- Email: [email protected]
The temperature-dependent release of volatile inorganic species from Victorian brown coals and German lignites under CO2 and H2O gasification conditions
Fuel 158 (2015) 72–80
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
The temperature-dependent release of volatile inorganic species from Victorian brown coals and German lignites under CO2 and H2O gasification conditions Joanne Tanner a, Marc Bläsing b,⇑, Michael Müller b, Sankar Bhattacharya a a b
Department of Chemical Engineering, Monash University, Wellington Rd, Clayton 3800, Australia Institute for Energy Research (IEF-2), Leo-Brandt-Str. 1, 52425 Jülich, Germany
h i g h l i g h t s The Na, Cl and S species released during gasification of low rank coal are reported. A release mechanism including the role and influence of CO2 and steam is proposed. The 2-phase mechanism is similar but distinct under CO2 and steam atmospheres. The release of volatile Na and Cl species essentially independent of gas atmosphere. The S species release mechanism depends directly on CO2 and steam concentration.
a r t i c l e
i n f o
Article history: Received 28 October 2014 Received in revised form 2 April 2015 Accepted 28 April 2015 Available online 8 May 2015 Keywords: Brown coal Lignite Gasification Volatile inorganics
a b s t r a c t To promote the use of abundant low rank coal resources, and to assist with the design, development and optimisation of hot gas cleaning and downstream processes, it is necessary to understand the release mechanisms of volatile inorganic species from low rank coals under high temperature gasification conditions. Although a significant amount of work has been reported under combustion and oxygen lean gasification conditions, these studies do not sufficiently explain the role of the gasification reagents CO2 and H2O in the mechanisms. Therefore, gasification experiments under 20% CO2 and 20% H2O in He were conducted at 1100 °C, 1200 °C and 1400 °C for two Victorian brown coals and four Rhenish lignites. Hot gas analysis was conducted by online molecular beam mass spectrometry to determine the intensity, relative quantity and timing of the release of volatile species of interest. Two overlapping phases were clearly observed from the results – devolatilisation and char gasification. Major species detected were 23Na+, 34H2S+, 35Cl+, 36HCl+, 39K+/39NaO+, 58NaCl+, 60COS+/60NaCl+, and 64SO+2. The release during devolatilisation was essentially independent of the bulk gas atmosphere and constituted the majority of Na and Cl species. The release of S-species occurred predominantly during gasification under both CO2 and H2O atmospheres by similar but distinct mechanisms, and was directly affected by the absence, presence and concentration of the gasification reagents. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Coal accounted for over half of the increase in global energy use over the past decade and is predicted to play a major role in all future projection scenarios to 2035 [1]. It is therefore important to continue the development of cost-effective, environmentally responsible coal processes, such as high efficiency gasification, for the conversion of this abundant resource to power and ⇑ Corresponding author. Tel.: +49 2461 61 1574; fax: +49 2461 61 3699. E-mail address: [email protected] (M. Bläsing). http://dx.doi.org/10.1016/j.fuel.2015.04.071 0016-2361/Ó 2015 Elsevier Ltd. All rights reserved.
products. In particular, low rank coals are of interest for use in advanced coal-to-product gasification technologies for countries with significant reserves, such as Australia and Germany. Low rank coals, as well as those of higher rank, often contain significant amounts of inorganic species including Na, Cl and S. During gasification and other high temperature processes, these species undergo decomposition and reaction, resulting in a complex mixture of volatile inorganic products known to cause slagging and fouling deposits, corrosion and pollution problems [2]. It is therefore important to design, develop and optimise high temperature gas cleaning systems, downstream conversion processes,
73
J. Tanner et al. / Fuel 158 (2015) 72–80
and environmental controls to enhance the overall gasification process efficiency. A comprehensive knowledge of the mechanisms of release and subsequent interactions between volatile inorganic species, including the role of the gasification reagents, is therefore required to facilitate the prediction of detrimental gas-phase components based on parent coal properties. Investigations into the modes of occurrence and release of volatile inorganics at high temperatures have historically been conducted using a variety of modelling and experimental methods [3,4]. Recently, online mass spectrometric techniques [5–8] have predominantly been used to measure in situ decomposition and reaction products from high temperature gasification and combustion processes. In particular, molecular beam mass spectrometry (MBMS) yields time dependent, online, simultaneous data pertaining to the release of volatile inorganic species under high temperature conditions of industrial interest [9]. Recent investigations in the area of the release of volatile inorganic species from low rank coals under gasification conditions have focused on the effects of temperature, pressure, oxygen partial pressure and steam addition [6,10–14]. However, the influence of different bulk gas atmospheres on the release mechanisms is still unknown. This study therefore uses the online MBMS technique to investigate the influence of temperature and bulk gas atmosphere on the in situ reaction products from high temperature gasification of two Victorian brown coals and four Rhenish lignites. In particular, a greater understanding of the role of the primary gasification reagents, CO2 and H2O, in the release and reaction mechanisms of the volatile Na-, Cl- and S-species under entrained flow conditions is sought. The results of this study are further compared with previous results obtained under O2 and O2/steam atmospheres [10,11,13] where appropriate.
Table 1 Results of the chemical analysis of the coals under investigation. HKN-S
HKN-S+
HKS
HKT
LY
MOR
Proximate analysis of the coals (mass %) Moisture (air dried) 20.02 19.76 Ash (dry basis) 3.58 4.23 Volatiles (dry basis) 51.43 53.17
20.45 6.59 52.45
10.27 13.20 55.24
11.16 7.99 48.25
14.92 3.59 49.31
Chemical composition (dry basis, mass %) C 65.8 65.8 H 4.81 4.98 N 0.78 0.84 S 0.205 0.508 Cl 0.01 0.03 Al 0.024 0.025 Fe 0.18 0.22 Ca 0.79 0.90 Mg 0.30 0.34 K 0.016 0.016 Na 0.19 0.19 Si 0.009 0.020
62.0 4.89 0.69 0.365 0.02 0.099 0.40 1.2 0.43 0.018 0.42 0.60
57.3 4.18 0.75 0.478 0.01 1.2 0.20 1.1 0.39 0.069 0.22 2.7
50.1 4.3 0.445 0.207 0.092 0.204 0.209 0.133 0.114 0.014 0.101 7.180
60.7 5.35 0.52 0.041 0.039 0.019 0.185 0.417 0.194 0.010 0.056 0.052
Elemental Ratios (molar basis) Na/Cl 28.88 Na/K 20.48 Ca/S 3.08 Na/Ca 0.42
27.48 38.29 2.58 0.62
31.05 5.56 1.90 0.34
1.68 12.40 0.52 1.32
2.21 9.51 8.16 0.23
11.33 19.68 1.42 0.36
HKx = Rhenish Lignites, Hambach; MOR = Morwell brown coal, Victoria.
LY = Loy
Yang
brown
coal,
Victoria;
likely an artefact of sampling, such as extraneous sand or clay from the interseam or overburden. The additional silica content may influence the release mechanism of sodium as discussed in previous work [14].
2.2. Experimental setup 2. Materials and methods 2.1. Fuel preparation Run-of-mine samples of four Rhenish lignites (HKN-S+, HKN-S, HKS and HKT) were supplied by RWE Power and two Victorian brown coals (Loy Yang – LY and Morwell – MOR) were supplied by the Loy Yang and Hazelwood Power Stations, respectively. Coals from these two regions were chosen due to their similarities in age (tertiary) and rank (low). The samples were air dried and pulverised in a mill. The Rhenish coals were sieved to a particle size range of <100 lm and the Victorian coals to 90–106 lm. All samples were thereafter preserved at room temperature under dry conditions. Of particular interest in this investigation was the difference in the inorganic composition of the coals, specifically regarding Na, K, S, Cl, Al, Si, and Ca. These elements and various compounds thereof, typically found in low rank coals, are known either for their volatile nature or for their influence on the volatilisation of other inherent or derived inorganic species released during the thermal treatment of coals. The chemical composition of the coals under investigation is given in Table 1. Chemical analysis was performed by the central division of analytical chemistry (ZCH) of the Forschungszentrum Jülich. The standard methods used and variance of analytical results are detailed elsewhere [6]. The chemical analysis of the Loy Yang sample showed anomalously high ash content – almost 8% on a dry basis (Table 1). Based on similar samples reported in recent literature, Loy Yang coal is expected to have an ash content in the range of 0.5–1.6% on a dry basis. Uncharacteristically high aluminium and silicon were also measured in the current sample in comparison to previously reported values for Loy Yang coal [7,15,16] and the presence of large amounts of silica was confirmed by XRD. These features are
The release of Na-, S-, and Cl-containing species during steam gasification was investigated under conditions analogous to those experienced by coal particles during industrial entrained flow gasification. The fixed-bed, batch configuration with continuous gas flow was deliberately chosen. In contrast, under continuous feed conditions the final product gas mixture would mask detection of the intermediate products, which is necessary for elucidation of the release mechanisms. Experiments were performed in a sealed horizontal reactor under controlled atmospheres of 20% CO2 in He or 20% H2O in He at temperatures of 1100 °C, 1200 °C and 1400 °C. Thermal cracking of organics tar species occurs in the reactor [11]. All parts of the reactor downstream of the reaction zone were maintained at temperatures above the condensation point of the species of interest, and the MBMS was installed downstream for online gas analysis. A simplified schematic of the coupled apparatus is shown in Fig. 1. The experiments were carried out in a high density alumina tube to prevent reaction of the tube walls with the released inorganic species. A total gas flow of 4.0 L/min comprising 20 vol.% CO2 in He or 20 vol.% H2O in He was maintained in the reactor. A background spectrum was recorded for 20 s and a platinum boat loaded with 100 ± 2 mg of prepared coal (dry basis) was rapidly inserted into the hot zone of the furnace. The sample was heated to the prescribed experimental temperature and the gaseous reaction products flowed to the end of the reactor to be analysed by the coupled MBMS. Six duplicates of each release measurement were performed with identical sub-samples and the results averaged to determine the experimental variance. The averaged results were analysed qualitatively by examining the released species detected, the shape of the individual release signals, and the relationships between species indicated by the intensity, timing and duration of various
74
J. Tanner et al. / Fuel 158 (2015) 72–80
Fig. 1. High temperature furnace and coupled MBMS apparatus with online data logging for batch experiments.
release peaks. Semi-quantitative analysis to determine correlations in the relative amounts of detected species was also performed by integrating the averaged signal intensity over time and normalising the peak areas to the background 46CO+2 or 17OH+ signal from 0 to 20 s for CO2 and steam experiments, respectively. In the case of this MBMS, a linear correlation between signal intensity and the absolute quantities of a species released cannot be assumed [9], however, relative results and comparative trends between different samples and temperatures could still be determined.
2.3. Molecular beam mass spectrometry for inorganics analysis MBMS is a reliable method for analysing inorganic gases produced under high temperatures and pressures, with many such applications reported using differential pressure to generate the molecular beam [17]. In each case, highly reactive, condensable species are effectively quenched in the high vacuum environment, such that no further reaction or condensation is possible [9]. A short explanation of the coupled MBMS system used for this investigation is provided here, and detailed description given elsewhere [18]. The MBMS consists of three differentially pumped chambers. Hot product gas from the reactor enters the first chamber through the front nozzle orifice where it undergoes supersonic free jet expansion due to the differential pressure between the reactor and the first chamber of the apparatus. The central section of the gas stream, now under free molecular flow, is directed by a skimmer into the second evacuated chamber. A third, high vacuum chamber follows, in which a small proportion of the molecules are ionised and deflected into the quadrupole mass analyser where they are separated based on mass-to-charge ratio (m/z). The resultant signals are amplified and recorded by the software as a function of time and m/z. During the ionisation process, sufficient energy to break intra-molecular bonds can be imparted. Hence, detected species in the MBMS measurements may represent the released species themselves and any fragment ions thereof. The species and fragments of interest in this investigation were 23Na+, 34O+2/34H2S+, 35 + Cl , 36HCl+, 39K+/39NaO+, 58NaCl+, 60COS+/60NaCl+, 64SO+2 and 74 KCl+. Furthermore, 17OH+, 30CO+ and 46CO+2 signals were recorded to monitor the progress of the devolatilisation and gasification reactions, and to determine the role of CO2, H2O and derived species in the mechanisms of release. The quadrupole system of the MBMS has a resolution of ±1 amu. Hence, for three of the measured signals listed above, two species or fragments of interest having the same mass-to-charge ratio (m/z) are recorded in a single signal. In each case, under the current experimental conditions, there was an indisputably dominant species. Oxygen was deliberately excluded from the system, therefore the trends in m/z = 34 were attributed to the release of 34H2S+. The ratio of inherent Na/K for these coals is high, and the 74KCl+ signal was barely discernible from the background noise, therefore changes in m/z = 39 were assumed to represent 39NaO+. For
m/z = 60, the delayed release peak coupled with the relatively low abundance of the 60NaCl+ isotope indicate that this signal was predominantly due to 60COS+. Hereafter, these m/z ratios will be attributed exclusively to the above-stated dominant species and discussed as such. 3. Results and discussion A complex set of simultaneous, interrelated reactions between solid carbon, H2O, CO2, H2 and CO occur in a commercial gasifier. From the point of view of a single coal particle, beginning at low initial temperatures, drying and partial combustion produce the H2O, CO2 and heat required to drive the key endothermic, heterogeneous gasification reactions. As the particle temperature increases, volatiles species are released, and the Boudouard reaction (Eq. (1)) and steam-gasification (Eqs. (2) and (3)) simultaneously deplete the remaining char, generating the desirable syngas components, H2 and CO, as well as the by-product CO2:
CðsÞ þ CO2ðgÞ $ 2COðgÞ
DH ¼ þ172:7 kJ=mol
CðsÞ þ H2 OðgÞ $ H2ðgÞ þ COðgÞ
DH ¼ þ131:5 kJ=mol
CðsÞ þ 2H2 OðgÞ $ CO2ðgÞ þ 2H2ðgÞ
DH ¼ þ88:0 kJ=mol
ð1Þ ð2Þ ð3Þ
A portion of the CO generated is combusted in the gas phase, (Eq. (4)):
COðgÞ þ 1=2O2ðgÞ ! CO2ðgÞ
DH ¼ 283 kJ=mol
ð4Þ
And the water–gas shift reaction also influences the final syngas composition, with its rate and equilibrium dependent on the reactor temperature:
COðgÞ þ H2 OðgÞ $ H2ðgÞ þ CO2ðgÞ
DH ¼ 40:6 kJ=mol
ð5Þ
The conditions used in this study, i.e. gasification under 20% CO2 or 20% H2O, were therefore chosen for their relevance to industrial conditions and the need to determine the potential influence of these reagents on the release of volatile inorganic species during coal gasification. These reactions are temperature dependent, as are the evolution and subsequent gas phase reactions of inherent volatile inorganic species; hence the effect of temperature in the range of commercial gasification is also of interest. 3.1. Qualitative analysis The general qualitative trends discussed here were observed for all six coals. To illustrate these trends, representative MBMS spectra from measurements performed using HKN-S coal at 1400 °C are presented in Figs. 2 and 3. In each of these plots, the intensity scale for a particular species is consistent to allow for direct comparison; however the scales have been varied between species to more clearly show the low-intensity signals. The intensity-time spectra for the inorganic species (Fig. 2) clearly show the rapid initial phase of the reaction.
J. Tanner et al. / Fuel 158 (2015) 72–80
75
Fig. 2. Averaged intensity-time spectra for inorganic species released from HKN-S coal at 1400 °C under 20% CO2 (left) and 20% steam (right) in helium.
Fig. 3. Intensity-time spectra for sulphur-containing and other non-inorganic species released from HKN-S coal at 1400 °C under 20% CO2 (left) and 20% steam (right) in helium.
Devolatilisation, marked in the intensity-time plots by two vertical dotted lines, was taken to begin with sample insertion at time = 0 and end with the completion of volatile alkali release several seconds later. In all cases, the duration of the devolatilisation phase was longer under 20% CO2 than 20% H2O due to the higher rate of char gasification by steam in comparison to CO2. Steam gasification occurs at approximately 3 times the rate of CO2 gasification
under the equivalent conditions [19], and therefore results in faster exposure of volatile compounds and more rapid devolatilisation thereof. The duration of the devolatilisation phase also decreased with increasing temperature due to increased vapour pressure and mass transfer rate of the volatile species to the particle surface, and higher reaction rates of the carbonaceous material surrounding the volatile species. With the exception of Na, all inorganic
76
J. Tanner et al. / Fuel 158 (2015) 72–80
species of interest were detected during the devolatilisation phase under both CO2 and H2O conditions with similar orders of magnitude of the corresponding signal intensity. Overall, increased release intensity with increasing temperature was observed. This trend is attributed to increased volatility and increased rates of formation and decomposition reactions at higher temperatures. These mechanisms act in combination to expose additional inorganic material within the particle, which also contributes to the increasingly shorter devolatilisation times due to increased exposure and availability in the gas phase of the reactive inorganic species. The intensity-time spectra for the sulphur-containing and other non-inorganic species are shown in Fig. 3. As for devolatilisation, gasification under both steam and CO2 was taken to begin at time = 0, however, the products from these slower, heterogeneous reactions were not immediately apparent due to the diluting effect of the initial rapid release of highly volatile species. Hence two reaction phases, devolatilisation and gasification, may be clearly identified from the qualitative results. The relative intensities of the 17OH+, 30CO+ and 46CO+2 signals (Fig. 3) reflect the dominant char gasification reaction depending on the experimental atmosphere. As devolatilisation ceases, the rate of gasification increases as indicated by the sharp decrease in the 17OH+ and 46CO+2 signals for steam and CO2 experiments respectively. Under both atmospheres, a corresponding increase in 30CO+ was observed due to evolution of CO from the steam-gasification and Boudouard reactions (Eqs. (1) and (2)). Additionally, an increase in 17OH+ was observed under CO2 conditions due to drying of the coal, and an analogous increase in 46CO+2 under steam conditions due to oxidation, gasification and shift reactions (Eqs. (3)–(5)). Under H2O-rich conditions, the duration of the gasification phase decreased with increasing temperature, as evidenced by the length of time taken for the 17OH+ signal to return to the initial baseline value during the steam gasification experiments (Fig. 4). The peak shape for 17OH+ also changed, becoming sharper and more compressed with increasing temperature. This indicates that the rate of consumption of 17OH+, and by extension the rate of the steam gasification reaction, increased with increasing temperature. In addition to the dominant gasification reaction itself, an increase in the rate of devolatilisation, cracking and direct steam reforming further contribute to the decrease in 17OH+ with increasing temperature. A similar trend was observed for 30CO+ and 46CO+2 during experiments under CO2, which are products of the heterogeneous gasification reaction and of subsequent secondary gas-phase reaction, respectively. Furthermore, the higher rate of gasification
Fig. 4. Averaged intensity-time spectra for 17OH+ released from HKN-S coal at 1100 °C, 1200 °C and 1400 °C under 20% steam in helium. Y-axis scale is constant for all sub-plots.
reactions at higher temperatures resulted in a decreased reaction time due to the finite amount of reagents, and the intensity of the product signals decreased with time as the carbonaceous material in the sample was depleted. The release and subsequent reactions of sulphur-containing species from coal during devolatilisation and char gasification occur by distinct but related mechanisms under CO2 and steam conditions. The sulphide and aliphatic sulphur-containing functional groups in the coal structure decompose readily during the initial devolatilisation phase [8,11,20], and a portion thereof react with coal-hydrogen and coal-oxygen to produce H2S, SO2 and COS. Some of the aliphatic sulphur is also reincorporated into the char matrix as more stable heterocyclic structures [21]. As devolatilisation ceases and the slower char gasification reactions begin to dominate the further release of gaseous species, strongly bound organic sulphur remaining in the char matrix as aromatic and heterocyclic species is decomposed and released, predominantly as COS, with some SO2. The conditions under which the gasification is occurring, H2O- or CO2-rich, hereafter dictate the subsequent mechanistic steps and the final relative amounts and timing of the released S-containing species during the gasification phase. Due to the rapid evolution of gaseous species from the coal during devolatilisation, which results in a local dilution effect around the particles, S-species decomposition and reactions occurring during the devolatilisation phase were observed to be essentially independent of the reaction atmosphere. This can be seen by comparing, for example, the intensity and timing of the initial H2S shoulder under steam gasification conditions, which closely matches that of the H2S peak under CO2 gasification (Fig. 3). The H2S–COS equilibrium (Eq. (6)) is governed during devolatilisation by the partial pressure of H2O (pH2O) within the particle and near the particle surface from steam generated by residual moisture in the coal.
H2 SðgÞ þ CO2ðgÞ $ COSðgÞ þ H2 OðgÞ
ð6Þ
As char gasification proceeds under CO2 gasification conditions, sulphur released as COS from the solid matrix undergoes no further reaction and was detected after a short delay. Under H2O gasification conditions, however, pH2O near the particle surface is sufficiently high during the gasification phase to influence the H2S–COS equilibrium. This equilibrium shift results in the parallel, delayed release of H2S and CO2 and significant reduction in release of COS under steam gasification conditions as shown in Fig. 3. Some of the sulphur species released during devolatilisation are also captured by inherent CaO, releasing additional H2O and CO2:
CaOðsÞ þ COSðgÞ $ CaSðsÞ þ CO2ðgÞ
ð7Þ
CaOðsÞ þ H2 SðgÞ $ CaSðsÞ þ H2 OðgÞ
ð8Þ
FactSage equilibrium calculations indicate that the forward reaction with inherent CaO is thermodynamically favourable, therefore CaS will be formed preferentially and stabilised by the initially low local pH2O during the devolatilisation phase under both atmospheres. The products of these two reactions are visible as a small initial peak in the 46CO+2 spectra under steam gasification conditions, and as part of the initial 17OH+ spectra under CO2 conditions. Sulphur captured as CaS during devolatilisation reacts via different pathways under different gasification conditions. During steam gasification, pH2O surrounding the particle increases gradually towards the end of the devolatilisation phase and the equilibrium shown in Eq. (8) is shifted to the left, contributing to the observed secondary release of H2S. Under CO2-rich conditions, the mechanism proceeds via a two-step process to release CO and SO2:
J. Tanner et al. / Fuel 158 (2015) 72–80
CaSðsÞ þ 4CO2ðgÞ $ CaSO4ðsÞ þ 4COðgÞ
ð9Þ
CaSO4ðsÞ ! CaOðsÞ þ SO2ðgÞ þ 1=2O2ðgÞ
ð10Þ
The CO released as a result of Eq. (9) under CO2-rich conditions is visible as a shoulder on the right of the 30CO+ peak, slightly preceding and overlapping the delayed SO2 release. The moderate, slightly delayed release of SO2 observed under steam conditions is attributed to the decomposition of inherent sulphates and char-S compounds during the gasification phase. This also occurs but is masked in the case of CO2 gasification by the SO2 from sulphate decomposition. Sodium may compete with calcium in coals having substoichiometric amounts of sulphur. Reaction (11) is enhanced by the presence of steam, which increases the concentration of NaOH in the gas phase. Thermodynamic calculations indicate that Eq. (11) proceeds to completion above 600 °C. Eq. (12) is temperature dependent within the range of this investigation. At 1100 °C, Na2S is partially converted to Na2SO4. At 1400 °C, Na2S is also only partially converted, however, the equilibrium products are as per Eq. (12) at this temperature. Therefore, under steam gasification conditions, coals with higher Na to Ca ratio will not only undergo increased competition for the capture of S, but result in a reduced release of Na and S species overall, particularly at lower temperatures, due to the formation of the somewhat stable sulphides.
CaSðsÞ þ 2NaOHðgÞ ! CaOðsÞ þ Na2 SðsÞ þ H2 OðgÞ
ð11Þ
Na2 SðsÞ þ H2 OðgÞ $ 2NaOðgÞ þ H2 SðgÞ
ð12Þ
The proposed mechanism for sulphur release is somewhat cyclic in nature, particularly with regards to the capture by Ca and Na, resulting in the observed prolonged release of SO2 under CO2 condition and H2S under steam gasification conditions. Ca, in particular, is continually converted between oxide, sulphide and sulphate forms to facilitate the release of various sulphur species. Once all the carbonaceous material has been gasified, the Ca remains in the residual ash as CaO according to Eqs. (7), (9) and (10) in the CO2-rich case and Eqs. (8) and (11) in the H2O-rich case. Similarly, there will also be a portion of the Na remaining in the ash as Na2S under steam gasification conditions according to Eq. (12). The difference in intensity-time spectra for the release of 64SO+2 under CO2- and H2O-rich atmospheres at various temperatures is shown in Fig. 5. In general, the relative amount of 64SO+2 increased with increasing temperature and was higher under CO2 conditions. These observations support the proposed mechanisms of release of SO2 under CO2-rich and H2O-rich conditions outlined above, both of which begin with a small amount of sulphur released as SO2
Fig. 5. Averaged intensity-time spectra for scale is constant for all sub-plots.
64
77
during the devolatilisation phase, and result in some volatilised sulphur intermediates being captured as metal sulphides. Under CO2 gasification conditions, captured sulphide sulphur is predominantly released as SO2 via the metal sulphate during the gasification phase (Eqs. (9) and (10)). This capture and release mechanism is temperature dependent within the range of this investigation. At 1100 °C, capture rates are low and the small amount of sulphate which is formed is stable, therefore the small release of SO2 at low temperatures is attributed almost solely to devolatilisation products, and the metal remains in sulphate form in the residual ash. At 1400 °C, the rates of sulphur capture and sulphate decomposition increase such that the two step sulphate mechanism proceeds more rapidly, and a significant delayed release of SO2 is observed. Additionally, equilibrium calculations indicate that while CaSO4 will completely decompose at 1400 °C, Na2SO4 is more stable and a significant portion will remain. This indicates that coals which capture sulphur as sodium sulphate or which contain native Na2SO4 will permanently capture both sodium and sulphur under CO2 gasification conditions at these temperatures. Under steam gasification conditions, sulphide sulphur reacts either directly with steam, or with NaOH. In both cases, the dominant sulphur-containing product is H2S, hence the relatively low release of SO2 in comparison with equivalent-temperature CO2 conditions. Another feature of the SO2 released under H2O-rich conditions in Fig. 5 is the secondary peak that occurred at 1100 °C, but was minimal at higher temperatures and did not occur under CO2-rich conditions. At temperatures just below 1100 °C, CaSO4 is relatively stable. Considering the endothermic nature of the char gasification reactions, and the higher reaction rate for steam gasification in comparison to CO2 gasification, it is possible that the temperature of the particle under H2O-rich conditions was slightly decreased during gasification of the carbonaceous material. Once the carbonaceous material was exhausted and the concentration of H2O returned to the baseline value, residual sulphates in the solid matrix were decomposed, releasing SO2 as observed. At higher temperatures, the rate of gasification increases, CaSO4 decomposes readily, and any slight delay was therefore masked by an overlap of the secondary SO2 release and the initial release from devolatilisation. 3.2. Semi-quantitative analysis Fig. 6 shows the relative amounts of SO2 released during the devolatilisation and gasification periods for each coal under H2O-rich conditions at various temperatures. In general, increasing initial release and decreasing secondary release with increasing
SO+2 released from HKN-S coal at 1100 °C, 1200 °C and 1400 °C under 20% CO2 (left) and 20% steam (right) in helium. Y-axis
78
J. Tanner et al. / Fuel 158 (2015) 72–80
Fig. 6. Averaged, normalised peak area of primary and secondary SO2 released from Victorian and Rhenish coals at 1100 °C, 1200 °C and 1400 °C under 20% steam gasification conditions.
temperature were observed. As temperature increases, functional group decomposition and volatilisation increases, resulting in the increased initial release. This depletes the finite supply of sulphur, resulting in the observed decreasing trend in secondary release at higher temperatures. Furthermore, the total release of sulphur as SO2 under H2O-rich conditions also generally decreased with increasing temperature due to increased permanent capture, reaction with NaOH, and release as H2S instead of SO2 under H2O-rich conditions. As discussed above, the observed trends in the release of SO2 with temperature, and the partitioning between the devolatilisation and gasification phases may be largely attributed to the relative stability and decomposition of sulphates under various conditions. These results further indicate that it may therefore be possible to control the release of SO2 under certain gasification conditions, either by controlling the reaction temperature or steam concentration in a continuous system, thereby avoiding the final release of SO2 observed once the steam concentration returns to baseline values in Fig. 5. However, further investigation is required to determine the influence of configuration, continuous operation and multi-reagent systems on the entrained flow gasification process. The averaged, normalised peak areas for selected inorganic species released at 1400 °C are presented in Fig. 7. The data presented have been scaled as noted in the legend to account for large differences in the orders of magnitude and for the variation in instrument sensitivity between different species. Therefore, while the relative release of each species under different conditions is directly comparable, the absolute values for different species are not.
Analogous experiments were also performed at 1100 °C and 1200 °C. The direction of the equilibrium reactions discussed above is generally opposite for CO2- and H2O-rich environments; therefore, the temperature dependence and release trends with temperature were contrary. Overall, the release of inorganic species at 1400 °C was higher under CO2 conditions, with the exception of 23 Na+ and 34H2S+. These trends in the semi-quantitative data reflect the proposed release mechanisms described above. The results are somewhat contrary to previous investigations which showed a significant decrease in the release of alkali metals and an increase in the release of alkali metal hydroxides under low-O2 gasification conditions in the presence of steam [13]. It has previously been shown that Na not associated with Si is present as either NaCl or carboxylate Na available for mobilisation, volatilisation or reaction [22]. The 23Na+ signal is therefore attributed directly to elemental sodium or to ionised fragments of Na-containing species released from the coal. In general, the release of 23Na+ was negligible under CO2-gasification conditions due to the higher rates of conversion to carbonate form and subsequent capture by inherent silicates and aluminosilicates [23,24] in the CO-rich environment created by the Boudouard reaction (Eq. (1)). Under steam gasification conditions, the trend in 23Na+ followed the concentration of inherent sodium in the parent coals. The observation of Na release under steam conditions only is due to a combination of the increased formation and subsequent fragmentation of NaOH, the competitive nature of the mechanism of S-capture and release shown in Eqs. (11) and (12), and the lower rate of capture of sodium as silicates and aluminosilicates. It has also been shown [25] that the presence of steam lowers the melting point of Na2CO3, leading to increased release of metallic Na. This effect may also explain the absence of 23Na+ release observed under CO2 conditions due to increased permanent capture of Na by silicates, according to:
2NaOHðgÞ þ COðgÞ $ Na2 CO3ðlÞ þ H2ðgÞ
ð13Þ
Na2 CO3ðlÞ þ xSiO2ðsÞ $ Na2 O xSiO2ðsÞ þ CO2ðgÞ
ð14Þ
36
+
35
+
The release of HCl and Cl followed the trend of inherent Cl in the parent coals for both CO2 and H2O conditions. The total release for each Cl species was lower under H2O-rich conditions, within the experimental variance. The majority of Cl and HCl were released during the initial devolatilisation phase, after which the HCl is involved in gas-phase equilibrium with NaCl via NaOH (Eq. (15)).
NaOHðgÞ þ HClðgÞ $ NaClðgÞ þ H2 OðgÞ
ð15Þ
Previous work under combustion and gasification conditions has shown that the capture of Na by aluminosilicates affects the release of Na and Cl species [12,13]. During devolatilisation under
Fig. 7. Averaged, normalised peak areas for inorganic species released from Victorian and Rhenish coals at 1400 °C under 20% CO2 (left) and 20% steam (right) in helium.
J. Tanner et al. / Fuel 158 (2015) 72–80
steam gasification conditions, the NaOH–NaCl equilibrium favours the hydroxide form. However, as gasification proceeds and NaCl is captured as silicates and aluminosilicates by Eqs. (16) and (17) [26], the equilibrium shifts to the right. NaOH and NaCl are consumed, resulting in the observed decrease in 39NaO+ and 58NaCl+ release under H2O-rich conditions.
2NaCl þ xAlSiy Oz þ H2 O ! Na-aluminosilicate þ 2HCl
ð16Þ
2NaCl þ xSiO2 þ H2 O ! Na-silicate þ 2HCl
ð17Þ
The release trends for the sulphur-containing species H2S and SO2 do not follow the inherent S content of the parent coals due to the variety and different reaction pathways of the various forms of sulphur present. According to the proposed mechanisms, coals relatively low in sulphides and aliphatic S-containing functional groups should exhibit lower release of initial H2S, COS and SO2 during the devolatilisation phase. Those with lower aromatic and heterocyclic S should exhibit lower release of H2S and COS during the gasification phase under H2O- and CO2-rich conditions, respectively. This summary is, however, complicated by the nature of the mechanism. The cyclic release, reincorporation, capture and competition steps which sulphur species undergo during the devolatilisation phase are interrelated, and the subsequent steps in the gasification mechanisms for each atmosphere include several equilibrium steps, which are both temperature and atmosphere dependent. It is therefore not possible to determine the ratio of the inherent forms of sulphur in a coal sample by these MBMS measurements alone. H2S release was generally shown to be higher under H2O-rich conditions, with the exception of LY. This is predominantly due to the artificially high Si content in the LY sample leading to increased capture of Na and hence reduced NaOH and H2S release. The formation of Na2S is therefore affected, influencing the release of H2S according to Eqs. (11) and (12). LY also has low Ca content, minimising the potential for CaS formation and reaction with H2O to release H2S via Eq. (8). The opposite effect is evident for HKN-S+, where low inherent Al + Si, high Ca and high S resulted in proportionally higher release of H2S. For HKS, the release of 39NaO+ was higher than expected due to low Al and Si in the parent coal resulting in low capture of Na. HKS also has the highest native Na content, indicating that NaOH will be higher, and therefore a greater proportion of S converted to H2S via Eqs. (11) and (12), hence the proportionally lower release of SO2. These results are in general agreement with trends observed in previous work under combustion and low-O2 gasification conditions in which it was shown that H2S release is in high negative correlation with Ca/S ratio under gasification and combustion conditions [10]. Previous studies [27] have also shown that the amount of sulphur fixed in ash was a function of the Ca/S ratio and independent of the form of sulphur. In the case of the six coals investigated here, those with higher Ca/S generally released less S-containing species than coals with lower Ca/S, confirming the previous group’s results. HKS and HKN-S have low total S content and relatively high Na and Ca, increasing the chance of capture as CaS in the absence of excess CO2, as reflected by the lower release of SO2 under H2O-rich conditions. 4. Conclusion The release and reaction mechanisms of volatile inorganic species from Victorian brown coals and Rhenish lignites at 1100 °C, 1200 °C and 1400 °C under 20% CO2 and 20% H2O in helium were investigated. In particular, the influence of temperature and the role of the gasification reagents CO2 and H2O in the release mechanisms themselves were studied, and the results compared with previous
79
investigations under similar conditions. The release and subsequent reactions of volatile species from coal during devolatilisation and char gasification were shown to occur by distinct but related mechanisms under CO2 and H2O conditions. Two overlapping phases were clearly observed from the results – devolatilisation and char gasification. The Na, Cl and S species released and subsequent reactions which occurred during devolatilisation were dependent primarily on the particle temperature and rate of char gasification, and were essentially independent of the reaction atmosphere itself. During the gasification phase, however, the release mechanisms for the Na, Cl and S species directly involved CO2 and H2O, and the absence, presence and concentration of the gasification reagent affected the release trends, intensity and timing. It is predominantly the equilibrium reactions in the proposed mechanism that were influenced by CO2 and H2O. In most cases, the equilibrium reactions exhibited opposite shifts for CO2 and H2O-rich environments; therefore, the temperature dependence and release trends with temperature were contrary. Overall, the release of inorganic species was higher under CO2 conditions. However, the release of gaseous Na was negligible under CO2-gasification conditions due to enhanced capture as aluminosilicates. Under steam gasification conditions, the trend in Na release followed the concentration of inherent sodium in the parent coals. The release of Cl species followed the trend of inherent Cl in the parent coals for both CO2 and H2O conditions. The release trends for the sulphur-containing species H2S and SO2 did not follow the inherent S content of the parent coals and H2S release was higher under H2O-rich conditions due to the dependence of sulphur capture on the Ca/S ratio and CaS–CaO equilibrium. It was shown that the inherent forms of sulphur in the parent coals affect the released sulphur-containing species, however, it was not possible to determine the ratio of the inherent forms of sulphur in a coal sample by these MBMS measurements alone. The proposed mechanisms also involve the competitive capture of S by Na and Ca. Ca acts as a catalyst of sorts, continually cycling between oxide, sulphide and sulphate forms to facilitate the release of various sulphur species. This mechanism may also result in the permanent capture of Na, Ca and S in the solid residue remaining after gasification, depending on the presence and concentration of CO2 and H2O and the composition of the parent coal. It may therefore be possible to partially control the release of some volatile species by controlling either the reaction temperature or steam concentration. Acknowledgements The authors would like to acknowledge Brown Coal Innovation Australia (BCIA), Bundesministerium für Wirtschaft und Energie (FKZ 0327773) and the Go8-DAAD Joint Research Scheme for providing the funding for this collaborative research project. References [1] OECD/IEA. World energy outlook. France: International Energy Agency; 2011. [2] Huffman GP, Huggins FE. Reactions and transformations of coal mineral matter at elevated temperatures. In: Mineral matter and ash in coal. American Chemical Society; 1986. p. 100–13. [3] Gottwald U, Monkhouse P, Wulgaris N, Bonn B. In-situ study of the effect of operating conditions and additives on alkali emissions in fluidised bed combustion. Fuel Process Technol 2002;75:215–26. [4] Monkhouse P. On-line spectroscopic and spectrometric methods for the determination of metal species in industrial processes. Prog Energy Combust Sci 2011;37:125–71. [5] Bläsing M, Müller M. Release of alkali metal, sulfur, and chlorine species during high-temperature gasification of coal and coal blends in a drop tube reactor. Energy Fuels 2012;26:6311–5. [6] Bläsing M, Müller M. Release of alkali metal, sulphur, and chlorine species from high temperature gasification of high- and low-rank coals. Fuel Process Technol 2013;106:289–94.
80
J. Tanner et al. / Fuel 158 (2015) 72–80
[7] van Eyk PJ, Ashman PJ, Alwahabi ZT, Nathan GJ. The release of water-bound and organic sodium from Loy Yang coal during the combustion of single particles in a flat flame. Combust Flame 2011;158:1181–92. [8] Yan J, Yang J, Liu Z. SH radical: the key intermediate in sulfur transformation during thermal processing of coal. Environ Sci Technol 2005;39:5043–51. [9] Wolf KJ, Müller M, Hilpert K, Singheiser L. Alkali sorption in second-generation pressurized fluidized-bed combustion. Energy Fuels 2004;18:1841–50. [10] Bläsing M, Melchior T, Müller M. Influence of temperature on the release of inorganic species during high temperature gasification of Rhenish lignite. Fuel Process Technol 2011;92:511–6. [11] 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. [12] Bläsing M, Müller M. Influence of pressure on the release of inorganic species during high temperature gasification of coal. Fuel 2011;90:2326–33. [13] Bläsing M, Müller M. Investigations on the influence of steam on the release of sodium, potassium, chlorine, and sulphur species during high temperature gasification of coal. Fuel 2012;94:137–43. [14] Tanner J, Bläsing M, Müller M, Bhattacharya S. Influence of temperature on the release of inorganic species from Victorian brown coals and German lignites under CO2 gasification conditions. Energy Fuels 2014;28:6289–98. [15] Hayashi J-I, Li C-Z. Structure and properties of Victorian brown coal. In: Li C-Z, editor. Advances in the science of Victorian brown coal. Elsevier; 2004. [16] Sakaguchi M, Laursen K, Nakagawa H, Miura K. Hydrothermal upgrading of Loy Yang brown coal – effect of upgrading conditions on the characteristics of the products. Fuel Process Technol 2008;89:391–6.
[17] Drowart J, Goldfinger P. Investigation of inorganic systems at high temperature by mass spectrometry. Angew Chem Int Ed Engl 1967;6:581–96. [18] Wolf KJ, Untersuchungen zur Freisetzung und Einbindung von Alkalien bei der reduzierenden Druckwirbelschichtverbrennung (Investigations of the release and sorption of alkali metals in pressurized fluidized bed combustion under reducing conditions), PhD Thesis: RWTH Aachen; 2003 [In German]. [19] Walker Jr PL, Rusinko Jr F, Austin LG. Gas reactions of carbon. In: Eley DD, Selwood PW, Weisz PB, editors. Advances in catalysts, 11. New York: Academic Press; 1959. p. 133–221. [20] Khan MR. Prediction of sulphur distribution in products during low temperature coal pyrolysis and gasification. Fuel 1989;68:1439–49. [21] Calkins WH. The chemical forms of sulfur in coal: a review. Fuel 1994;73:475–84. [22] Brockway DJ, Ottrey AL, Higgins RS. Inorganic constituents. In: Durie RA, editor. The science of Victorian brown coal. Butterworth-Heinemann; 1991. p. 597–650 [chapter 11]. [23] Kosminski A, Ross DP, Agnew JB. Transformations of sodium during gasification of low-rank coal. Fuel Process Technol 2006;87:943–52. [24] Kosminski A, Ross DP, Agnew JB. Reactions between sodium and kaolin during gasification of a low-rank coal. Fuel Process Technol 2006;87:1051–62. [25] Byung Ho S, Sang Done K. Catalytic activity of alkali and iron salt mixtures for steam-char gasification. Fuel 1993;72:797–803. [26] Wei X, Huang J, Liu T, Fang Y, Wang Y. Transformation of alkali metals during pyrolysis and gasification of a lignite. Energy Fuels 2008;22:1840–4. [27] Schafer HNS. Functional groups and ion exchange properties. In: Durie RA, editor. The science of Victorian brown coal. Butterworth-Heinemann; 1991. p. 323–57 [chapter 7].
Contents lists available at ScienceDirect
Fuel journal homepage: www.elsevier.com/locate/fuel
The temperature-dependent release of volatile inorganic species from Victorian brown coals and German lignites under CO2 and H2O gasification conditions Joanne Tanner a, Marc Bläsing b,⇑, Michael Müller b, Sankar Bhattacharya a a b
Department of Chemical Engineering, Monash University, Wellington Rd, Clayton 3800, Australia Institute for Energy Research (IEF-2), Leo-Brandt-Str. 1, 52425 Jülich, Germany
h i g h l i g h t s The Na, Cl and S species released during gasification of low rank coal are reported. A release mechanism including the role and influence of CO2 and steam is proposed. The 2-phase mechanism is similar but distinct under CO2 and steam atmospheres. The release of volatile Na and Cl species essentially independent of gas atmosphere. The S species release mechanism depends directly on CO2 and steam concentration.
a r t i c l e
i n f o
Article history: Received 28 October 2014 Received in revised form 2 April 2015 Accepted 28 April 2015 Available online 8 May 2015 Keywords: Brown coal Lignite Gasification Volatile inorganics
a b s t r a c t To promote the use of abundant low rank coal resources, and to assist with the design, development and optimisation of hot gas cleaning and downstream processes, it is necessary to understand the release mechanisms of volatile inorganic species from low rank coals under high temperature gasification conditions. Although a significant amount of work has been reported under combustion and oxygen lean gasification conditions, these studies do not sufficiently explain the role of the gasification reagents CO2 and H2O in the mechanisms. Therefore, gasification experiments under 20% CO2 and 20% H2O in He were conducted at 1100 °C, 1200 °C and 1400 °C for two Victorian brown coals and four Rhenish lignites. Hot gas analysis was conducted by online molecular beam mass spectrometry to determine the intensity, relative quantity and timing of the release of volatile species of interest. Two overlapping phases were clearly observed from the results – devolatilisation and char gasification. Major species detected were 23Na+, 34H2S+, 35Cl+, 36HCl+, 39K+/39NaO+, 58NaCl+, 60COS+/60NaCl+, and 64SO+2. The release during devolatilisation was essentially independent of the bulk gas atmosphere and constituted the majority of Na and Cl species. The release of S-species occurred predominantly during gasification under both CO2 and H2O atmospheres by similar but distinct mechanisms, and was directly affected by the absence, presence and concentration of the gasification reagents. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Coal accounted for over half of the increase in global energy use over the past decade and is predicted to play a major role in all future projection scenarios to 2035 [1]. It is therefore important to continue the development of cost-effective, environmentally responsible coal processes, such as high efficiency gasification, for the conversion of this abundant resource to power and ⇑ Corresponding author. Tel.: +49 2461 61 1574; fax: +49 2461 61 3699. E-mail address: [email protected] (M. Bläsing). http://dx.doi.org/10.1016/j.fuel.2015.04.071 0016-2361/Ó 2015 Elsevier Ltd. All rights reserved.
products. In particular, low rank coals are of interest for use in advanced coal-to-product gasification technologies for countries with significant reserves, such as Australia and Germany. Low rank coals, as well as those of higher rank, often contain significant amounts of inorganic species including Na, Cl and S. During gasification and other high temperature processes, these species undergo decomposition and reaction, resulting in a complex mixture of volatile inorganic products known to cause slagging and fouling deposits, corrosion and pollution problems [2]. It is therefore important to design, develop and optimise high temperature gas cleaning systems, downstream conversion processes,
73
J. Tanner et al. / Fuel 158 (2015) 72–80
and environmental controls to enhance the overall gasification process efficiency. A comprehensive knowledge of the mechanisms of release and subsequent interactions between volatile inorganic species, including the role of the gasification reagents, is therefore required to facilitate the prediction of detrimental gas-phase components based on parent coal properties. Investigations into the modes of occurrence and release of volatile inorganics at high temperatures have historically been conducted using a variety of modelling and experimental methods [3,4]. Recently, online mass spectrometric techniques [5–8] have predominantly been used to measure in situ decomposition and reaction products from high temperature gasification and combustion processes. In particular, molecular beam mass spectrometry (MBMS) yields time dependent, online, simultaneous data pertaining to the release of volatile inorganic species under high temperature conditions of industrial interest [9]. Recent investigations in the area of the release of volatile inorganic species from low rank coals under gasification conditions have focused on the effects of temperature, pressure, oxygen partial pressure and steam addition [6,10–14]. However, the influence of different bulk gas atmospheres on the release mechanisms is still unknown. This study therefore uses the online MBMS technique to investigate the influence of temperature and bulk gas atmosphere on the in situ reaction products from high temperature gasification of two Victorian brown coals and four Rhenish lignites. In particular, a greater understanding of the role of the primary gasification reagents, CO2 and H2O, in the release and reaction mechanisms of the volatile Na-, Cl- and S-species under entrained flow conditions is sought. The results of this study are further compared with previous results obtained under O2 and O2/steam atmospheres [10,11,13] where appropriate.
Table 1 Results of the chemical analysis of the coals under investigation. HKN-S
HKN-S+
HKS
HKT
LY
MOR
Proximate analysis of the coals (mass %) Moisture (air dried) 20.02 19.76 Ash (dry basis) 3.58 4.23 Volatiles (dry basis) 51.43 53.17
20.45 6.59 52.45
10.27 13.20 55.24
11.16 7.99 48.25
14.92 3.59 49.31
Chemical composition (dry basis, mass %) C 65.8 65.8 H 4.81 4.98 N 0.78 0.84 S 0.205 0.508 Cl 0.01 0.03 Al 0.024 0.025 Fe 0.18 0.22 Ca 0.79 0.90 Mg 0.30 0.34 K 0.016 0.016 Na 0.19 0.19 Si 0.009 0.020
62.0 4.89 0.69 0.365 0.02 0.099 0.40 1.2 0.43 0.018 0.42 0.60
57.3 4.18 0.75 0.478 0.01 1.2 0.20 1.1 0.39 0.069 0.22 2.7
50.1 4.3 0.445 0.207 0.092 0.204 0.209 0.133 0.114 0.014 0.101 7.180
60.7 5.35 0.52 0.041 0.039 0.019 0.185 0.417 0.194 0.010 0.056 0.052
Elemental Ratios (molar basis) Na/Cl 28.88 Na/K 20.48 Ca/S 3.08 Na/Ca 0.42
27.48 38.29 2.58 0.62
31.05 5.56 1.90 0.34
1.68 12.40 0.52 1.32
2.21 9.51 8.16 0.23
11.33 19.68 1.42 0.36
HKx = Rhenish Lignites, Hambach; MOR = Morwell brown coal, Victoria.
LY = Loy
Yang
brown
coal,
Victoria;
likely an artefact of sampling, such as extraneous sand or clay from the interseam or overburden. The additional silica content may influence the release mechanism of sodium as discussed in previous work [14].
2.2. Experimental setup 2. Materials and methods 2.1. Fuel preparation Run-of-mine samples of four Rhenish lignites (HKN-S+, HKN-S, HKS and HKT) were supplied by RWE Power and two Victorian brown coals (Loy Yang – LY and Morwell – MOR) were supplied by the Loy Yang and Hazelwood Power Stations, respectively. Coals from these two regions were chosen due to their similarities in age (tertiary) and rank (low). The samples were air dried and pulverised in a mill. The Rhenish coals were sieved to a particle size range of <100 lm and the Victorian coals to 90–106 lm. All samples were thereafter preserved at room temperature under dry conditions. Of particular interest in this investigation was the difference in the inorganic composition of the coals, specifically regarding Na, K, S, Cl, Al, Si, and Ca. These elements and various compounds thereof, typically found in low rank coals, are known either for their volatile nature or for their influence on the volatilisation of other inherent or derived inorganic species released during the thermal treatment of coals. The chemical composition of the coals under investigation is given in Table 1. Chemical analysis was performed by the central division of analytical chemistry (ZCH) of the Forschungszentrum Jülich. The standard methods used and variance of analytical results are detailed elsewhere [6]. The chemical analysis of the Loy Yang sample showed anomalously high ash content – almost 8% on a dry basis (Table 1). Based on similar samples reported in recent literature, Loy Yang coal is expected to have an ash content in the range of 0.5–1.6% on a dry basis. Uncharacteristically high aluminium and silicon were also measured in the current sample in comparison to previously reported values for Loy Yang coal [7,15,16] and the presence of large amounts of silica was confirmed by XRD. These features are
The release of Na-, S-, and Cl-containing species during steam gasification was investigated under conditions analogous to those experienced by coal particles during industrial entrained flow gasification. The fixed-bed, batch configuration with continuous gas flow was deliberately chosen. In contrast, under continuous feed conditions the final product gas mixture would mask detection of the intermediate products, which is necessary for elucidation of the release mechanisms. Experiments were performed in a sealed horizontal reactor under controlled atmospheres of 20% CO2 in He or 20% H2O in He at temperatures of 1100 °C, 1200 °C and 1400 °C. Thermal cracking of organics tar species occurs in the reactor [11]. All parts of the reactor downstream of the reaction zone were maintained at temperatures above the condensation point of the species of interest, and the MBMS was installed downstream for online gas analysis. A simplified schematic of the coupled apparatus is shown in Fig. 1. The experiments were carried out in a high density alumina tube to prevent reaction of the tube walls with the released inorganic species. A total gas flow of 4.0 L/min comprising 20 vol.% CO2 in He or 20 vol.% H2O in He was maintained in the reactor. A background spectrum was recorded for 20 s and a platinum boat loaded with 100 ± 2 mg of prepared coal (dry basis) was rapidly inserted into the hot zone of the furnace. The sample was heated to the prescribed experimental temperature and the gaseous reaction products flowed to the end of the reactor to be analysed by the coupled MBMS. Six duplicates of each release measurement were performed with identical sub-samples and the results averaged to determine the experimental variance. The averaged results were analysed qualitatively by examining the released species detected, the shape of the individual release signals, and the relationships between species indicated by the intensity, timing and duration of various
74
J. Tanner et al. / Fuel 158 (2015) 72–80
Fig. 1. High temperature furnace and coupled MBMS apparatus with online data logging for batch experiments.
release peaks. Semi-quantitative analysis to determine correlations in the relative amounts of detected species was also performed by integrating the averaged signal intensity over time and normalising the peak areas to the background 46CO+2 or 17OH+ signal from 0 to 20 s for CO2 and steam experiments, respectively. In the case of this MBMS, a linear correlation between signal intensity and the absolute quantities of a species released cannot be assumed [9], however, relative results and comparative trends between different samples and temperatures could still be determined.
2.3. Molecular beam mass spectrometry for inorganics analysis MBMS is a reliable method for analysing inorganic gases produced under high temperatures and pressures, with many such applications reported using differential pressure to generate the molecular beam [17]. In each case, highly reactive, condensable species are effectively quenched in the high vacuum environment, such that no further reaction or condensation is possible [9]. A short explanation of the coupled MBMS system used for this investigation is provided here, and detailed description given elsewhere [18]. The MBMS consists of three differentially pumped chambers. Hot product gas from the reactor enters the first chamber through the front nozzle orifice where it undergoes supersonic free jet expansion due to the differential pressure between the reactor and the first chamber of the apparatus. The central section of the gas stream, now under free molecular flow, is directed by a skimmer into the second evacuated chamber. A third, high vacuum chamber follows, in which a small proportion of the molecules are ionised and deflected into the quadrupole mass analyser where they are separated based on mass-to-charge ratio (m/z). The resultant signals are amplified and recorded by the software as a function of time and m/z. During the ionisation process, sufficient energy to break intra-molecular bonds can be imparted. Hence, detected species in the MBMS measurements may represent the released species themselves and any fragment ions thereof. The species and fragments of interest in this investigation were 23Na+, 34O+2/34H2S+, 35 + Cl , 36HCl+, 39K+/39NaO+, 58NaCl+, 60COS+/60NaCl+, 64SO+2 and 74 KCl+. Furthermore, 17OH+, 30CO+ and 46CO+2 signals were recorded to monitor the progress of the devolatilisation and gasification reactions, and to determine the role of CO2, H2O and derived species in the mechanisms of release. The quadrupole system of the MBMS has a resolution of ±1 amu. Hence, for three of the measured signals listed above, two species or fragments of interest having the same mass-to-charge ratio (m/z) are recorded in a single signal. In each case, under the current experimental conditions, there was an indisputably dominant species. Oxygen was deliberately excluded from the system, therefore the trends in m/z = 34 were attributed to the release of 34H2S+. The ratio of inherent Na/K for these coals is high, and the 74KCl+ signal was barely discernible from the background noise, therefore changes in m/z = 39 were assumed to represent 39NaO+. For
m/z = 60, the delayed release peak coupled with the relatively low abundance of the 60NaCl+ isotope indicate that this signal was predominantly due to 60COS+. Hereafter, these m/z ratios will be attributed exclusively to the above-stated dominant species and discussed as such. 3. Results and discussion A complex set of simultaneous, interrelated reactions between solid carbon, H2O, CO2, H2 and CO occur in a commercial gasifier. From the point of view of a single coal particle, beginning at low initial temperatures, drying and partial combustion produce the H2O, CO2 and heat required to drive the key endothermic, heterogeneous gasification reactions. As the particle temperature increases, volatiles species are released, and the Boudouard reaction (Eq. (1)) and steam-gasification (Eqs. (2) and (3)) simultaneously deplete the remaining char, generating the desirable syngas components, H2 and CO, as well as the by-product CO2:
CðsÞ þ CO2ðgÞ $ 2COðgÞ
DH ¼ þ172:7 kJ=mol
CðsÞ þ H2 OðgÞ $ H2ðgÞ þ COðgÞ
DH ¼ þ131:5 kJ=mol
CðsÞ þ 2H2 OðgÞ $ CO2ðgÞ þ 2H2ðgÞ
DH ¼ þ88:0 kJ=mol
ð1Þ ð2Þ ð3Þ
A portion of the CO generated is combusted in the gas phase, (Eq. (4)):
COðgÞ þ 1=2O2ðgÞ ! CO2ðgÞ
DH ¼ 283 kJ=mol
ð4Þ
And the water–gas shift reaction also influences the final syngas composition, with its rate and equilibrium dependent on the reactor temperature:
COðgÞ þ H2 OðgÞ $ H2ðgÞ þ CO2ðgÞ
DH ¼ 40:6 kJ=mol
ð5Þ
The conditions used in this study, i.e. gasification under 20% CO2 or 20% H2O, were therefore chosen for their relevance to industrial conditions and the need to determine the potential influence of these reagents on the release of volatile inorganic species during coal gasification. These reactions are temperature dependent, as are the evolution and subsequent gas phase reactions of inherent volatile inorganic species; hence the effect of temperature in the range of commercial gasification is also of interest. 3.1. Qualitative analysis The general qualitative trends discussed here were observed for all six coals. To illustrate these trends, representative MBMS spectra from measurements performed using HKN-S coal at 1400 °C are presented in Figs. 2 and 3. In each of these plots, the intensity scale for a particular species is consistent to allow for direct comparison; however the scales have been varied between species to more clearly show the low-intensity signals. The intensity-time spectra for the inorganic species (Fig. 2) clearly show the rapid initial phase of the reaction.
J. Tanner et al. / Fuel 158 (2015) 72–80
75
Fig. 2. Averaged intensity-time spectra for inorganic species released from HKN-S coal at 1400 °C under 20% CO2 (left) and 20% steam (right) in helium.
Fig. 3. Intensity-time spectra for sulphur-containing and other non-inorganic species released from HKN-S coal at 1400 °C under 20% CO2 (left) and 20% steam (right) in helium.
Devolatilisation, marked in the intensity-time plots by two vertical dotted lines, was taken to begin with sample insertion at time = 0 and end with the completion of volatile alkali release several seconds later. In all cases, the duration of the devolatilisation phase was longer under 20% CO2 than 20% H2O due to the higher rate of char gasification by steam in comparison to CO2. Steam gasification occurs at approximately 3 times the rate of CO2 gasification
under the equivalent conditions [19], and therefore results in faster exposure of volatile compounds and more rapid devolatilisation thereof. The duration of the devolatilisation phase also decreased with increasing temperature due to increased vapour pressure and mass transfer rate of the volatile species to the particle surface, and higher reaction rates of the carbonaceous material surrounding the volatile species. With the exception of Na, all inorganic
76
J. Tanner et al. / Fuel 158 (2015) 72–80
species of interest were detected during the devolatilisation phase under both CO2 and H2O conditions with similar orders of magnitude of the corresponding signal intensity. Overall, increased release intensity with increasing temperature was observed. This trend is attributed to increased volatility and increased rates of formation and decomposition reactions at higher temperatures. These mechanisms act in combination to expose additional inorganic material within the particle, which also contributes to the increasingly shorter devolatilisation times due to increased exposure and availability in the gas phase of the reactive inorganic species. The intensity-time spectra for the sulphur-containing and other non-inorganic species are shown in Fig. 3. As for devolatilisation, gasification under both steam and CO2 was taken to begin at time = 0, however, the products from these slower, heterogeneous reactions were not immediately apparent due to the diluting effect of the initial rapid release of highly volatile species. Hence two reaction phases, devolatilisation and gasification, may be clearly identified from the qualitative results. The relative intensities of the 17OH+, 30CO+ and 46CO+2 signals (Fig. 3) reflect the dominant char gasification reaction depending on the experimental atmosphere. As devolatilisation ceases, the rate of gasification increases as indicated by the sharp decrease in the 17OH+ and 46CO+2 signals for steam and CO2 experiments respectively. Under both atmospheres, a corresponding increase in 30CO+ was observed due to evolution of CO from the steam-gasification and Boudouard reactions (Eqs. (1) and (2)). Additionally, an increase in 17OH+ was observed under CO2 conditions due to drying of the coal, and an analogous increase in 46CO+2 under steam conditions due to oxidation, gasification and shift reactions (Eqs. (3)–(5)). Under H2O-rich conditions, the duration of the gasification phase decreased with increasing temperature, as evidenced by the length of time taken for the 17OH+ signal to return to the initial baseline value during the steam gasification experiments (Fig. 4). The peak shape for 17OH+ also changed, becoming sharper and more compressed with increasing temperature. This indicates that the rate of consumption of 17OH+, and by extension the rate of the steam gasification reaction, increased with increasing temperature. In addition to the dominant gasification reaction itself, an increase in the rate of devolatilisation, cracking and direct steam reforming further contribute to the decrease in 17OH+ with increasing temperature. A similar trend was observed for 30CO+ and 46CO+2 during experiments under CO2, which are products of the heterogeneous gasification reaction and of subsequent secondary gas-phase reaction, respectively. Furthermore, the higher rate of gasification
Fig. 4. Averaged intensity-time spectra for 17OH+ released from HKN-S coal at 1100 °C, 1200 °C and 1400 °C under 20% steam in helium. Y-axis scale is constant for all sub-plots.
reactions at higher temperatures resulted in a decreased reaction time due to the finite amount of reagents, and the intensity of the product signals decreased with time as the carbonaceous material in the sample was depleted. The release and subsequent reactions of sulphur-containing species from coal during devolatilisation and char gasification occur by distinct but related mechanisms under CO2 and steam conditions. The sulphide and aliphatic sulphur-containing functional groups in the coal structure decompose readily during the initial devolatilisation phase [8,11,20], and a portion thereof react with coal-hydrogen and coal-oxygen to produce H2S, SO2 and COS. Some of the aliphatic sulphur is also reincorporated into the char matrix as more stable heterocyclic structures [21]. As devolatilisation ceases and the slower char gasification reactions begin to dominate the further release of gaseous species, strongly bound organic sulphur remaining in the char matrix as aromatic and heterocyclic species is decomposed and released, predominantly as COS, with some SO2. The conditions under which the gasification is occurring, H2O- or CO2-rich, hereafter dictate the subsequent mechanistic steps and the final relative amounts and timing of the released S-containing species during the gasification phase. Due to the rapid evolution of gaseous species from the coal during devolatilisation, which results in a local dilution effect around the particles, S-species decomposition and reactions occurring during the devolatilisation phase were observed to be essentially independent of the reaction atmosphere. This can be seen by comparing, for example, the intensity and timing of the initial H2S shoulder under steam gasification conditions, which closely matches that of the H2S peak under CO2 gasification (Fig. 3). The H2S–COS equilibrium (Eq. (6)) is governed during devolatilisation by the partial pressure of H2O (pH2O) within the particle and near the particle surface from steam generated by residual moisture in the coal.
H2 SðgÞ þ CO2ðgÞ $ COSðgÞ þ H2 OðgÞ
ð6Þ
As char gasification proceeds under CO2 gasification conditions, sulphur released as COS from the solid matrix undergoes no further reaction and was detected after a short delay. Under H2O gasification conditions, however, pH2O near the particle surface is sufficiently high during the gasification phase to influence the H2S–COS equilibrium. This equilibrium shift results in the parallel, delayed release of H2S and CO2 and significant reduction in release of COS under steam gasification conditions as shown in Fig. 3. Some of the sulphur species released during devolatilisation are also captured by inherent CaO, releasing additional H2O and CO2:
CaOðsÞ þ COSðgÞ $ CaSðsÞ þ CO2ðgÞ
ð7Þ
CaOðsÞ þ H2 SðgÞ $ CaSðsÞ þ H2 OðgÞ
ð8Þ
FactSage equilibrium calculations indicate that the forward reaction with inherent CaO is thermodynamically favourable, therefore CaS will be formed preferentially and stabilised by the initially low local pH2O during the devolatilisation phase under both atmospheres. The products of these two reactions are visible as a small initial peak in the 46CO+2 spectra under steam gasification conditions, and as part of the initial 17OH+ spectra under CO2 conditions. Sulphur captured as CaS during devolatilisation reacts via different pathways under different gasification conditions. During steam gasification, pH2O surrounding the particle increases gradually towards the end of the devolatilisation phase and the equilibrium shown in Eq. (8) is shifted to the left, contributing to the observed secondary release of H2S. Under CO2-rich conditions, the mechanism proceeds via a two-step process to release CO and SO2:
J. Tanner et al. / Fuel 158 (2015) 72–80
CaSðsÞ þ 4CO2ðgÞ $ CaSO4ðsÞ þ 4COðgÞ
ð9Þ
CaSO4ðsÞ ! CaOðsÞ þ SO2ðgÞ þ 1=2O2ðgÞ
ð10Þ
The CO released as a result of Eq. (9) under CO2-rich conditions is visible as a shoulder on the right of the 30CO+ peak, slightly preceding and overlapping the delayed SO2 release. The moderate, slightly delayed release of SO2 observed under steam conditions is attributed to the decomposition of inherent sulphates and char-S compounds during the gasification phase. This also occurs but is masked in the case of CO2 gasification by the SO2 from sulphate decomposition. Sodium may compete with calcium in coals having substoichiometric amounts of sulphur. Reaction (11) is enhanced by the presence of steam, which increases the concentration of NaOH in the gas phase. Thermodynamic calculations indicate that Eq. (11) proceeds to completion above 600 °C. Eq. (12) is temperature dependent within the range of this investigation. At 1100 °C, Na2S is partially converted to Na2SO4. At 1400 °C, Na2S is also only partially converted, however, the equilibrium products are as per Eq. (12) at this temperature. Therefore, under steam gasification conditions, coals with higher Na to Ca ratio will not only undergo increased competition for the capture of S, but result in a reduced release of Na and S species overall, particularly at lower temperatures, due to the formation of the somewhat stable sulphides.
CaSðsÞ þ 2NaOHðgÞ ! CaOðsÞ þ Na2 SðsÞ þ H2 OðgÞ
ð11Þ
Na2 SðsÞ þ H2 OðgÞ $ 2NaOðgÞ þ H2 SðgÞ
ð12Þ
The proposed mechanism for sulphur release is somewhat cyclic in nature, particularly with regards to the capture by Ca and Na, resulting in the observed prolonged release of SO2 under CO2 condition and H2S under steam gasification conditions. Ca, in particular, is continually converted between oxide, sulphide and sulphate forms to facilitate the release of various sulphur species. Once all the carbonaceous material has been gasified, the Ca remains in the residual ash as CaO according to Eqs. (7), (9) and (10) in the CO2-rich case and Eqs. (8) and (11) in the H2O-rich case. Similarly, there will also be a portion of the Na remaining in the ash as Na2S under steam gasification conditions according to Eq. (12). The difference in intensity-time spectra for the release of 64SO+2 under CO2- and H2O-rich atmospheres at various temperatures is shown in Fig. 5. In general, the relative amount of 64SO+2 increased with increasing temperature and was higher under CO2 conditions. These observations support the proposed mechanisms of release of SO2 under CO2-rich and H2O-rich conditions outlined above, both of which begin with a small amount of sulphur released as SO2
Fig. 5. Averaged intensity-time spectra for scale is constant for all sub-plots.
64
77
during the devolatilisation phase, and result in some volatilised sulphur intermediates being captured as metal sulphides. Under CO2 gasification conditions, captured sulphide sulphur is predominantly released as SO2 via the metal sulphate during the gasification phase (Eqs. (9) and (10)). This capture and release mechanism is temperature dependent within the range of this investigation. At 1100 °C, capture rates are low and the small amount of sulphate which is formed is stable, therefore the small release of SO2 at low temperatures is attributed almost solely to devolatilisation products, and the metal remains in sulphate form in the residual ash. At 1400 °C, the rates of sulphur capture and sulphate decomposition increase such that the two step sulphate mechanism proceeds more rapidly, and a significant delayed release of SO2 is observed. Additionally, equilibrium calculations indicate that while CaSO4 will completely decompose at 1400 °C, Na2SO4 is more stable and a significant portion will remain. This indicates that coals which capture sulphur as sodium sulphate or which contain native Na2SO4 will permanently capture both sodium and sulphur under CO2 gasification conditions at these temperatures. Under steam gasification conditions, sulphide sulphur reacts either directly with steam, or with NaOH. In both cases, the dominant sulphur-containing product is H2S, hence the relatively low release of SO2 in comparison with equivalent-temperature CO2 conditions. Another feature of the SO2 released under H2O-rich conditions in Fig. 5 is the secondary peak that occurred at 1100 °C, but was minimal at higher temperatures and did not occur under CO2-rich conditions. At temperatures just below 1100 °C, CaSO4 is relatively stable. Considering the endothermic nature of the char gasification reactions, and the higher reaction rate for steam gasification in comparison to CO2 gasification, it is possible that the temperature of the particle under H2O-rich conditions was slightly decreased during gasification of the carbonaceous material. Once the carbonaceous material was exhausted and the concentration of H2O returned to the baseline value, residual sulphates in the solid matrix were decomposed, releasing SO2 as observed. At higher temperatures, the rate of gasification increases, CaSO4 decomposes readily, and any slight delay was therefore masked by an overlap of the secondary SO2 release and the initial release from devolatilisation. 3.2. Semi-quantitative analysis Fig. 6 shows the relative amounts of SO2 released during the devolatilisation and gasification periods for each coal under H2O-rich conditions at various temperatures. In general, increasing initial release and decreasing secondary release with increasing
SO+2 released from HKN-S coal at 1100 °C, 1200 °C and 1400 °C under 20% CO2 (left) and 20% steam (right) in helium. Y-axis
78
J. Tanner et al. / Fuel 158 (2015) 72–80
Fig. 6. Averaged, normalised peak area of primary and secondary SO2 released from Victorian and Rhenish coals at 1100 °C, 1200 °C and 1400 °C under 20% steam gasification conditions.
temperature were observed. As temperature increases, functional group decomposition and volatilisation increases, resulting in the increased initial release. This depletes the finite supply of sulphur, resulting in the observed decreasing trend in secondary release at higher temperatures. Furthermore, the total release of sulphur as SO2 under H2O-rich conditions also generally decreased with increasing temperature due to increased permanent capture, reaction with NaOH, and release as H2S instead of SO2 under H2O-rich conditions. As discussed above, the observed trends in the release of SO2 with temperature, and the partitioning between the devolatilisation and gasification phases may be largely attributed to the relative stability and decomposition of sulphates under various conditions. These results further indicate that it may therefore be possible to control the release of SO2 under certain gasification conditions, either by controlling the reaction temperature or steam concentration in a continuous system, thereby avoiding the final release of SO2 observed once the steam concentration returns to baseline values in Fig. 5. However, further investigation is required to determine the influence of configuration, continuous operation and multi-reagent systems on the entrained flow gasification process. The averaged, normalised peak areas for selected inorganic species released at 1400 °C are presented in Fig. 7. The data presented have been scaled as noted in the legend to account for large differences in the orders of magnitude and for the variation in instrument sensitivity between different species. Therefore, while the relative release of each species under different conditions is directly comparable, the absolute values for different species are not.
Analogous experiments were also performed at 1100 °C and 1200 °C. The direction of the equilibrium reactions discussed above is generally opposite for CO2- and H2O-rich environments; therefore, the temperature dependence and release trends with temperature were contrary. Overall, the release of inorganic species at 1400 °C was higher under CO2 conditions, with the exception of 23 Na+ and 34H2S+. These trends in the semi-quantitative data reflect the proposed release mechanisms described above. The results are somewhat contrary to previous investigations which showed a significant decrease in the release of alkali metals and an increase in the release of alkali metal hydroxides under low-O2 gasification conditions in the presence of steam [13]. It has previously been shown that Na not associated with Si is present as either NaCl or carboxylate Na available for mobilisation, volatilisation or reaction [22]. The 23Na+ signal is therefore attributed directly to elemental sodium or to ionised fragments of Na-containing species released from the coal. In general, the release of 23Na+ was negligible under CO2-gasification conditions due to the higher rates of conversion to carbonate form and subsequent capture by inherent silicates and aluminosilicates [23,24] in the CO-rich environment created by the Boudouard reaction (Eq. (1)). Under steam gasification conditions, the trend in 23Na+ followed the concentration of inherent sodium in the parent coals. The observation of Na release under steam conditions only is due to a combination of the increased formation and subsequent fragmentation of NaOH, the competitive nature of the mechanism of S-capture and release shown in Eqs. (11) and (12), and the lower rate of capture of sodium as silicates and aluminosilicates. It has also been shown [25] that the presence of steam lowers the melting point of Na2CO3, leading to increased release of metallic Na. This effect may also explain the absence of 23Na+ release observed under CO2 conditions due to increased permanent capture of Na by silicates, according to:
2NaOHðgÞ þ COðgÞ $ Na2 CO3ðlÞ þ H2ðgÞ
ð13Þ
Na2 CO3ðlÞ þ xSiO2ðsÞ $ Na2 O xSiO2ðsÞ þ CO2ðgÞ
ð14Þ
36
+
35
+
The release of HCl and Cl followed the trend of inherent Cl in the parent coals for both CO2 and H2O conditions. The total release for each Cl species was lower under H2O-rich conditions, within the experimental variance. The majority of Cl and HCl were released during the initial devolatilisation phase, after which the HCl is involved in gas-phase equilibrium with NaCl via NaOH (Eq. (15)).
NaOHðgÞ þ HClðgÞ $ NaClðgÞ þ H2 OðgÞ
ð15Þ
Previous work under combustion and gasification conditions has shown that the capture of Na by aluminosilicates affects the release of Na and Cl species [12,13]. During devolatilisation under
Fig. 7. Averaged, normalised peak areas for inorganic species released from Victorian and Rhenish coals at 1400 °C under 20% CO2 (left) and 20% steam (right) in helium.
J. Tanner et al. / Fuel 158 (2015) 72–80
steam gasification conditions, the NaOH–NaCl equilibrium favours the hydroxide form. However, as gasification proceeds and NaCl is captured as silicates and aluminosilicates by Eqs. (16) and (17) [26], the equilibrium shifts to the right. NaOH and NaCl are consumed, resulting in the observed decrease in 39NaO+ and 58NaCl+ release under H2O-rich conditions.
2NaCl þ xAlSiy Oz þ H2 O ! Na-aluminosilicate þ 2HCl
ð16Þ
2NaCl þ xSiO2 þ H2 O ! Na-silicate þ 2HCl
ð17Þ
The release trends for the sulphur-containing species H2S and SO2 do not follow the inherent S content of the parent coals due to the variety and different reaction pathways of the various forms of sulphur present. According to the proposed mechanisms, coals relatively low in sulphides and aliphatic S-containing functional groups should exhibit lower release of initial H2S, COS and SO2 during the devolatilisation phase. Those with lower aromatic and heterocyclic S should exhibit lower release of H2S and COS during the gasification phase under H2O- and CO2-rich conditions, respectively. This summary is, however, complicated by the nature of the mechanism. The cyclic release, reincorporation, capture and competition steps which sulphur species undergo during the devolatilisation phase are interrelated, and the subsequent steps in the gasification mechanisms for each atmosphere include several equilibrium steps, which are both temperature and atmosphere dependent. It is therefore not possible to determine the ratio of the inherent forms of sulphur in a coal sample by these MBMS measurements alone. H2S release was generally shown to be higher under H2O-rich conditions, with the exception of LY. This is predominantly due to the artificially high Si content in the LY sample leading to increased capture of Na and hence reduced NaOH and H2S release. The formation of Na2S is therefore affected, influencing the release of H2S according to Eqs. (11) and (12). LY also has low Ca content, minimising the potential for CaS formation and reaction with H2O to release H2S via Eq. (8). The opposite effect is evident for HKN-S+, where low inherent Al + Si, high Ca and high S resulted in proportionally higher release of H2S. For HKS, the release of 39NaO+ was higher than expected due to low Al and Si in the parent coal resulting in low capture of Na. HKS also has the highest native Na content, indicating that NaOH will be higher, and therefore a greater proportion of S converted to H2S via Eqs. (11) and (12), hence the proportionally lower release of SO2. These results are in general agreement with trends observed in previous work under combustion and low-O2 gasification conditions in which it was shown that H2S release is in high negative correlation with Ca/S ratio under gasification and combustion conditions [10]. Previous studies [27] have also shown that the amount of sulphur fixed in ash was a function of the Ca/S ratio and independent of the form of sulphur. In the case of the six coals investigated here, those with higher Ca/S generally released less S-containing species than coals with lower Ca/S, confirming the previous group’s results. HKS and HKN-S have low total S content and relatively high Na and Ca, increasing the chance of capture as CaS in the absence of excess CO2, as reflected by the lower release of SO2 under H2O-rich conditions. 4. Conclusion The release and reaction mechanisms of volatile inorganic species from Victorian brown coals and Rhenish lignites at 1100 °C, 1200 °C and 1400 °C under 20% CO2 and 20% H2O in helium were investigated. In particular, the influence of temperature and the role of the gasification reagents CO2 and H2O in the release mechanisms themselves were studied, and the results compared with previous
79
investigations under similar conditions. The release and subsequent reactions of volatile species from coal during devolatilisation and char gasification were shown to occur by distinct but related mechanisms under CO2 and H2O conditions. Two overlapping phases were clearly observed from the results – devolatilisation and char gasification. The Na, Cl and S species released and subsequent reactions which occurred during devolatilisation were dependent primarily on the particle temperature and rate of char gasification, and were essentially independent of the reaction atmosphere itself. During the gasification phase, however, the release mechanisms for the Na, Cl and S species directly involved CO2 and H2O, and the absence, presence and concentration of the gasification reagent affected the release trends, intensity and timing. It is predominantly the equilibrium reactions in the proposed mechanism that were influenced by CO2 and H2O. In most cases, the equilibrium reactions exhibited opposite shifts for CO2 and H2O-rich environments; therefore, the temperature dependence and release trends with temperature were contrary. Overall, the release of inorganic species was higher under CO2 conditions. However, the release of gaseous Na was negligible under CO2-gasification conditions due to enhanced capture as aluminosilicates. Under steam gasification conditions, the trend in Na release followed the concentration of inherent sodium in the parent coals. The release of Cl species followed the trend of inherent Cl in the parent coals for both CO2 and H2O conditions. The release trends for the sulphur-containing species H2S and SO2 did not follow the inherent S content of the parent coals and H2S release was higher under H2O-rich conditions due to the dependence of sulphur capture on the Ca/S ratio and CaS–CaO equilibrium. It was shown that the inherent forms of sulphur in the parent coals affect the released sulphur-containing species, however, it was not possible to determine the ratio of the inherent forms of sulphur in a coal sample by these MBMS measurements alone. The proposed mechanisms also involve the competitive capture of S by Na and Ca. Ca acts as a catalyst of sorts, continually cycling between oxide, sulphide and sulphate forms to facilitate the release of various sulphur species. This mechanism may also result in the permanent capture of Na, Ca and S in the solid residue remaining after gasification, depending on the presence and concentration of CO2 and H2O and the composition of the parent coal. It may therefore be possible to partially control the release of some volatile species by controlling either the reaction temperature or steam concentration. Acknowledgements The authors would like to acknowledge Brown Coal Innovation Australia (BCIA), Bundesministerium für Wirtschaft und Energie (FKZ 0327773) and the Go8-DAAD Joint Research Scheme for providing the funding for this collaborative research project. References [1] OECD/IEA. World energy outlook. France: International Energy Agency; 2011. [2] Huffman GP, Huggins FE. Reactions and transformations of coal mineral matter at elevated temperatures. In: Mineral matter and ash in coal. American Chemical Society; 1986. p. 100–13. [3] Gottwald U, Monkhouse P, Wulgaris N, Bonn B. In-situ study of the effect of operating conditions and additives on alkali emissions in fluidised bed combustion. Fuel Process Technol 2002;75:215–26. [4] Monkhouse P. On-line spectroscopic and spectrometric methods for the determination of metal species in industrial processes. Prog Energy Combust Sci 2011;37:125–71. [5] Bläsing M, Müller M. Release of alkali metal, sulfur, and chlorine species during high-temperature gasification of coal and coal blends in a drop tube reactor. Energy Fuels 2012;26:6311–5. [6] Bläsing M, Müller M. Release of alkali metal, sulphur, and chlorine species from high temperature gasification of high- and low-rank coals. Fuel Process Technol 2013;106:289–94.
80
J. Tanner et al. / Fuel 158 (2015) 72–80
[7] van Eyk PJ, Ashman PJ, Alwahabi ZT, Nathan GJ. The release of water-bound and organic sodium from Loy Yang coal during the combustion of single particles in a flat flame. Combust Flame 2011;158:1181–92. [8] Yan J, Yang J, Liu Z. SH radical: the key intermediate in sulfur transformation during thermal processing of coal. Environ Sci Technol 2005;39:5043–51. [9] Wolf KJ, Müller M, Hilpert K, Singheiser L. Alkali sorption in second-generation pressurized fluidized-bed combustion. Energy Fuels 2004;18:1841–50. [10] Bläsing M, Melchior T, Müller M. Influence of temperature on the release of inorganic species during high temperature gasification of Rhenish lignite. Fuel Process Technol 2011;92:511–6. [11] 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. [12] Bläsing M, Müller M. Influence of pressure on the release of inorganic species during high temperature gasification of coal. Fuel 2011;90:2326–33. [13] Bläsing M, Müller M. Investigations on the influence of steam on the release of sodium, potassium, chlorine, and sulphur species during high temperature gasification of coal. Fuel 2012;94:137–43. [14] Tanner J, Bläsing M, Müller M, Bhattacharya S. Influence of temperature on the release of inorganic species from Victorian brown coals and German lignites under CO2 gasification conditions. Energy Fuels 2014;28:6289–98. [15] Hayashi J-I, Li C-Z. Structure and properties of Victorian brown coal. In: Li C-Z, editor. Advances in the science of Victorian brown coal. Elsevier; 2004. [16] Sakaguchi M, Laursen K, Nakagawa H, Miura K. Hydrothermal upgrading of Loy Yang brown coal – effect of upgrading conditions on the characteristics of the products. Fuel Process Technol 2008;89:391–6.
[17] Drowart J, Goldfinger P. Investigation of inorganic systems at high temperature by mass spectrometry. Angew Chem Int Ed Engl 1967;6:581–96. [18] Wolf KJ, Untersuchungen zur Freisetzung und Einbindung von Alkalien bei der reduzierenden Druckwirbelschichtverbrennung (Investigations of the release and sorption of alkali metals in pressurized fluidized bed combustion under reducing conditions), PhD Thesis: RWTH Aachen; 2003 [In German]. [19] Walker Jr PL, Rusinko Jr F, Austin LG. Gas reactions of carbon. In: Eley DD, Selwood PW, Weisz PB, editors. Advances in catalysts, 11. New York: Academic Press; 1959. p. 133–221. [20] Khan MR. Prediction of sulphur distribution in products during low temperature coal pyrolysis and gasification. Fuel 1989;68:1439–49. [21] Calkins WH. The chemical forms of sulfur in coal: a review. Fuel 1994;73:475–84. [22] Brockway DJ, Ottrey AL, Higgins RS. Inorganic constituents. In: Durie RA, editor. The science of Victorian brown coal. Butterworth-Heinemann; 1991. p. 597–650 [chapter 11]. [23] Kosminski A, Ross DP, Agnew JB. Transformations of sodium during gasification of low-rank coal. Fuel Process Technol 2006;87:943–52. [24] Kosminski A, Ross DP, Agnew JB. Reactions between sodium and kaolin during gasification of a low-rank coal. Fuel Process Technol 2006;87:1051–62. [25] Byung Ho S, Sang Done K. Catalytic activity of alkali and iron salt mixtures for steam-char gasification. Fuel 1993;72:797–803. [26] Wei X, Huang J, Liu T, Fang Y, Wang Y. Transformation of alkali metals during pyrolysis and gasification of a lignite. Energy Fuels 2008;22:1840–4. [27] Schafer HNS. Functional groups and ion exchange properties. In: Durie RA, editor. The science of Victorian brown coal. Butterworth-Heinemann; 1991. p. 323–57 [chapter 7].