- Email: [email protected]
Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54
JOEI309_proof ■ 27 February 2017 ■ 1/11
Journal of the Energy Institute xxx (2017) 1e11
Contents lists available at ScienceDirect
Journal of the Energy Institute journal homepage: http://www.journals.elsevier.com/journal-of-the-energyinstitute
Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition Q2
Wei Li a, Shiyuan Li a, b, *, Mingxin Xu a, b, Xin Wang a, b a b
Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China University of Chinese Academy of Sciences, Beijing 100049, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 December 2016 Received in revised form 9 February 2017 Accepted 13 February 2017 Available online xxx
In order to investigate the behavior of limestone sulfation under oxy-fuel circulating fluidized bed (CFB) combustion condition, experiments were conducted in a 50 kW oxy-fuel CFB system under the O2/CO2 and air combustion conditions. A small cage, containing limestone particles, was dipped into the dense zone of the CFB combustor during the experiments. The calcination of limestone, pore structure of the product layer, and calcium conversion were studied. It was found that the increasing of temperature would promote the calcination of limestone and the high concentration of CO2 would inhibit calcination of limestone. The formation process of the product layer was completely different between the direct and indirect sulfation, while it was almost the same during the indirect sulfation under the oxy-fuel and air combustion. However, both the temperature and gas compositions played important roles in determining the pore structures of the product layer during the limestone indirect sulfation process. Under the O2/CO2 combustion condition, the calcium conversion of indirect sulfation was always higher than that of direct sulfation. The highest final calcium conversion after 60 min was found at 900 C under the O2/CO2 combustion condition. © 2017 Published by Elsevier Ltd on behalf of Energy Institute.
Keywords: Oxy-fuel Circulation fluidized bed Sulfation behavior Product layers
1. Introduction The combustion of fossil fuels for power generation is a main contributor to greenhouse gas emissions. Oxy-fuel combustion is widely considered as one of the most promising technologies to address the emissions of greenhouse gas [1e3]. With this technology, the CO2 concentration in the flue gas may be up to 90%. Therefore, an easy CO2 capture and storage becomes possible. Compared with pulverized coal (PC) oxy-fuel combustion, oxy-fuel circulating fluidized bed (CFB) combustion is believed to be a better choice for CO2 capture because of several unique advantages. Some well-known advantages include fuel flexibility, low NOx emission, and efficient sulphur removal. For the above reasons, oxy-fuel CFB combustion technology has attracted further attention from investigators [4e14]. During the typical oxy-fuel combustion process, a part of flue gas would be recycled to the furnace, thus SO2 in the flue gas would gradually accumulate. The experimental results of the 0.8 MW oxy-CFB units at Canmet Energy [15] showed that the high SO2 concentration would cause the corrosion of the tube and device due to the sulphate deposition after a long period operation. In addition, the high SO2 concentration would be harmful for CO2 compression, purification and transportation [16]. The removal of the SO2 is necessary during the oxy-fuel CFB combustor operation. Limestone sorbent injection is a relatively simple and low-cost process for SO2 capture. The operating temperature of CFBCs can vary greatly with fuels used. For example, more reactive fuels such as lignite are fired at 800e850 C, and fuels such as char are fired at 900e950 C. Moreover, because circulation of solids in the combustor can help to an effective control of the temperatures, an oxy-fuel CFB boiler can be operated under the oxygen concentration of range from 21% to above 50%. Therefore, the limestone can be surrounded by CO2 concentration ranging from 50% to above 90% during their stage in the CFB combustor. Thus, according to the equilibrium CO2 pressure over limestone on temperature, as suggested by Baker [17], the sulfation reaction of limestone can proceed via two different routes under the oxy-fuel CFB combustion condition, as shown in Fig. 1.
* Corresponding author. Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China. Fax: þ86 10 82543119. E-mail address: [email protected] (S. Li). http://dx.doi.org/10.1016/j.joei.2017.02.005 1743-9671/© 2017 Published by Elsevier Ltd on behalf of Energy Institute.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
Q1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 2/11
2
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 1. Equilibrium CO2 partial pressure over limestone.
In the zone under the curve in the figure, the operating condition leads to a previous calcination of CaCO3, and then sulfation of CaO, which is known as the indirect sulfation: CaCO3(s) / CaO(s) þ CO2(g)
(1)
CaO(s) þ SO2(g) þ 0.5 O2(g) / CaSO4(s)
(2)
In the zone above the curve, the sulfation of limestone will occur between CaCO3 and SO2 directly, being necessary lower temperatures to operate under indirect sulfation for the certain CO2 concentration. The direct sulfation step can be represented by the following formula: CaCO3(s) þ SO2(g) þ 0.5O2(g) / CaSO4(s) þ CO2(g)
(3)
Some works about the sulfation of limestone under oxy-fuel combustion condition have been done with TGA and fix-bed combustor. Most of them are related with direct sulfation. Liu et al. [18] suggested that the direct sulfation reaction can often proceed at a fairly high rate even at high conversions and consequently enables better sorbent utilization. However, Garcia-Labiano et al. [19] hold the contrast opinion, the sulfation conversion reached by the limestone under the indirect sulfation was always higher than that under the direct sulfation. The TGA and fix-bed combustor has many limitations owing to its absence of those existing in fluidized bed, such as simultaneous calcination and sulfation, attrition, thermical shock, and crackle. Up to now, only few works have been done about the sulfation of limestone with the fluidized bed reactors [20,21]. These studies are concerned with the effects of the operation parameters, such as temperature, inlet oxygen concentration, particle size etc., on the limestone desulfurization characterization, but there are still many unknowns. The aim of the paper is to study the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition. The calcination of CaCO3, pore structure of product layer, and calcium conversion under the air and oxy-fuel combustion conditions were investigated in an oxy-CFB combustor. 2. Experimental 2.1. Experimental system The experimental system consisted of a circulating fluidized bed combustor, a cyclone, a U-valve, a flue gas cooler, a bag filter, a fuel and sorbent feed unit, a gas supply unit, and a measurement and data acquisition system. Fig. 2 shows a schematic diagram of the installation. The stainless steel combustor has a height of 3250 mm and an inside diameter of 100 mm. The combustor is equipped on the outside with three electrical heaters and a solid cooler to adjust the heat duty of 30e50 kW. Coal is fed to the combustor by means of a screw feeder located just above the distributor. Another screw feeder controls the sorbent fed to the combustor. The reactant gases, air, CO2 and O2, are supplied from cylinders by mass flow controllers to simulate typical gas compositions (O2/CO2, O2/ N2) entering into the combustor in oxy-fuel combustion conditions. The oxygen concentration of inlet gas ranges from 21% to 50%. The measurement system consists of thermocouples, pressure sensors, flow meters, and gas analyzers. The temperatures and pressures in the system are measured at distances of 125 mm, 475 mm, 825 mm, 1525 mm, 2925 mm, and 3100 mm along the combustor above the distributor, and at the cyclone and the loop seal. The oxygen concentration of the flue gas is measured by a zirconia oxygen analyzer. The concentration of CO2, CO, SO2, and NOx in the flue gas is monitored on-line by an FTIR analyzer (GASMTE DX4000) located after the cyclone. 2.2. Procedure In order to investigate the limestone sulfation behavior for oxy-fuel circulating fluidized bed combustion, a small cage made from stainless steel net containing limestone particles was dipped into the dense zone of the CFB combustor during the experiment, as can be seen in Fig. 3. The cage has a height of 20 mm and a diameter of 20 mm. The aperture mesh of the stainless steel is 1 mm. All the limestone Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 3/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
3
Fig. 2. Schematic diagram of the experimental apparatus.
particles in the cage were sized to a diameter in a range of 2e3 mm, and the bed material in the furnace was sized to a diameter in a range of 0.1e0.5 mm, which ensured that the bed materials could pass through the holes of the cage freely while limestone particles could not escape from the cage. In this case, it was assumed that the reaction occurred under the condition of unchanged SO2 concentration because the quantity of limestone was small, about 5 g. The cage was taken out after the variation of reaction time (15 min, 30 min, 45 min and 60 min), and the products were analyzed with X-ray fluorescence (XRF), nitrogen adsorption analysis, and scanning electron microscope (SEM). From these analyses, the composition and microstructure of the product layer with the variation of reaction time were obtained. 2.3. Materials Datong coal, a typical bituminous in China, was used in all tests. The results of fuel analysis are shown in Tables 1 and 2. The coal was sieved to a diameter between 0.1 and 1 mm. Silicon sand of 2.5 kg with particle diameters between 0.1 and 0.5 mm was used as the bed
Fig. 3. The stainless steel cage.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 4/11
4
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Table 1 Fuel analysis for Datong coal. LHV (MJ kg1)
Ultimate analysis (wt%)
Qnet,ar
FCar
Mar
Aar
Var
Car
Har
Oar
Sar
Nar
22.61
44.38
2.2
26.05
27.37
58.08
3.73
8.58
0.32
1.04
Proximate analysis (wt%)
Table 2 The composition of the ash of Datong coal. Composition
SiO2
Al2O3
Fe2O3
CaO
MgO
TiO2
SO3
P2O5
K2O
Na2O
Content (wt%)
45.23
37.83
4.02
5.42
0.66
1.62
2.50
0.18
0.32
0.14
material. The composition of limestone in the stainless steel cage is shown in Table 3. The particle size distributions for coal and bed material are shown in Fig. 4. 3. Result and discussion In order to investigate the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, experiments were conducted in a circulating fluidized bed in oxy-fuel combustion condition with inlet oxygen concentrations of 40% and the temperature range from 850 C to 950 C. In addition, the experiments were also carried out in air combustion condition with the temperature of 850 C to study the effect of the reaction atmosphere for the indirect sulfation. All the tests were conducted with the oxygen concentration in the flue gas of about 6 ± 1%. Fig. 5 shows the average pressure drop along the height of the combustor. Fig. 6 shows the combustion temperature profile along the height of the combustor. The combustion temperature and pressure drop were all stable during the experiments. 3.1. The calcination of limestone The contents of C, O and Ca in product were obtained by XRF analysis. According the equilibrium of elements, the ratios of CaO and CaCO3 could be calculated. Fig. 7 gives the ratios of CaO in the products under different combustion conditions. Under the O2/CO2 combustion condition, there was no CaO detected in the products at the temperature of 850 C. It proves that the CaCO3 reacts with SO2 directly without calcination; However, CaO is found at the temperature of 900 C and 950 C, indicating that the reaction of sulfation switches from direct to indirect route as the temperature rises from 850 C to above 900 C, with the inlet oxygen concentration of 40%. As can be expected, the limestone would be decomposed under the air combustion condition. Under the O2/CO2 combustion condition, the calcination reaction rate of limestone is fast at 950 C with the period of 30 min, while it costs about 45 min at 900 C. Besides, the final percent of CaO at 950 C after 60 min is about 68%, which is higher than that at 900 C. Under the air combustion condition, the calcination process of limestone takes about 45 min, and the final ratio of CaO is about 62%, which is higher than that at 900 C and lower than that at 950 C under the O2/CO2 combustion condition. It can be proved that the calcination of limestone is normally determined by the CO2 partial pressure and temperature in the system together. The increasing of temperature would promote the calcination of limestone, which has been proved by many researches [22e24], while the high concentration of CO2 would inhibit calcination of limestone [25,26]. Khinast et al. [26] stated that CO2 may affect the sorption at CaO/CaCO3/CO2 interface, and result in the lower decompose reaction rate of limestone. 3.2. The pore structure of the product layers Particle surface of the unreacted CaCO3 is non-porous and looks smooth (shown in Fig. 8). The SEM photographs of product layers under different combustion conditions and reaction time are shown in Figs. 9e12, from which the formation process of the product layer during the different route of limestone sulfation can be seen. The SEM photographs with variation of reaction times from 15 to 60 min under air combustion condition at 850 C are shown in Fig. 9. Comparing Fig. 9(a) with Fig. 4, it can be noticed that the pores appeared in the surface of the product in 15 min, and the microstructures of the product layers seems porous. It is caused by the CO2 released from the surface of limestone during the process of calcination. Before 45 min, the microstructures of the product layers become more and more porous as reaction time increases. The porous product layer is benefit to the sulfation reaction of limestone. But, visible sinter can be observed from the photograph in 60 min, the product layer seems dense and less porous. It would prevent any further reaction of the SO2 with the CaO. The SEM photographs with variation of reaction times from 15 to 60 min under O2/CO2 combustion condition at 850 C are shown in Fig. 10. As can be seen in Fig. 10, the surface of the product layer does not exhibit a porous structure, which is notably different from that under air combustion condition (Fig. 9). At 15 min, the CaSO4 crystal emerges in clusters, and the product layer seems rough. As the reaction process, the CaSO4 grows, then progressively covers the particle surface. Finally, a continuous product layer formed, which is very dense and less porous (Fig. 10(d)), similar with the surface of unreacted CaCO3. Table 3 Limestone (CaCO3) composition. Composition
CaO
SiO2
Al2O3
Fe2O3
K2O
TiO2
SrO
Cr2O3
Content (wt%)
54.87
2.06
0.81
0.53
0.30
0.07
0.41
0.04
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 5/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 4. The particle size distributions for (a) DT coal and (b) bed material.
Fig. 5. The average pressure drop along the height of the combustor (900 C).
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 6/11
6
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 6. The combustion temperature profile along the height of the combustor (O2/CO2).
Fig. 7. Ratios of CaO of the products.
Figs. 11 and 12 show the SEM photographs with variation of reaction times from 15 to 60 min under O2/CO2 combustion at 900 C and 950 C, respectively. The formation process is similar with that under the air combustion mode. But the microstructure of the products layers in 60 min under the O2/CO2 combustion condition is denser and less porous than that under the air combustion condition. This reveals that the structure of the product layer depends on the reaction temperature and atmosphere. The result of nitrogen adsorption analysis of the product layer under different combustion condition and reaction time is shown in Fig. 13, which confirms the observation from the SEM photographs. The nitrogen adsorption analysis can provide the quantitative results of the pore structure of the product layer. Under the air combustion condition, from the former analysis, we can know that the limestone reacts with SO2 through the indirect sulfation. At the beginning, the calcination of the CaCO3 and the sulfation of CaO occur simultaneous, and the calcination of the CaCO3 is
Fig. 8. SEM photographs of particle surface of the unreacted CaCO3.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 7/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 9. SEM photographs of the product layer in different reaction times under air combustion condition (850 C).
Fig. 10. SEM photographs of the product layer in different reaction times under O2/CO2 combustion condition (850 C).
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
7
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 8/11
8
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 11. SEM photographs of the product layer in different reaction times under O2/CO2 combustion condition (900 C).
Fig. 12. SEM photographs of the product layer in different reaction times under O2/CO2 combustion condition (950 C).
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 9/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
9
dominant. During the process of calcination, the CO2 gas released, causing the porosity product layer. As a result, the BET surface area is larger than that of the unreacted CaCO3, as given in Figs. 13 and 9. As the calcination process of limestone, more and more CaO formed and CO2 released in the product. As a result, the BET surface area increased with increasing the reaction time. When the calcination is completed in 45 min, the sulfation and sinter of CaO play the dominant role. There are two reasons that could explain the significant decrease of the BET surface area from 45 min to 60 min. As the sulfation reaction proceeds, the pore structure becomes poor due to that the molar volume of the CaCO3 is higher than that of the CaO, and as a result, the external pores are plugged and a thin layer is formed, as can be seen in Fig. 9(d). In addition, the sinter of CaO would also lead to the modification of product layer pore structure, and result in a lower BET surface area due to the loss of micro-porosity. It can be concluded that the pore structures of the product layer were determined by the calcination, sulfation, and sinter during the indirect sulfation process. Under the O2/CO2 combustion condition, the sulfation of limestone occurs between CaCO3 and SO2 directly without calcination at 850 C, which is known as direct sulfation. At earlier stages, the amount of the product formed is only sufficient to cover a fraction of the limestone surface and a continuous product layer has not been formed. The CaSO4 crystal would make the product layer porous, and it is the reason that the BET surface area in 15 min is higher than that of the unreacted limestone. As the reaction proceeds, a continuous product would be formed. As a result, the product layer looks non-porous, and the BET surface area decreases as the reaction time increases, as can be seen in Fig. 13. As Stewart et al. [27] suggested, all the visible surfaces of the CaCO3 particles can be seen covered by a layer of crystals packed tightly. However, our observation was different from the same research [28,29] which suggests that the product layer formed by the direct sulfation reaction is porous. As the temperature increases to above 900 C, the sulfation of limestone would switch to indirect route under O2/CO2 combustion condition. From Fig. 13, it can be obtained that the variation tendency of BET surface curves at 900 C and 950 C under O2/CO2 combustion condition is qualitatively similar with that at 850 C under the air combustion condition. It can also be seen from the SEM photographs (Figs. 9, 11 and 12). Comparing the SEM photographs and BET surface areas of the indirect and direct sulfation product layer under the O2/CO2 combustion condition, we can see obvious different formation process. The pore structure of the indirect sulfation product layer was always better than that of the direct sulfation product layer before 45 min. The difference can be attributed to the calcination of the CaCO3 during the indirect sulfation. Although the formation of the indirect sulfation product layer under the air combustion condition and O2/CO2 combustion condition is similar, the BET surface areas were a little different, as shown in Fig. 13. It is suggested that the temperature and gas compositions all play important roles in determining the pore structures of the product layer during the limestone indirect sulfation process. Under the O2/CO2 combustion condition, the BET surface areas at 950 C are higher than that at 900 C in earlier stage (15 min and 30 min), while it is reversed in the final stage (45 min and 60 min). The higher calcination reaction rate and more CO2 gas release at higher temperature can promote the pore structure in the earlier stage. Otherwise, the higher temperature would also enhance the sinter of CaO, resulting in the poor structure in the final stage. It can also be seen in Fig. 13 that the BET surface area under the O2/CO2 combustion condition is lower than that under the air combustion condition. It can be attributed to that the higher CO2 partial pressure under the O2/CO2 combustion condition would inhibit the calcination of CaCO3 [25,26] and enhance the sinter of the CaO [30e32]. 3.3. The calcium conversion Both indirect and direct sulfation of limestone is the gasesolid reaction, which is performed in two regimes [33e37]. The first regime is fast and controlled by chemical reaction and/or gas diffusion through the porous product layer. The second regime where the rate is slower as the control switches to gas diffusion process once the pore structure of the product layer becomes dense and less porous. Fig. 14 shows the calcium conversion of limestone under different combustion condition and reaction time. (Calcium conversion ¼ (the amount of the CaCO3 in the product that is converted to CaSO4)/(the amount of the CaCO3 of the unreacted limestone). During the direct sulfation process (850 C, O2/CO2), it is observed that the rate of calcium conversion of limestone declines gradually with reaction time. There is no obvious difference in calcium conversion after 45 min. The reduction in the reaction rate should be caused by a decrease of the BET surface area, as can be seen in Fig. 13, which results in the decrease of gas diffusion rate through the product layer.
Fig. 13. The nitrogen adsorption analysis of the product layer.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 10/11
10
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 14. The calcium conversion of limestone under different combustion condition.
During the indirect sulfation process (900 C and 950 C, O2/CO2; 850 C, Air), the rate of calcium conversion first increases and then decreases markedly. Before 45 min, the sulfation is controlled by chemical reaction and/or gas diffusion through the porous product layer. The increasing of the BET surface area is benefit to enhance the sulfation reaction rate. Once the product become dense (after 45 min), a relatively non-porous product exerts a moderate resistance to gas penetration, and reaction rate becomes dramatically reduced at this stage. Several observations were obtained based on results shown in Fig. 14. First, the reaction rate and calcium conversion of indirect sulfation are always higher than that of direct sulfation. The result is similar with the research by Garcia-Labiano et al. [19]. Second, under the O2/CO2 combustion condition, the reaction rate and calcium conversion at 950 C are higher than that at 900 C in earlier stage (15 min and 30 min), however, it is reversed in the final stage (45 min and 60 min). In our experiment, the highest final calcium conversion after 60 min was found at 900 C under the O2/CO2 combustion condition, which is also observed by Garcia-Labiano et al. [19] and de Diego et al. [20]. Third, the reaction rate and calcium conversion under the air combustion condition are higher than that under the O2/CO2 combustion condition. It is worth to point out that these observations are similar with the results of the BET surface area of the product layer. It indicated that the pore structure of the product plays an important role in the reaction rate and calcium conversion of the limestone sulfation. 4. Conclusion In order to study the limestone sulfation behavior under oxy-fuel combustion condition for oxy-fuel circulating fluidized bed, experiments were conducted in a 50 kW oxy-fuel CFBC system under the O2/CO2 and air combustion condition. The temperature ranges from 850 C to 950 C. A small cage, containing limestone particles, was dipped into the dense zone of the CFB combustor during the tests. The cage was taken out after the variation of reaction time. The calcination of limestone, pore structure of the product layer, and calcium conversion were studied. The sulfation mechanism switches from direct to indirect route as the temperature rises from 850 C to above 900 C with the inlet oxygen concentration of 40% under the O2/CO2 combustion condition. It was found that the increasing of temperature would promote the calcination of limestone and the high concentration of CO2 would inhibit calcination of limestone. The formation process of the product layer is completely different between the direct and indirect sulfation and is almost the same during the indirect sulfation under the oxy-fuel and air combustion. Both the temperature and gas compositions play important roles in determining the pore structures of the product layer during the limestone indirect sulfation process. The pore structure of the indirect sulfation product layer was always better than that of the direct sulfation product layer before 45 min. The calcium conversion of indirect sulfation is always higher than that of direct sulfation. And the highest final calcium conversion after 60 min was found at 900 C under the O2/CO2 combustion condition. Acknowledgments The authors gratefully acknowledge the support by the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant XDA07030200) and the International Science & Technology Cooperation Program of China (Grant No. 2014DFG61680). References [1] B.J.P. Buhre, L.K. Elliott, C.D. Sheng, R.P. Gupta, T.F. Wall, Oxy-fuel combustion technology for coal-fired power generation, Prog. Energy Combust. Sci. 31 (4) (2005) 283e307. [2] Y. Tan, E. Croiset, M.A. Douglas, K.V. Thambimuthu, Combustion characteristics of coal in a mixture of oxygen and recycled flue gas, Fuel 85 (4) (2006) 507e512. [3] M.B. Toftegaard, J. Brix, P.A. Jensen, P. Glarborg, A.D. Jensen, Oxy-fuel combustion of solid fuels, Prog. Energy Combust. Sci. 36 (2010) 581e625. [4] R.I. Singh, R. Kumar, Current status and experimental investigation of oxy-fired fluidized bed, Renew. Sustain. Energy Rev. 61 (2016) 398e420. [5] H.I. Mathekga, B.O. Oboirien, B.C. North, A review of oxy-fuel combustion in fluidized bed reactors, Int. J. Energy Res. 40 (7) (2016) 878e902. € h€ [6] K. Myo anen, T. Hypp€ anen, T. Pikkarainen, T. Eriksson, A. Hotta, Near zero CO2 emissions in coal firing with oxy-fuel circulating fluidized bed boiler, Chem. Eng. Technol. 32 (2009) 355e363. [7] L. Jia, Y. Tan, C. Wang, E.J. Anthony, Experimental study of oxy-fuel combustion and desulfurization in a mini-CFBC, Energy Fuels 21 (2007) 3160e3164.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
JOEI309_proof ■ 27 February 2017 ■ 11/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
11
[8] L. Jia, Y. Tan, D. McCalden, Y. Wu, I. He, R. Symonds, E.J. Anthony, Commissioning of 0.8 MWth CFBC for oxy-fuel combustion, Int. J. Greenh. Gas Control 7 (2012) 240e243. [9] L. Duan, H. Sun, C. Zhao, W. Zhou, X. Chen, Coal combustion characteristics on an OXY-CFB combustor with warm flue gas recycle, Fuel 127 (2014) 47e51. [10] T. Czakiert, W. Muskala, S. Jankowska, G. Krawczyk, P. Borecki, Combustible matter conversion in an oxy-fuel circulating fluidized-bed (CFB) environment, Energy Fuels 26 (2012) 5437e5445. [11] T. Czakiert, K. Sztekler, S. Karski, D. Markiewicz, W. Nowak, Oxy-fuel circulating fluidized bed combustion in a small pilot-scale test rig, Fuel Process. Technol. 91 (2010) 1617e1623. [12] M. Lupion, R. Diego, L. Loubeau, B. Navarrete, CIUDEN CCS project: status of the CO2 capture technology development plant in power generation, Energy Proc. 4 (2011) 5639e5646. [13] L. Tan, S. Li, W. Li, E. Shou, Q. Lu, Effects of oxygen staging and excess oxygen on O2/CO2 combustion with a high oxygen concentration in a circulating fluidized bed, Energy Fuels 28 (3) (2014) 2069e2075. [14] W. Li, S. Li, Q. Ren, L. Tan, H. Li, J. Liu, Q. Lu, Study of oxy-fuel coal combustion in a 0.1 MWth circulating fluidized bed at high oxygen concentrations, Energy Fuels 28 (2) (2014) 1249e1254. [15] Y. Tan, L. Jia, Y. Wu, E.J. Anthony, Experiences and results on a 0.8 MWth oxy-fire operation pilot-scale circulating fluidized bed, Appl. Energy 92 (2012) 343e347. [16] R. Stanger, T. Wall, Sulphur impacts during pulverized coal combustion in oxy-fuel technology for carbon capture and storage, Prog. Energy Combust. Sci. 37 (2011) 69e88. [17] E.H. Baker, The calcium oxideecarbon dioxide system in the pressure range 1e300 atmospheres, J. Chem. Soc. 6 (1962) 464e470. [18] H. Liu, S. Katagiri, U. Kaneko, K. Okazaki, Sulfation behavior of limestone under high CO2 concentration in O2/CO2 coal combustion, Fuel 79 (8) (2000) 945e953. [19] F. Garcia-Labiano, A. Rufas, L. de Diego, M. de las Obras-Loscertales, P. Gaýan, A. Abad, J.A. Gayan, Calcium-based sorbents behavior during sulphation at oxy-fuel fluidised bed combustion conditions, Fuel 11 (2009) 3100e3108. [20] L.F. de Diego, A. Rufas, F. Garcia-Labiano, M. de las Obras-Loscertales, A. Abad, P. Gaýan, J.A. Gayan, Optimum temperature for sulphur retention in fluidised beds working under oxy-fuel combustion conditions, Fuel 114 (2013) 106e113. n ~ ez, I. Guedea, I. Bolea, L.I. Díez, L.M. Romeo, Experimental study of SO2 and NOx emissions in fluidized bed oxy-fuel combustion, Fuel Process. Technol. 106 [21] C. Lupia (2013) 587e594. [22] Q. Zhong, I. Bjerle, Calcination kinetics of limestone and the microstructure of nascent CaO, Thermochim. Acta 223 (1993) 109e120. [23] C. Chen, C. Zhao, C. Liang, K.L. Pang, Calcination and sintering characteristics of limestone under O2/CO2 combustion atmosphere, Fuel Process. Technol. 88 (2007) 171e178. [24] L. Wang, S. Li, E.G. Eddings, Fundamental study of indirect vs direct sulfation under fluidized bed conditions, Ind. Eng. Chem. Res. 54 (14) (2015) 3548e3555. [25] J.S. Dennis, A.N. Hayhurst, The effect of CO2 on the kinetics and extent of calcination of limestone and dolomite particles in fluidised beds, Chem. Eng. Sci. 42 (1987) 2361e2372. [26] J. Khinast, G.F. Krammer, C. Brunner, G. Staudinger, Decomposition of limestone: the influence of CO2 and particle size on the reaction rate, Chem. Eng. Sci. 51 (1996) 623e634. [27] M.C. Stewart, V. Manovic, E.J. Anthony, A. Macchi, Enhancement of indirect sulfation of limestone by steam addition, Environ. Sci. Technol. 44 (2011) 8781e8786. [28] M.J.H. Snow, J.P. Longwell, A.F. Sarofim, Direct sulfation of calcium carbonate, Ind. Eng. Chem. Res. 27 (1988) 268e273. [29] M.R. Hajaligol, J.P. Longwell, A.F. Sarofim, Analysis and modeling of direct sulfation of CaCO3, Ind. Eng. Chem. Res. 27 (1988) 2203e2210. [30] C. Hanson, C. Tullin, Sintering of calcium carbonate and lime mud in a carbon dioxide atmosphere, J. Pulp Paper Sci. 22 (9) (1996) 327e331. [31] D. Beruto, A.W. Searcy, A.W. Searcy, CO2 catalyzed surface area and porosity changes in high-surface area CaO aggregates, J. Am. Ceram. Soc. 67 (1984) 512e515. [32] H. Borgwardt Robert, Calcium oxide sintering in atmospheres containing water and carbon dioxide, Ind. Eng. Chem. Res. 28 (4) (1989) 493e500. [33] C. Wang, L. Jia, Y. Tan, E.J. Anthony, The effect of water on the sulphation of limestone, Fuel 89 (2010) 2628e2632. lvarez, Analysis of the direct sulfation of calcium carbonate, Thermochim. Acta 242 (1994) 161e172. [34] A.B. Uuh Fuertes, G. Velasco, E. Fuente, T. a [35] J. Zheng, J.G. Yates, P.N. Rowe, A model for desulphurisation with limestone in a fluidised coal combustor, Chem. Eng. Sci. 37 (1982) 167e174. [36] C.B. Wang, L. Chen, The effect of steam on simultaneous calcination and sulfation of limestone in CFBB, Fuel 175 (2016) 164e171. [37] H. Wang, S. Guo, D.Y. Liu, L. Yang, X. Wei, S.H. Wu, Understanding the impacts of water vapor on CaO sulfurization in a laboratory-scale fluidized bed, Energy Fuels 30 (9) (2016) 7108e7117.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
JOEI309_proof ■ 27 February 2017 ■ 1/11
Journal of the Energy Institute xxx (2017) 1e11
Contents lists available at ScienceDirect
Journal of the Energy Institute journal homepage: http://www.journals.elsevier.com/journal-of-the-energyinstitute
Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition Q2
Wei Li a, Shiyuan Li a, b, *, Mingxin Xu a, b, Xin Wang a, b a b
Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China University of Chinese Academy of Sciences, Beijing 100049, China
a r t i c l e i n f o
a b s t r a c t
Article history: Received 9 December 2016 Received in revised form 9 February 2017 Accepted 13 February 2017 Available online xxx
In order to investigate the behavior of limestone sulfation under oxy-fuel circulating fluidized bed (CFB) combustion condition, experiments were conducted in a 50 kW oxy-fuel CFB system under the O2/CO2 and air combustion conditions. A small cage, containing limestone particles, was dipped into the dense zone of the CFB combustor during the experiments. The calcination of limestone, pore structure of the product layer, and calcium conversion were studied. It was found that the increasing of temperature would promote the calcination of limestone and the high concentration of CO2 would inhibit calcination of limestone. The formation process of the product layer was completely different between the direct and indirect sulfation, while it was almost the same during the indirect sulfation under the oxy-fuel and air combustion. However, both the temperature and gas compositions played important roles in determining the pore structures of the product layer during the limestone indirect sulfation process. Under the O2/CO2 combustion condition, the calcium conversion of indirect sulfation was always higher than that of direct sulfation. The highest final calcium conversion after 60 min was found at 900 C under the O2/CO2 combustion condition. © 2017 Published by Elsevier Ltd on behalf of Energy Institute.
Keywords: Oxy-fuel Circulation fluidized bed Sulfation behavior Product layers
1. Introduction The combustion of fossil fuels for power generation is a main contributor to greenhouse gas emissions. Oxy-fuel combustion is widely considered as one of the most promising technologies to address the emissions of greenhouse gas [1e3]. With this technology, the CO2 concentration in the flue gas may be up to 90%. Therefore, an easy CO2 capture and storage becomes possible. Compared with pulverized coal (PC) oxy-fuel combustion, oxy-fuel circulating fluidized bed (CFB) combustion is believed to be a better choice for CO2 capture because of several unique advantages. Some well-known advantages include fuel flexibility, low NOx emission, and efficient sulphur removal. For the above reasons, oxy-fuel CFB combustion technology has attracted further attention from investigators [4e14]. During the typical oxy-fuel combustion process, a part of flue gas would be recycled to the furnace, thus SO2 in the flue gas would gradually accumulate. The experimental results of the 0.8 MW oxy-CFB units at Canmet Energy [15] showed that the high SO2 concentration would cause the corrosion of the tube and device due to the sulphate deposition after a long period operation. In addition, the high SO2 concentration would be harmful for CO2 compression, purification and transportation [16]. The removal of the SO2 is necessary during the oxy-fuel CFB combustor operation. Limestone sorbent injection is a relatively simple and low-cost process for SO2 capture. The operating temperature of CFBCs can vary greatly with fuels used. For example, more reactive fuels such as lignite are fired at 800e850 C, and fuels such as char are fired at 900e950 C. Moreover, because circulation of solids in the combustor can help to an effective control of the temperatures, an oxy-fuel CFB boiler can be operated under the oxygen concentration of range from 21% to above 50%. Therefore, the limestone can be surrounded by CO2 concentration ranging from 50% to above 90% during their stage in the CFB combustor. Thus, according to the equilibrium CO2 pressure over limestone on temperature, as suggested by Baker [17], the sulfation reaction of limestone can proceed via two different routes under the oxy-fuel CFB combustion condition, as shown in Fig. 1.
* Corresponding author. Institute of Engineering Thermophysics, Chinese Academy of Sciences, Beijing 100190, China. Fax: þ86 10 82543119. E-mail address: [email protected] (S. Li). http://dx.doi.org/10.1016/j.joei.2017.02.005 1743-9671/© 2017 Published by Elsevier Ltd on behalf of Energy Institute.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
Q1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 2/11
2
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 1. Equilibrium CO2 partial pressure over limestone.
In the zone under the curve in the figure, the operating condition leads to a previous calcination of CaCO3, and then sulfation of CaO, which is known as the indirect sulfation: CaCO3(s) / CaO(s) þ CO2(g)
(1)
CaO(s) þ SO2(g) þ 0.5 O2(g) / CaSO4(s)
(2)
In the zone above the curve, the sulfation of limestone will occur between CaCO3 and SO2 directly, being necessary lower temperatures to operate under indirect sulfation for the certain CO2 concentration. The direct sulfation step can be represented by the following formula: CaCO3(s) þ SO2(g) þ 0.5O2(g) / CaSO4(s) þ CO2(g)
(3)
Some works about the sulfation of limestone under oxy-fuel combustion condition have been done with TGA and fix-bed combustor. Most of them are related with direct sulfation. Liu et al. [18] suggested that the direct sulfation reaction can often proceed at a fairly high rate even at high conversions and consequently enables better sorbent utilization. However, Garcia-Labiano et al. [19] hold the contrast opinion, the sulfation conversion reached by the limestone under the indirect sulfation was always higher than that under the direct sulfation. The TGA and fix-bed combustor has many limitations owing to its absence of those existing in fluidized bed, such as simultaneous calcination and sulfation, attrition, thermical shock, and crackle. Up to now, only few works have been done about the sulfation of limestone with the fluidized bed reactors [20,21]. These studies are concerned with the effects of the operation parameters, such as temperature, inlet oxygen concentration, particle size etc., on the limestone desulfurization characterization, but there are still many unknowns. The aim of the paper is to study the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition. The calcination of CaCO3, pore structure of product layer, and calcium conversion under the air and oxy-fuel combustion conditions were investigated in an oxy-CFB combustor. 2. Experimental 2.1. Experimental system The experimental system consisted of a circulating fluidized bed combustor, a cyclone, a U-valve, a flue gas cooler, a bag filter, a fuel and sorbent feed unit, a gas supply unit, and a measurement and data acquisition system. Fig. 2 shows a schematic diagram of the installation. The stainless steel combustor has a height of 3250 mm and an inside diameter of 100 mm. The combustor is equipped on the outside with three electrical heaters and a solid cooler to adjust the heat duty of 30e50 kW. Coal is fed to the combustor by means of a screw feeder located just above the distributor. Another screw feeder controls the sorbent fed to the combustor. The reactant gases, air, CO2 and O2, are supplied from cylinders by mass flow controllers to simulate typical gas compositions (O2/CO2, O2/ N2) entering into the combustor in oxy-fuel combustion conditions. The oxygen concentration of inlet gas ranges from 21% to 50%. The measurement system consists of thermocouples, pressure sensors, flow meters, and gas analyzers. The temperatures and pressures in the system are measured at distances of 125 mm, 475 mm, 825 mm, 1525 mm, 2925 mm, and 3100 mm along the combustor above the distributor, and at the cyclone and the loop seal. The oxygen concentration of the flue gas is measured by a zirconia oxygen analyzer. The concentration of CO2, CO, SO2, and NOx in the flue gas is monitored on-line by an FTIR analyzer (GASMTE DX4000) located after the cyclone. 2.2. Procedure In order to investigate the limestone sulfation behavior for oxy-fuel circulating fluidized bed combustion, a small cage made from stainless steel net containing limestone particles was dipped into the dense zone of the CFB combustor during the experiment, as can be seen in Fig. 3. The cage has a height of 20 mm and a diameter of 20 mm. The aperture mesh of the stainless steel is 1 mm. All the limestone Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 3/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
3
Fig. 2. Schematic diagram of the experimental apparatus.
particles in the cage were sized to a diameter in a range of 2e3 mm, and the bed material in the furnace was sized to a diameter in a range of 0.1e0.5 mm, which ensured that the bed materials could pass through the holes of the cage freely while limestone particles could not escape from the cage. In this case, it was assumed that the reaction occurred under the condition of unchanged SO2 concentration because the quantity of limestone was small, about 5 g. The cage was taken out after the variation of reaction time (15 min, 30 min, 45 min and 60 min), and the products were analyzed with X-ray fluorescence (XRF), nitrogen adsorption analysis, and scanning electron microscope (SEM). From these analyses, the composition and microstructure of the product layer with the variation of reaction time were obtained. 2.3. Materials Datong coal, a typical bituminous in China, was used in all tests. The results of fuel analysis are shown in Tables 1 and 2. The coal was sieved to a diameter between 0.1 and 1 mm. Silicon sand of 2.5 kg with particle diameters between 0.1 and 0.5 mm was used as the bed
Fig. 3. The stainless steel cage.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 4/11
4
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Table 1 Fuel analysis for Datong coal. LHV (MJ kg1)
Ultimate analysis (wt%)
Qnet,ar
FCar
Mar
Aar
Var
Car
Har
Oar
Sar
Nar
22.61
44.38
2.2
26.05
27.37
58.08
3.73
8.58
0.32
1.04
Proximate analysis (wt%)
Table 2 The composition of the ash of Datong coal. Composition
SiO2
Al2O3
Fe2O3
CaO
MgO
TiO2
SO3
P2O5
K2O
Na2O
Content (wt%)
45.23
37.83
4.02
5.42
0.66
1.62
2.50
0.18
0.32
0.14
material. The composition of limestone in the stainless steel cage is shown in Table 3. The particle size distributions for coal and bed material are shown in Fig. 4. 3. Result and discussion In order to investigate the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, experiments were conducted in a circulating fluidized bed in oxy-fuel combustion condition with inlet oxygen concentrations of 40% and the temperature range from 850 C to 950 C. In addition, the experiments were also carried out in air combustion condition with the temperature of 850 C to study the effect of the reaction atmosphere for the indirect sulfation. All the tests were conducted with the oxygen concentration in the flue gas of about 6 ± 1%. Fig. 5 shows the average pressure drop along the height of the combustor. Fig. 6 shows the combustion temperature profile along the height of the combustor. The combustion temperature and pressure drop were all stable during the experiments. 3.1. The calcination of limestone The contents of C, O and Ca in product were obtained by XRF analysis. According the equilibrium of elements, the ratios of CaO and CaCO3 could be calculated. Fig. 7 gives the ratios of CaO in the products under different combustion conditions. Under the O2/CO2 combustion condition, there was no CaO detected in the products at the temperature of 850 C. It proves that the CaCO3 reacts with SO2 directly without calcination; However, CaO is found at the temperature of 900 C and 950 C, indicating that the reaction of sulfation switches from direct to indirect route as the temperature rises from 850 C to above 900 C, with the inlet oxygen concentration of 40%. As can be expected, the limestone would be decomposed under the air combustion condition. Under the O2/CO2 combustion condition, the calcination reaction rate of limestone is fast at 950 C with the period of 30 min, while it costs about 45 min at 900 C. Besides, the final percent of CaO at 950 C after 60 min is about 68%, which is higher than that at 900 C. Under the air combustion condition, the calcination process of limestone takes about 45 min, and the final ratio of CaO is about 62%, which is higher than that at 900 C and lower than that at 950 C under the O2/CO2 combustion condition. It can be proved that the calcination of limestone is normally determined by the CO2 partial pressure and temperature in the system together. The increasing of temperature would promote the calcination of limestone, which has been proved by many researches [22e24], while the high concentration of CO2 would inhibit calcination of limestone [25,26]. Khinast et al. [26] stated that CO2 may affect the sorption at CaO/CaCO3/CO2 interface, and result in the lower decompose reaction rate of limestone. 3.2. The pore structure of the product layers Particle surface of the unreacted CaCO3 is non-porous and looks smooth (shown in Fig. 8). The SEM photographs of product layers under different combustion conditions and reaction time are shown in Figs. 9e12, from which the formation process of the product layer during the different route of limestone sulfation can be seen. The SEM photographs with variation of reaction times from 15 to 60 min under air combustion condition at 850 C are shown in Fig. 9. Comparing Fig. 9(a) with Fig. 4, it can be noticed that the pores appeared in the surface of the product in 15 min, and the microstructures of the product layers seems porous. It is caused by the CO2 released from the surface of limestone during the process of calcination. Before 45 min, the microstructures of the product layers become more and more porous as reaction time increases. The porous product layer is benefit to the sulfation reaction of limestone. But, visible sinter can be observed from the photograph in 60 min, the product layer seems dense and less porous. It would prevent any further reaction of the SO2 with the CaO. The SEM photographs with variation of reaction times from 15 to 60 min under O2/CO2 combustion condition at 850 C are shown in Fig. 10. As can be seen in Fig. 10, the surface of the product layer does not exhibit a porous structure, which is notably different from that under air combustion condition (Fig. 9). At 15 min, the CaSO4 crystal emerges in clusters, and the product layer seems rough. As the reaction process, the CaSO4 grows, then progressively covers the particle surface. Finally, a continuous product layer formed, which is very dense and less porous (Fig. 10(d)), similar with the surface of unreacted CaCO3. Table 3 Limestone (CaCO3) composition. Composition
CaO
SiO2
Al2O3
Fe2O3
K2O
TiO2
SrO
Cr2O3
Content (wt%)
54.87
2.06
0.81
0.53
0.30
0.07
0.41
0.04
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 5/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 4. The particle size distributions for (a) DT coal and (b) bed material.
Fig. 5. The average pressure drop along the height of the combustor (900 C).
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
5
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 6/11
6
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 6. The combustion temperature profile along the height of the combustor (O2/CO2).
Fig. 7. Ratios of CaO of the products.
Figs. 11 and 12 show the SEM photographs with variation of reaction times from 15 to 60 min under O2/CO2 combustion at 900 C and 950 C, respectively. The formation process is similar with that under the air combustion mode. But the microstructure of the products layers in 60 min under the O2/CO2 combustion condition is denser and less porous than that under the air combustion condition. This reveals that the structure of the product layer depends on the reaction temperature and atmosphere. The result of nitrogen adsorption analysis of the product layer under different combustion condition and reaction time is shown in Fig. 13, which confirms the observation from the SEM photographs. The nitrogen adsorption analysis can provide the quantitative results of the pore structure of the product layer. Under the air combustion condition, from the former analysis, we can know that the limestone reacts with SO2 through the indirect sulfation. At the beginning, the calcination of the CaCO3 and the sulfation of CaO occur simultaneous, and the calcination of the CaCO3 is
Fig. 8. SEM photographs of particle surface of the unreacted CaCO3.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 7/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 9. SEM photographs of the product layer in different reaction times under air combustion condition (850 C).
Fig. 10. SEM photographs of the product layer in different reaction times under O2/CO2 combustion condition (850 C).
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
7
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 8/11
8
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 11. SEM photographs of the product layer in different reaction times under O2/CO2 combustion condition (900 C).
Fig. 12. SEM photographs of the product layer in different reaction times under O2/CO2 combustion condition (950 C).
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 9/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
9
dominant. During the process of calcination, the CO2 gas released, causing the porosity product layer. As a result, the BET surface area is larger than that of the unreacted CaCO3, as given in Figs. 13 and 9. As the calcination process of limestone, more and more CaO formed and CO2 released in the product. As a result, the BET surface area increased with increasing the reaction time. When the calcination is completed in 45 min, the sulfation and sinter of CaO play the dominant role. There are two reasons that could explain the significant decrease of the BET surface area from 45 min to 60 min. As the sulfation reaction proceeds, the pore structure becomes poor due to that the molar volume of the CaCO3 is higher than that of the CaO, and as a result, the external pores are plugged and a thin layer is formed, as can be seen in Fig. 9(d). In addition, the sinter of CaO would also lead to the modification of product layer pore structure, and result in a lower BET surface area due to the loss of micro-porosity. It can be concluded that the pore structures of the product layer were determined by the calcination, sulfation, and sinter during the indirect sulfation process. Under the O2/CO2 combustion condition, the sulfation of limestone occurs between CaCO3 and SO2 directly without calcination at 850 C, which is known as direct sulfation. At earlier stages, the amount of the product formed is only sufficient to cover a fraction of the limestone surface and a continuous product layer has not been formed. The CaSO4 crystal would make the product layer porous, and it is the reason that the BET surface area in 15 min is higher than that of the unreacted limestone. As the reaction proceeds, a continuous product would be formed. As a result, the product layer looks non-porous, and the BET surface area decreases as the reaction time increases, as can be seen in Fig. 13. As Stewart et al. [27] suggested, all the visible surfaces of the CaCO3 particles can be seen covered by a layer of crystals packed tightly. However, our observation was different from the same research [28,29] which suggests that the product layer formed by the direct sulfation reaction is porous. As the temperature increases to above 900 C, the sulfation of limestone would switch to indirect route under O2/CO2 combustion condition. From Fig. 13, it can be obtained that the variation tendency of BET surface curves at 900 C and 950 C under O2/CO2 combustion condition is qualitatively similar with that at 850 C under the air combustion condition. It can also be seen from the SEM photographs (Figs. 9, 11 and 12). Comparing the SEM photographs and BET surface areas of the indirect and direct sulfation product layer under the O2/CO2 combustion condition, we can see obvious different formation process. The pore structure of the indirect sulfation product layer was always better than that of the direct sulfation product layer before 45 min. The difference can be attributed to the calcination of the CaCO3 during the indirect sulfation. Although the formation of the indirect sulfation product layer under the air combustion condition and O2/CO2 combustion condition is similar, the BET surface areas were a little different, as shown in Fig. 13. It is suggested that the temperature and gas compositions all play important roles in determining the pore structures of the product layer during the limestone indirect sulfation process. Under the O2/CO2 combustion condition, the BET surface areas at 950 C are higher than that at 900 C in earlier stage (15 min and 30 min), while it is reversed in the final stage (45 min and 60 min). The higher calcination reaction rate and more CO2 gas release at higher temperature can promote the pore structure in the earlier stage. Otherwise, the higher temperature would also enhance the sinter of CaO, resulting in the poor structure in the final stage. It can also be seen in Fig. 13 that the BET surface area under the O2/CO2 combustion condition is lower than that under the air combustion condition. It can be attributed to that the higher CO2 partial pressure under the O2/CO2 combustion condition would inhibit the calcination of CaCO3 [25,26] and enhance the sinter of the CaO [30e32]. 3.3. The calcium conversion Both indirect and direct sulfation of limestone is the gasesolid reaction, which is performed in two regimes [33e37]. The first regime is fast and controlled by chemical reaction and/or gas diffusion through the porous product layer. The second regime where the rate is slower as the control switches to gas diffusion process once the pore structure of the product layer becomes dense and less porous. Fig. 14 shows the calcium conversion of limestone under different combustion condition and reaction time. (Calcium conversion ¼ (the amount of the CaCO3 in the product that is converted to CaSO4)/(the amount of the CaCO3 of the unreacted limestone). During the direct sulfation process (850 C, O2/CO2), it is observed that the rate of calcium conversion of limestone declines gradually with reaction time. There is no obvious difference in calcium conversion after 45 min. The reduction in the reaction rate should be caused by a decrease of the BET surface area, as can be seen in Fig. 13, which results in the decrease of gas diffusion rate through the product layer.
Fig. 13. The nitrogen adsorption analysis of the product layer.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65
JOEI309_proof ■ 27 February 2017 ■ 10/11
10
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
Fig. 14. The calcium conversion of limestone under different combustion condition.
During the indirect sulfation process (900 C and 950 C, O2/CO2; 850 C, Air), the rate of calcium conversion first increases and then decreases markedly. Before 45 min, the sulfation is controlled by chemical reaction and/or gas diffusion through the porous product layer. The increasing of the BET surface area is benefit to enhance the sulfation reaction rate. Once the product become dense (after 45 min), a relatively non-porous product exerts a moderate resistance to gas penetration, and reaction rate becomes dramatically reduced at this stage. Several observations were obtained based on results shown in Fig. 14. First, the reaction rate and calcium conversion of indirect sulfation are always higher than that of direct sulfation. The result is similar with the research by Garcia-Labiano et al. [19]. Second, under the O2/CO2 combustion condition, the reaction rate and calcium conversion at 950 C are higher than that at 900 C in earlier stage (15 min and 30 min), however, it is reversed in the final stage (45 min and 60 min). In our experiment, the highest final calcium conversion after 60 min was found at 900 C under the O2/CO2 combustion condition, which is also observed by Garcia-Labiano et al. [19] and de Diego et al. [20]. Third, the reaction rate and calcium conversion under the air combustion condition are higher than that under the O2/CO2 combustion condition. It is worth to point out that these observations are similar with the results of the BET surface area of the product layer. It indicated that the pore structure of the product plays an important role in the reaction rate and calcium conversion of the limestone sulfation. 4. Conclusion In order to study the limestone sulfation behavior under oxy-fuel combustion condition for oxy-fuel circulating fluidized bed, experiments were conducted in a 50 kW oxy-fuel CFBC system under the O2/CO2 and air combustion condition. The temperature ranges from 850 C to 950 C. A small cage, containing limestone particles, was dipped into the dense zone of the CFB combustor during the tests. The cage was taken out after the variation of reaction time. The calcination of limestone, pore structure of the product layer, and calcium conversion were studied. The sulfation mechanism switches from direct to indirect route as the temperature rises from 850 C to above 900 C with the inlet oxygen concentration of 40% under the O2/CO2 combustion condition. It was found that the increasing of temperature would promote the calcination of limestone and the high concentration of CO2 would inhibit calcination of limestone. The formation process of the product layer is completely different between the direct and indirect sulfation and is almost the same during the indirect sulfation under the oxy-fuel and air combustion. Both the temperature and gas compositions play important roles in determining the pore structures of the product layer during the limestone indirect sulfation process. The pore structure of the indirect sulfation product layer was always better than that of the direct sulfation product layer before 45 min. The calcium conversion of indirect sulfation is always higher than that of direct sulfation. And the highest final calcium conversion after 60 min was found at 900 C under the O2/CO2 combustion condition. Acknowledgments The authors gratefully acknowledge the support by the Strategic Priority Research Program of the Chinese Academy of Sciences (Grant XDA07030200) and the International Science & Technology Cooperation Program of China (Grant No. 2014DFG61680). References [1] B.J.P. Buhre, L.K. Elliott, C.D. Sheng, R.P. Gupta, T.F. Wall, Oxy-fuel combustion technology for coal-fired power generation, Prog. Energy Combust. Sci. 31 (4) (2005) 283e307. [2] Y. Tan, E. Croiset, M.A. Douglas, K.V. Thambimuthu, Combustion characteristics of coal in a mixture of oxygen and recycled flue gas, Fuel 85 (4) (2006) 507e512. [3] M.B. Toftegaard, J. Brix, P.A. Jensen, P. Glarborg, A.D. Jensen, Oxy-fuel combustion of solid fuels, Prog. Energy Combust. Sci. 36 (2010) 581e625. [4] R.I. Singh, R. Kumar, Current status and experimental investigation of oxy-fired fluidized bed, Renew. Sustain. Energy Rev. 61 (2016) 398e420. [5] H.I. Mathekga, B.O. Oboirien, B.C. North, A review of oxy-fuel combustion in fluidized bed reactors, Int. J. Energy Res. 40 (7) (2016) 878e902. € h€ [6] K. Myo anen, T. Hypp€ anen, T. Pikkarainen, T. Eriksson, A. Hotta, Near zero CO2 emissions in coal firing with oxy-fuel circulating fluidized bed boiler, Chem. Eng. Technol. 32 (2009) 355e363. [7] L. Jia, Y. Tan, C. Wang, E.J. Anthony, Experimental study of oxy-fuel combustion and desulfurization in a mini-CFBC, Energy Fuels 21 (2007) 3160e3164.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
JOEI309_proof ■ 27 February 2017 ■ 11/11
W. Li et al. / Journal of the Energy Institute xxx (2017) 1e11
11
[8] L. Jia, Y. Tan, D. McCalden, Y. Wu, I. He, R. Symonds, E.J. Anthony, Commissioning of 0.8 MWth CFBC for oxy-fuel combustion, Int. J. Greenh. Gas Control 7 (2012) 240e243. [9] L. Duan, H. Sun, C. Zhao, W. Zhou, X. Chen, Coal combustion characteristics on an OXY-CFB combustor with warm flue gas recycle, Fuel 127 (2014) 47e51. [10] T. Czakiert, W. Muskala, S. Jankowska, G. Krawczyk, P. Borecki, Combustible matter conversion in an oxy-fuel circulating fluidized-bed (CFB) environment, Energy Fuels 26 (2012) 5437e5445. [11] T. Czakiert, K. Sztekler, S. Karski, D. Markiewicz, W. Nowak, Oxy-fuel circulating fluidized bed combustion in a small pilot-scale test rig, Fuel Process. Technol. 91 (2010) 1617e1623. [12] M. Lupion, R. Diego, L. Loubeau, B. Navarrete, CIUDEN CCS project: status of the CO2 capture technology development plant in power generation, Energy Proc. 4 (2011) 5639e5646. [13] L. Tan, S. Li, W. Li, E. Shou, Q. Lu, Effects of oxygen staging and excess oxygen on O2/CO2 combustion with a high oxygen concentration in a circulating fluidized bed, Energy Fuels 28 (3) (2014) 2069e2075. [14] W. Li, S. Li, Q. Ren, L. Tan, H. Li, J. Liu, Q. Lu, Study of oxy-fuel coal combustion in a 0.1 MWth circulating fluidized bed at high oxygen concentrations, Energy Fuels 28 (2) (2014) 1249e1254. [15] Y. Tan, L. Jia, Y. Wu, E.J. Anthony, Experiences and results on a 0.8 MWth oxy-fire operation pilot-scale circulating fluidized bed, Appl. Energy 92 (2012) 343e347. [16] R. Stanger, T. Wall, Sulphur impacts during pulverized coal combustion in oxy-fuel technology for carbon capture and storage, Prog. Energy Combust. Sci. 37 (2011) 69e88. [17] E.H. Baker, The calcium oxideecarbon dioxide system in the pressure range 1e300 atmospheres, J. Chem. Soc. 6 (1962) 464e470. [18] H. Liu, S. Katagiri, U. Kaneko, K. Okazaki, Sulfation behavior of limestone under high CO2 concentration in O2/CO2 coal combustion, Fuel 79 (8) (2000) 945e953. [19] F. Garcia-Labiano, A. Rufas, L. de Diego, M. de las Obras-Loscertales, P. Gaýan, A. Abad, J.A. Gayan, Calcium-based sorbents behavior during sulphation at oxy-fuel fluidised bed combustion conditions, Fuel 11 (2009) 3100e3108. [20] L.F. de Diego, A. Rufas, F. Garcia-Labiano, M. de las Obras-Loscertales, A. Abad, P. Gaýan, J.A. Gayan, Optimum temperature for sulphur retention in fluidised beds working under oxy-fuel combustion conditions, Fuel 114 (2013) 106e113. n ~ ez, I. Guedea, I. Bolea, L.I. Díez, L.M. Romeo, Experimental study of SO2 and NOx emissions in fluidized bed oxy-fuel combustion, Fuel Process. Technol. 106 [21] C. Lupia (2013) 587e594. [22] Q. Zhong, I. Bjerle, Calcination kinetics of limestone and the microstructure of nascent CaO, Thermochim. Acta 223 (1993) 109e120. [23] C. Chen, C. Zhao, C. Liang, K.L. Pang, Calcination and sintering characteristics of limestone under O2/CO2 combustion atmosphere, Fuel Process. Technol. 88 (2007) 171e178. [24] L. Wang, S. Li, E.G. Eddings, Fundamental study of indirect vs direct sulfation under fluidized bed conditions, Ind. Eng. Chem. Res. 54 (14) (2015) 3548e3555. [25] J.S. Dennis, A.N. Hayhurst, The effect of CO2 on the kinetics and extent of calcination of limestone and dolomite particles in fluidised beds, Chem. Eng. Sci. 42 (1987) 2361e2372. [26] J. Khinast, G.F. Krammer, C. Brunner, G. Staudinger, Decomposition of limestone: the influence of CO2 and particle size on the reaction rate, Chem. Eng. Sci. 51 (1996) 623e634. [27] M.C. Stewart, V. Manovic, E.J. Anthony, A. Macchi, Enhancement of indirect sulfation of limestone by steam addition, Environ. Sci. Technol. 44 (2011) 8781e8786. [28] M.J.H. Snow, J.P. Longwell, A.F. Sarofim, Direct sulfation of calcium carbonate, Ind. Eng. Chem. Res. 27 (1988) 268e273. [29] M.R. Hajaligol, J.P. Longwell, A.F. Sarofim, Analysis and modeling of direct sulfation of CaCO3, Ind. Eng. Chem. Res. 27 (1988) 2203e2210. [30] C. Hanson, C. Tullin, Sintering of calcium carbonate and lime mud in a carbon dioxide atmosphere, J. Pulp Paper Sci. 22 (9) (1996) 327e331. [31] D. Beruto, A.W. Searcy, A.W. Searcy, CO2 catalyzed surface area and porosity changes in high-surface area CaO aggregates, J. Am. Ceram. Soc. 67 (1984) 512e515. [32] H. Borgwardt Robert, Calcium oxide sintering in atmospheres containing water and carbon dioxide, Ind. Eng. Chem. Res. 28 (4) (1989) 493e500. [33] C. Wang, L. Jia, Y. Tan, E.J. Anthony, The effect of water on the sulphation of limestone, Fuel 89 (2010) 2628e2632. lvarez, Analysis of the direct sulfation of calcium carbonate, Thermochim. Acta 242 (1994) 161e172. [34] A.B. Uuh Fuertes, G. Velasco, E. Fuente, T. a [35] J. Zheng, J.G. Yates, P.N. Rowe, A model for desulphurisation with limestone in a fluidised coal combustor, Chem. Eng. Sci. 37 (1982) 167e174. [36] C.B. Wang, L. Chen, The effect of steam on simultaneous calcination and sulfation of limestone in CFBB, Fuel 175 (2016) 164e171. [37] H. Wang, S. Guo, D.Y. Liu, L. Yang, X. Wei, S.H. Wu, Understanding the impacts of water vapor on CaO sulfurization in a laboratory-scale fluidized bed, Energy Fuels 30 (9) (2016) 7108e7117.
Please cite this article in press as: W. Li, et al., Study on the limestone sulfation behavior under oxy-fuel circulating fluidized bed combustion condition, Journal of the Energy Institute (2017), http://dx.doi.org/10.1016/j.joei.2017.02.005