Influence of pressure on the release of inorganic species during high temperature gasification of coal

Influence of pressure on the release of inorganic species during high temperature gasification of coal

Fuel 90 (2011) 2326–2333 Contents lists available at ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel Influence of pressure on the ...

1MB Sizes 1 Downloads 36 Views

Fuel 90 (2011) 2326–2333

Contents lists available at ScienceDirect

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

Influence of pressure on the release of inorganic species during high temperature gasification of coal Marc Bläsing ⇑, Michael Müller 1 Institute for Energy Research (IEK-2), Leo-Brandt-Str. 1, 52425 Jülich, Germany

a r t i c l e

i n f o

Article history: Received 23 December 2010 Received in revised form 3 February 2011 Accepted 4 February 2011 Available online 17 February 2011 Keywords: Coal Gasification Pressure Release of alkali metal, sulphur, chlorine species

a b s t r a c t Alkali metal, sulphur, and chlorine species released during coal gasification are of concern, because they can lead to problems in colder parts of the plant. Therefore, hot gas cleaning technology is recently under development. This clean-up strategy requires a comprehensive knowledge of the release characteristics of inorganic compounds. The principal objective of this work was to provide details of the influence of pressure on the release of key chemical species, e.g. sodium, potassium, sulphur, and chlorine. Hence, a total of 19 different coals were investigated in lab-scale gasification experiments in an electrical heated pressurised furnace at absolute pressures of 2, 4, and 6 bar in an atmosphere of He/7.5v%O2 at 1325 °C. Hot gas analysis was carried out by molecular beam mass spectrometry. The quantitative results showed a decreasing release of 34H2S+, 36HCl+, 39K+/39NaO+, 58 74 NaCl+, 64 SOþ KCl+ with increasing pressure. The discussion was supported by thermodynamic 2 , and calculations. Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction The world’s energy demand is on the rise. The International Energy Agency published in their reference scenario of the ‘‘World Energy Outlook 2009’’ an increasing demand in electricity by 76% in 2007–2030, requiring 4800 GW of additional capacity [1]. Furthermore, they report the ongoing dominant role of fossil fuels in energy production worldwide, with coal as one of the major fuels with a share of the global power generation of 44% in 2030. The development of cleaner, more efficient techniques in next coal power plant generation is getting increasingly important [2]. Recently, coal-based combined cycle power generation systems are developed or are under development with the aim to increase the efficiency [3,4]. An promising coal utilisation process is the integrated coal gasification combined cycle (IGCC), which is able to use a broad band of different solid fuels with high efficiency. Additionally, it offers the possibility of carbon sequestration [4–7]. During entrained flow coal gasification at temperatures up to 1500 °C and pressure up to 3.0 MPa a complex mixture of solid, liquid, and gaseous phases is formed. Of special interest are vapour alkali metal, sulphur and chlorine species. In general, these species are highly volatile and they can

⇑ Corresponding author. Tel.: +49 2461 61 1574; fax: +49 2461 61 3699. E-mail addresses: [email protected]e (M. Bläsing), [email protected] de (M. Müller). 1 Tel.: +49 2461 61 6812; fax: +49 2461 61 3699. 0016-2361/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2011.02.013

lead to problems when they reach colder parts of the plant, where they form sticky and corrosive layers. As a consequence, this can lead to several problems in downstream plant components [8]. Both the increase in efficiency and the decrease of the amount of harmful species can be reached by hot fuel gas cleaning [4]. This clean-up strategy requires a comprehensive knowledge of the release characteristics of alkali metal, potassium, sulphur, and chlorine species. Despite research efforts during the last decades there are a lot of open questions regarding the release and the underlying release mechanisms. Therefore, a large body of knowledge on the release of the above mentioned species during gasification was already published. However, there is still a lack of knowledge regarding the conditions relevant for the IGCC, especially high temperature and elevated pressure. Little literature can be found on this topic. Oleschko and Müller [9] reported the decrease of NaCl and SO2 in flue gas from coal combustion with increasing pressure. Reichelt [10] reported the decrease of alkali metal concentration under elevated pressure. However, their experiments where done under oxidising conditions. Sathe et al. [11] reported a strong influence of pressure on the release of alkali metal and alkaline earth metal species during devolatilisation of Victorian brown coal. However, further investigations on the release of harmful species under the conditions of the IGCC process have to be carried out. Therefore, lab-scale experiments were done with a broad range of different coals at 1325 °C furnace temperature and an absolute pressure of 2, 4, and 6 bar. The results complement current analytical capabilities and data sets.

2327

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

2. Experimental 2.1. Fuel preparation Samples from 10 hard coals and 9 lignites were collected and prepared for analysis. STD-1 to STD-5 and K2-3 were mined in Germany, STN-1 and STN-2 in Spitsbergen, K2-4 in Colombia, and K2-5 in Spain. The lignites were mined open cast in Germany. The coals were dried and grinded in a mill. Particles with a size smaller 100 lm were stored under dry conditions at room temperature. The chemical analysis of the coals under investigation is given in Tables 1 and 2. The coals were chosen for investigation of a broad range of different coals. 2.2. Experimental setup The gasification experiments were carried out in an electrically heated, pressurised furnace. For in situ, online determination of the composition of the hot gas composition the furnace was coupled to a molecular beam mass spectrometer. A schematic of the experimental setup is given in Fig. 1. The gasification reactor mainly consisted of a heated flow channel housed in a furnace with two independent heating zones which was built in a pressure vessel. For the flow channel, a high density alumina tube was used to prevent the tube walls from reactions with the released species. The inner diameter of the flow channel was 25 mm, the total length of the tube was 350 mm. A mixture of He/7.5v%O2 flowed continuously through the channel. Helium was selected as the carrier gas. Because of the low atomic mass, it leads to the highest signal intensities in the MBMS [12]. The sample boat was positioned in the hot part of the pressurised furnace by a horizontally displaceable alumina rod coupled to a stainless steel rod. The gas flow was controlled by BROOKS gas flow

controllers and the pressure by a BROOKS pressure controller. The reaction zone was heated by a SiC heating element. The highest attainable temperature in this zone is 1325 °C at 6 bar. In order to prevent condensation of inorganic species of interest all parts downstream the reaction zone were kept at temperatures above 825 °C. A typical experimental run consisted of the following steps. At the start of the experiment a platinum sample boat loaded with 50 mg of coal was inserted into the cold sample inlet of the heated flow channel (about 70 °C). A gas flow of 4 l/min He and 0.324 l/min O2 corresponding to 92.5% He and 7.5% O2 was fed into the reactor to simulate a gasification like environment. When the pressure reached the desired value of 2, 4, or 6 bar the background signal was recorded. Of special interest is the 34 Oþ 2 -signal, which is used for normalisation of the data as will be explained in the result section. Then the sample boat was placed in the hot reaction zone by a horizontally displaceable rod. The MBMS consists of three differentially pumped chambers as shown in Fig. 2. The hot reaction products flowed to the end of the reactor where they entered the MBMS through a nozzle with 0.1 mm in diameter. Due to immediately supersonic free jet expansion of the hot gas into the first high vacuum chamber (102 mbar), the species were cooled far below room temperature in microseconds, attained free molecular flow and therefore, formed a molecular beam [12]. The core of the expanded gases is extracted by a conical skimmer of 1 mm diameter and directed into the third chamber. There a hot filament emits electrons with an electron energy of 50 eV and an emission current of 1 mA. Every 104–103 molecule is ionised by electron impact. After passing the deflector, the ions are filtered in a quadrupole mass analyser and detected by an off axis electron multiplier. The amplified signal is recorded by a computer and software package as a function of time and mass-to-charge ratio. In order to be able to monitor the gasification process with sufficient temporal resolution, 10 scans per

Table 1 Chemical composition of the hard coal samples (mass%).

C H O N S Cl Al Fe Ca Mg K Na Si

K2-3

K2-4

K2-5

STD-1

STD-2

STD-3

STD-4

STD-5

STN-1

STN-2

79.4 2.76 10.5 0.978 0.9 0.127 3.06 0.69 0.3 0.16 0.73 0.3 4.57

65.5 4.72 21.9 1.241 0.49 0.011 1.84 0.73 0.21 0.17 0.22 0.32 4.89

31.6 2.60 34.3 0.73 0.98 0.032 7.81 2.35 0.72 0.45 1.93 0.19 16.1

83.6 4.22 6.26 1.72 0.75 0.116 0.96 0.52 0.15 0.094 0.16 0.053 1.4

80.3 4.72 8.47 1.65 0.77 0.156 0.91 0.55 0.43 0.2 0.16 0.088 1.6

78.4 4.98 10.62 1.68 0.89 0.185 0.90 0.43 0.26 0.11 0.17 0.072 1.3

65.2 3.67 16.38 1.36 0.77 0.136 3.30 1.30 0.78 0.49 0.75 0.16 5.7

59.8 4.14 20.90 1.41 0.94 0.237 3.10 1.30 0.77 0.44 0.87 0.19 5.9

78.8 5.54 8.86 1.61 0.64 0.015 0.64 0.60 0.72 0.21 0.11 0.25 1.60

74.8 4.52 15.3 1.25 1.25 0.009 1.35 0.99 1.09 0.25 0.17 0.50 2.46

Table 2 Chemical composition of the lignite samples (mass%).

C H O N S Cl Al Fe Ca Mg K Na Si

K2-1

K2-2

K3-1

K3-2

K3-3

HKN-S

HKN-S+

HKS

HKT

54.8 5.604 33.8 0.574 0.24 0.035 0.069 0.58 1.39 0.37 0.014 0.16 0.25

56.9 5.563 33.2 0.585 0.3 0.037 0.042 0.49 1.26 0.51 0.021 0.39 0.22

51.0 5.134 37.6 0.5 1.35 0.022 0.27 1.3 1.01 0.26 0.061 0.008 1.33

55.5 4.861 37.6 0.577 0.28 0.023 0.051 0.8 1.03 0.36 0.014 0.017 0.22

54.6 5.03 37.2 0.636 0.31 0.0207 0.06 0.66 0.89 0.28 0.016 0.017 0.18

65.8 4.81 28.1 0.78 0.205 0.01 0.034 0.25 1.0 0.37 0.02 0.22 0.01

65.8 4.98 26.6 0.84 0.508 0.025 0.034 0.25 1.2 0.44 0.023 0.22 0.023

62.0 4.89 28.0 0.69 0.365 0.023 0.12 0.48 1.4 0.48 0.024 0.22 0.72

57.3 4.18 27.5 0.75 0.478 0.011 1.5 0.28 1.3 0.47 0.085 0.23 3.6

2328

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

Fig. 1. Schematic of the experimental setup.

second were acquired. To ensure reproducibility, six samples were measured. In preliminary measurements mass spectra from 10 to 150 amu were scanned. Key chemical species detected by MBMS 74 were 34H2S+, 36HCl+, 39K+/39NaO+, 58NaCl+, 64 SOþ KCl+. NaOH 2 , and is without a doubt a key specie, however 40NaOH+ is not easily detectable by MBMS, because it has the same mass-to-charge ratio as 40Ar+. This results in an overlapping of the two peaks. Argon is always present in the gas phase to a small extent because the oxygen which has been used for the experiments is slightly contaminated with Argon due to its production process.

3. Results The batch experiments simulate the gasification of coal under pressure. As an example of the qualitative results of the MBMSmeasurements, spectra of the Norwegian hard coal STN-2 are depicted in Fig. 3. Key species were 30CO+, 34H2S+, 36HCl+, 39 + 39 74 K / NaO+, 58NaCl+, 64 SOþ KCl+ which were mainly re2 , and leased during devolatilisation phase. Shortly after sample insertion the sample reached the required temperature for devolatilisation. During devolatilisation volatile organic and inorganic matter were released and reacted immediately with oxygen. This caused to a lack of oxygen as shown by the sharp drop off of the 34 Oþ 2 -signal-intensity and an increase of

the signal-intensity of 30CO+ (Fig. 3). However, the change-over from devolatilisation to char reactions phase is gradual. The majority of the release occurred during the devolatilisation phase with high intensity. Therefore, quantification of the spectra was performed only for the devolatilisation phase. Quantification was performed through normalisation of the peak area during the devolatilisation phase to the 34 Oþ 2 -signal of the first 20 s of the experimental run, during which time the steady oxygen concentration led to a steady signal as shown in Fig. 3. The quantitative results are depicted in Figs. 4–9. Both the hard coals and the lignites under investigation showed a decrease of the averaged, normalised peak areas of 34H2S+ with increasing pressure (Fig. 4), e.g. the hard coal STD-5 showed a very strong decrease of 34H2S+ of 66.0% at 4 bar, and 87.2% at 6 bar. In contrast, the lignite K2-1 showed a smaller influence of pressure on the release of 34H2S+ of 29.5% at 4 bar and 34.0% at 6 bar. Additionally, the release of 64 SOþ 2 decreased with increasing pressure, e.g. K2-5 by 65.8% at 4 bar and 88.4% at 6 bar (Fig. 5). The averaged peak areas of 36HCl+ decreased for both the hard coals and the lignites (Fig. 6). The variance of the decrease is high, e.g. STD-3 showed a decrease of 3.5% at 4 bar and 42.4% at 6 bar, whether K2-5 showed a decrease of 52.8% at 4 bar and 85.5% at 6 bar. The averaged peak areas of 58NaCl+ and 74KCl+ decreased with increasing pressure (Figs. 7 and 8). STN-2 showed a very strong decrease of 58NaCl+ of 88.7% at 4 bar and 95.6% at 6 bar. Additionally, the decrease of 74KCl+ was strong for STN-2 75.8% at 4 bar and 85.1% at 6 bar, too. The averaged peak areas of 39K+/39NaO+ decreased with increasing pressure over a broad range (Fig. 9). K2-3 showed a moderate decrease of 7.5% at 4 bar and 26.2% at 6 bar, whereas STD-1 showed a strong decrease of 75.8% at 4 bar and 85.5% at 6 bar.

4. Discussion Regarding the release of 36HCl+, 58NaCl+, and 64 SOþ 2 , the experimental results are comparable to the results reported by Oleschko

Fig. 2. Schematic representation of the MBMS system.

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

Fig. 3. Intensity-time-profiles of several inorganic compounds released during gasification experiments with the Norwegian hard coal STN-2 at 2, 4, and 6 bar.

Fig. 4. Averaged, normalised peak areas of

34

H2S+ at 1325 °C and 2–6 bar.

Fig. 5. Averaged, normalised peak areas of

64

SOþ 2 at 1325 °C and 2–6 bar.

2329

2330

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

Fig. 6. Averaged, normalised peak areas of

Fig. 7. Averaged, normalised peak areas of

Fig. 8. Averaged, normalised peak areas of

and Müller [9] and, specifically regarding NaCl, to the results of Reichelt [10]. Oleschko and Müller [9] explained the decreasing release of 36HCl+, 58NaCl+, and 64 SOþ 2 with increasing pressure by pointing out that these species have a lower concentration in the gas phase. The increase in absolute pressure for a given partial pressure of a vapour compound, which is in equilibrium in a condensed phase, leads to a lower concentration in the gas phase. However, the influence is assumed to be rather small in the pressure range of the pressure experiments (2–6 bar). Additional explanations are now provided. In principle, there are two main

36

HCl+ at 1325 °C and 2–6 bar.

58

NaCl+ at 1325 °C and 2–6 bar.

74

KCl+ at 1325 °C and 2–6 bar.

mechanisms: first, the reduced formation of volatile compounds, which includes the explanation of Oleschko and Müller [9], and second, the enhanced capture of volatilised compounds. However, the two mechanisms occur at the same time, and the fraction of S released/captured is in variance regarding the conditions and the coal under investigation. The investigation of Sathe et al. [11] of the release of alkali and alkaline earth metals during devolatilisation of Victorian brown coal support the proposed classification. However, their experiments have been done at much lower temperature.

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

Fig. 9. Averaged, normalised peak areas of

The influence of pressure on the release of 34H2S+ and 64 SOþ 2 can be explained by the shift of several reactions. The decomposition equilibrium of FeS2 is shifted to the precursor side by increasing pressure. Therefore, smaller amounts of gaseous S species were

2331

41 + 41

K / NaO+ at 1325 °C and 2–6 bar.

formed with increasing pressure. However, the shift is rather slight under the present experimental conditions, as shown by thermodynamic calculations and as assumed from the results of Xu and Kumagai [13]. To describe the present results satisfactorily, further

Fig. 10. Thermodynamically stable compounds of Na at 1325 °C, 2–6 bar and oxygen to fuel ratio = 0.5.

2332

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

Fig. 11. Thermodynamically stable compounds of K at 1325 °C, 2–6 bar and oxygen to fuel ratio = 0.5.

explanations are needed. A likely explanation of the release trend of 34H2S+ is the following: The formation of H2S and char-bound S was described by Zevenhoven and Kilpinen [14], as shown by way of example in the non-stoichiometric Eqs. (1) and (2). Reaction 1 is suppressed under increasing pressure. Furthermore, H2S can react with the remaining char to form stable thiophenic structures (Eq. (2)). The reaction of released H2S with char is enhanced under increased pressure. As a result, smaller amounts of 34H2S+ were formed under increased pressure; this is in agreement with the present results and the results of Nichols et al. [15].

of char increased with increasing pressure. Additionally, Gadiou et al. [17] reported that increasing pressure leads to an increasing residence time of volatile products in the remaining char. In summary, the retarded mass transfer leads to the residence time of the released species increasing with increasing pressure; this effect can enhance capture reactions as shown in the following equations:

Fuel-S ! H2 S þ Char þ Volatiles

ð1Þ

Fuel-S ! Char-S þ Volatiles

ð2Þ

The release of 36HCl+ decreased for all coals under investigation. Regarding the reactions in Eqs. (5) and (6), the increasing pressure caused a shift in the equilibrium to the precursor side. Additionally, the retarded mass transfer out of the char could cause a reduced release of 36HCl+ during the devolatilisation phase.

The released amount of the alkali metal species 41K+/41NaO+, NaCl+, and 74KCl+ decreases with increasing pressure for both hard coals and lignites. A likely explanation is the following. The residence time of the volatilised species in the remaining coal is very important for the capture. Thus, the influence of pressure on the residence time needs to be discussed. Yang et al. [16] investigated the influence of pressure on coal devolatilisation (up to 50 bar). They found that the majority of the volatiles evolved at 400–800 °C. The devolatilisation rate decreased and the amount

NaCl þ SiO2 þ H2 O $ Na2 O  SiO2 þ HCl

ð3Þ

NaCl þ H2 O þ xAlSiy Oz $ Na-aluminosilicates þ HCl

ð4Þ

58

CaCl2 þ H2 O $ CaO þ 2HCl

ð5Þ

CaCl2 þ SiO2 þ H2 O $ CaSiO3 þ 2HCl

ð6Þ

To support the discussion from a thermodynamic point of view calculations were made using FactSage 5.4.1, a common program for the calculation of systems in thermodynamic equilibrium using the principle of minimisation of the Gibbs energy.

M. Bläsing, M. Müller / Fuel 90 (2011) 2326–2333

Input for the modelling were the gasification conditions of the experiments, the fuel composition (Tables 1 and 2) and an oxygen to fuel ratio (k) of 0.5. Further information on the calculation procedure can be found in Bläsing and Müller [18,19]. The results of thermodynamic modelling are shown in Figs. 10 and 11. The most stable alkali metal compounds predicted by thermodynamic calculations are alkali metal chlorides, alkali metal hydroxides, atomic alkali metal species, and alkali metal aluminosilicates. Alkali metal aluminosilicates are by far the most abundant alkali metal compound for all hard coals and for the silica- and alumina-rich lignite HKT. For example, Na fixed in NaAlSi3O8 accounts for 72.8–80.1% of the total amount of Na in STD-1. For the lignites except K3-1 and HKT, alkali metal hydroxides are the most abundant alkali metal species, e.g. K fixed in hydroxide accounted for 80.2–86.5% of the total K in HKN-S. The formation of alkali metal aluminosilicates and alkali metal hydroxides is, in most cases, slightly enhanced by increasing pressure. The formation of alkali metal chlorides and metallic species, in most cases, slightly decreases with increasing pressure, e.g. a decrease of 20% of KCl is predicted for STN-2 under pressure increasing from 2 to 6 bar, and a decrease of 46% of NaCl is predicted for similarly increasing pressures. The results of the thermodynamic modelling confirm the assumption that increased pressure leads to retarded mass transfer and to enhanced capture of NaCl and KCl. 5. Conclusions The influence of pressure on the release of Na-, K-, S-, and Cl-species was investigated in lab-scale gasification experiments. 19 different coals were gasified in He/7.5%O2 at 1325 °C. The hot gasification products were analysed by molecular beam mass spectrometry. Key chemical species detected by the mass spectrometer were 34H2S+, 36HCl+, 39K+/39NaO+, 58NaCl+, 64 SOþ 2, 74 KCl+. For further quantitative data analysis the averaged, normalised peak areas were calculated. In general, the influence of pressure on the release of this species was found very strong. All coals under investigation showed a decreasing release of the vapour species of interest with increasing pressure. Likely explanations include shifts in reaction balances, a retarded mass transfer and as a result an enhanced capture due to longer residence time of the volatilised compounds. In principle, the thermodynamic calculations prove the trend of decreasing alkali chloride release and the enhanced formation of alkali aluminosilicates.

2333

Acknowledgments The work described in this paper has been done in the framework of the HotVeGas-EM Project supported by Bundesministerium für Wirtschaft und Technologie (FKZ 0327773C). Part of the coals were kindly supplied by RWE Power AG and Vattenfall Europe AG. References [1] http://www.worldenergyoutlook.org/docs/weo2009/WEO2009_es_german.pdf. [2] Beer J. High efficiency electric power generation: the environmental role. Prog Energy Combust Sci 2007;33:107–34. [3] Wolk RH, McDaniel J. High efficiency coal fuelled power generation. Energy Convers Manage 1992;33:705–12. [4] Müller M, Pavone D, Rieger M, Abraham R. Hot fuel gas cleaning in IGCC at gasification temperature. In: Fourth international conference on clean coal technologies, Dresden, Germany; May 2009. [5] Mondol JD, McIlveen-Wright D, Rezvani S, Huang Y, Hewitt N. Technoeconomic evaluation of advanced IGCC lignite coal fuelled power plants with CO2 capture. Fuel 2009;88:2495–506. [6] Newcomer A, Jay A. Storing syngas lowers the carbon price for profitable coal gasification. Environ Sci Technol 2007;41:7974–9. [7] Newby RA, Bannister RL. Advanced hot gas cleaning systems for coal gasification processes. Trans ASME 1994;116:338–44. [8] Bakker W. High temperature corrosion in gasifiers. Mater Res 2004;7:53–9. [9] Oleschko H, Müller M. Influence of coal combustion and operating conditions on the release of alkali species during combustion of hard coal. Energy Fuels 2007;21:3240–8. [10] Reichelt T. Freisetzung gasförmiger Alkaliverbindungen bei atmosphärischer und druckaufgeladener Verbrennung, Fortschritt-Berichte VDI, Düsseldorf, ISBN 3-18-368703-8; 2001. [11] Sathe C, Hayashi J, Li CZ, Chiba T. Release of alkali and alkaline earth metallic species during rapid devolatilisation of a Victorian brown coal at elevated pressure. Fuel 2003;82:1491–7. [12] Wolf KJ. Untersuchungen zur Freisetzung und Einbindung von Alkalien bei der re-duzierenden Druckwirbelschichtverbrennung. PhD-thesis, RWTH Aachen, Aachen; 2003. [13] Xu WC, Kumagai M. Sulfur transformation during rapid hydrodevolatilisation of coal under pressure by using a continuos free fall pyrolyzer. Fuel 2003;82:245–54. [14] Zevenhoven R, Kilpinen P. Control of pollutants in flue gases and fuel gases. ebook 2002. [15] Nichols KM, Hedman PO, Smoot LD, Blackham AU. Fate of coal-sulphur in a laboratory-scale coal gasifier. Fuel 1989;68:243–8. [16] Yang H, Chen H, Ju F, Yan R, Zhang S. Influence of pressure on coal devolatilisation and char gasification. Energy Fuels 2007;21:3165–70. [17] Gadiou G, Bouzidi Y, Prado G. The devolatilisation of millimetre sized coal particles at high heating rate: the influence of pressure on the structure and reactivity of the char. Fuel 2002;81:2121–30. [18] Bläsing M, Müller M. Mass spectrometric investigations on the release of inorganic species during gasification and combustion of German hard coals. Combust Flame 2010;157:1374–81. [19] Bläsing M, Müller M. Mass spectrometric investigations on the release of inorganic species during gasification and combustion of Rhenish lignite. Fuel 2010;89:2417–24.