Fuel 82 (2003) 1085–1090 www.fuelfirst.com Biodesulphurization as a complement to the physical cleaning of coalq Olegario Martı́neza, Carlos Dı́eza, Nick Milesb, Chandu Shahb, Antonio Morána,* a Department of Chemical Engineering, Natural Resources Institute, University of León, Avenida de Portugal 41, León 24071, Spain Department of Chemical, Environmental and Mining Engineering, The University of Nottingham, University Park, Nottingham NG7 2RD, UK b Received 12 June 2002; accepted 19 December 2002; available online 6 February 2003 Abstract Physical and biological processes have been combined with a view to reduce the sulphur and ash content of finely ground coal. The coal used was a semianthracite from the North Spain Coal Field. A sample of several kg in a coal/water suspension with a w/w concentration of 15% was subjected to a physical treatment combining cyclone and flotation separation processes. Representative samples were taken from each of five size fractions: original feed and the physical separation products, which were screened through water into sub-samples classified by particle size. Each fraction from the physical separation and the screening was analysed for ash and sulphur. The elimination yield of the hydrocyclone was 22% of ash and 21% of sulphur. The figure for ash raised to 41% when the hydrocyclone and flotation equipment were both used, with no change in sulphur elimination. In turn, the biodesulphurization treatment applied in addition to the two processes raised the respective yields to 59 and 42%. The change in the calorific value of the coal was hardly significant, while sulphur emissions, expressed as g S/GJ were reduced by 51%. Additionally, an assessment was made of the thermal behaviour of the various samples from the physical and biological treatment by means of programmed temperature combustion analysis carried out by a thermogravimetric equipment. q 2003 Elsevier Science Ltd. All rights reserved. Keywords: Biodesulphurization; Combustion; Thermogravimetry 1. Introduction Today coal accounts for over 25% of the world’s primary energy consumption (approximately 2200 million tonnes petroleum equivalent [Mtpe] in 1998) and 15% of that of the European Union (about 214 Mtpe. for the same year), while the figure is over 75% in countries such as China and South Africa. The share of coal in the world’s energy balance is forecast to increase by about 30% over today’s figures by 2005. Coal will, therefore, play an important part in the short and medium term, given that the proven reserves make up 80% of those of fossil fuels, with sufficient available for 250 years, or 800 years if we take into account the theoretical reserves in places such as South Africa, Brazil and Eastern Europe. The demand for coal in Spain has remained constant at around 19 Mtpe over the last decade, despite a reduction in industrial and domestic use of over 50% over the same period. In the present decade, although energy use will * Corresponding author. Tel.: þ 34-987-291841; fax: þ 34-987-291839. E-mail address: dfqamp@unileon.es (A. Morán). q Published first on the web via Fuelfirst.com—http://www.fuelfirst.com increase by 15%, a decrease of about 40% in the use of coal as an energy source is expected, in the face of a sharp increase in the use of renewable energy sources and, above all, natural gas [1]. Coal burning in Spain’s major power stations accounts for 90% of consumption and makes up 30% of electricity generation, this use of coal being the cause of nearly 70% of sulphur dioxide emissions into the atmosphere. This gas is the main cause of acid rain and so legislation is continually reducing the permitted levels of emission. The European Directive for Large Combustion Plants (88/609/EEC) proposes a 37% reduction of sulphur dioxide emissions for Spain by 2003 with regard to 1980 levels. Emission reduction can be tackled by means of physical, chemical and biological processes, which can be applied before or after the coal is burnt. The techniques used in the present study, based on a combination of physical and biological processes, are for implementing during the preparation of the coal, before it is burnt. The prime aim of this study is to assess the improvement brought about by the biological treatment of the samples in processes of physical separation with regard to the reduction of ash and sulphur content. We also seek to assess 0016-2361/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0016-2361(03)00014-0 1086 O. Martı́nez et al. / Fuel 82 (2003) 1085–1090 the possible changes that these physical and biological processes bring about in the behaviour of the coal during combustion. 2. Experimental The coal used came from the washing plant of Hullera Vasco-Leonesa S.A. It is a semianthracite with a relatively high sulphur content (2.4%). The first treatment was in a hydrocyclone. The treatment of finely ground coal in hydrocyclones offers better separation possibilities than the use of densemedium cyclones, as these pose problems of separation of products from the medium, with consequently higher separation costs [2]. Separation occurs as a result of the centrifugal force to which the coal pulp is subjected in the hydrocyclone, whereby the separation of the heavier particles from the lighter ones is brought about by the former creating of a descending vortex, while the latter form an ascending one. The hydrocyclone used was 7 cm long by 5 cm in diameter [2]. The hydrocyclone light fraction exit stream was then subjected to bubble flotation, giving two exit streams from the flotation equipment. During the process, samples of coal pulp are taken simultaneously from all streams except tailings. The study of each of these streams was carried out by means of fractioning and screening through water into sub-samples of different particle sizes (M : . 125 mm; N : 125 . t (mm) . 63; P : 63 . t (mm) . 38; R : 38 . t (mm) . 20; S : , 20 mm), which were then dried and analysed. Both the initial coal and the four exit streams—from the hydrocyclone and the flotation unit (Fig. 1)—were subjected to biodesulphurization treatment in shaken bioreactors. Treatment, which lasted 10 days, was carried out by inoculating the coal with a culture rich in desulphurizing micro-organisms that had previously been adapted to this type of coal and kept at pH ¼ 20 and 30 8C [3 – 5]. The thermogravimetric analysis, carried out to determine the behaviour in combustion of the coal samples that have been treated, consisted of burning a small amount, about 25 mg, in a constant air stream in a thermobalance, with a heating gradient of 50 8C/min [6,8]. During the operation, the thermogravimetric equipment registered mass loss and rates of mass loss when the heating rate was regulated in the combustion stage until it became quasi-isothermic, thereby avoiding ignition of the sample and allowing for a better resolution of the peak representing the loss rate in the curve corresponding to derived thermogravimetry (high-resolution thermogravimetry). 3. Results The results obtained are expressed in three clearly differentiated groups: those concerning the yield obtained in physical separation, those of the biological process, and the behaviour of the coal in combustion. 3.1. Physical separation yields The results of analyses of the various samples involved in physical separation are shown in Table 1. Results are overall for the samples, both on being fed into the hydrocyclone and for the products of separation and the floating particles from the process of separation by flotation. The analyses of each of the screened sub-samples are also given. In the hydrocyclone, the separation yield of the light fraction is nearly 80% and is greater for the smaller particle sizes, as may be seen in Fig. 2, with a concentration of the heavy fraction exiting from the hydrocyclone in larger particles sizes. Likewise, the reduction in ash content in the light fraction was greater in the fractions of larger particles sizes (Fig. 2). Overall sulphur reduction was 21%, rising to 39% in the fraction with the coarsest particle size. Furthermore, the heavy fraction, a little over 20% of the hydrocyclone feed, corresponded to low-quality coal, with a greater ash and sulphur content, especially in the fractions of finer particle size, although such fractions were small. Considering that for the flotation unit, the yield of the light fraction (floating particles) is nearly 90%, and although Fig. 1. Diagram of the physical and biological separation equipment used. O. Martı́nez et al. / Fuel 82 (2003) 1085–1090 1087 Table 1 Results of physical separation for various particle sizes Overall Ma Na Pa Ra Sa Feed Coal mass (kg) Ash (%d.m.b) Sulphur (%d.m.b) 10.00 24.1 2.40 1.06 25.0 2.31 2.60 18.1 2.14 1.80 19.4 2.26 1.28 20.0 2.49 3.26 32.7 2.67 Light fraction Coal mass (kg) Ash (%d.m.b) Sulphur (%d.m.b) 7.67 18.9 1.89 0.23 5.4 1.41 1.80 7.7 1.63 1.49 11.0 1.78 1.09 14.9 1.79 3.07 31.7 2.17 Heavy fraction Coal mass (kg) Ash (%d.m.b) Sulphur (%d.m.b) 2.33 38.6 3.20 0.83 28.6 2.39 0.80 39.2 2.72 0.31 48.7 3.15 0.19 52.3 5.32 0.19 47.0 7.50 Floating particles Coal mass (kg) Ash (%d.m.b) Sulphur (%d.m.b) 6.87 14.2 1.90 0.21 5.2 1.39 1.62 7.3 1.63 1.43 9.3 1.77 1.05 11.4 1.87 2.55 23.2 2.37 a M : .125 mm; N : 125 . t (mm) . 63; P : 63 . t (mm) . 38; R : 38 . t (mm) . 20; S : ,20 mm. b d.m.: dry matter. Fig. 3. (K) Yields for the recuperation of coal, (S) the reduction of ash content and (A) the reduction of sulphur in flotation equipment. Particles sizes: M : .125 mm; N : 125 . s (mm) . 63; P : 125 . s (mm) . 63; R : 38 . s (mm) . 20; S : ,20 mm. tendency as for the hydrocyclone used alone, as this makes a greater contribution in the overall process. For its part, the total reduction of sulphur content is 21%, the same as for the hydrocyclone alone, affecting basically fractions of greater particle size. It may therefore be said that the hydrocyclone has a great effect on the reduction yields of ash and sulphur. it is slightly higher for intermediate particle sizes, the degree of separation is very similar regardless of particle size (Fig. 3). As for the reduction in ash content, the same figure shows a greater decrease for the smaller particle sizes, which are the most affected. The reduction in sulphur content is nil in the flotation unit. Even in the lightest fractions, it seems that the sulphur content increases. The reason for this must lie in the surface properties of pyrites, whereby it tends to float with the coal, especially with fractions of finer size. Fig. 4 shows the global separation yields for the hydrocyclone and flotation equipment used together. Coal recuperation is almost 70%, this yield increasing in the case of smaller particle sizes, basically owing to the cyclonic separation. The reduction in ash content follows the same The results for the biodesulphurization of coal in stirred bioreactors are shown in Table 2, where ash content is observed to reduce between 22 and 30%, owing to the acid medium used in the process [3,5]. At the same time, a reduction occurs in the sulphur content, this reduction being between 16 and 27%, with higher yields from cleaner coal samples (light fraction and floating matter). The table also shows the values of sulphur emissions, expressed in g S/GJ, bearing in mind that all the sulphur will be released into the atmosphere as sulphur dioxide during combustion. Given that the biodesulphurization process dissolves mineral matter, the coal becomes more concentrated, which Fig. 2. (K) Yields for the recuperation of coal, (W) the reduction of ash content and (A) the reduction of sulphur in hydrocyclone. Particles sizes: M : .125 mm; N : 125 . s (mm) . 63; P : 125 . s (mm) . 63; R : 38 . s (mm) . 20; S : ,20 mm. Fig. 4. (K) Yields for the recuperation of coal, (S) the reduction of ash content and (A) the reduction of sulphur in hydrocyclone and flotation equipment together. Particles sizes: M : . 125 mm; N : 125 . s (mm) . 63; P : 125 . s (mm) . 63; R : 38 . s (mm) . 20; S : ,20 mm. 3.2. Yields of the biological process 1088 O. Martı́nez et al. / Fuel 82 (2003) 1085–1090 Table 2 Reduction in ash and sulphur content and in sulphur emissions in the biological process Feed Treated feed Yield(%) Light fraction Treated light fraction Yield (%) Floating matter Treated floating matter Yield (%) a %Ash (d.s.a) %S (d.s.a) PCS (MJ/kg, d.s.a) Emission (g S/GJ) 24.1 18.4 23.6 18.9 14.7 22.2 14.2 9.9 30.2 2.40 2.02 15.7 1.89 1.56 17.5 1.90 1.40 26.7 25.76 28.05 – 27.60 28.46 – 29.90 30.70 – 932 720 22.7 685 548 20.0 635 456 28.2 Dry sample. results in a higher calorific value [7], expressed as a function of dry matter. When expressed as a function of dry matter free of mineral matter, however, it is not greatly affected by the biological process [6,8]. Fig. 5 shows the results of the comparative yields of the physical and biological processes, together with the improvement brought about by the latter. Biological treatment of the feed leads to a reduction of 23% in sulphur emission. When the light fraction from the hydrocyclone was treated biologically, the ash and sulphur contents and sulphur emissions were reduced, respectively, by 39, 35 and 41%, great improvements on the reductions of the physical process applied alone. Regarding the fraction from the flotation unit, the biological process enhanced the reduction of sulphur content and emissions. When the physical and biological process were both used, the yields reached 59% for ash content reduction, 42% for sulphur content reduction and 51% for the decrease in sulphur emissions. temperature of the volatile matter and the ignition of the char [9]. Furthermore, this technique, in combination with mass spectrometry (TG – MS) has made it possible to identify both the intermediaries of combustion and its end products, together with their relationship to the coal rank [12 –14]. The change in the combustibility of the different coal samples was determined by comparing their combustion profiles obtained by the thermobalance. Fig. 6a shows the enhanced combustibility of the biodesulphurized coal in comparison with the original coal. This improvement consists in a drop in the coal’s initial combustion temperature, Tv ; due to a decrease in the length of the oxygen chemisorption stage, during which the sample slightly increased its mass. Consequently, the temperature of highest mass loss rate, Tmax ; is reached earlier, as is the final temperature of combustion, Tf ; i.e. the coal burns at a lower temperature. Biological treatment brings about minor structural changes in the coal, especially on the surface, linked with the reduction in the content of mineral matter and pyritic sulphur, which may favour the chemisorption of oxygen and the onset of the release of volatile matter at a lower temperature. Primary devolatization, therefore, occurs between 400 and 500 8C with the release of tar volatiles, a process which begins in the biodesulphurized sample at 380 8C. On the other hand, in biodesulphurized coal, the drop in mineral matter content 3.3. Combustibility of the coal Thermogravimetry (TG) has often been used in recent years, especially programmed-temperature combustion curves, for the study of coal combustibility and its dependence on the coal’s rank, mineral content and maceral composition [9 – 11]. Combustion is in two simultaneous stages, devolatization and oxidation of the char, characterized, respectively, by the ignition Fig. 5. Reduction yields for ash content and emissions for the physical and biological processes used together. Fig. 6. Variation of the combustibility of the coal samples. Changes produced in the physical and biological treatment compared with the original samples. O. Martı́nez et al. / Fuel 82 (2003) 1085–1090 reduces its catalytic effect, thus slowing down combustion. Furthermore, in the treated sample, a secondary peak is seen to disappear at approximately 750 8C, corresponding to the decomposition of the carbonates in the mineral matter, present in the original sample and absent in the treated sample owing to the acid medium used during biodesulphurization. The presence or absence of carbonates was determined by chemical analysis and the comparison of devolatization profiles of original and treated coal [6,8]. As for the change in the hydrocyclone, Fig. 6b shows this to be negligible, except for the fact that the reduction in ash content occurring in this physical process affects the carbonate content of the sample obtained and therefore the secondary peak in the combustion profile. For this samplepart of the light fraction-treated in the biodesulphurization process, as in the previous case, the initial combustion temperature, Tv ; is reached earlier, probably owing to the increase in oxygen content [8], which forces down the sample’s capacity for the chemisorption of oxygen. The behaviour of this biodeulphurized sample is similar to that of the first one treated. The sample of floating matter obtained in the flotation unit was compared with the biodesulphurized sample from the same source and with the feedstock of the flotation equipment, i.e. the light fraction from the hydrocyclone. The results concerning the change in combustibility of the samples are shown in Fig. 7a, where the feed sample and 1089 the floated sample are seen to behave very similarly, fundamentally owing to the latter making up 90% of the former. As regards the biodesulphurized coal, as in the previous cases, the three characteristic temperatures of the combustion profile, Tv ; Tmax and Tf ; are reached earlier, although to a lesser extent than for the above showed samples. In short, the reduction of the ash content due to both the physical and biological processes, and especially to the surface change brought about by diodesulphurization in an acid medium, noticeably improves the non-isothermic reactivity (combustibility) of the coal [8]. On the other hand, the sample corresponding to the heavy fraction from the hydrocyclone undergoes a drop in combustibility in comparison with the feed sample, shown by the combustion profiles in Fig. 7b. In the case of this sample, initial combustion temperature is reached later than for the original sample, as is the temperature of maximum weight loss rate, Tmax ; owing to the higher ash content. Additionally and consequently, a peak occurs at high temperatures corresponding to the decomposition of the carbonates in the mineral matter, which is greater than in the original sample, owing to the concentration of the mineral matter taking place in the heavy phase from the hydrocyclone. Besides, the desulphurized sample displays a behaviour in line with its enhanced combustibility seen in the other desulphurized fractions, borne out by the advanced reaching of the characteristic temperatures Tv ; Tmax and Tf ; as well as by the disappearance of the secondary peak associated with the decomposition of the carbonates. In this case, however, the value of the highest mass loss rate, DTGmax ; is lower, as, despite biodesulphurization lowering the ash content, it is still higher than in the original sample fed into the hydrocyclone. 4. Conclusions The following conclusions may be drawn from the experiments and the foregoing discussion of the results: Fig. 7. Variation of the combustibility of floating and heavy fraction samples. Changes produced in the physical and biological treatment compared with the original samples. † The two physical methods used together, hydrocyclone and flotation equipment, bring about a reduction of 20.8% in the sulphur content and 41.1% in the ash content. Combined with the biological treatment, they reduce the sulphur content by up to 41.7% and the ash content by up to 58.9%. † An increase occurs in the calorific value of the samples treated in both the physical and biological processes. Its value expressed as a proportion of pure fuel shows that its variation in the treated coal samples is barely significant. Linked with this minimal change is the major reduction in sulphur emissions, expressed as g S/GJ, which reaches 51.1% when the physical and biological processes are both used. † Regarding the combustibility of the coal, the samples submitted to the physical process are barely affected, 1090 O. Martı́nez et al. / Fuel 82 (2003) 1085–1090 while the biological treatment always leads to an improvement in combustibility, shown by the earlier reaching of the characteristic temperatures corresponding to combustion profiles (Tv ; Tmax ; Tf ), as well as the disappearance of the secondary peak associated with the thermal decomposition of the carbonates, both of which are factors making for an enhanced combustibility. References [1] OECD/IEA. Energy policies of IEA countries. Spain 2001 review; 2001. [2] Rubiera F, Hall ST, Shah CL. Fuel 1997;76:1187. [3] Aller A, Martı́nez O, De Linaje JA, Méndez R, Morán A. Fuel Process Technol 2001;69:45. [4] Morán A, Cara J, Miles N, Shah C. Fuel 2002;81:299. [5] Martı́nez O, Aller A, Alonso J, Gómez E, Morán A. In: Pajares JA, Tascón JM, editors. Coal science and technology 24, vol. II. Amsterdam: Elsevier; 1995. p. 1749. [6] Rubiera F, Morán A, Martı́nez O, Fuente E, Pis JJ. Fuel Process Technol 1997;52:165. [7] Morán A, Aller A, Cara J, Martı́nez O, Encinas JP, Gómez E. Fuel Process Technol 1997;52:155. [8] Rubiera F, Arenillas A, Martı́nez O, Morán A, Fuente E, Pis JJ. Environ Sci Technol 1999;33:476. [9] Crelling JC, Hippo EJ, Woerner BA, West Jr DP. Fuel 1992; 71:151. [10] Smith SE, Neavel RC, Hippo EJ, Miller RN. Fuel 1981;60:458. [11] Crelling JC, Skorupska NM, Marsh H. Fuel 1988;67:781. [12] Wang W, Brown SD, Thomas KM, Crelling JC. Fuel 1994;73:341. [13] González de Andrés AI, Thomas KM. Fuel 1994;73:635. [14] Varey JE, Hindmarsh CJ, Thomas KM. Fuel 1996;75:164.