Assessment of Uruguayan Oil Shales: physicochemical, thermal and morphological characterization

Fuel 234 (2018) 347–357

Contents lists available at ScienceDirect

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

Full Length Article

Assessment of Uruguayan Oil Shales: physicochemical, thermal and morphological characterization

T

⁎

Martín Torresa, , Jorge Castiglionia, Luis Yermánc, Leopoldo Suescund, Bruno Contib, Manuela Morales Demarcob, Pablo Gristob, Patrice Portugaua, Andrés Cuñaa Área de Fisicoquímica, DETEMA, Facultad de Química – Universidad de la República, Gral. Flores 2124 CC 1157, 11800 Montevideo, Uruguay Gerencia de Exploración y Producción, Administración Nacional de Combustibles, Alcohol y Portland, Av, Libertador Brig. Gral. Lavalleja y Paraguay, CP 11100 Montevideo, Uruguay c School of Civil Engineering, University of Queensland, St Lucia Campus, Advanced Engineering Building, Brisbane, QLD 4072, Australia d Laboratorio de Cristalografía, Química del Estado Sólido y Materiales, Area Física, DETEMA, Facultad de Química – Universidad de la República, Gral. Flores 2124 CC 1157, 11800 Montevideo, Uruguay a

b

A R T I C LE I N FO

A B S T R A C T

Keywords: Oil shale Thermal analysis Kinetics Combustion

Oil shales from Mangrullo Formation (Uruguay) was assessed as a potential fuel for energy production. The assessment is based on a comprehensive characterization of the material from a thermal perspective. Co-combustion of this type of fuel with biomass waste can enhance the combustion performance and decrease hazardous gas emissions. The low heating value (3.2 MJ kg−1) of the Uruguayan oil shale indicates that it could be mixed with biomass to obtain better results. Stratified sampling of drill cores was used to obtain a representative sample. The morphological and structural characteristics of the oil shale were studied by X-ray diffraction and scanning electron microscopy. Chemical composition (hydrocarbons and minerals) of oil shales was investigated by X-ray diffraction, X-ray fluorescence, proximate and ultimate analysis and Fourier transform infrared spectroscopy. Rock Eval analysis was performed to measure richness and maturity of the Uruguayan oil shale, the results of the TOC content (8.93%) and the hydrogen index (525 mg HC/ g TOC) indicated that it constitutes an excellent source rock. The kinetic of the combustion and thermal properties of oil shale were studied. The activation energy of the different reaction stages was calculated using the Flynn-Wall-Ozawa model, then we obtained the pre-exponencial factor and the reaction order optimizing the theoretical model using the experimental data. Results indicated that the combustion of the oil shale exhibits multiple reaction stages with activations energies that varies between 152.2 and 316.4 kJ mol−1.

1. Introduction Oil shales (OS) are fine grain sedimentary rocks that contain organic material of high molecular weight, called kerogen, disseminated in its inorganic matrix [1,2]. These rocks were formed by the accumulation of fine grain sediments and organic matter (such as plankton, algae or rest of plants) in anaerobic conditions, at the bottom of a lake or a sea. During the Early Permian age (Artinskian), a large inland sea developed in the territories that we know today as Uruguay, southern Brazil, Paraguay, South Africa and southern Namibia [3]. The restricted conditions of the sea allowed the accumulation and preservation of high quantities of organic matter, generating the deposition of OS and carbonates. In Uruguay these lithologies are grouped in Mangrullo Formation

⁎

[4], called Irati Formation in Brazil [5] and Whithill Formation in South Africa [6]. The Mangrullo Formation is located in northeastern Uruguay, in the departments of Cerro Largo, Rivera and Tacuarembó. Between 1970 and 1990, the National Oil Company of Uruguay (Administración Nacional de Combustibles, Alcohol y Portland-ANCAP) drilled 392 stratigraphic wells in that region to assess the quality of the OS. The studies performed by ANCAP were limited to the identification and quantification of different wells. Samples from the wells were collected for classification and chemical analysis. In some cases, the samples were characterized from a physicochemical point of view. However, these analyses were limited by the technology available at the time. These analyses were made to evaluate the volume of hydrocarbons that can be obtained through pyrolysis of the OS. Through this assessment, it was estimated that approximately 277 million barrels of

Corresponding author. E-mail address: [email protected] (M. Torres).

https://doi.org/10.1016/j.fuel.2018.07.031 Received 2 April 2018; Received in revised form 4 July 2018; Accepted 6 July 2018 0016-2361/ © 2018 Elsevier Ltd. All rights reserved.

Fuel 234 (2018) 347–357

M. Torres et al.

oil (mmbo), could be obtained by pyrolysis [7]. In the last few years, different researches about new applications for OS were reported internationally. Besides the generation of hydrocarbons through pyrolysis, a possible application of OS, used in some countries, is the generation of energy by the direct combustion of the material. Recent works reported that the co-combustion of OS or bituminous coal with biomass have a good performance for energy production with a decreases of the contaminants released to the environment [8–11]. Nowadays, the OS has industrial applications in China and Estonia. In China only for oil production by pyrolysis, but in Estonia also for energy production by combustion using pulverized firing and circulating fluidizing bed [12–18]. Taking into consideration the OS resources that develops in Uruguay and this new and attractive way of energy generation through co-combustion, we propose a comprehensively study of the applications of this resource. To reach the objective of the investigation proposed, it is important to cover the fundamental aspects related with engineering processes and chemistry, focusing on energy production through cocombustion of OS with biomass waste. In this work, the results of physicochemical, thermal and morphological characterization of the Uruguayan OS are presented. The general objective is evaluate the combustion properties of the Uruguayan OS for the possible future application in the co-combustion with biomass. This is the first time the kinetic of Uruguayan OS combustion is investigated. The results of this work could generate valuable information in other topics such as economic, geology and environmental aspects of the resource, to be consider in a potential exploitation of the resource.

Fig. 2. Stratigraphy of a representative wells.

Sample criteria was based on selecting wells whose second bituminous seam was at a depth up to 35 m and with high oil (>5%) and a low sulfur (<5%). This distance from the ground level takes into consideration a potential mining of the resource. Three wells fulfill that criteria. For this study, the well with the second OS seam closer to the ground level was chosen (approximately 12,5 m, location in Fig. 3).

2. Methodology 2.1. Raw material: selection and classification A set of rock samples (drill cores) corresponding to wells drilled by ANCAP was used (see Fig. 1). The stratigraphy of the wells (Fig. 2) shows two distinctive organic-rich levels (called bituminous seams), separated by a bed of siltstone and limestones. The deeper of these two bituminous seams always shows a higher organic content. A rigorous analysis of the wells was made to select samples for this work considering the future study of the co-combustion of OS with biomass. The selection of samples was made taking into consideration the depth of the OS levels and physicochemical data (Fischer Assay and Sulfur content) generated by ANCAP in previous studies.

2.2. Total organic carbon analysis, Rock Eval pyrolysis, Vitrinite Reflectance and Kerogen Microscopy The total organic carbon (TOC) analysis was performed using 0.15 g of sample. The sample was treated with HCl with the purpose of removing carbonates and then was filtered using fiberglass paper. The

Fig. 1. Map of Uruguay showing the onshore sedimentary basins (in yellow). The black triangle indicates the area explored by ANCAP. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. Map of Uruguay with the location of the selected well indicated in red. Coordinates UTM22S (WGS84) x = 222288, y = 6450242. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 348

Fuel 234 (2018) 347–357

M. Torres et al.

M-5900LV equipment under 20 kV of voltage supply. The elemental composition of different zones was made with an Energy Dispersive Spectrometer probe (NORAN instruments EDS poor advantage). Transmission electron microscopy (TEM) was carried out with an electronic microscope HRSTEM JEOL 2100, with a voltage supply of 200 kV, LaB6 filament with a CCD GATAN Orius 100 camera. Also, the elemental composition of the sample in different levels was analyzed by an EDS probe Oxford XMax 65T, silicon Drift detector of 65 mm2.

residue and the paper were dried and later introduced in a LECO C230CH carbon analyser at 1000 °C. The Rock–Eval pyrolysis technique followed in this study is based on the methodology described by Espitalié et al. [19,20] and NuñezBetelu & Baceta [21]. This technique provides data on the quantity, type, and thermal maturity of the associated organic matter. This technique was performed using 100 mg (± 0.1 mg) of sample, then the sample was placed into stainless steel crucibles. After, the sample was placed into the oven under the following temperature program in an inert atmosphere: 3–4 min at 300 °C, then pyrolyzed at 25 °C min−1 to 600 °C. For the analysis a Rock-Eval pyrolyzer model SRA TPH, was used. The pyrolysis values collected on the computer are presented in a table that includes values such as Tmax , S1, S2, S3, TOC, HI, OI. The definition of these parameters could be found in the works of Espitalié et al. [19,20] and Nuñez-Betelu & Baceta [21]. The hydrogen and oxygen index were used as indicators of kerogen type in the Van Krevelen diagram. For the kerogen microscopy, the rock samples were grounded to pass a 40-mesh sieve and then treated with a 30% hydrochloric acid solution. Then the samples are washed to neutrality and treated with a 70% hydrofluoric acid (HF) solution with the aim of remove the carbonate and silicate mineral phases. The sample was stirred intermittently for 24 h, filtered and a second 70% HF quote was added and the digestion repeated as explained [22]. The sample was then rinsed to neutrality and the organic matter that floats on a heavy liquid, was removed and prepared as a kerogen slide and a vitrinite plug. Then the sample was examined under halogen light, the plug reveals the percentage of vitrinite reflectance (%Ro) that is used to determine the coal rank and shale thermal maturity. This data is then cross-referenced with the Thermal Alteration Index (TAI) to determine the thermal maturity of the samples based on the colour of the palynomorphs. [23,24].

2.5. Fourier transform infrared spectroscopy Fourier transform infrared (FTIR) spectra were obtained with a pellet of sample diluted in KBr at 5% in a range of 4000–400 cm−1 with a Shimadzu, IR Prestige-21 Fourier transform infrared spectrophotometer.

2.6. Proximate and ultimate analysis The quantification of the percentage content of hydrogen, carbon, nitrogen and sulfur was performed with CHNS/O Thermo Scientific FLASH 2000 instruments. Oxygen was determined by difference. The proximate analysis was performed using the norm UNIT NBR 8112/ 1986. Ceramic capsules and an Isotemp muffle furnace of Fischer Scientific were used for this analysis.

2.7. Heating value The High heating value (HHV) was determined with a Calorimetric bomb, model Parr 1341, equipped with a 6672 precision thermometer. The determination was performed with 0.5 g of sample. For each determination the correction for heat of formation of acid was applied according to the calorimetric bomb manual [29]. The lower heating value (LHV) was determined from the HHV and the results from the elemental analysis.

2.3. X-ray diffraction and Rietveld analysis Samples of OS and Oil Shale Ash (OSA) were analyzed by X-ray powder diffraction, with the aim to perform a quantitative phase analysis by the internal standard method [25] and Rietveld analysis [26]. Y2O3 powder (Sigma-Aldrich, 99.999% fired in air at 1200°C for 72 h) was used as internal standard for determination of the amorphous content. X-ray powder diffraction data was collected on a Rigaku ULTIMA IV Powder Diffractometer in θ –θ geometry operating with CuKα (40 kV/30 mA) X-ray sealed tube and a curved Ge diffracted beam monochromator with a focal distance of 285 mm. A sample of OS and other of OSA were weighted and ground-mixed with a weighted amount of Y2O3. Weight percentages of Y2O3 in OS and OSA were 14.945% and 6.410% respectively. Native and standard-added samples were placed in a glass sample holder and mounted in the center of the diffractometer. Data was collected in the 7.5–90o 2θ range in steps of 0.02 o in 2θ and 15 s/step. Phase identification was performed using Crystallographica Search-Match 3.1.0.2 Software and the Powder Diffraction File PDF-2 (ICDD-2013). After phase identification, data was fitted by the Rietveld method using GSAS/EXPGUI suite of programs [27,28] for quantitative phase analysis. Refinements were performed using a shifted Chebyshev polynomial (21 elements) for each pattern and modified Thompsos-Cox-Hastings pseudo-Voight peak shape corrected for asymmetry. A common sample shift was refined for all the phases in the same pattern. Gaussian GU and Lorentzian LX and LY parameters and Lij anisotropic broadening parameter for Muscovite (see below) were refined and constrained among the same phase in different patterns.

2.8. X Ray fluorescence (XRF) The XRF technique was performed using a spectrometer, Spectro Xepos, equipped with three secondary targets (Mo, Al2O3, HOPG) controlled by a computer system and a Silicon drift detector with a 140 eV resolution (Mn Kα line) and a 8 μm Moxtek Dura – Berilium window. An X-ray tube operating up to 50 kV and 3.3 A produces the primary photon beam. X-ray line intensities and elemental concentrations were calculated using the software Spectro X Lab Pro. The Compton model was used for the calculations. In the spectra, chemical elements were identified and quantified with different level of accuracy and precision. Reference materials as IAEA SOIL 7 and NIST 2704 were used for instrument calibration.

2.9. Gamma-ray spectrometry Gamma-ray spectrometry was performed using an HPGe coaxial semiconductor detector of relative efficiency of 32% and FWHM resolution of 1.85 keV, both for the 1332 keV photons of 60 Co. The detector was shielded using a lead shielding assembly (thickness 100 mm) and was coupled to a gamma-ray spectrometric system controlled by computer with Canberra Genie 2000 software. To achieve equilibrium with the radon daughter radionuclides the samples were stored in closed vials wrapped with aluminum foil for 1 month before the measurement was carried out. The counting time was 54.000 s. As calibration source was used a soil spiked with QCYB41. Reference materials IAEA 375 was used as a standard for calibration.

2.4. Scanning electron microscopy and transmission electron microscopy Scanning electron microscopy (SEM) was performed with a JEOL JS 349

Fuel 234 (2018) 347–357

M. Torres et al.

2.10. Thermogravimetric Analysis (TGA) and Differential Thermal Analysis (DTA) TGA analysis was performed using a SHIMADZU TGA-50 thermal analyzer at heating rates of 5, 10, 20, 30, 40 and 50 °C min−1 with temperatures ranging from 25 °C to 950 °C. Compressed air at 50 mL min−1 was used for combustion experiments. About 15.0 ± 0.2 mg sample was weighed and placed in a platinum crucible. The OS sample was micronized and passed through a 50 mesh sieve to eliminate the effect of temperature distribution in the sample [30]. Kinetic analysis was performed to further study the kinetic parameters of OS combustion. The fundamental rate equation used in all heterogeneous solid state reactions can be expressed as the Arrhenius equation:

Fig. 4. A) OSA sieved with a 75 μm grid. B) OSA sieved and mixed with agglomerant. C),D) OSA in cone molds.

E ⎛ α⎞

⎛ dα ⎞ = k (T ) f (α ) = Ae−⎝ RT ⎠ f (α ) ⎝ dt ⎠

(1)

Where t is the time, Eα is the apparent activation energy, A is the pre-exponential factor, T is the reaction temperature, R is the universal gas constant, f(α ) is the reaction model which describes the dependence of the reaction rate on the extent of reaction and α is conversion degree which is described as:

α=

mo−mt mo−m∞

Fig. 5. OSA cones placed on a ceramic tray ready for thermal treatment.

(2)

Where m o , mt and m∞ were the initial sample mass, sample mass at time t and sample mass at the end of reaction, respectively. The activation energy Eα of the OS was calculated by the FlynnWall-Ozawa(FWO) [31,32] method using the Eq. 3:

AEα ⎞ E log (β ) = log ⎛ −2.315−0.4567⎛ α ⎞ ⎝ RT ⎠ ⎝ RG (α ) ⎠ ⎜

triangular ash cones (Fig. 4). The temperature was selected according to the surrounding operational conditions of an industrial boiler [33]. Ash cones were then placed into the analyzer and heated to 1400 °C at a heating rate of 8 °C min−1 for measuring with a flow rate of 2–3 mL/ min. Four characteristic temperatures were measured for OSA: deformation temperature (DT), softening temperature (ST), hemispherical temperature (HT), and flow temperature (FT). As shown in Fig. 5, each cone was mounted on a ceramic tray and placed into a high temperature furnace selecting an analytical method with a predefined furnace atmosphere (oxidizing or reducing). A high resolution digital camera collects images after the furnace temperature reaches the method defined starting point. The equipment records automatically all the deformation temperatures of the sample. Fusion temperatures were measured at four defined points under both reducing and oxidizing conditions. The DT is the temperature at which the point of the cone begins to round, ST is the temperature at which the base of the cone is equal to its height, HT is the temperature at which the base of the cone is twice its height and FT is reached when the cone has spread to a fused mass no more than 1.6 mm in height. The absolute density of the OS and OSA was measured using a gas pycnometer (Gas Pycnometer, Model AccuPyc II 1340 Series, Micromeritics, GA, USA). Before the determination, the samples were dried at 105 °C for 24 h.

⎟

(3)

Where β is the heating rate, A is the pre-exponential factor, Eα is the activation energy, R is the universal gas constant, G(α ) is the integrated form of the reaction model and T is the temperature of the reaction. By plotting logβ vs 1/T at selected α , it is possible to get the slope at different α values. Eα values corresponding to α values can be obtained for the different processes that are taking place during the combustion of the OS. The DTA analysis was performed using a SHIMADZU DTA-50 at two heating rates (5 and 50 °C min−1) with an air flow rate of 50 mL min−1 from 25 °C to 950 °C. The experiment was conducted with 15 ± 0.2 mg of sample and the same weight of α -Al2O3 was used as reference. For this analysis, platinum crucibles were used. This analysis was made with the objective of knowing the thermal behaviour of the different processes that are taking place. For the calculation of the Eα error, we had obtained the Eα at different α values using the FWO model. For each process the Eα values obtained are not strictly constant, due to this fact the Eα values reported are an average value plus-minus the ‘S’ value times 2 and divided by the square root of the data amount.

S ⎛ ⎞ ⎜X ± k (n−1) ⎟ ⎝ ⎠

3. Results and discussion 3.1. Physicochemical analysis Table 1 shows the results of C, H, S, O, N, ash, volatile matter, fixed carbon, HHV and LHV obtained for the Uruguayan OS. The moisture content for the Uruguayan OS was (9.1 ± 2.0)% on wet basis. Comparing the C content for different OS, the result indicated that the Uruguayan OS shows similar carbon content than the reported for Nongan (China) and for Kentucky (USA), with a difference of 0.4% and 1% respectively. Moreover, comparing the Uruguayan OS with the Brazilian Irati formation, the Uruguayan shows a difference of 7.2% in this value. The Uruguayan OS presents ash content similar to the reported for Brazil, but higher than the Estonian and Huadian OS, and also lower than the Kentucky and Nongan OS. The Uruguayan OS present a S content 2.4% and 1.3% greater than the reported for Estonia and Kentucky respectively, nevertheless is comparable with the reported for Irati and Huadian. These similar characteristics with the

(4)

Where X is the average of the Eα values calculated for each α , k is a constant whose value is 2, S is the standard deviation of the Eα values and n is the number of values used to obtain X . The same methodology was used for the other kinetic values reported. 2.11. Ash fusion analysis and absolute density determination Ash fusion temperatures were measured by a LECO’s AF700 ash fusibility determinator under an oxidizing (air compressed) and reducing (50% CO/ 50% CO2) atmosphere according to ASTMD1857. Oil shale ash (OSA), obtained at 700 °C in air atmosphere was sieved using a 75 μm and mixed with dextrin solution and then shaped into 350

Fuel 234 (2018) 347–357

M. Torres et al.

Table 1 Chemical properties of the Uruguayan OS compared with other OS of the world. Region

Mangrulloa (Uruguay) Huadianb (China) Iratic (Brazil) Nonganb (China) Estoniad, e, f Kentuckyg (USA)

Ultimate analysis (wt%)

Proximate analysis (wt%)

Calorific value (MJ kg−1)

C

H

Oγ

N

S

VM

FCφ

Ash

HHV

LHV

9.3

2.3

11.6

0.2

4.0

24.0

3.4

72.6

3.2

2.8

29.2

4.3

4.1

0.6

4.9

39.3

3.8

56.9

*

13.1

16.5

1.8

0.3

0.5

4.0

*

*

76.9

*

*

9.7

1.0

–

0.3

5.7

9.7

1.6

89.3

*

2.9

27.4 – 30.4 10.3

2.7 – 3.0 1.3

23.6 – 27.7 1.7

0.07 – 0.1 0.4

1.6 2.7

47.5 – 49.6 *

1.1 – 1.3 *

49.3 – 50.7 83.6

8.6 – 11.9 *

8.5 – 11.0 *

VM: volatile matter, FC: Fixed carbon. All values are expressed in dry basis. φ : Determined by difference: FC = 100%−VM%−Ash%. γ:

Determined by difference: %O = 100% - C% - H% - N% - S% - Ash%.

References: a – obtained in this work, b – [34], c – [35], d – [36], e – [37], f – [12] g – [38]. *: no data reported.

gas prone) but, in the location analyzed, they are just entering the oil window, having generated a small amount of free hydrocarbons (S1). Regarding to a possible future application of the OS by pyrolysis technologies, the pyrolyzed carbon (PC) that is defined as (S1 + S2)/10 could be used as an indicator of the amount of carbon that could be pyrolised. For the Uruguayan OS, these value is 4.894 (% g HC/grock ). Comparing this result with others reported for the Irati Formation of the Parana Basin [40], we can see similarities on the PC, for example the average value obtained of the data reported for the geochemical unit H are closer to the value obtained in the present work, the value is 4.162 (% g HC/grock ). According to this work, this comparison indicates that the Uruguayan OS has an excellent concentration of organic matter with excellent potential for oil generation. Furthermore, the data obtained indicates that the Uruguayan OS could be attractive for develops studies focused on pyrolysis technologies. A Search-Match procedure allowed the identification of the main phases in OS and OSA samples independently (Table 3). Due to the nature of the material analyzed (OS), the exact phase composition it was impossible to determinate using the available data, since a complete mineralogical and chemical characterization of the sample was not available. Main minerals found in the Search-Match procedure (quartz, muscovite, pyrite, potassium-iron hydroxyl sulphate) were consistent with previously described minerals in OS materials from the region [4]. Minor phases, such as a feldspar mineral in the albite-microcline series was not uniquely identified since the X-ray fluorescence analysis available could not identify the presence of Na and K. Diffraction peaks with maximum intensity above 1.5 times the background signal were identified, crystallographic information was

Brazilian OS could be explained due to the fact that both were deposited in the same region, age and depositional environment. The Uruguayan OS has a low HHV and LHV which implies that it is not appropriate to generate energy through direct combustion preferred cocombustion to enhance heating value and diminish hazardous emissions. The similarity of the LHV of the Uruguayan shale compared to Nongan OS could be explained based on similar carbon content. The Uruguayan OS has an higher oxygen and hydrogen content compared with Nongan, Irati and Kentucky suggesting that there is more amount of hydrocarbon and oxygenated compounds in the Uruguayan OS. With the aim of evaluate the hydrocarbon generation potential, Rock–Eval analysis was performed and it is reported in Table 2, OS has a high total organic content (TOC > 5%), representing an excellent potential source rock [23,39]. However, the S1 value shows a low amount of free hydrocarbons and the S2 value indicates a high hydrocarbon generation potential. Tmax value from Rock Eval pyrolysis was abnormally low and out of the interpretative range without giving confident information about the thermal maturity of the OS. The Hydrogen Index (HI) and Oxygen Index (OI) when are plotted in the Van Krevelen Diagram show a dominantly type II kerogen (oil and gas prone), indicating a marine depositional environment for the OS. The Thermal Alteration Index (TAI), based on coloration of pollen and spores, shows values of 2+/3- indicating that the sample is in the early oil window. Furthermore, the Vitrinite Reflectance analysis (Ro%) shows a mean value of 0.61. This Ro% value also suggest an early mature stage of the OS [23]. Overall, these results indicates that the OS of Mangrullo Formation represents a good quality source rock (oil and

Table 2 Organic richness and hydrocarbon generation potential results. TOC (%)

TIC (%)

S1 (mgHC/grock )

S2 (mgHC/grock )

S3 (mgCO2/grock )

Tmax (°C)

S1/TOC

HI (mg HC/g TOC)

OI (mg CO2/g TOC)

(8.93 ± 0.45)

(0.37 ± 0.02)

(2.02 ± 0.46)

(46.92 ± 4.69)

(0.48 ± 0.07)

(399 ± 3)

23

525

20

TOC: total organic carbon. TIC: total inorganic carbon. HI: hydrogen index. OI: oxygen index. 351

Fuel 234 (2018) 347–357

M. Torres et al.

majoritarian phases identified. Fig. 6 shows the fit of the data for native samples of OS and OSA. Fits of standard-added patterns and details on the weight percentages obtained for standard-added patterns are included in the Supplementary Information. The FTIR spectrum of the OS is shown in Fig. 7. Various functional groups were identified in our interpretation of the FTIR spectra of raw Uruguayan OS. It consists of stretching and bending vibrations from the aliphatic and aromatic groups of kerogen. At 3420 cm−1 we could identify the H-O–H stretch vibrations of the absorbed water of the sample. Kerogen shows a set of absorption bands at 2926 cm−1, 2855 cm−1 and 1464 cm−1 (aliphatic C-H stretching), at 1636 cm−1 (C = C stretching) and at 1034 cm−1 due to the sulfoxide stretching (S = O) [43–45]. Regarding the inorganic matrix, the peak associated with the presence of silicates at 1879 cm−1 could be [46]. The peaks at 3620 cm−1 (OH stretching, crystalline hydroxyl) and 916 cm−1 (OH deformation, linked to 2Al3+) are attributed to the Al-O–H present in the Muscovite like a possible crystalline phase [47,48]. Those last bands together with the presence of 527 cm−1 and 470 cm−1 bands could be attributed to Montmorillonite [49]. The bands at 1163 cm−1 and 426 cm−1 are associated with the presence of pyrite [43,49]. Finally, the peaks at 778 cm−1 and 470 cm−1 can be attributed to feldspar, bonds from Si-OSi and Si-O respectively [49]. Furthermore, is important note that comparing the results obtained by DRX with FTIR analysis, various functional groups observed at the solid surface are in agreement with some crystalline phases obtained by DRX, especially those of inorganic minerals. SEM analysis was performed with an uncrushed sample of a representative Uruguayan oil shale particle (Fig. 8). Table 6 shows the atomic composition of the different zones (A, B and C) analyzed by SEM-EDS. Results showed that the OS has heterogeneous composition with zones with high carbon concentration (between 30–40%) an others with no carbon but high concentration of others elements such as Si, Al, Fe and S. The TEM analysis (shown in the Supplementary Material) confirms the heterogeneity in the composition of the OS. In addition, the HR-TEM shows that the OS is composed by both crystalline and amorphous phases. The results related to XRF analysis of the OS are presented in Table 7. The most important metals identified through this analysis were Al, Si, K and Fe. Silicon was the most abundant element with a content of 18.8% consistent with XRD and FTIR results that may indicate presence of quartz (SiO2) in addition to clay minerals. Having in mind the content of some base metals in the OS combustion ashes (Cu, Pb, Zn), it may be worth to analyses their potential to produce them as raw material. According to [50], copper, one of the most important base metals, can be mined in deposits with concentrations up to 0.5%, and approximately 0.3% when other valuable metals

Table 3 Phases identified in OS and OSA samples (native and Y2O3 added powder patterns). Main phases

OS

OSA

Quartz,SiO2 Muscovite Pyrite K(Fe, Al)3(SO4)2(OH)6 Hematite, Fe2O3 Na,K Feldspar

✓ ✓ ✓ ✓ – ✓

✓ ✓ – – ✓ ✓

✓ – ✓ ✓

– ✓ – –

Minor phases (less than 0.5% weight) Nontronite or Illite Cristobalite Zeolite Gypsum CaSO4.2H2O

Table 4 Fit quality indicators parameters for the Rietveld analysis. Parameter

OS

OS + Y2O3

OSA

OSA + Y2O3

(%) Rp (%) wRp RBragg

8.22 10.49 8.06

7.77 10.55 5.74

9.06 11.81 6.01

7.80 10.11 4.34

Combined χ 2 Novariables

5.641

4.473

81

84

Table 5 Calculated weight percentage of majoritarian phases identified in OS and OSA after quantitative phase analysis with estimated uncertainty. Phase

wt% in OS

wt% in OSA

Quartz, SiO2 Muscovite Na,K Feldspar Pyrite K(Fe0.7 Al 0.3 )3(SO4)2(OH)6 Hematite, Fe2O3 Amorphous and minority phases Total Fe w%

22.0 ± 3.0 25.0 ± 3.0 2.5 ± 0.9 0.5 ± 0.3 9.4 ± 2.7 – 40.6 ± 3.0 2.5 ± 1.0

27.0 ± 3.0 45.0 ± 3.0 14.0 ± 3.0 – – 3.9 ± 1.5 10.1 ± 3.0 3.1 ± 1.0

obtained from AMCSD [41] and Crystallography Open Database [42] open databases in CIF format to allow a quantitative phase analysis by the Rietveld Method. A simultaneous fit of native and standard-added datasets with equal relative proportions of the main phases identified in Table 3 was performed. The fit quality indicator parameters are shown in Table 4. Table 5 contains the calculated weight percentages of the

Fig. 6. Rietveld fit of OS and OSA. 352

Fuel 234 (2018) 347–357

M. Torres et al.

Fig. 7. FTIR spectra of the Uruguayan OS.

Table 8 Results of ash fusion analysis. Temperature(°C) Atmosphere

IT

ST

HT

FT

Oxidizing Reducing

1112 1024

1317 1168

1346 1269

1371 1394

All the temperatures have an error of ± 5 °C.

should consider that combustion ashes would be ready for metal benefaction, with no need of mining activities. It would be also interesting to study other potential applications for combustion ashes, such as the extraction and refinement of rare earth elements, if present in substantial concentrations [51]. It may be worth to study more profoundly the presence of heavy metals in the OS of Mangrullo Formation, and in particular in OS combustion ashes, with regards to their environmental risk and toxicity. The chemical structure of those heavy metals may be relevant to their associated risk. The relatively high sulfur content of Mangrullo OS, anticipates a significant concentration of sulfur oxides associated to combustion air emissions. The absolute density value determined for OSA was slightly higher than for OS: (2.687 ± 0.004) g cm−3 and (2.237 ± 0.001) g cm−3 respectively. These values could be used as reference for large fired boilers designs in the heat and mass transfer balances. The increase of the density value is due to the decomposition of the organic matter whose density is less than obtained for water. The sample of OSA contain inorganic material and the density of the oxides formed are greater than the water density, thus the density values increase. The value of the OS density obtained in this work, according to Q.Wang et.al [52], indicates that the Uruguayan OS is not suitable for being retorted to yield oil, and is expected in future analysis obtain low aliphaticity

Fig. 8. SEM micrograph of Uruguayan OS. A, B and C shows the analyzed zones by EDS.

Table 6 Atomic composition of the different zones analyzed by SEM-EDS. Weight (wt%) Zones

C

O

Mg

Al

Si

S

K

Ca

Ti

Fe

Zn

Cu

Pd

A B C

0.0 32.1 0.0

50.5 46.2 43.0

1.6 0.2 0.0

7.4 1.9 3.6

24.3 5.9 10.8

7.9 1.9 6.7

0.0 0.7 1.1

0.0 0.7 3.2

0.0 0.3 0.4

6.2 9.6 31.1

0.6 0.0 0.0

0.0 0.5 0.0

1.5 0.0 0.0

The values are approximated because it is a semi-quantitative analysis.

are present. This author affirms that lead and zinc are usually mined together, due to the fact that most of their deposits contain both base metals, with cut-off grades between 6 to 10% of combined metal content. Results of XRF for Mangrullo OS indicate actual concentrations of base metals well below those cut-offs, but a complete economic analysis Table 7 X-ray Fluorescence results. Element

Al

Si

Sample/units OS OSA

S

K

Fe

Cu

Zn

Rb

18.8 28.1

5.4 1.9

Y

Zr

Ba

Pb

21.3 31.0

120 179

214 297

25.7 40.3

μg/g

wt% 3.3 6.4

Sr

<3 <3

4.9 7.2

39.8 59.1

77.1 116

All reported values are given with an error of 10% of the value. 353

106 155

151 222

Fuel 234 (2018) 347–357

M. Torres et al.

Fig. 9. TGA curves at different heating rates of OS in air.

Fig. 10. DrTGA curves at different heating rates of OS in air.

The high HT and FT is more characteristic of sialic ashes. On the other hand, low HT or FT are typical of calsialic, ferricalsialic and ferrisialic ashes [53]. The grater content of minerals with a high melting point like the silicate causes an increase in the FT of the OSA. During the heating in an oxidizing atmosphere, the OSA cone suffer a volume rise, which means that the sample reacts with the atmosphere during oxidation. After the initial volume growth, the sample decreases its volume and the process continues until the FT is reached. This effect is shown in videos supplied as Supplementary Material. In all cases except FT, the observed temperature under reducing conditions is lower than the corresponding temperature under oxidizing

and average methylene chain length. The information provided by the ash fusion analysis could be useful for boiler design. This analysis gives the maximum operation temperature at which ash deposition does not occur in large fired boilers. Thus, it avoids slagging problems on combustion chamber and pipe surfaces declining the heat transfer efficiency. The results of the ash fusion analysis are summarized in Table 8. The relationship between ash-fusion temperature (FT) and mineral and chemical composition of OSA is in agreement with the results of DRX analysis. The composition obtained by DRX can explain the behavior of the OSA with the temperature. 354

Fuel 234 (2018) 347–357

M. Torres et al.

Fig. 11. DTA (filled line) coupled with TGA (dashed line) for two heating rates. Table 9 TGA reaction intervals, peak temperatures (T′max ) and mass losses of OS combustion (except for initial water loss). Second stages

Third stages

Fourth stages

β (°C/min)

′ a Tmax (°C)

loss

Interval

Interval

(°C)

(wt%)

(°C)

′ a Tmax (°C)

loss

(wt%)

′ a Tmax (°C)

loss

(°C)

5 10 20 30 40 50

230–533 250–546 255–567 262–571 274–572 273–576

387 402 414 423 437 and 453 451

15.2 15.0 15.4 15.3 15.5 15.6

533–635 546–653 567–670 571–676 572–683 576–686

593 612 629 639 649 649

2.8 2.7 2.6 2.2 2.2 2.0

635–924 653–944 670–940 676–949 683–942 686–948

679 693 710 721 732 741

4.7 4.7 4.6 4.7 4.6 4.7

a

Interval

(wt%)

The temperature at which the maximum mass loss rate is reached.

namely: water evaporation, volatile matter release/decomposition, combustion of fixed carbon and decomposition of inorganic minerals. The sample exhibits similar patterns of thermal decomposition with increasing temperature at varying heating rates.The overall mass losses during the heating process at different heating rates are almost the same, indicating that the heating rates have little effect on the total mass loss. During the heating process, a slight weight loss occurred between room temperature and approximately 180 °C, which is attributed to the loss of moisture, including free water and interlayer water from clay minerals. The water evaporation is an endothermic process and is in agreement with the DTA curve (Fig. 11a and b). In presence of oxygen a thermochemical process combining thermal and oxidation effects develops simultaneously, shown in Table 9 and Fig. 9. The second stage that appear between 230–580 °C is mainly the thermal oxidative degradation of the volatile matter in the OS, which results in oxidation of some organic matter directly within the sample and released and burnt a significant amount of low molecular weight hydrocarbons. The chemical reaction in this range is complex with a composition change. This stage exhibits the maximum weight loss of the overall process and a exothermic behavior (Fig. 11a and b). The third stage for OS in the temperature range 530–690 °C is due to the oxidation of the fixed carbon. The fourth stage is associated with the decomposition of the inorganic material of the mineral matrix at temperatures above 635 °C, the destruction of the ash compound begins (mineral decomposition) which is apparent for oil shales with high ash contents [56]. This process is generally exothermic, this effect could be

conditions. This effect can be attributed to the iron present in the sample [54]. For the Uruguayan OS, the maximum operation temperature when its burned in a boiler should be less than 1317 °C under oxidizing atmosphere. These value obtained is consistent with the operational temperature range of a large fired boiler. Regarding to gamma ray spectroscopy results for the OS and OSA samples shows the presence of radionuclides that exist naturally in the Earth’s crust. These radionuclides are mainly isotopes of 232 Th, 238 U series and their naturally decay series and also of 40 K. After reaching the secular equilibrium, the calculated activity of 232 Th from 212 Pb (238.63 keV) and the activity of 226 Ra from 214 Bi (609.31 keV) were (39.8 ± 3.0) Bq kg−1, (66.9 ± 4.5) Bq kg−1. The activity of 40 K was (526 ± 85) Bq kg−1. For OSA the results were 232 Th (54.6 ± 3.5) Bq kg−1, 226 Ra (100 ± 6) Bq kg−1 and 40 K (679 ± 100) Bq kg−1. All the values are expressed per kilogram of OS. If we compare these values obtained with the emission of coffee, coal ash and uranium, the values are 1000 Bq kg−1, 2000 Bq kg−1 and 25 x 106 Bq kg−1 respectively [55]. It should be noted that these results for the OS and OSA indicate that both are not dangerous for the environment regarding its radioactivity. 3.2. Thermal and kinetic analysis of OS combustion The experimental TGA curves of OS at different heating rates are shown on Fig. 9. The heating process could be divided into four stages 355

Fuel 234 (2018) 347–357

M. Torres et al.

Table 11 shows the mass loss during the heating of the OS for the different stages obtained using the model developed and the values calculated from the experimental data. In spite of the fact that exist a little difference between the total experimental mass loss and the total mass loss obtained from the model, maximum difference of 11%, and having in mind the global process (stages II, III and IV), the developed model fit well to the heating process of the OS. This model could be a good tool to design combustion equipment that uses the OS as fuel.

Table 10 Kinetics parameters obtained for the different stages of the Uruguayan OS combustion. Stage

Ea (kJ mol−1)

Log(A)

n

II III IV

(152.2 ± 8.67) (293.3 ± 17.3) (316.4 ± 39.4)

(9.99 ± 0.28) (16.26 ± 0.40) (15.54 ± 0.74)

(3.83 ± 0.21) (1.86 ± 0.12) (4.03 ± 0.23)

seen in Fig. 11a and b. It should be noted that the weight loss for volatile matter (stage two), fixed carbon (stage tree) and overall mass loss are according with the values obtained in the proximate analysis. Fig. 10 shows DrTGA curves. When the heating rate is low, the OS particles are gradually heated, this effect generate a good heat transfer to the inner portions. According the heating rate increases, the total temperature change between the particle surface and the inner portions grows, this effect shortened the combustion times of the OS. On the other hand, the increases of the heating rate causes that the mass and heat transfer resistance rising up. This demonstrate the inertia effect of the devolatization observed in Fig. 9 and Fig. 10. Due to this fact, the ′ values of Tmax increases. Nevertheless, the OS underwent a heavy thermal shock according the heating rate rising up, this fact causes that the reaction rate with the oxygen presented in the inner portions was improved and the combustion intensity was enhanced. Furthermore, the DrTGA curves fluctuates as the temperature increases, showing multi sub-peaks in the second stage and becoming more obvious as heating rate increases [34]. The activation energy of each stage during the heating process for the Uruguayan OS was obtained from the slope using the FWO method at different α values, then we calculate the average activation energy. Futhermore, we have supposed a reaction order for the different reaction stages and then we obtained the pre-exponencial factor for the Arrhenius model. With these values we proceed to build the theoretical TGA curve, then we minimize the difference between the experimental data and the theoretical values varying the reaction order for each heating rate. The results of the average values for each stage of the Ea, pre-exponencial factor (A) and reaction order (n) for the Uruguayan OS are summarized on Table 10. The values obtained for the Eα successively increased from the stages II to IV, indicating that the combustion reaction becomes less favourable from a kinetic point of view, as the reaction proceeded. The error values are similar to others reported in the literature [43]. The low activation energy for the stage two, could be attribute to the volatile decomposition, then the following processes may require more activation energy to burn, such as fixed carbon. The Eα value for the stage IV was the higher, suggesting that OS needed higher temperature to burnout. It was also observed that the kinetic results are close to the other research findings in the literature from the point of view of OS combustion [10,34,56,57].

4. Conclusions In this work morphological, crystallochemical and physicochemical characteristics of the Uruguayan OS were studied. Kinetics of the combustion process of the OS were also determined. The material shows a heterogeneous distribution of their components with a wide range of different metals. The Rock Eval analysis allowed characterize the OS like an excellent source rock just entering the oil window, having generated a small amount of free hydrocarbons. The Uruguayan OS presented low HHV (3.2 MJ kg−1) and multiple combustion stages. The result of the HHV suggest that the organic matter is heavy diluted in the inorganic matrix. This is one of the reasons whereby in a future will be attractive develops further studies for the Uruguayan OS. It could be used as a solid fuel for energy generation by mixing it with other solid fuels, such as biomass waste, in cocombustion facilities. This allow the energy recovery. Apart from studying the particular application of Mangrullo OS as a fuel for co-combustion processes, and identifying some of the challenges of this application based on the chemical properties, this work allowed to detect a series of necessary knowledges, which may evolve into research ideas or projects, for the valorization of the Mangrullo OS and their combustion by-products. Considering the characteristics of Uruguayan OS together with the increasing industrial technologies and costs, a new research for understand the behavior of co-combustion of OS with biomass need to be proposed.

Acknowledgments The present work was carried out with the support of Agencia Nacional de Investigación e Innovación (ANII project FSE–1–2016–1–131635). M. Torres thanks the Uruguayan ANII for the grant received (POS–NAC–2016–1–129893) and the PEDECIBAQUIMICA for the economic support. The authors thank to Dr. Pedro Curto of IIMPI, Faculty of Engineering, Universidad de la República for the Ash Fusion Analysis, Dr. Jossano Saldanha Marcuzzo from Instituto Nacional de Pesquisas Espaciais (Brazil) for the density measurements and PHC. María del Rosario Odino from Laboratorio de Tecnogestión for the XRF and Gamma-ray analysis.

Table 11 Values obtained using the parameters obtained with the FWO model and compared with the experimental values. Stage II

III

IV

Total mass ratio

°C min−1

mmod (mg)

m exp (mg)

mmod (mg)

m exp (mg)

mmod (mg)

m exp (mg)

mT , mod /mT , exp

5 10 20 30 40 50

13.9 13.9 14.4 14.3 14.6 14.7

15.2 15.0 15.4 15.3 15.5 15.6

2.1 2.4 2.4 1.4 1.6 2.0

2.8 2.7 2.6 2.2 2.2 2.0

4.3 4.2 4.3 4.5 4.4 4.4

4.7 4.7 4.6 4.7 4.6 4.7

0.89 0.92 0.94 0.91 0.92 0.95

mmod :mass loss predicted with the model, m exp : mass loss experimental measured, mT , mod : total mass loss predicted with the model, mT , exp :total mass loss experimental measured. 356

Fuel 234 (2018) 347–357

M. Torres et al.

Appendix A. Supplementary data [30]

Supplementary data associated with this article can be found, in the online version, athttps://doi.org/10.1016/j.fuel.2018.07.031.

[31]

References

[32]

[1] Scouten CS, Speight JG. Fuel science and technology handbook. Part IV: Oil Shale. Chapters, pages 25–31; 1990. [2] Allred VD. Oil shale processing technology. 1; 1982. [3] Anderson ME, Long John A, Evans Fiona J, Almond John E, Theron Johannes N, Bender PA. Biogeographic affinities of middle and late devonian fishes of south africa. Records Western Australian Museum, Supplement 1999;57:157–68. [4] Goso H, Bossi Cenozoico JC. Geoloǵa del Uruguay. Montevideo: Universidad de la República; 1966. p. 259–305. [5] Sérgio Estanislau do Amaral. Geologia e petrologia da formação irati (permiano) no estado de são paulo. Boletim IGA 1971;2:03–81. [6] Cole DI, Basson WA. Whitehill formation. Catalogue of South African Lithostratigraphic Units, 3:51–52; 1991. [7] ANCAP. Administración Nacional de Combustibles Alcohol y Portland.https:// exploracionyproduccion.ancap.com.uy/innovaportal/v/2435/3/innova.front/oilshales.html, 2018. [Online; accessed 20-February-2018]. [8] Yang Yu, Xiaofeng Lu, Wang Quanhai. Investigation on the co-combustion of low calorific oil shale and its semi-coke by using thermogravimetric analysis. Energy Convers Manage 2017;136:99–107. [9] Fan Yunlong, Zhaosheng Yu, Fang Shiwen, Lin Yan, Lin Yousheng, Liao Yanfen, Ma Xiaoqian. Investigation on the co-combustion of oil shale and municipal solid waste by using thermogravimetric analysis. Energy Convers Manage 2016;117:367–74. [10] Yao Zhongliang, Ma Xiaoqian, Wang Zihuan, Chen Limei. Characteristics of cocombustion and kinetic study on hydrochar with oil shale: a thermogravimetric analysis. Appl Therm Eng 2017;110:1420–7. [11] Zhou Chuncai, Liu Guijian, Wang Xudong, Qi Cuicui. Co-combustion of bituminous coal and biomass fuel blends: Thermochemical characterization, potential utilization and environmental advantage. Bioresour Technol 2016;218:418–27. [12] Pihu Tonu, Konist Alar, Neshumayev Dmitri, Loo Lauri, Molodtsov Artjom, Valtsev Aleksandr. Full-scale tests on the co-firing of peat and oil shale in an oil shale fired circulating fluidized bed boiler. Oil Shale 2017;34(3):250–62. [13] Konist Alar, Loo Lauri, Valtsev Aleksandr, Maaten Birgit, Siirde Andres, Neshumayev Dmitri, Pihu Tõnu. Calculation of the amount of estonian oil shale products from combustion in regular and oxy-fuel mode in a cfb boiler. Oil Shale 2014;31(3):211. [14] Väli E, Valgma I, Reinsalu E. Usage of estonian oil shale. Oil Shale 2008;25(2S):101–14. [15] Hotta A, Parkkonen R, Hiltunen M, Arro H, Loosaar J, Parve T, Pihu T, Prikk A, Tiikma T. Experience of estonian oil shale combustion based on cfb technology at narva power plants. Oil Shale 2005;22(4). [16] Uibu Mai, Uus Mati, Kuusik Rein. Co2 mineral sequestration in oil-shale wastes from estonian power production. J Environ Manage 2009;90(2):1253–60. [17] Jiang XM, Han XX, Cui ZG. New technology for the comprehensive utilization of chinese oil shale resources. Energy 2007;32(5):772–7. [18] Jiang XM, Han XX, Cui ZG. Progress and recent utilization trends in combustion of chinese oil shale. Progress Energy Combust Sci 2007;33(6):552–79. [19] Espitalie Jean, Madec ,Marcel, Tissot B, Mennig JJ, Leplat, Paul et al. Source rock characterization method for petroleum exploration. In: Offshore Technology Conference. Offshore Technology Conference, 1977. [20] Espitalie J, Deroo G, Marquis F. La pyrolyse rock-eval et ses applications. deuxième partie. Revue de l’Institut français du Pétrole 1985;40(6):755–84. [21] Nunez-Betelu L, Baceta JI. Basics and application of rock-eval/toc pyrolysis &58; an example from the uppermost paleocene/lowermost eocene in the basque basin, western pyrenees. Munibe Ciencias Naturales, 46(unknown):43–62, 1994. [22] Senftle Joseph T, Landis Charles R, McLaughlin Renee L. Organic petrographic approach to kerogen characterization. Organic geochemistry Springer; 1993. p. 355–74. [23] McCarthy Kevin, Rojas Katherine, Niemann Martin, Palmowski Daniel, Peters Kenneth, Stankiewicz Artur. Basic petroleum geochemistry for source rock evaluation. Oilfield Rev 2011;23(2):32–43. [24] Pearson DL. Pollen/spore color standard. Phillips Petroleum Company Exploration Projects Section (reproduced in Traverse, A., 1988. Palaeopalynology, Plate 1. Unwin Hyman, Boston), 1984. [25] Pring Allan. Pecharsky, vk & zavalij, py 2003. fundamentals of powder diffraction and structural characterization of minerals. Geolog Mag 2005;142(3). pp. 305–305. [26] Rietveld HaM. A profile refinement method for nuclear and magnetic structures. J Appl Crystallography 1969;2(2):65–71. [27] Larson AC, Von Dreele RB. Report laur 86–748. New Mexico, USA: Los Alamos National Laboratory; 2000. [28] Brian H. Toby. Expgui, a graphical user interface for gsas. J Appl Crystallography 2001;34(2):210–3. [29] Calorimetric Bomb Parr 1341 Manual.https://www.parrinst.com/products/oxygen-

[33]

[34]

[35]

[36] [37]

[38] [39] [40]

[41] [42] [43]

[44]

[45]

[46]

[47] [48]

[49]

[50] [51] [52] [53]

[54] [55]

[56]

[57]

357

bomb-calorimeters/1341-plain-jacket-bomb-calorimeter/documents/. [Online; accessed 24-May-2018]. Kok and İlker Şengüler Mustafa Versan, Şengüler İlker. Geological and thermal characterization of eskişehir region oil shales. J Therm Anal Calorimetry April 2014;116(1):367–72. Ozawa Takeo. A new method of analyzing thermogravimetric data. Bull Chem Soc Jpn 1965;38(11):1881–6. Jiang Longbo, Yuan Xingzhong, Li Hui, Xiao Zhihua, Liang Jie, Wang Hou, Zhibin Wu, Chen Xiaohong, Zeng Guangming. Pyrolysis and combustion kinetics of sludgecamphor pellet thermal decomposition using thermogravimetric analysis. Energy Convers Manage 2015;106:282–9. Latva-Somppi Jouko, Kauppinen Esko I, Valmari Tuomas, Ahonen Petri, Gurav Abhijit S, Kodas Toivo T, Johanson Bo. The ash formation during co-combustion of wood and sludge in industrial fluidized bed boilers. Fuel Process Technol 1998;54(1–3):79–94. Bai Fengtian, Sun Youhong, Liu Yumin, Li Qiang, Guo Mingyi. Thermal and kinetic characteristics of pyrolysis and combustion of three oil shales. Energy Convers Manage 2015;97:374–81. Carter Scott D, Robl Thomas L, Taulbee Darrell N, Rubel Aurora M. Testing of an irati oil shale in a multi-stage fluidized bed retorting process. Fuel 1991;70(11):1347–51. Yörük Can Rüstü, Meriste Tõnis, Trikkel Andres, Kuusik Rein. Oxy-fuel combustion of estonian oil shale: kinetics and modeling. Energy Procedia 2016;86:124–33. Lauri Loo, Alar Konist, Dmitri Neshumayev, Tõnu Pihu, Birgit Maaten, and Andres Siirde. Ash and flue gas from oil shale oxy-fuel circulating fluidized bed combustion. 11:1218, 05 2018. Carter Scott D, Taulbee Darrell N. Fluidized bed steam retorting of kentucky oil shale. Fuel Processing Technol 1985;11(3):251–72. Peters Kenneth E, Cassa Mary Rose. Applied source rock geochemistry: chapter 5: Part ii. essential elements. 1994. dos Santos Rosane, Euzébio Darlly Erika, dos Reis Silva, Bergamaschi Sérgio, Martins Maria Virǵnia Alves Martins Alves, Rodrigues René, et al. Oil generation potential assessment and paleoenvironmental interpretation of irati formation (lower permian) in northwestern of paraná basin (brazil). J Sedimentary Environ 2016;1(2):254–67. American mineralogist crystal structure database. =http://rruff.geo.arizona.edu/ AMS/amcsd.php, 2018. Crystallography open database. =http://www.crystallography.net/cod/, 2018. [Online; accessed 16-January-2018]. Bai Fengtian, Sun Youhong, Liu Yumin, Liu Baochang, Guo Mingyi, Lü Xiaoshu, Guo Wei, Li Qiang, Hou Chuanbin, Wang Qiuwen. Kinetic investigation on partially oxidized huadian oil shale by thermogravimetric analysis. Oil Shale 2014;31(4):377. Altun N Emre, Hwang Jiann-Yang, Hicyilmaz Cahit. Enhancement of flotation performance of oil shale cleaning by ultrasonic treatment. Int J Mineral Process 2009;91(1):1–13. Grice Kliti, Stefan, SchoutenBlokker Peter, Derenne Sylvie, Largeau Claude, Nissenbaum Arie, Sinninghe Damsté Jaap S. Structural and isotopic analysis of kerogens in sediments rich in free sulfurised botryococcus braunii biomarkers. Organic Geochem 2003;34(3):471–82. Adams Mike J, Awaja Firas, Bhargava Suresh, Grocott Stephen, Romeo Melissa. Prediction of oil yield from oil shale minerals using diffuse reflectance infrared fourier transform spectroscopy. Fuel 2005;84(14–15):1986–91. RRUFF.http://rruff.info/Muscovite/R040108, 2018. [Online; accessed 28February-2018]. Saikia Bhaskar J, Parthasarathy Gopalakrishnarao. Fourier transform infrared spectroscopic characterization of kaolinite from assam and meghalaya, northeastern india. J Modern Phys 2010;1(04):206. Sun Youhong, Bai Fengtian, Liu Baochang, Liu Yumin, Guo Mingyi, Guo Wei, Wang Qiuwen, Lü Xiaoshu, Yang Fang, Yang Yang. Characterization of the oil shale products derived via topochemical reaction method. Fuel 2014;115:338–46. Pohl Walter L. Economic geology: principles and practice. John Wiley & Sons; 2011. Mohamed Hussain Azeez, Campos Leonel. Oil shale ash utilization in industrial processes as an alternative raw material, 2016. Wang Qian, Hou YC, Wu WZ, Liu Qing, Liu ZY. The structural characteristics of kerogens in oil shale with different density grades. Fuel 2018;219:151–8. Vassilev Stanislav V, Kitano Kunihiro, Takeda Shohei, Tsurue Takashi. Influence of mineral and chemical composition of coal ashes on their fusibility. Fuel Process Technol 1995;45(1):27–51. Combustion Laboratory. Combustion Laboratory.https://www.et.byu.edu/larryb/ Ash 2018. [Online; accessed 12-March-2018]. World Nuclear Association.http://www.world-nuclear.org/information-library/ safety-and-security/radiation-and-health/radiation-and-life.aspx, 2018. [Online; accessed 12-March-2018]. Kok Mustafa Versan, Ozgur Emre. Combustion performance and kinetics of oil shales. Energy Sources, Part A: Recovery, Utilization, Environ Effects 2016;38(8):1039–47. Han XX, Jiang XM, Cui ZG. Thermal analysis studies on combustion mechanism of oil shale. J Therm Anal Calorimetry 2006;84(3):631–6.