Fuel 111 (2013) 201–210

Contents lists available at SciVerse ScienceDirect

Fuel

journal homepage: www.elsevier .com/locate / fuel

Hydrothermal carbonization of sewage sludge for energy productionwith coal

0016-2361/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fuel.2013.04.052

⇑ Corresponding author at: Minerals, Metals and Materials Technology Centre & Department of Civil and Environmental Engineering, National University of SingaEngineering Drive 1, Singapore 117576, Singapore. Tel.: +65 6516 5135; fax: +65 6872 5483.

E-mail address: ceerbala@nus.edu.sg (R. Balasubramanian).

Ganesh K. Parshetti a,b, Zhengang Liu a,b, Akshay Jain c, M.P. Srinivasan c,Rajasekhar Balasubramanian a,b,⇑a Minerals, Metals and Materials Technology Centre, National University of Singapore, EA Engineering Drive 1, Singapore 117576, Singaporeb Department of Civil and Environmental Engineering, National University of Singapore, EA Engineering Drive 1, Singapore 117576, Singaporec Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore 117576, Singapore

h i g h l i g h t s

� Energy rich carbonaceous char wassynthesized by HTC of urban sewagesludge.� The hydrothermally prepared char

was characterized by a suite ofanalytical techniques.� Co-combustion of upgraded sludge

char with raw coal as well asupgraded coal was studied.� The upgraded char can be used for

energy production with minimalgreenhouse gas emission.

g r a p h i c a l a b s t r a c t

a r t i c l e i n f o

Article history:Received 28 January 2013Received in revised form 31 March 2013Accepted 22 April 2013Available online 9 May 2013

Keywords:Hydrothermal carbonizationFuelCo-combustionSludge charTGA–FTIR

a b s t r a c t

Hydrothermal carbonization using subcritical water (250 �C, 8–10 MPa and 15 min reaction time) wasinvestigated to recover solid carbonaceous fuel i.e. sludge char (HT-SL) from urban sewage sludge. Thecarbonaceous HT-SL had an energy–density of 15.82 MJ kg�1. For achieving maximum waste-to-energyconversion, the co-combustion of HT-SL with low rank Indonesian coal (LRIC) and hydrothermallyupgraded LRIC (HT-LRIC) was investigated using a thermogravimetric analyzer (TGA) and the emissioncharacteristics of gaseous pollutants were determined by using coupled Fourier transform infrared spec-troscopy (FT-IR). To gain insights into the physico-chemical and microstructure properties, carbonaceousfuel were characterized by proximate, ultimate, field emission scanning electron microscopy, FT-IR, X-raydiffraction, Brunauer, Emmett and Teller and inductively coupled plasma optical emission spectrometryanalysis. Conventional TGA and kinetic parameters such as activation energy of various LRIC, HT-LRIC andHT-SL blends were also determined. This fundamental study provides a basic insight into co-combustionof HT-SL with LRIC and HT-LRIC, which forms a scientific basis for the efficient utilization of sewagesludge as an energy source while minimizing greenhouse gas emissions.

� 2013 Elsevier Ltd. All rights reserved.

pore, EA

202 G.K. Parshetti et al. / Fuel 111 (2013) 201–210

1. Introduction

In recent years, an increasing awareness of sustainability issueshas led to an impetus for efficiency improvement, hazard minimiza-tion and utilization of renewable waste resources [1]. Sewagesludge from wastewater treatment plants is a potential energy re-source despite containing high levels of pollutants [2,3]. In Singa-pore, around 80,000 tons of sewage sludge is generated each yearfrom 6 wastewater reclamation plants [4]. In the global context,the amount of sewage sludge generated is expected to increase sig-nificantly, considering the rapid population growth, urbanizationand industrialization [5]. Various option have been explored inmany countries around the world to deal with the accumulationof municipal sewage sludge, including incineration, land filling, landapplication, road surfacing, conversion to fertilizer, and compres-sion into building blocks [6,7]. However, due to strict environmentaland legislative constraints, there is an urgent need for developingmore efficient and environmentally friendly approaches for sludgeutilization. One of such approaches is to produce solid fuels fromsludge through proper pre-treatment, which can reduce depen-dence on the dwindling supply of fossil fuels while simultaneouslysolving the environmental problems associated with sludge dis-posal. Treatment of sludge using hydrothermal carbonization(HTC) can be an attractive alternative to conventional sludge dis-posal technologies. A number of studies have recently been con-ducted on HTC covering a wide range of biomass feedstock’s (e.g.,glucose, cellulose, lignin, and starch), fruit shells, sawdust, ricestraw, municipal solid waste, etc. [8–11]. Results from these studiesindicated a significant fraction of carbon remains within the hydro-thermally prepared char during the HTC process, suggesting pro-duction of highly densified fuels and mitigation of greenhouse gasemissions as well. There have been some studies on the hydrother-mal treatment of sewage sludge, as reported in the literature. Doteet al. [12] reported on the hydrothermal liquidization of dewateredsewage sludge. The dewatered sludge was liquidized at tempera-tures above 150–175 �C, and the viscosity of the liquidized sludgewas reduced to that of concentrated sludge. Bougrier et al. [13] em-ployed hydrothermal treatment to reduce sewage sludge quantityand improve energy yield using anaerobic digestion. Nges and Liu[14] evaluated effects of hydrothermal pre-treatment on the dewa-tered-sewage sludge from a municipal wastewater treatment plantin order to improve its biodegradability through anaerobic diges-tion. Park et al. [2] studied effects of hydrothermally pretreatedsewage sludge on the stability and dispersibility of slurry fuel usingpulverized coal. Yanagida et al. [15] showed that dewatered sewagesludge (approximately 80% water) was liquidized by hydrothermaltreatment in order to make coal–water paste (CWP) for use in apressurized–fluidized-bed-combustion (PFBC) power plant. Re-cently, Escala et al. [16] found that HTC of sewage sludge enhanceddewaterability of the sludge, which resulted in a notable lower con-sumption of energy invested in conventional drying processes.However, there has been no previous systematic investigation onthe HTC of sewage sludge for co-combustion with coal and hydro-thermally treated coal with real-time analysis of gaseous pollutants.HTC takes advantage of high-moisture contents of sludge feedstocksince water is used as the reaction medium under high temperature(250–350 �C) and pressure (5–15 MPa). These conditions produce ahighly reactive solvation environment and avoid an energeticallycostly phase change associated with feedstock drying [9]. Complexbiomolecules decompose and reform into a variety of compoundsthat partition into a self-separating solid phase when conditions re-turn to ambient temperature and pressure.

Recently published reports in the literature have shown that theco-combustion of sewage sludge with coal may be viable from theenergetic, economic and environmental points of view, especially

when it is carried out in existing infrastructures [17–20]. Co-firingof coal with the sludge is a promising approach to improve the en-ergy output, achieve the environmentally friendly use of coal byreducing greenhouse gas emissions and utilizing waste residuesinstead of land-filling [21]. However, co-combustion of coal andsewage sludge causes considerable changes in combustion charac-teristics because the fuel characteristics of sewage sludge differgreatly from those of coal [22,23]. Moreover, the flame may notbe stable because of high volatile matter in dried sludge, whichmay change its combustion profiles [24,25]. Because of these rea-sons, coal-fired power plants prefer to co-combust coal with sludgethat has been carbonized at a low temperature to lower the con-tents of volatile matter [26–29]. Low-rank coals such as ligniteand sub-bituminous coals are frequently carbonized and upgradedby hydrothermal treatment; the upgraded coal by the HTC processhas lower moisture and oxygen contents compared with raw coal[30,31]. Several European countries carry out co-combustion ofdried and/or dehydrated sewage sludge in coal power plants[32–35]. In order to predict the behavior of carbonized sludge(e.g. hydrothermally prepared char) during co-combustion withnatural coal and hydrothermally upgraded coal, it is very impor-tant to understand its physico-chemical, structural, thermal andcombustion characteristics.

Thermogravimetric analysis (TGA) provides the preliminaryknowledge of initial and final temperatures for the combustion ofsolid fuels as well as other relevant data such as maximum reactiv-ity temperature and char burnout percentage [10,11]. This infor-mation can then be used to forecast combustion efficiency,residence time, excess air, boiler design, etc. The use of Fouriertransform infrared spectroscopy (FTIR) in conjunction with TGAcan provide insights into characterization of gaseous products aris-ing from thermal degradation of various solid materials [4]. In thisstudy, urban sewage sludge was processed through the HTC pro-cess to obtain energy densified carbonaceous sludge char (HT-SL)material. Further co-combustion of HT-SL with low rank Indone-sian coal (LRIC) and hydrothermally upgraded LRIC (HT-LRIC)was studied with an aim to identify optimal conditions for achiev-ing a maximum waste-to-energy conversion. The physical, chemi-cal, structural and thermal characteristics of carbonaceous fuelswere assessed. In addition, a systematic examination of the emis-sion of gaseous pollutants during co-combustion was also con-ducted using a TGA–FTIR system with different proportions ofHT-SL in its blends with LRIC and HT-LRIC.

2. Experimental

2.1. Materials

The secondary sewage sludge from Ulu Pandan (Singapore)sewage treatment plant was collected for this study. The collectedsludge (pH: 6.6) was immediately transferred to the lab and storedin a plastic box not beyond 24 h at 4 �C prior to use. LRIC was ob-tained from a coal mining company in Jakarta, Indonesia. LRIC wasmanually chopped into small pieces to promote homogeneousmixing and facilitate their effective up-gradation during the HTCtreatment.

2.2. HTC: apparatus and procedure

HTC of sludge and LRIC was carried out in a 500 mL Parr stirredpressure batch reactor (model 4575), as shown in Fig. S1 (see Sup-plementary Information). The HTC study was conducted at a tem-perature of 250 �C and a pressure of (8–10 MPa) with a constantreaction time of 15 min. In a typical run, 250 ml of sludge and

G.K. Parshetti et al. / Fuel 111 (2013) 201–210 203

25 g of LRIC was dispersed in 250 mL of de-ionized water in thereactor. The reactor was sealed and heated up to the desired reac-tion temperature in an electric furnace. After the desired residencetime, the reactor was rapidly cooled to quench the reaction. Theresidence time was defined as the time taken for HTC to be com-pleted at the desired reaction temperature, excluding preheatingand cooling time. When the reactor temperature reached the roomtemperature, the reactor was opened to collect the solid carbona-ceous products. The solid carbonaceous hydrothermally preparedchar product was recovered by filtration, washed with deionizedwater and dried. The char product was weighed independentlyand stored in a sealed container until analysis.

2.3. Characterization

The 5E-MAG6600 automatic proximate analyzer was used todetermine the weight percentages (wt%) of moisture, volatile mat-ter, fixed carbon and ash in LRIC, HT-LRIC, dry sludge and HT-SL.The elementary analyzer (vario MACRO cube, ELEMENTAR, Ger-many) was also used to carry out the ultimate analysis to deter-mine the weight percentages of chemical elements (carbon,hydrogen, nitrogen, oxygen and sulfur) in the LRIC, HT-LRIC, drysludge and HT-SL. Trace elements were determined using micro-wave digestion following inductively coupled plasma optical emis-sion spectroscopy ((ICP-OES), Perkin Elmer, ELAN 6100, USA)analysis. The surface morphology of the all materials was deter-mined by field emission scanning electron microscopy ((FE-SEM),JEOL JSM-6700F Oxford Inca Energy 400). Nitrogen adsorption–desorption isotherms were collected at 77 K using a N2 adsorptionanalyzer (Quantachrome instruments, NOVA 4200e). FT-IR spectraof LRIC, HT-LRIC and HT-SL were recorded using a BIO-RAD, Excal-ibur series spectrometer with a resolution of 4 cm�1. Wide angle X-ray diffraction (XRD) on samples was conducted using an X-ray dif-fractometer (Bruker NEW D8 ADVANCE, Germany) with Ni-filteredCu Ka radiation (k = 1.54 Å) and operated at an accelerating voltageand an emission current of 40 kV and 40 mA, respectively. The XRDpatterns were acquired over the 2h range from 10� to 70�. In thepresent study, in order to interpret the thermogravimetric (TG)and differential thermogravimetric (DTG) profiles, the solid sampleweight loss and the rate of weight loss of all the samples were per-formed in the furnace of the Perkin Elmer TGA7 system operated atgas flow (mixture of N2(70%) and O2(30%)), of 100 ml min�1 whenheating from 25 to 900 �C at a rate of 20 �C min�1. The LRIC, HT-LRIC and HT-SL samples were ground to micron-sized particlesand grounded samples were placed upon a sieve that passed parti-cles of less than 150 lm (micron). Thus, the particle size of theLRIC, HT-LRIC and HT-SL particles and their blends was approxi-mately 150 lm for TGA analysis. The TG and DTG profiles wereanalyzed to find the key combustion parameters. The ignition tem-perature and burnout temperatures were determined using theTG–DTG (the first derivative of TG curve) method based on previ-ous publications in the literature [10,36]. In brief, the ignition tem-perature was determined based on the temperature at which theDTG had its peak value and the corresponding slope to the inter-section with respect to the TG profile. The burnout temperaturewas detected based on the mass stabilization. The gas productsfrom the combustion of samples in TGA were determined using acoupled FTIR spectrometer (Perkin–Elmer FTIR Spectrum 2000).While other conditions were kept the same as those used in TGA,the sample was controlled at �20 mg to provide enough gas foranalysis. To minimize the secondary reaction, the gas products re-leased from thermo-gravimetric balance were swept immediatelyto a gas cell, followed by detection using the FTIR spectrometer.The transfer line and the head of the TGA balance were heated ata constant temperature of 200 �C, to avoid the condensation of vol-atile decomposition products. The scanning range of FTIR was

4000–400 cm�1, in terms of wave number. The resolution andthe sensitivity were set at 4 cm�1 and 1, respectively. The FTIRspectra of the gas products were collected continuously with thebaseline corrected.

3. Results and discussion

3.1. Fuel characteristics

The characteristics of the LRIC, HT-LRIC, dry sludge and carbo-naceous HT-SL produced from sludge after the HTC treatment wereinvestigated comprehensively to identify the solid material, espe-cially the carbon component and inorganic elements. Results ob-tained from the proximate and ultimate analysis and the heatingvalues of LRIC, HT-LRIC, dry sludge and HT-SL are shown inTable S1 (see Supplementary Information). The ultimate analysisrevealed the high contents of carbon and oxygen in the HT-SL(i.e., 40.08 and 52.45 wt%, respectively), supporting the high organ-ic content. The calorific value (15.82 MJ kg�1) of HT-SL was rela-tively higher than that of dry sludge (10.83 MJ kg�1) and close tothose of biomass fuels (i.e., 15–20 MJ kg�1) [37]. Calorific valuesof 9.70 and 14.77 MJ kg�1 were reported previously for the carbon-ized sludge obtained at 300–500 �C in pyrolyser [23,38]. The HT-LRIC produced had the energy–density of 22.46 MJ kg�1 which ishigher than that of the LRIC (18.18 MJ kg�1), suggesting that thehydrothermal treatment results in further coalification of LRIC.As expected, the ash content of the HT-SL was higher (40.46%)and its fixed carbon content was lower (13.89%) after the HTCtreatment. High ash content and low fixed carbon are the charac-teristics of the carbonized products of sewage sludge as has beenreported previously [2,38]. The ash content of the HT-SL was signif-icantly higher, probably due to progressive concentration of miner-als and destructive volatilization of lignocellulosic matters as theprocessing temperature increased. The content of the volatile mat-ter in HT-SL (41.72%) was significantly lower than that of the driedsludge (62.70%). A decreased content of the volatile matter in theorganic component of the HT-SL probably resulted in an increasedheating value of 15.82 MJ kg�1. When LRIC was subjected to hydro-thermal up-grading, the moisture (13.56%), ash (2.10%) and volatilematter (48.08%) contents in the samples decreased slightly, and thefixed carbon (41.74%) content increased notably as a result of car-bonization. Inorganic elements in LRIC, HT-LRIC, dry sludge andHT-SL were determined using ICP-OES. The results are summarizedin Table S2 (see Supplementary Information). A slight reduction inthe contents of trace elements in LRIC and SL was noticed after HTCtreatment presumably due to dissolution of metal ions. A numberof trace elements were not in the detectable range in dry sludgeand HT-SL, which may be due to their lower contents in the sewagesludge, resulting from the separation of industrial and municipalwastewaters for their treatment [7]. Combustion of the sewagesludge leads to a higher stability of heavy metals in the ashes ascompared to the parent sludge [39]. The remaining char, withthe heavy metals strongly bound in their matrix, such ashes afterproper mineralogical, physical and chemical characterization, canbecome a raw material for brick manufacture, agricultural reme-dial [40] or even adsorbents [41]. Therefore the sludge carbonizedchar should not be considered as a waste, but as a low cost by-product with the added-value [41] or burnt, to supply the energyfor pyrolysis [42].

3.2. Fuel microstructure properties

The morphology of carbonaceous HT-SL, LRIC and HT-LRIC wasstudied by using FE-SEM (Fig. S2 (see Supplementary Informa-tion)). The FE-SEM image of HT-SL (Fig. S2a) shows that a wide

0

25

50

75

100

m/m

0 (%)

Temperature (oC)

HT-SL LRIC HT-LRIC LRIC + HT-SL (10%) LRIC + HT-SL (20%) LRIC + HT-SL (30%) HT- LRIC + HT-SL (10%) HT- LRIC + HT-SL (20%) HT- LRIC + HT-SL (30%)

(a)0 200 400 600 800

0 200 400 600 80010

0

-10

-20

-30

-40

-50

-60

-70

DTG

(mg/

min

)

Temperature (oC)

HT-SL LRIC HT-LRIC LRIC + HT-SL (10%) LRIC + HT-SL (20%) LRIC + HT-SL (30%) HT-LRIC + HT-SL (10%) HT-LRIC + HT-SL (20%) HT-LRIC + HT-SL (30%)

(b)

Fig. 1. (a) TG curves for LRIC, HT-LRIC, HT-SL and their blends. All samples wereheated at 20 �C min�1. (b) DTG curves for LRIC, HT-LRIC, HT-SL and their blends. Allsamples were heated at 20 �C min�1.

204 G.K. Parshetti et al. / Fuel 111 (2013) 201–210

variety of pores are present in HT-SL along with fibrous structure.Sludge mainly contains extracellular polymeric substances (EPS)and inorganic material [2]. EPS, mainly composed of protein andpolysaccharide, exists in flocs and cellular tissues of sewage sludge,and contributes to the aggregation of flocculated sludge particles. Itreflects that flocs and cellular tissues were broken by the thermalenergy and high pressure during HTC treatment due to a seriesof chemical reactions involving hydrolysis, dehydration, decarbox-ylation, polymerization, and aromatization [8–11], and protein inEPS was probably denatured and destroyed at high temperaturesfollowed by agglomeration into larger particles by binding ontowhatever solid structures that remained in the suspension. Thegrains observed from SEM images could also be associated to thesintering of the inorganic material of the biomass due to thermaltreatment. Thus, HTC of sewage sludge overall leads to the forma-tion of grains of different particle sizes and heterogeneous surfacemorphology. FE-SEM images shown in Fig. S2b and c also revealchanges between the raw coal and the hydrothermally-upgradedcoal including formation of larger individual grains in the upgradedcoal, presumably due to collapse of micro-pores by partial melting,or softening. These images support a reaction sequence in whichportions of LRIC were solubilized, carbonized in solution and reas-sembled onto solid components in the milieu. A transformationthat accompanied this process was a color change from brown LRICto a distinctly blackened HT-LRIC.

Surface functional groups of HT-SL, LRIC and HT-LRIC materialwere investigated by FTIR spectroscopy (Fig. S3 (see Supplemen-tary Information)). The characteristic band observed was theabsorbance peak at 800–1200 cm�1 due to mineral ash content at1000–1100 cm�1 [38] for HT-SL. The strength of CAH bond(2800–3000 cm�1) and OAH (3200–3610 cm�1) bond in HT-SL isless than that of LRIC and HT-LRIC because of high ash content[38]. The FTIR spectra of both pristine LRIC and HT-LRIC were sim-ilar over the entire wave number range. An absorption peak attrib-utable to hydrophobic C@C (aromatic stretching vibrations)functional groups of the coal [38] is visible at about 1620 cm�1 inall three samples. Broad and strong absorption peaks attributableto hydroxyl groups of LRIC, HT-LRIC and HT-SL are present at3200–3600 cm�1. Moreover, the presence of oxygenated functionalgroups in HT-SL is suggested by the bands at: 3000–3700 (a wideband attributed to OAH stretching vibrations in hydroxyl or car-boxyl groups), 2300–2350 and 1460 cm�1 (CAO stretching vibra-tions in hydroxyl, ester or ether and OAH bending vibrations)[43]. The results provided more direct evidence that HT-SL con-tained organic compounds as also evidenced by ultimate analysis.Taking into account the chemical characteristics of the HT-SL prod-uct (i.e., the high concentration of oxygenated functional groups), itcan be envisaged that this type of char material may constitute anexcellent precursor for the preparation of activated carbons bycontrolled chemical activation [44]. The XRD patterns (Fig. S4(see Supplementary Information)) for the LRIC, HT-LRIC and HT-SL were studied. In the LRIC and HT-SL, the presence of quartz(SiO2) (at 2h = 26.60� and 21.36�) can be observed from the XRDspectra [45]. As can be seen from this figure, there is a peak at2h = 20.66�, which probably indicated the presence of c-Al2O3

[45] in LRIC and HT-SL. The XRD patterns revealed the presenceof quartz as main mineral component in LRIC and HT-SL. In con-trast, the HT-LRIC obtained at 250 �C was no longer crystalline, ascan be deduced from the absence of reflections (at 2h = 26.60�,21.36� and 20.66�) in the XRD pattern. These results suggest thatthe metal compounds amount was reduced from the LRIC duringthe HTC process. Qualitatively, it was observed that the HT-SLand HT-LRIC were fluffy and high energy density powders, whichcan be easily compacted. The surface areas of the BET were 11, 8and 1 m2 g�1 for HT-SL, HT-LRIC and LRIC, respectively. While porevolumes for HT-SL, HT-LRIC and LRIC were 0.070, 0.032 and

0.005 cm3 g�1 respectively. The adsorption isotherms for HT-SL,LRIC and HT-LRIC in this study are shown in Fig. S5 (see Supple-mentary Information). The hydrothermally prepared char sampleshad relatively small BET surface areas due to low porosity. Undersuch conditions, the specific surface area values calculated onlycorrespond to the external surface [46]. It has also been reportedpreviously that the porosity of hydrothermally prepared chars islow [46]. Improvements in the porosity and the surface area ofhydrothermally prepared chars are therefore necessary to enabletheir use as hydrogen storage or electrical energy storage (superca-pacitors). Such improvements have been achieved with a combina-tion of thermal and chemical activation that resulted in a surfacearea of >2000 m2 g�1 [47].

3.3. Thermal and co-combustion analysis

Fig. 1a and b show the TG and DTG profiles of LRIC, HT-LRIC, HT-SL and their blends respectively. The blend refers to the mixing ofHT-SL with LRIC and HT-LRIC in the proportion of 10%, 20% and 30%of the former. Table 1 shows the results of thermogravimetric anal-ysis for LRIC, HT-LRIC, HT-SL and their blends. As shown in TG pro-file (Fig. 1a), the combustion process can be divided into threestages for interpretation. The temperature of the first stage rangingfrom ambient room temperature at the beginning of the analysis toabout 200 �C, corresponding to a loss of moisture and of the very

Table 1Combustion parameters of LRIC, HT-LRIC, HT-SL and their blends.

Sample To (�C)a Tf (�C)b DTG peak temperature (�C) DTG Rmax (mg/min)c Burnout (%)

LRIC 335 411 354 35.40 88.0HT-LRIC 310 435 327 34.50 94.0HT-SL 291 455 300 48.79 49.0LRIC + HT-SL (10%) 321 393 329 51.82 85.5LRIC + HT-SL (20%) 314 410 321 56.59 80.0LRIC + HT-SL (30%) 308 370 314 61.24 77.0HT-LRIC + HT-SL (10%) 305 444 326 21.93 93.0HT-LRIC + HT-SL (20%) 298 450 308 29.40 94.5HT-LRIC + HT-SL (30%) 296 453 307 32.36 94.5

a To = ignition temperature.b Tf = burnout temperature.c DTG Rmax = maximum weight loss rate.

G.K. Parshetti et al. / Fuel 111 (2013) 201–210 205

light volatile materials. The second stage began at 200 �C andended at about 460 �C, and this stage reflected mainly the decom-position of the organic fraction of the respective materials [20]. Thethird stage began at 460 �C and ended at about 900 �C. In this stage,the weight loss of all the samples was relatively small, mainly be-cause of the ash content. The results showed that the major weightlosses for HT-SL, LRIC and HT-LRIC occur at 290–460, 330–415 and310–435 �C, respectively. The DTG profile (Fig. 1b) for LRIC and HT-LRIC showed only one major peak, which could be due to thehomogeneity of coal. A previous study suggested that the presenceof a single peak at relatively high temperatures in the LRIC and HT-LRIC could be due to the decomposition of complex organic matter(polyaromatic hydrocarbons, organic polymers) taking place athigh temperatures [48]. There also exists a small peak related tocarbonate decomposition in LRIC at around 440 �C [48]. On theother hand, the DTG profile for HT-SL displayed two peaks. Thedecomposition peaks of the HT-SL are comparable to those of var-ious carbonaceous materials which have been reported elsewhere.The decomposition at 290–340 �C was found to be similar to thedecomposition of hemicellulose [49,50], cellulose [49,50], micro al-gae [51], leather [52], and some aliphatic amino acid [53], whereasthe decomposition at higher temperature, 420–460 �C, is thoughtto be due to the decomposition of the complex and/or aromaticstructures such as lignin [48] and some plastics [54]. The totalburnout percentage for HT-SL (49%) was lower due to its highash contents, whereas the burnout of LRIC and HT-LRIC was 88%and 94%, respectively. The increased burnout efficiency in HT-LRICcan be attributed to reduction in the ash content of LRIC followingthe hydrothermal treatment process.

To determine the interactive effect of HT-SL on combustioncharacteristics of LRIC and HT-LRIC, co-combustion experimentswere conducted in three different HT-SL blending ratios of 10%,20%, and 30%. As shown in Table 1, as the blending ratios of HT-SL increased, the ignition temperature of the all blended solid fuelsdecreased. The ignition temperature of the coal/HT-SL blends fol-lowed the weighted percentage of HT-SL and hence exhibited anadditive behavior. Higher blending ratios of HT-SL with LRIC andHT-LRIC will result in a very low ignition temperature causing apotential fire hazard, and hence care must be taken to ensure thatreduction in the ignition temperature should be within the allow-able range for the presently available equipment [36]. An improve-ment in the burnout temperature of the LRIC/HT-SL blends wasobserved as compared to LRIC combustion alone, but did not followthe weighted percentage. However, an increase in the burnouttemperature of the HT-LRIC/HT-SL was observed with an increasein HT-SL percentage, exhibiting an additive behavior. A similarobservation with an increase in the burnout temperatures has beenreported by other researchers for co-combustion of hydrother-mally prepared char with coal [11,36,55]. In the blended fuel, itappears that HT-SL and coal burn independently. The HT-SL reactsfaster, reducing the oxygen concentration required for coal and

thus decreasing the incremental effect of improvement in theburnout, while the blending ratio was increased. The change inboth the ignition temperature and burnout temperatures occursdue to the influence of the HT-SL on the combustion characteristicsof coal, leading to changes in the fuel reactivity [11]. The decreas-ing burnout percentage of the LRIC/HT-SL in the range of 10–30%blends (Table 1) with the proportion of HT-SL could arise due tothe masking effect of the high ash content at the later stage of com-bustion. However, the blends containing HT-LRIC/HT-SL in therange of 20–30% showed slightly higher burnout percentage thanthe individual LRIC. This increased efficiency may be attributedto the higher reactivity of the organic matrix HT-LRIC/HT-SL. Fur-ther, as can be seen from Table 1, the DTG mass loss rate (DTG Rmax)for all blended fuels gradually increased with an increase in theHT-SL in the blended fraction. In addition, the temperature atwhich the peak weight loss rate occurred was reduced when theblending ratio was increased. In general, the reactivity is directlyproportional to the height of the DTG peak and inversely propor-tional to the peak temperature [11]. In this case, the DTG Rmax in-creased and the peak temperature decreased upon blending HT-SL with coal, confirming an increase in reactivity. These results alsoshow that the combustion behavior of the coal-HT-SL blends wasinfluenced by the quantity of HT-SL added.

3.4. Kinetic parameters

Studies on TGA profiles contribute to enhancing the knowledgeof the kinetics of the HT-SL/ LRIC and HT-SL/HT-LRIC co-combus-tion process and, therefore, to establishing the optimum opera-tional conditions for better utilization of hydrothermallyprepared chars. Combustion kinetic parameters were evaluatedfor the temperature ranges as mentioned in Table 2. The tempera-ture dependence of heterogeneous solid-state reactions may be de-scribed by the Arrhenius equation:

dadT¼ A

be�E=RTð1� aÞn ð1Þ

where n stands for the reaction order and A and E represent the pre-exponential factor and the activation energy of the reaction, respec-tively and mass change a is defined as,

a ¼ m0 �mt

m0 �mfð2Þ

where m0 represents the initial sample mass in an experiment, mt

denotes the current sample mass, and mf specifies the final samplemass.

Eq. (1) can be expressed in an integral form as follows:

Z a

0

dað1� aÞn

¼ ab

Z T

T0

e�E=RT dT ð3Þ

Table 2Summary of kinetic parameters – activation energy.

Sample Temperaturerange

Mechanism a b Correlationcoefficient

Pre-exponential factor (A) (min�1) Activation energy (E) (kJ mol�1)

LRIC 335 411 F1 2.39 10028 0.99 2.19 � 106 83.37HT-LRIC 310 435 F1 0.35 8659 0.96 2.45 � 105 71.99HT-SL 291 340 F1 9.08 13111 0.92 2.30 � 109 109.00

420 455 D4 �2.32 2102 0.86 9.41 � 109 17.47LRIC + HT-SL (10%) 321 393 F1 �2.04 8899 0.94 2.31 � 104 73.98LRIC + HT-SL (20%) 314 410 F1 �2.67 6476 0.95 8.96 � 103 53.84LRIC + HT-SL (30%) 308 370 F1 �2.73 6319 0.82 8.26 � 103 52.53HT-LRIC + HT-SL

(10%)305 444 F1 �2.33 6948 0.98 1.34 � 104 57.76

HT-LRIC + HT-SL(20%)

298 450 F1 �4.65 5354 0.97 1.02 � 104 44.51

HT-LRIC + HT-SL(30%)

296 453 F1 �4.88 5131 0.98 7.80 � 102 42.65

206 G.K. Parshetti et al. / Fuel 111 (2013) 201–210

Using the approximation introduced by Coats [56], Eq. (3) canbe integrated and rearranged into the following logarithmic form.

lngðaÞT2

� �¼ ln

ARbE

1� 2RTE

� �� �� E

RTð4Þ

A plot of ln[g(a)/T2] against reciprocal of temperature shouldgive a straight line with a slope equal to �E/R. g(a) is the integralfunction of conversion. Apparent activation energy E and the pre-exponential factor A can be calculated by using the slope and inter-cept obtained from Eq. (4). The function, g(a), depends on themechanism controlling the reaction, the size and the shape of thereacting particles. Table S3 (see Supplementary Information)shows the expressions of g(a) for the basic model functions usuallyemployed for the kinetic study of solid-state reactions. By means ofthese functions, it was possible to identify possible reaction mech-anisms governing the process of thermal oxidation of the solidsamples from the TG curves. The form of g(a) that gave a straightline with a highest correlation coefficient was taken as the functionof the model that would best represent the kinetics of mass loss foreach separate reaction. Model F1 presented the highest correlationcoefficients for all the solid fuel samples studied except for the sec-ond stage (420–455 �C) of HT-SL combustion, suggesting that thechemical first order reaction is the most effective mechanism forcombustion. The effective solid-state mechanisms were found tobe F1 for the first stage of HT-SL combustion and D4 for the secondstage combustion based on the magnitude of correlation coeffi-cients obtained (Table 2). Activation energies associated with thecombustion of LRIC and HT-LRIC were found to be 83.37 and71.99 kJ mol�1, respectively. As a general observation, it may benoted that activation energies of LRIC decreased with the hydro-thermal treatment. Pre-exponential factors (logA) of the abovesamples were estimated to be 2.19 � 106 and 2.45 � 105 min�1,respectively. HT-SL had two combustion stages with the activationenergies associated with the initial step (291–340 �C) and the sec-ond step (420–455 �C) being 109.00 and 17.47 kJ mol�1, respec-tively. As the activation energy of HT-SL in the first reaction washigher than that in the second reaction, the main decompositionof HT-SL most likely occurred in the first step in the temperaturerange of 291–340 �C. This finding is consistent with what has beenreported previously in the literature for sludge pyrolysis with theactivation energy being in the range of 17–332 kJ mol�1 [54,57].Activation energies of the LRIC/HT-SL blended samples (LRIC/HT-SL (10–30%)) were found to be in the range of 73.98–52.53 kJ mol�1. As reported in Table 2, the activation energies ofthe blend samples were lower than that of LRIC (83.37 kJ mol�1).As for other HT-LRIC/HT-SL blends (HT-LRIC/HT-SL (10–30%)),activation energies were in the range of 57.76–42.65 kJ mol�1,respectively. The activation energy of the HT-LRIC/HT-SL (30%,

42.65 kJ mol�1) blend showed the lowest activation energy. Asthe activation energy represents the critical energy needed to startcombustion [58], the coal-HT-SL with the lowest activation energyis preferred.

3.5. TGA–FTIR analysis of gas products

TGA has demonstrated to be a very valuable technique forstudying the combustion of a wide range of solid samples whileFTIR is one of the most powerful techniques for gas analysis. There-fore, TGA coupled with FTIR was used in this study to measureweight loss and evolution of the corresponding volatiles for LRIC,HT-LRIC, HT-SL and their blends at a heating rate of 20 �C min�1.The volatiles identified with TG-FTIR analysis in the current studywere CH4 (2978 cm�1), CO (2119 cm�1), CO2 (2360 cm�1), andsome H2O (3750 cm�1) based on the FTIR data reported in the lit-erature [4,59]. The absorbance bands of water (at 3750 cm�1)might be due to the dehydration of the sludge by losing both freewater and chemically bonded water, previously evidenced by TGAanalysis. Gases such as NOx, including NO (1762 cm�1), NO2

(1520 cm�1) and SO2 (1342 cm�1) were not detectable, suggestingthat the concentrations of these gases released were negligible[59]. An FTIR three-dimensional stack plot of LRIC and HT-SL inthe temperature range of 25–900 �C, maintained with TGA, isshown in Fig. 2a and b, respectively. It can be observed fromFig. 2a and b that, at the temperature between 200 and 550 �C,CO2 was observed with a major abundance for LRIC, while theCO2 absorbance band was observed at temperatures around 210–330 and 350–520 �C for HT-SL. The latter two absorbance bandsobserved for HT-SL can be attributed to the release of volatilesand decomposition of organic compounds, respectively, whichare consistent with the two peaks of the weight loss rates observedin the DTG curve (Fig. 1b) occurring at 290 and 450 �C, respectively.Meanwhile, another strong peak of CO (2119 cm�1) in LRIC and HT-SL appeared, possibly due to incomplete oxidation of combustiblecarbonaceous compounds [4]. The release of CH4 was observed inthe spectrum for HT-SL at 200–300 �C which is attributed to cata-lytic effects of alkali and alkaline earth metals present in the HT-SLash as shown in Table S2 (see Supplementary Information) [60].The decomposition of these organics (CH4) at higher temperatureled to the further releasing of CO2 in gas products. The obscureabsorbance bands of water vapor at 3700–3550 cm�1 were ob-served in both LRIC and HT-SL, which might be due to the dehydra-tion of both free water and chemically-bonded water in coal aswell as sludge [4]. As the combustion temperature increased above600 �C, the release of pollutant gases was almost negligible, indi-cating the completion of combustion.

500 1000 1500 2000 2500 3000 3500 4000

0.000

0.042

0.084

0.126

0.168

494.30

900.0

1189.6

1710.8

2203.0

H2OCH4

CO2

CO

CO2

Wavenumber (cm -1)

Abso

rban

ce u

nits

Time (s)

(a)

500 1000 1500 2000 2500 3000 3500 4000

0.000

0.033

0.066

0.099

0.132

494.30

900.0

1189.6

1710.8

2203.0

CO2

(b)

H2O

CH4

CO2CO2

CO

CO2

Wavenumber (cm-1)Ab

sorb

ance

uni

ts

Time (s)

Fig. 2. Three-dimensional (3D) spectral plot of the gases produced from the combustion of (a) 100% LRIC and (b) 100% HT-SL at heating rate of 20 �C min�1 in TGA–FTIR.

G.K. Parshetti et al. / Fuel 111 (2013) 201–210 207

Fig. 3a shows the emission profile of CO2 from the combustionof LRIC, HT-LRIC, HT-SL and their blends at different temperatures.The CO2 emission profiles for LRIC, HT-LRIC and their blends at thetemperature of 200–550 �C showed one peak, which might corre-spond to the individual peak of the weight loss rate in the DTGcurves (Fig. 1b). However, two CO2 peaks for HT-SL were observedover the same range of temperatures, one in the range of 200–310 �C and the other in the range of 320–530 �C. The two peaksmight be attributed to the heterogeneity of the sludge samples.The first peak between temperatures of 200–310 �C could be re-lated to the HT-SL peak of the weight loss rate in the DTG curveoccurring at 300 �C, and the peak between temperatures 320 and530 �C to the DTG curve peak occurring at 458 �C (Fig. 1b). The sec-ond peak reflecting a strong emission of CO2 was more prominentthan the first peak. This observation suggests that the CO2 emissionwas relatively higher during the char burning which occurred inthe temperature range of 320–530 �C. For the HT-LRIC and HT-LRIC/HT-SL blends, the CO2 emission peak shifted to lower temper-atures compared to the LRIC and LRIC/HT-SL blends. For the LRIC/HT-SL blends, the emission profile peak of CO2 was centeredaround 425 �C. However, for the HT-LRIC/HT-SL blends, the emis-sion profile peak was centered around 350 �C. Fig. 3a also showsthat the CO2 released had its maximum value for LRIC based onthe absorbance value. The magnitude of the CO2 emission was

generally about 8–10 times higher than that of other detectedgases (CO and CH4 as presented in Fig. 3b and c) in the presentstudy. The integral areas under the CO2 emission curves for LRIC,HT-LRIC and HT-SL were 24.4, 16.0 and 15.5 respectively. TheCO2 emission was mainly influenced by the contents of fuel inthe blends. In addition, it should be noted that the CO2 emissiondecreased in proportion to the amount of HT-SL used in the blendsand the HT-LRIC/HT-SL (30%) blend showed the maximum CO2

reduction. This was confirmed by the integral areas under theemission curves, which were 19.3 for LRIC/HT-SL (10%), 18.7 forLRIC/HT-SL (20%), 16.8 for LRIC/HT-SL (30%), 15.9 for HT-LRIC/HT-SL (10%), 15.4 for HT-LRIC/HT-SL (20%), 14.0 for HT-LRIC/HT-SL (30%).

The CO emission profiles for the fuel samples are shown as afunction of combustion temperature in Fig. 3b. The highest COemission occurred at about 425 �C for the LRIC. All profiles showeda major single peak except that of HT-SL. The two CO emissionpeaks from HT-SL could be related to the two peaks observed inthe DTG curves (Fig. 1b). The initial stage of CO release was prob-ably caused by the decarboxylation reaction of alkyl side-chainscontaining the carbonyl functional groups of lignocellulosic bio-mass, while the release of CO at the high temperatures was prob-ably due to secondary reactions [61,62]. The CO emission of allfuel blends started around 200 �C and showed differences in the

0 200 400 600 800 1000

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0.0035

0.0040

Abso

rban

ce

Temperature (oC)

LRIC HT-LRIC HT-SL LRIC + HT-SL (10%) LRIC + HT-SL (20%) LRIC + HT-SL (30%) HT-LRIC + HT-SL (10%) HT-LRIC + HT-SL (20%) HT-LRIC + HT-SL (30%)

(b)

0 200 400 600 800 1000

0.00

0.05

0.10

0.15

0.20Ab

sorb

ance

Temperature (oC)

LRIC HT-LRIC HT-SL LRIC + HT-SL (10%) LRIC + HT-SL (20%) LRIC + HT-SL (30%) HT-LRIC + HT-SL (10%) HT-LRIC + HT-SL (20%) HT-LRIC+ HT-SL (30%)

(a)

0 200 400 600 800 1000

0.0000.0020.0040.0060.0080.0100.0120.0140.0160.0180.0200.0220.024

Abso

rban

ce

Temperature (oC)

LRIC HT-LRIC HT-SL LRIC + HT-SL (10%) LRIC + HT-SL (20%) LRIC + HT-SL (30%) HT-LRIC + HT-SL (10%) HT-LRIC + HT-SL (20%) HT-LRIC + HT-SL (30%)

(c)

Fig. 3. Evolution of (a) CO2, (b) CO and (c) CH4 from the combustion of LRIC, HT-LRIC, HT-SL and blends at heating rate of 20 �C min�1 in TGA–FTIR.

208 G.K. Parshetti et al. / Fuel 111 (2013) 201–210

peak position/height due to differences in the relative contents ofthe main carbonaceous constituents. The LRIC/HT-SL and HT-LRIC/HT-SL blends showed CO emission peak around 420 and350 �C, respectively. The results revealed that the blending of HT-SL with LRIC and HT-LRIC decreased CO emissions considerably,with the latter blend showing the highest CO reduction (30%).The CH4 emission profiles for all the fuel samples and blends areshown in Fig. 3c. All profiles demonstrate dispersed peaks in thetemperature range of 180–500 �C when the volatile matter andchar were burnt. The HT-SL showed an emission peak at 220 �Cand the LRIC at 423 �C, suggesting two burning processes involvingvolatile materials and fixed carbon, respectively. The release of CH4

from LRIC in the fixed carbon stage was much lower than the re-lease of CH4 for HT-SL in the volatile material stage. The high emis-sion of CH4 was probably due to the release of methoxyl groups,resulting mainly from lignocellulosic decomposition in HT-SL[62]. The CH4 emission beyond 500 �C was almost negligible withthe combustion temperature rising up to 900 �C. It appeared thatthe HT-LRIC/HT-SL (30%) blend produced the lowest total emissionin CH4. The CH4 emission was hardly detectable while using HT-SL/LRIC/HT-LRIC blends, which may be due to the oxidation of hydro-carbons during co-combustion.

4. Conclusions

The HTC of sewage sludge provided a densified carbonaceoussolid product HT-SL with a calorific value of 15.82 MJ kg�1 and un-

ique physico-chemical and microstructure characteristics. Thestudy on co-combustion of carbonaceous HT-SL with raw LRICand its upgraded version (HT-LRIC) following hydrothermal treat-ment provided several attractive benefits: (1) Combustion behav-ior of the LRIC/HT-SL and HT-LRIC/HT-SL blends was influencedby the fraction of HT-SL in the blend; (2) the emission productsof LRIC and HT-LRIC and those with co-combustion with HT-SLare mainly composed of CO2, CO and CH4 and their concentrationwas reduced with HT-SL addition; (3) the study showed that com-bining HT-SL with LRIC/HT-LRIC can improve the quality of com-bustion emissions; and (4) this work also suggests that the HTCprocess has potential for conversion of waste sludge to energy den-sified solid fuels and carbonaceous functional materials for use in avariety of applications, including energy generation through co-combustion with coal.

Acknowledgments

The authors are thankful to the Singapore Economic Develop-ment Board (EDB), National Environment Agency (ETRP grant)and the Minerals, Metals and Materials Technology Centre(M3TC), National University of Singapore for the financialsupport extended to this project. The authors are also thankfulto Ms. Ni Xiping, Research Officer, A�STAR Institute of MaterialsResearch and Engineering (IMRE), Singapore for technicalassistance.

G.K. Parshetti et al. / Fuel 111 (2013) 201–210 209

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.fuel.2013.04.052.

References

[1] Patel AD, Meesters K, Uil HD, Jong ED, Blok K, Patel MK. Sustainabilityassessment of novel chemical processes at early stage application to biobasedprocesses. Energy Environ Sci 2012;5:8430–44.

[2] Park SJ, Bae JS, Lee DW, Ra HW, Hong JC, Choi YC. Effects of hydrothermallypretreated sewage sludge on the stability and dispersibility of slurry fuel usingpulverized coal. Energy Fuel 2011;25:3934–9.

[3] Bourgel C, Veron E, Poirier J, Defoort F, Seiler JM, Peregrina C. Behavior ofphosphorus and other inorganics during the gasification of sewage sludge.Energy Fuel 2011;25:5707–17.

[4] Shao J, Yan R, Chen H, Wang B, Lee DH, Liang DT. Pyrolysis characteristics andkinetics of sewage sludge by thermogravimetry Fourier transform infraredanalysis. Energy Fuel 2008;22:38–45.

[5] Cao YC, Pawlowski A. Sewage sludge-to-energy approaches based on anaerobicdigestion and pyrolysis: brief overview and energy efficiency assessment.Renew Sustain Energy Rev 2012;16:1657–65.

[6] Fytili D, Zabaniotou A. Utilization of sewage sludge in EU application of old andnew methods – a review. Renew Sustain Energy Rev 2008;12:116–40.

[7] Ivanov VN, Wang JY, Stabnikova ST, Tay L, Tay JH. Microbiological monitoringin the biodegradation of sewage sludge and food waste. J Appl Microbiol2004;96:641–7.

[8] Sevilla M, Fuertes AB. The production of carbon materials by hydrothermalcarbonization of cellulose. Carbon 2009;47:2281–9.

[9] Berge ND, Ro KS, Mao J, Flora JRV, Chappell MA, Bae S. Hydrothermalcarbonization of municipal waste streams. Environ Sci Technol2011;45:5696–703.

[10] Liu ZG, Quek A, Hoekman SK, Srinivasan MP, Balasubramanian R.Thermogravimetric investigation of hydrochar-lignite co-combustion.Bioresour Technol 2012;123:646–62.

[11] Parshetti GK, Hoekman SK, Balasubramanian R. Chemical, structural andcombustion characteristics of carbonaceous products obtained byhydrothermal carbonization of palm empty fruit bunches. Bioresour Technol2013;135:683–9.

[12] Dote Y, Yokoyama S, Minowa T, Masuta T. Thermochemical liquidization ofdewatered sewage sludge. Biomass Bioenergy 1993;4:243–8.

[13] Bougrier C, Delgenes JP, Carrere H. Combination of thermal treatments andanaerobic digestion to reduce sewage sludge quantity and improve biogasyield. Process Safe Environ 2006;84:280–4.

[14] Nges IA, Liu J. Effects of anaerobic pre-treatment on the degradation ofdewatered-sewage sludge. Renew Energy 2009;34:1795–800.

[15] Yanagida T, Fujimoto S, Minowa T. Application of the severity parameter forpredicting viscosity during hydrothermal processing of dewatered sewagesludge for a commercial PFBC plant. Bioresour Technol 2010;101:2043–5.

[16] Escala M, Zumbühl T, Koller Ch, Junge R, Krebs R. Hydrothermal carbonizationas an energy-efficient alternative to established drying technologies forsewage sludge: a feasibility study on a laboratory scale. Energy Fuel2013;27:454–60.

[17] Stasta P, Boron J, Bebar L, Stehlik P, Oral J. Thermal processing of sewagesludge. Appl Therm Eng 2006;26:1420–6.

[18] Cartmell E, Gostelow P, Riddell-Black D, Simms N, Oakey J, Morris J, et al.Biosolids – a fuel or a waste? An integrated appraisal of five co-combustionscenarios with policy analysis. Environ Sci Technol 2006;40:649–58.

[19] Otero M, Calvo LF, Gil MV, Garcia AI, Moran A. Co-combustion of differentsewage sludge and coal: a non-isothermal thermogravimetric kinetic analysis.Bioresour Technol 2008;99:6311–9.

[20] Xiao H, Ma X, Liu K. Co-combustion kinetics of sewage sludge with coal andcoal gangue under different atmospheres. Energy Convers Manage2010;51:1976–80.

[21] Robinson AL, Rhodes JS, Keith DW. Assessment of potential carbondioxidereductions due to biomass–coal cofiring in the United States. EnvironSci Technol 2003;37:5081–9.

[22] Kim JK, Lee HD. Investigation on the combustion possibility of dry sewagesludge as a pulverized fuel of thermal power plant. J Ind Eng Chem2010;16:510–6.

[23] Park SW, Jang CH. Characteristics of carbonized sludge for co-combustion inpulverized coal power plants. Waste Manage 2011;31:523–9.

[24] Tsuji H, Hashimoto N, Shirai H, Makino H. A study on combustioncharacteristics of dried sludge pellets by use of a pulverized coal combustiontest furnace. J Japan Inst Energy 2009;88:422–9.

[25] Kim MY, Kim JK, Lee HD, Kim S. Combustion characteristics of sewage sludge-bituminous coal mixture. In: Proceeding of the Korean Society of IndustrialEngineering Chemistry. Seoul, South Korea; 2009. p. 93–6.

[26] Amari T, Tanaka M, Koga Y, Okuno S, Tajima A. Biomass solid fuel from sewagesludge with pyrolysis and co-firing in coal power plant. In: Proceedings of theJapan Society of Mechanical Engineers, Symposium on EnvironmentalEngineering. Japan; 2006.

[27] Youichi K, Yuuki E, Hiroshi O, Kazuaki K, Takeshi A, Kimitoshi O. Biomass solidfuel production from sewage sludge with pyrolysis and co-firing in coal powerplant. Mitsubishi Heavy Ind Tech Rev 2007;44:43–6.

[28] Tomiyama K, Katoh D, Yamazaki S, Saitoh M, Kudo M. Sludge recycling systemby carbonization. Mitsui Zosen Tech Rev 2007;190:39–44.

[29] Shimura S, Kabuto T, Kitabayashi M, Shimizu K. Biomass fuel use of carbonizedsewage sludge. Electric Furnace Steel 2007;78:73–8.

[30] Katalambula H, Gupta R. Low-grade coals: a review of some prospectiveupgrading technologies. Energy Fuel 2009;23:3392–405.

[31] Nonaka M, Hirajima T, Sasaki K. Upgrading of low rank coal and woodybiomass mixture by hydrothermal treatment. Fuel 2011;90:2578–84.

[32] Werther J, Ogada T. Sewage sludge combustion. Prog Energy Combust Sci1999;25:55–116.

[33] Luts D, Devoldere K, Laethem B, Bartholomeeusen W, Ockier P. Coincinerationof dried sewage sludge in coal-fired power plants: a case study. Water SciTechnol 2000;42:259–68.

[34] Stelmach S, Wasielewski R. Co-combustion of dried sewage sludge and coal ina pulverized coal boiler. J Mater Cycl Waste Manage 2008;10:110–5.

[35] Hansson J, Berndes G, Johnsson F, Kjarstad J. Co-firing biomass with coal forelectricity generation: an assessment of the potential in EU27. Energy Policy2009;37:1444–55.

[36] Muthuraman M, Namioka T, Yoshikawa K. Characteristics of co-combustionand kinetic study on hydrothermally treated municipal solid waste withdifferent rank coals: a thermogravimetric analysis. Appl Energy2010;87:141–8.

[37] Ebling JM, Jenkins BM. Physical and chemical properties of biomass fuels.Trans ASAE 1985;28:898–902.

[38] Park SW, Jang CH. Effects of carbonization and solvent-extraction on change infuel characteristics of sewage sludge. Bioresour Technol 2011;102:8205–10.

[39] Albertson OE, Bruno JM, Copeland B, Crumm J, Homoki GG, Koch CM, et al.Sludge incineration: thermal destruction of residues. In: Proceedings of thesymposium: Water Environment Federation, Alexandria, USA; 1992.

[40] Otero M, Díez C, Calvo LF, García AI, Morán A. Analysis of the co-combustion ofsewage sludge and coal by TG-MS. Biomass Bioenergy 2002;22:319–29.

[41] Khiari B, Marias F, Zagrouba F, Vaxelaire J. Analytical study of the pyrolysisprocess in a wastewater treatment pilot station. Desalination2004;167:39–47.

[42] Rumphorst MP, Ringel HD. Pyrolysis of sewage sludge and use of pirólisis coke.J Anal Appl Pyrol 1994;28:137–55.

[43] Araujo-Andrade C, Ruiz F, Martinez-Mendoza JR, Terrones H. Infrared andRaman spectra, conformational stability, ab initio calculations of structure, andvibrational assignment of alpha and beta glucose. J Mol Struct THEOCHEM2005;714:143–6.

[44] Wang LL, Guo YP, Zou B, Rong CG, Ma XY, Qu YN, et al. High surface areaporous carbons prepared from hydrochars by phosphoric acid activation.Bioresour Technol 2011;102:1947–50.

[45] Fang P, Cen C, Chen D, Tang Z. Carbonaceous adsorbents prepared from sewagesludge and its application for Hg0 adsorption in simulated flue gas. Chin JChem Eng 2010;18:231–8.

[46] Fuertes AB, Arbestain MC, Sevilla M, Macia-Agullo JA, Fiol S, Lopez R, et al.Chemical and structural properties of carbonaceous products obtained bypyrolysis and hydrothermal carbonisation of corn stover. Aust J Soil Res2010;48:618–26.

[47] Sevilla M, Fuertes AB, Mokaya R. High density hydrogen storage insuperactivated carbons from hydrothermally carbonized renewable organicmaterials. Energy Environ Sci 2011;4:1400–10.

[48] Folgueras MB, Diaz RM, Xiberta J. Pyrolysis of blends of different types ofsewage sludge with one bituminous coal. Energy 2005;30:1079–91.

[49] Jindarom C, Meeyoo V, Rirksomboon T, Rangsunvigit P. Thermochemicaldecomposition of sewage sludge in CO2 and N2 atmosphere. Chemosphere2007;67:1477–84.

[50] Müller-Hagedon M, Bockhorn H, Krebs L, Mü ller U. A comparative kineticstudy on the pyrolysis of three different wood species. J Anal Appl Pyrol2003;68-69:231–49.

[51] Peng W, Wu Q, Tu P, Zhao N. Pyrolytic characteristics of microalgae asrenewable energy source determined by thermogravimetric analysis.Bioresour Technol 2001;80:1–7.

[52] Heikkinen JM, Hordijk JC, Jong WD, Spliethoff H. Thermogravimetry as a tool toclassify waste components to be used for energy generation. J Anal Appl Pyrol2004;71:883–900.

[53] Tang W, Wang C, Chen D. An investigation of the pyrolysis kinetics of somealiphatic amino acids. J Anal Appl Pyrol 2006;75:49–53.

[54] Chen X, Jeyaseelan SJ. Study of sewage sludge pyrolysis mechanism andmathematical modeling. J Environ Eng 2001;126:585–93.

[55] Muthuraman M, Namioka T, Yoshikawa K. A comparison of co-combustioncharacteristics of coal with wood and hydrothermally treated municipal solidwaste. Bioresour Technol 2010;101:2477–82.

[56] Coats AW, Redfern JP. Kinetic parameters from thermogravimetric data.Nature 1964;201:68–9.

[57] Conesa J, Marcilla A, Prats D, Rodriguez-Pastor M. Kinetic study of the pyrolysisof sewage sludge. Waste Manage Res 1997;15:293–305.

[58] Idris SS, Rahman NA, Ismail K, Alias AB, Rashid ZA, Aris MJ. Investigation onthermochemical behaviour of low rank Malaysian coal, oil palm biomass andtheir blends during pyrolysis via thermogravimetric analysis (TGA). BioresourTechnol 2010;101:4584–92.

210 G.K. Parshetti et al. / Fuel 111 (2013) 201–210

[59] Wang C, Wu Y, Liu Q, Yang H, Yin F. Analysis of the behavior of pollutant gasemissions during wheat straw/coal cofiring by TG–FTIR. Fuel Process Technol2011;92:1037–41.

[60] Nola GD, Jong WD, Spliethoff H. TG–FTIR characterization of coal andbiomass single fuels and blends under slow heating rate conditions:partitioning of the fuel-bound nitrogen. Fuel Process Technol 2010;91:103–15.

[61] Cao J, Xiao G, Xu X, Shen D, Jin B. Study on carbonization of lignin by TG–FTIRand high-temperature carbonization reactor. Fuel Process Technol2013;106:41–7.

[62] Giuntoli J, de Jong W, Arvelakis S, Spliethoff H, Verkooijen AHM. Quantitativeand kinetic TG–FTIR study of biomass residue pyrolysis: dry distiller’s grainswith solubles (DDGS) and chicken manure. J Anal Appl Pyrolysis2009;85:301–12.