 °C (Stage I) with remarkable evolution of H₂O, CO₂, and CO corresponding to the devolatilization and combustion of most volatiles; and finally an apparent weight loss of 20.21% between 400 and 500 °C (Stage II), mainly due to the combustion of fixed carbon along with a significant CO₂ yield. The m_b of SD (1.93%) was slightly lower than the corresponding ash content (2.3%) shown in Table 1 due to ash transformation at high temperatures.

The positive and negative values of DSC in Figure 2 represent endothermic and exothermic conditions, respectively. For both SS and SD, the removal of moisture below 180 °C was an endothermic process, and the combustion of volatiles and char corresponded to the distinct exothermic peaks of Stage I and Stage II, respectively. Additionally, the decomposition of CaCO₃ during Stage III of SS was an endothermic process. With the temperature further increasing, both SS and SD presented an endothermic process, indicating the existence of mineral transformation in ash. During the whole combustion process, the heat release of SS was 52% less than that of SD, which was consistent with the different calorific values as shown in Table 1. In addition, SS presented a higher heat release ratio of Stage I/Stage II (2.09) than SD (0.90), which mainly resulted from the higher fixed carbon /volatile ratio of SD.

SS and SD presented similar gas evolution properties during Stage I and Stage II. During Stage I, the evolution of H₂O, CO, and CO₂ of SD was mainly caused by the degradation and combustion of hemicellulose, cellulose, and lignin, and the release of H₂O, CO, and CO₂ of SS was mainly due to the decomposition and combustion of lipid, protein, and carbohydrate. SD presented a stronger intensity of H₂O and CO during Stage I than SS due to its higher volatile and hydration contents, suggesting a stronger local reducing atmosphere during devolatilization. During Stage II, the evolution of CO and CO₂ of both SD and SS was mainly due to the combustion of fixed carbon (C-rich materials by aromatic cyclization and graphitization). However, the complex compositions and high ash content in SS resulted in a wider temperature zone of decomposition and higher char combustion temperatures. In addition, for both SS and SD, the conversion of hydrogen to H₂O mainly took place during Stage I, whereas the conversion of carbon to CO₂/CO took place throughout the whole combustion process, which made it reasonable to evaluate the combustion status by the conversion ratio of carbon.

3.1.2. FTIR Analysis

To study the residue evolution properties of SS and SD, FTIR analyses of SS, SD, and their residues at various temperatures were conducted. The mid-infrared spectra of the original SS and SD are shown in Figure 4, and the mid-infrared spectra of residues at multiple temperatures are illustrated in Figure 5. The characteristic absorption bands observed in Figures 4 and 5 are presented in Table 3.¹⁹⁻²²

Figure 4.

Mid-infrared spectra of (a) SS and (b) SD.

Figure 5.

Mid-infrared spectra of (a) SS residues and (b) SD residues.

Table 3. Characteristic Absorption Bands Observed in the Mid-Infrared Spectra.

band (cm–1)	assignments
3600–3000	O–H stretching of hydroxyl groups or water
3000–2800	C–H stretching of aliphatic methylene or alkyl
2280–2210	C≡C stretching, C≡N stretching
1750–1700	C=O stretching of aldehydes, carboxylates, esters, ketones
1680–1640	C=O stretching of carboxylates (amide I band)
1597, 1509, 1375	C=C stretching of aromatic rings, alkenes
1570–1540	N–H in-plane (amide II band)
1510–1390	CO32− stretching of carbonates
1470–1350	C–H deformation of aliphatic compounds
1335–1240	C–O stretching of carboxylates, phenols, ethers
1160	C–O stretching of esters
1100–1000	C–OH stretching of carbohydrates
1080	Si–O stretching of silicates
1058	C–O stretching, glycosidic linkages between sugar units
1035	Si–O stretching of silicates
898	C–O–C stretching of glucose ring (cellulose)
872	CO2-3 out-of-plane bending of carbonates
712	CO32− in-plane bending of carbonates
668	O–H out-of-plane bending in C–OH alcoholic groups
462	Fe2O3

From the SS spectrum presented in Figure 4, there were abundant O–H (3600–3000 cm^–1) and C–H (2800–3000 cm^–1) bonds in SS, indicating the presence of considerable alcohol and aliphatic compounds. The distinct absorption peak at 1650 cm^–1 (C=O stretching) corresponded to carboxylates in lipid and protein. The peak at 1543 cm^–1 corresponded to the overlap of the C=C stretching vibration and the N–H in-plane bending vibration (amides II band). These amide structures mainly originated from abundant DNA and RNA of animal and plant cells and microorganisms in SS. In addition, the high ash content of SS resulted in the distinct band overlap of organic functional groups and inorganic components.¹⁹ The intense peaks at 1456 and 1417 cm^–1 were mainly owing to the combination of the CO_3²⁻ stretching vibration and the C–H deformation vibration, and the peaks at 1080 and 1035 cm^–1 corresponded to the combination of the Si–O stretching vibration and the C–O stretching vibration. The SS spectrum indicated the main organic components of lipid, protein, and carbohydrate²² and the main inorganic components of silica oxides (1080 and 1035 cm^–1), carbonates (1456, 1417, 872, and 712 cm^–1), and iron oxides (462 cm^–1).

From Figure 5a, SS mainly experienced organic evolution from 200 to 600 °C and inorganic evolution from 600 to 800 °C. The desorption of external water mainly occurred below 200 °C. With the temperature increasing from 200 to 400 °C, the quantities of O–H, C–H, and C–O decreased significantly by bond breaking²³ with the release of H₂O, CO₂, and CO, as shown in Figure 3a, and the cleavage of intramolecular H-bonds, the subsequent dehydration, and complex condensation reactions led to the structure rearrangement of carbon (aromatic cyclization and graphitization),²⁴ which resulted in the stronger absorption of C=C stretching (1597 cm^–1) and the absorption of C≡C and C≡N stretchings (2215 cm^–1) at 400 °C. When the temperature reached 600 °C, most amounts of C=C (1597 cm^–1) and C=O (1650 cm^–1) bonds were consumed by oxidation with the release of CO₂, and the residue presented significant features of ash. Then, carbonates (1453, 1417, 872, and 712 cm^–1; mainly CaCO₃) decomposed between 600 and 800 °C. The relative amounts of carbonates increased with the consumption of organics in SS and then decreased by decomposition.

From the SD spectrum in Figure 4, the intense absorption bands of O–H stretching (3600–3000 cm^–1), C–H stretching (2800–3000 cm^–1), and C=O stretching (1750–1640 cm^–1) were visible, indicating the presence of alcohol, aliphatic compounds, ketones, or short-chain carboxylic acids in SD. Peaks at 1597, 1509, and 1375 cm^–1 (C=C stretching vibration) corresponded to aromatic rings from lignin compounds. Peaks between 1470 and 1350 cm^–1 were due to the C–H deformation vibration of cellulose and lignin. Peaks between 1300 and 1100 cm^–1 resulted from the C–O stretching vibration of carboxylic acids, phenols, or ethers. Peaks between 1100 and 1000 cm^–1 (C–OH stretching vibration) indicated the presence of carbohydrates like cellulose and hemicellulose. SD was also characterized by the peaks at 1058 and 898 cm^–1 corresponding to O-glycosidic linkages between sugar units and glucose rings. The FTIR results verified the three main components of hemicellulose, lignin, and cellulose of SD.²⁵

From Figure 5b, SD mainly experienced similar residue evolution properties to SS, including the organic evolution from 200 to 500 °C and inorganic evolution from 500 to 800 °C. After the removal of external moisture below 200 °C, the amounts of O–H, C–H, and C–O–C of the glucose ring and O-glycosidic linkages between sugar units gradually decreased by decomposition and oxidization;²³ meanwhile, complex condensation and rearrangement reactions took place below 400 °C with the generation of C-rich compounds and the release of H₂O and CO₂, as shown in Figure 3b, resulting in the stronger C=C vibration (1597 cm^–1) and C≡C vibration (2215 cm^–1) at 400 °C. Compared to SS, SD exhibited a lower burnout temperature of about 500 °C, during which C=C (1597 cm^–1) and C=O (1737 cm^–1) got nearly exhausted, and strong inorganic peaks (Si–O stretching at 1088 and 1035 cm^–1, CO₃²⁻ vibration at 1450 and 872 cm^–1) became visible. The carbonates (mainly CaCO₃ and K₂CO₃) gradually decomposed with the further increase in temperature from 600 to 800 °C.

3.2. Combustion Characteristics of Blends

The TG, DTG, and DSC curves of all samples as well as the experimental and theoretical TG curves of blends are shown in Figure 6. The TG, DTG, and DSC curves of blends exhibited a significant correlation with the blending ratio.

Figure 6.

(a) TG, (b) DTG, (c) DSC, and (d) theoretical TG curves of blends (E, experimental value; T, theoretical value).

Blends presented three main weight-loss stages during the heating process: Stage I (devolatilization and combustion of most volatiles), Stage II (combustion of fixed carbon and residual volatiles), and Stage III (decomposition of carbonates), and the details are shown in Table 4. Blends presented higher initial temperatures of Stage I and lower final temperatures of Stage II than SS due to the addition of SD. With the SD proportion increasing, the weight loss of Stage I increased significantly, whereas the heat release ratio of Stage I/Stage II decreased, which was mainly due to the remarkable increase in the calorie-rich fixed carbon. Meanwhile, blends with higher SD proportions presented lower weight-loss ratios of Stage III due to the lower carbonate contents, and the increased heat release during combustion resulted in the earlier decomposition of carbonates during Stage III.

Table 4. Weight-Loss Stages of SS, SD, and Blends during the Heating Process.

	stage I			stage II			stage III
samples	T (°C)	TG (%)	DSC (%)	T (°C)	TG (%)	DSC (%)	T (°C)	TG (%)
SS	180–400	31.40	67.59	400–650	12.81	32.43	650–750	9.42
SS5SD5	200–400	49.72	57.12	400–620	18.93	42.94	620–750	5.45
SS3SD7	200–400	60.11	53.80	400–600	20.63	46.22	600–750	3.85
SS1SD9	220–400	68.15	50.11	400–600	19.32	49.91	600–750	1.64
SD	220–400	74.62	47.49	400–500	20.21	52.50

The combustion parameters were also calculated by the TG-DTG curves and are presented in Table 5. T_max is the temperature with the maximum mass loss rate; the ignition temperature (T_i) is defined as the intersection of the tangent at T_max and the horizontal curve at the temperature starting to lose weight concerning the TG curve; and the burnout temperature (T_b) is the temperature where the mass loss of samples reached 98% of the total mass loss. The combustibility index S is used as a criterion for fuel combustion characteristics. A bigger S indicates a better combustion characteristic, and S is defined as follows²⁶

1

where DTG_max denotes the maximum mass loss rate (%/min), DTG_mean is the average mass loss rate (%/min), T_i is the ignition temperature (K), and T_b is the burnout temperature (K).

Table 5. Combustion Characteristic Parameters of SS, SD, and Their Blends^a,b,c.

samples	Ti	Tb	mb	T1	DTG1	T2	DTG2	T3	DTG3	DTGmean	S × 10–8
	(°C)	(°C)	(%)	(°C)	(%/min)	(°C)	(%/min)	(°C)	(%/min)	(%/min)	(K–3·min–2)
SS	246	734	38.33	332	6.55	395	1.07	717	1.63	1.14	2.75
SS5SD5	247	716	21.51	340	15.05	449	1.42	696	0.95	1.55	8.72
SS3SD7	254	701	14.21	340	21.54	457	1.79	690	0.58	1.81	14.44
SS1SD9	262	547	6.15	343	25.06	480	2.39	670	0.35	3.05	32.60
SD	267	505	1.93	341	30.67	468	3.40			3.71	50.15

^aNotes: T₁, T₂, and T₃ are the temperatures according to DTG peaks of Stage I, Stage II, and Stage III, respectively.

^bDTG₁, DTG₂, DTG₃ are the DTG values at T₁, T₂, and T₃, respectively,.

^cDTG_mean is the average mass loss rate during the whole mass loss process.

As shown in Table 5, SS exhibited a lower T_i but much higher T_b and m_b than SD, and the DTG₁ and DTG_mean of SS were much smaller than those of SD due to their different components. The index S of SS (2.75 × 10^–8 K^–3·min^–2) was much smaller than that of SD (5.015 × 10^–7 K^–3·min^–2), which indicated that SD had better combustion characteristics than SS.

With the increase in SD proportion, m_b of blends decreased proportionally, and both DTG₁ and DTG_mean increased due to the increasing volatile content and combustible material content. Meanwhile, T_i increased slightly and T_b decreased significantly due to the higher T_i and lower T_b of SD. The increasing S indicated that blends with higher SD proportions exhibited better combustion properties.

In addition, the comparison of experimental and theoretical TG curves of blends is shown in Figure 6d. The theoretical values are calculated as follows

2

where TG_SD and TG_SS are the experimental TG values of SD and SS, respectively, and p_SD and p_SS are the mass proportions of SD and SS in blends, respectively.

The experimental TG curves before 340 °C were in good agreement with theoretical curves, which suggested that the decomposition of labile compounds was barely influenced by cocombustion. Nonetheless, the experimental weight-loss ratios were smaller than the theoretical values between 350 and 500 °C for all blends, especially for SS5SD5. During cocombustion, SS ash slowed down the devolatilization and char combustion of SD by decelerating heat transfer and oxygen diffusion, and therefore, more time or a higher temperature was required for the combustion process, which was undesirable in practical applications. In addition, blends presented higher burnout ratios than theoretically, and the experimental m_b values of SS5SD5, SS3SD7, and SS1SD9 were higher than the calculated values by 1.23, 1.38, and 0.56%, respectively. During cocombustion, volatiles in SD were released rapidly to generate loose structures with abundant pores in char, which promoted the oxygen diffusion in SS char and contributed to the higher burnout ratios of blends. In addition, the higher calorific value of SD increased the interior temperature of residues, which also promoted burnout and ash decomposition. On the whole, cocombustion of SS and SD showed positive effects on improving fuel properties, and SS3SD7 exhibited optimum interactions during combustion.

The gas release curves of blends during the heating process are demonstrated in Figure 7. Three blends presented similar evolution properties of H₂O, CO, and CO₂ to individual SS and SD. Most H₂O was released during Stage I, while CO and CO₂ presented considerable release yields during both Stage I and Stage II, which suggested that the conversion of hydrogen of blends mainly took place during Stage I and the conversion of carbon took place throughout the whole combustion process. In addition, the intensity of both H₂O and CO₂ got stronger with the SS proportion increasing.

Figure 7.

DTG curves and gas release curves of (a) SS5SD5, (b) SS3SD7, and (c) SS1SD9.

The carbon in the fuel was continuously converted to flue gas during the whole combustion process, and the carbon conversation ratios of SS, SD, and blends during different weight-loss stages are calculated as follows and illustrated in Table 6

3

where α_C1, α_C2, and α_C3 denote the carbon conversion ratio during Stages I, II, and III, respectively; m_C1, m_C2, and m_C3 denote the relative carbon amounts in both CO and CO₂ released during Stages I, II, and III, respectively; and m_C denotes the total relative carbon amount in both CO and CO₂ released during the whole heating process. The relative carbon amount in CO and CO₂ is obtained by the line integral of CO and CO₂ release curves versus temperature.

Table 6. Carbon Conversion Ratios of Samples during Different Weight-Loss Stages.

samples	SS	SS5SD5	SS3SD7	SS1SD9	SD
αC1/%	55.9	50.7	49.2	48.3	46.9
αC2/%	34.4	43.7	48.7	51.1	53.1
αC3/%	9.7	4.3	2.1	0.6
αC2/(αC1+αC2)	0.381	0.457	0.497	0.514	0.531
FCad/Cad	0.227	0.309	0.326	0.341	0.348

With the increase in SD proportion, α_C1 and α_C2 gradually increased, which was mainly due to the increasing volatile contents and fixed carbon contents in samples. The α_C2/(α_C1+α_C2) ratios were higher than the FC_ad/C_ad ratios for all samples, which demonstrated that there was a distinct overlap between the combustion of volatiles and fixed carbon. In addition, samples with higher SD proportions presented a much lower α_C3 due to the much lower carbonates in blends.

3.3. NO_x Emission Properties

The NO_x emission during combustion mainly resulted from three sources: fuel NO_x, thermal NO_x, and prompt NO_x,²⁷ but the thermal NO_x and prompt NO_x were mainly generated above 1500 °C. Therefore, the NO_x in this study was considered as fuel NO_x, which came from nitrogen in the fuel. A flue gas analyzer was used to monitor the concentrations of O₂, CO₂, CO, NO, and NO₂ during the isothermal combustion in a tube furnace, and NO_x concentration (c(NO_x)) was considered as the total concentration of NO and NO₂. According to the monitoring results, the CO concentration was maintained below 50 mg/m³ during the whole combustion process, indicating the complete combustion of samples.

During volatile combustion, fuel nitrogen was partly transformed to NO_x or N₂ by complex oxidation–reduction reactions and generated heterocyclic-N, amine-N, and nitrile-N compounds,²⁸ which were further released during char combustion. The NO_x could be roughly divided into volatile-NO_x released during the combustion of volatile and char-NO_x released during char combustion.¹⁶ To obtain the correlation between NO_x release and the combustion process, the conversation ratio of carbon in fuel to CO₂ (α_C) was studied, and α_C as a function of time t is calculated as follows

4

where m_t(CO₂) denotes the amount of CO₂ released before time t, and m₀(CO₂) denotes the total amount of CO₂ released during the whole combustion process. The amount of CO₂ is obtained by the line integral of CO₂ release curves versus time.

3.3.1. NO_x Emission Properties of SS and SD

The c(NO_x) and α_C as a function of time during the combustion of SS and SD at 800, 900, and 1000 °C are demonstrated in Figure 8. With the process of time, the carbon conversion proceeded accompanied by the successive release of volatile-NO_x and char-NO_x. With the temperature increasing, the carbon conversion of both SS and SD was accelerated, and the acceleration effect on SS was stronger than that on SD. SS always presented a much higher volatile-NO_x peak value than SD at different temperatures due to its high nitrogen content, and both SS and SD presented a much stronger volatile-NO_x release peak and a less distinct char-NO_x release peak, indicating the much higher amounts of volatile-NO_x than char-NO_x. With the temperature increasing from 800 to 1000 °C, the enhanced devolatilization contributed to an earlier volatile-NO_x peak with a higher peak value, and the release of char-NO_x also took place much earlier due to the acceleration of char combustion. The boundary between volatile-NO_x and char-NO_x became weak due to the stronger overlap of devolatilization and char combustion at higher temperatures.

Figure 8.

NO_x release curves and α_C-change curves during combustion of (a) SS and (b) SD.

To investigate the specific correlation between NO_x emission and the combustion process, the absolute NO_x yield per gram of sample (m(NO_x), mg/g) as a function of α_C and its derivative with respect to α_C (dm(NO_x)/dα_C, mg/g, representing the NO_x release amount per unit carbon conversion) are illustrated in Figure 9. Particularly, the total carbon could be divided into volatile carbon and fixed carbon according to Table 1, which contributed to the definition of an imaginary boundary (the dotted lines in Figure 9) between volatile-NO_x and char-NO_x.

Figure 9.

NO_x yield versus α_C during combustion of (a) SS and (b) SD.

As shown in Figure 9a, the dm(NO_x)/dα_C of SS presented a declining trend during volatile combustion (the α_C range of 0–70%) for all temperatures, indicating that more volatile nitrogen released during the earlier carbon conversion stage, whereas the dm(NO_x)/dα_C of SD in Figure 9b changed slightly during volatile combustion (the α_C range of 0–65%) for all temperatures, indicating that the volatile-NO_x of SD was released more uniformly. The nitrogen-containing materials in SS became more independent during the wastewater treatment process and were more likely to decompose with the temperature rising, while nitrogen in SD was mainly uniformly distributed in certain forms and tended to release accompanied by the consumption of volatiles.

The volatile-NO_x release properties varied with temperature. SS presented similar volatile-NO_x yields at 800 and 900 °C but exhibited dramatically lower volatile-NO_x yields at 1000 °C, whereas the volatile-NO_x yields of SD continued to decrease with the increase in temperature. The increasing temperature enhanced the reducing atmospheres of CO, H₂, and hydrocarbyls by the acceleration of devolatilization, which significantly promoted the conversion from fuel nitrogen to N₂ and contributed to volatile-NO_x reduction. In addition, the high volatile content enhanced the effects of temperature on devolatilization, and therefore, SD presented a wider volatile-NO_x-related temperature range.

Temperature also had impacts on char-NO_x of both SS and SD. There was a significant increase in dm(NO_x)/dα_C during the char combustion of SS at both 800 and 900 °C, whereas the corresponding dm(NO_x)/dα_C at 1000 °C exhibited a declining trend and presented much lower values than that at 800 and 900 °C. SS presented similar and considerable char-NO_x yields at 800 and 900 °C but showed a much lower char-NO_x amount at 1000 °C. The abundant metal oxides in SS such as CaO and Fe₂O₃ as shown in Table 2 had positive effects on the reduction of NO_x precursors by generating FeN_x and CaC_xN_y, and the transformation of fuel nitrogen to N₂ was also enhanced by Ca₂Fe₂O₅, especially at higher temperatures.^29,30 High temperature also accelerated the generation of CaO by carbonate decomposition and contributed to NO_x reduction. Meanwhile, SS ash with a high slagging potential^31,32 enhanced the fixing of nitrogen in char with the fusion of eutectic mixtures¹⁵ at 1000 °C. Therefore, combustion at 1000 °C presented much less char-NO_x yield.

Nonetheless, temperature had different impacts on char-NO_x of SD. The corresponding dm(NO_x)/dα_C of char-NO_x increased moderately during the whole char combustion process at 800 °C but rose sharply during the last char combustion stage after moderate increases at 900 and 1000 °C. The char-NO_x during the last 10% carbon conversion process increased remarkably with the temperature increasing, and SD presented higher char-NO_x yields at 900 and 1000 °C than at 800 °C. The increasing temperature enhanced the condensation process of combustible materials in SD and probably promoted the transfer of fuel nitrogen into char, which increased the char-NO_x release of SD.

The total NO_x yield and the conversion ratio of fuel nitrogen to NO_x (α_N) of SS and SD at different temperatures are shown in Figure 10, and α_N is calculated as follows

5

where m_N denotes the nitrogen amount in NO_x released during the whole combustion process and m_N0 denotes the total nitrogen amount in the fuel.

Figure 10.

Total NO_x yields and α_N of SS and SD at different temperatures.

The higher nitrogen content of SS as shown in Table 1 led to the higher NO_x yields of SS to a great extent, and the NO_x yields of SD at 800, 900 and 1000 °C were 27.0, 24.9, and 26.4% of that of SS, respectively. Nonetheless, the α_N values of SS at 800, 900, and 1000 °C were smaller than those of SD by 2.29, 1.72, and 1.70%, respectively. The high ash content of SS slowed down the oxygen diffusion and provided abundant metal oxides with positive nitrogen reduction effects, both of which contributed to the smaller α_N of SS. Furthermore, the ash slagging of SS at high temperatures also fixed more nitrogen in ash. In addition, the total NO_x yield of SD continued to decrease from 800 to 1000 °C, whereas the total NO_x yield of SS experienced a tiny increase from 800 to 900 °C and a remarkable decrease from 900 to 1000 °C, which suggested that the NO_x emission of higher volatile fuels was more sensitive to temperature.

3.3.2. NO_x Emission Properties of Blends

The composition of samples changed with the blending ratio, which had important impacts on NO_x emission properties during combustion. The NO_x release curves versus time, and NO_x yield and dm(NO_x)/dα_C curves versus carbon conversion ratio at 800, 900, and 1000 °C are demonstrated in Figure 11.

Figure 11.

NO_x release curves and NO_x yields versus α_C of blends: (a), (b), and (c) at 800 °C; (d), (e), and (f) at 900 °C; (g), (h), and (i) at 1000 °C.

The NO_x emission varied remarkably with the blending ratio. As shown in Figure 11, with the SD proportion increasing, the volatile-NO_x concentration reduced significantly, and the total NO_x yield decreased for all temperatures owing to the decreasing nitrogen content. However, the blending ratio exhibited uncertain effects on char-NO_x emission at different temperatures.

Blends with higher SD proportions presented lower dm(NO_x)/dα_C during char combustion and released less char-NO_x at both 800 and 900 °C. The char-NO_x peaks were visible at 800 °C for all blends as shown in Figure 11c, indicating the existence of a concentrated char-NO_x release process, whereas all blends exhibited relatively steady char-NO_x release during the whole char combustion process at 900 °C as shown in Figure 11f. In addition, at 1000 °C, the dm(NO_x)/dα_C during the early and middle stages of char combustion (the α_C range of 60–85%) decreased with the SD proportion increasing, but blends with higher SD proportions presented higher dm(NO_x)/dα_C during the last char combustion process (the α_C range of 90–100%), as shown in Figure 11i, because the char-NO_x of SD tended to release during the last char combustion stage at 1000 °C.

Furthermore, to investigate the effects of temperature on NO_x emission for a specific blend, the NO_x yields and dm(NO_x)/dα_C versus α_C of three blends at different temperatures are illustrated in Figure 12.

Figure 12.

NO_x yield versus α_C during the combustion of (a) SS5SD5, (b) SS3SD7, and (c) SS1SD9.

Both SS5SD5 and SS3SD7 presented similar volatile-NO_x changing rules with temperature to SS, and SS1SD9 exhibited similar volatile-NO_x changing properties with temperature to SD. The dm(NO_x)/dα_C during volatile combustion of SS5SD5 and SS3SD7 decreased slightly from 800 to 900 °C and then decreased significantly from 900 to 1000 °C. However, the volatile-NO_x of SS1SD9 continued to decrease dramatically from 800 to 1000 °C. The increase in volatile content strengthened the effects of temperature increase on volatile-NO_x release due to the enhanced local reducing atmosphere.

With the increase in temperature, the release of char-NO_x of three blends tended to move toward the earlier carbon conversion process, suggesting that the nitrogen in char was more sensitive to temperature than fixed carbon and the release of char-NO_x proceeded earlier than char combustion. The char-NO_x amount of SS5SD5 changed slightly from 800 to 900 °C and then decreased from 900 to 1000 °C, presenting similar properties to SS. The decrease of char-NO_x from 900 to 1000 °C was mainly due to the severer fusion of eutectic mixtures and the relatively stronger reduction effects of metal oxides in SS at 1000 °C. However, the char-NO_x yields of SS1SD9 continued to increase from 800 to 1000 °C, presenting similar properties to those of SD due to its high SD proportion, and the char-NO_x yield of SS3SD7 increased from 800 to 900 °C and changed slightly from 900 to 1000 °C, which was affected by both SS and SD.

3.3.3. Synergistic Effects on NO_x Emission

To evaluate the synergistic effects of cocombustion, the experimental and theoretical values of the total NO_x yield and α_N of three blends at different temperatures are compared in Figure 13. The theoretical NO_x yield (m(NO_x)_T) during cocombustion is calculated as follows

6

where m(NO_x)_SD and m(NO_x)_SS are the experimental values of m(NO_x) for SD and SS, respectively; and p_SD and p_SS are the mass proportions of SD and SS in blends, respectively. The theoretical α_N is calculated by eq 5 using the theoretical m_N corresponding to m(NO_x)_T.

Figure 13.

(a) Total NO_x yields and (b) α_N of blends (E, experimental value; T, theoretical value).

Compared to theoretical values, the NO_x yield of SS5SD5 decreased by 4.9, 5.7, and 5.5% at 800, 900, and 1000 °C, respectively, and the corresponding α_N decreased by 0.29, 0.34, and 0.26% at 800, 900, and 1000 °C, respectively. Cocombustion influenced the local atmosphere during devolatilization and had positive impacts on reducing volatile-NO_x from SS but negative impacts on volatile-NO_x reduction from SD. The NO_x from SS took up the vast majority of the total NO_x of SS5SD5, and therefore SS5SD5 presented an overall volatile-NO_x reduction effect. Meanwhile, compared to individual SS, the more fixed carbon (mainly from SD) strengthened the reducibility of char, and the SS ash in SS5SD5 decelerated the char combustion of SD and lengthened the lifetime of fixed carbon, both of which contributed to the reduction of NO_x from SS. In addition, the abundant metal oxides in SS ash promoted the reduction of NO_x from SD.

The NO_x yield of SS3SD7 decreased by 3.7, 4.2, and 4.8% compared to the theoretical values at 800, 900, and 1000 °C, respectively, and the α_N decreased by 0.23, 0.26, and 0.24% at 800, 900, and 1000 °C, respectively. During the combustion of SS3SD7, the NO_x from SS and SD presented approximate amounts. Compared to individual SS, the higher volatile content of SS3SD7 generated a stronger reducing atmosphere and promoted volatile-NO_x reduction from SS; the fixed carbon from SD and the considerable SS ash also contributed to NO_x reduction; meanwhile, the high SD proportion of SS3SD7 led to a much looser char structure, which enhanced oxygen diffusion and promoted NO_x emission. With the comprehensive effects of various factors, SS3SD7 presented a positive NO_x reduction performance but the synergistic effects were slightly weaker than those of SS5SD5.

Compared to theoretical values, the NO_x yield of SS1SD9 increased by 5.1% at 800 °C and decreased by 3.7% and 6.7% at 900 and 1000 °C, respectively, and α_N presented an increase of 0.36% at 800 °C but decreased by 0.25 and 0.38% at 900 and 1000 °C, respectively. During the combustion of SS1SD9, the NO_x from SD took up the majority of the total NO_x. SS1SD9 presented a much looser structure than SS during combustion, which promoted the oxidation of fuel nitrogen, although SS1SD9 presented a much stronger reducing atmosphere during devolatilization than individual SS. Meanwhile, the small amount of SS ash played little role in NO_x reduction. Therefore, SS1SD9 presented negative NO_x reduction effects at 800 °C. Nonetheless, the effects of reducing atmosphere on volatile-NO_x reduction became enhanced with an increase in temperature, and SS1SD9 showed a positive NO_x reduction performance at 900 and 1000 °C.

In conclusion, the synergistic effects on NO_x emission during cocombustion were complex and influenced by various factors. In this study, the NO_x reduction effects of SS1SD9 were highly dependent on combustion temperature, and SS5SD5 showed better NO_x reduction performances than SS3SD7 at all temperatures. Nonetheless, the NO_x yield of SS5SD5 remained high and the application of SS5SD5 was also severely restricted by the low calorific value and worse combustion characteristics. Synthetically considering the SS disposal capacity, combustion characteristic, and NO_x yield, SS3SD7 is more recommended in practical applications.

4. Conclusions

The thermal behaviors and NO_x release properties during combustion of SS, SD, and their blends were studied, and the main conclusions are summarized as follows.

• (1)All samples experienced dehydration, volatile combustion, char combustion, and ash transformation during the heating process. The conversion of hydrogen to H₂O mainly occurred during volatile combustion, whereas the conversion of carbon to CO₂/CO proceeded during the whole combustion process.
• (2)SD exhibited better combustion characteristics than SS, and with the SD proportion increasing, blends exhibited better combustion performances. Compared to theoretical conditions, the burnout ratio increased due to the looser structure of SD, but the combustion of char (mainly from SD) was decelerated by SS ash during cocombustion.
• (3)SS emitted more NO_x mainly due to its high nitrogen content than SD but showed a lower α_N owing to its high ash content and high metal oxide content like CaO and Fe₂O₃. Volatile nitrogen in SS existed in less stable forms than that in SD, and the volatile-NO_x release rate versus carbon conversion ratio presented a declining trend for SS but remained relatively stable for SD.
• (4)Samples with higher SD proportions released less NO_x due to the lower nitrogen contents. The increase in temperature was more likely to influence the NO_x emission of higher volatile samples. The NO_x yields of SD and SS1SD9 continued to decrease from 800 to 1000 °C, whereas the NO_x yields of SS, SS5SD5, and SS3SD7 changed slightly from 800 to 900 °C and decreased remarkably from 900 to 1000 °C.
• (5)Cocombustion exhibited synergistic effects on NO_x emission according to the difference between experimental and theoretical NO_x yields. The blending ratio and temperature affected the synergistic effects by affecting the local atmosphere during devolatilization, oxygen diffusion rate, and the reductivity of char. SS5SD5 and SS3SD7 always presented a positive NO_x reduction performance, and SS1SD9 had opposite NO_x reduction effects at different temperatures. SS5SD5 showed better overall NO_x reduction effects, but its NO_x yields remained high due to its high nitrogen content.

Synthetically considering the SS disposal capacity, combustion characteristic, and NO_x yield, an SS proportion around 30% in blends is more recommended in practical applications, and a study on the correlation between NO_x release and carbon conversion can lay the foundation for efficient initial NO_x control methods during combustion. In addition, due to the differences between bench-scale tests and complicated industrial situations, it is suggested to conduct some relevant actual field tests in the future to obtain a more comprehensive correlation between NO_x release and fuel status.

The authors declare no competing financial interest.