Science in China Series B: Chemistry © 2007

Science in China Press Springer-Verlag

Experimental and mechanism studies on simultaneous desulfurization and denitrification from flue gas using a flue gas circulating fluidized bed ZHAO Yi†, XU PeiYao, SUN XiaoJun & WANG LiDong School of Environmental Science & Engineering, North China Electric Power University, Baoding 071003, China

The oxidizing highly reactive absorbent was prepared from fly ash, industry lime, and an oxidizing additive M. Experiments of simultaneous desulfurization and denitrification were carried out in a flue gas circulating fluidized bed (CFB). The effects of influencing factors and calcium availability were also investigated on the removal efficiencies of desulfurization and denitrification. Removal efficiencies of 95.5% for SO2 and 64.8% for NO were obtained respectively under the optimal experimental conditions. The component of the spent absorbent was analyzed with chemical analysis methods. The results indicated that more nitrogen species appeared in the spent absorbent except sulfur species. A scanning electron microscope (SEM) and an accessory X-ray energy spectrometer were used to observe micro-properties of the samples, including fly ash, oxidizing highly reactive absorbent and spent absorbent. The simultaneous removal mechanism of SO2 and NO based on this absorbent was proposed according to the experimental results. flue gas circulating fluidized bed, simultaneous desulfurization and denitrification, oxidizing highly reactive absorbent, mechanism

In recent years, the interests in combined desulfurization and denitrification from flue gas increased rapidly. Many technologies have been proposed, among which the stage treatment technology is considered to be a mature one. In this traditional technology, a separate NOx control system, e.g., the selective catalytic reduction (SCR) or selective non-catalytic reduction (SNCR), should be installed to the back of the desulfurization equipment. Although it succeeded in combined removal of the SO2 and NO, it is not easy to achieve wide industrial application because of the large occupying area and high running cost. To reduce the cost of flue gas purification, development of new technologies and equipments of simultaneous flue gas desulfurization and denitrification has become the leading research direction in the air pollution control field. There are many investigations in the world, but most of them have technical and economic defects, and cannot develop to practicable technologies. www.scichina.com

www.springerlink.com

The technology of flue gas desulfurization with flue gas CFB[1] was firstly proposed by Lurgi Lentjes Bischoff (LLB) Company in Germany and developed rapidly by the comprehensive investigation for the process and the accumulation of engineering practice in the past 20 years. It has achieved increasing commercial application all over the world as it reduced the cost of investment and running to 50%－70% of that for wet process. The large-scale installation of flue gas CFB has been applied in the world and China, however, this technology lacks the ability of simultaneous denitrification. The research results show that the key point for simultaneous desulfurization and denitrification technique in CFB is to oxide NO into NO2 rapidly in the flue gas, the latter is easily soluble in water. There is less literaReceived April 29, 2005; accepted May 22, 2006 doi: 10.1007/s11426-007-0013-0 † Corresponding author (email: [email protected]) Supported by the Significant Pre-research Foundation of North China Electric Power University (D03-035)

Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

ture on the rapid oxidation of NO at low temperature. An investigation of simultaneous desulfurization and denitrification with oxidant/NaOH solutions was carried out by Chu and Chien et al.[2]. A technique to oxidize NO rapidly by adding hydroxyl radicals into the flue gas was proposed by Hori[3]. Based on the researches for many years, the oxidizing highly reactive absorbent has been exploited, of which the capability of simultaneous denitrification and denitrification was verified firstly by fixing bed and duct injection[4,5], and the satisfactory results were obtained. The experiments of simultaneous desulfurization and denitrification with simulated flue gas were done in a self-designed flue gas CFB, and the higher removal efficiencies for SO2 and NOx were achieved. The removal products of SO2 and NOx contained in spent absorbent were analyzed with chemical analysis methods. The micro-properties of fly ash, oxidizing highly reactive absorbent and the spent absorbent were characterized by a SEM and an accessory X-ray energy spectrometer. The mechanism of desulfurization and denitrification based on the oxidizing highly reactive absorbent was studied. The innovative research results have significant academic sense and application value.

1 Experimental 1.1 Experimental system and the experiment of simultaneous desulfurization and denitrification Experiments of simultaneous denitrification and desulfurization were carried out in a self designed experimental system, as shown in Figure 1. The key part of the fluidized bed reactor is a vertical cylinder with the length 4500 mm and inner diameter 250 mm, on which several temperature station points are located. In this experiment, the gas mixture containing SO2, NO, H2O and air was used to simulate actual flue gas and heated by an electric heater, in which SO2 and NO came from each cylinder. The experiment system operated at the negative pressure by using an induced draft fan. The pressure drop was 800 Pa in the reactor. The oxidizing highly reactive absorbent was fed to the reactor by a screw feeder. The feed-in amounts were controlled by adjusting the divergence of the feeder. Most of the solid materials discharged from the reactor were collected by a cyclone cleaner and re-circulated into the reactor. During the operation, the atomized drops of water re136

Figure 1 The experimental apparatus of flue gas CFB system. 1, Air inlet; 2, steel bottle of SO2; 3, steel bottle of NO; 4, glass-rotor flow meter; 5, buffer bottle; 6, electric heater; 7, fluidized bed reactor; 8, water tank; 9, high-pressure pump; 10, spray nozzle; 11, screw material-fed machine; 12, vortex dust remover; 13, material-circling leg; 14, gas analysis instrument; 15, induced-draft fan; 16, ash hopper.

sulting from a high-pressure pump were sprayed into the reactor from the bottom of the reactor to adjust the adequate flue gas humidity. The concentrations of SO2 and NOx at inlet and outlet were determined by a flue gas analyzer (MRU 95/3CD flue gas analyzer, Germany) hand-running. The flue gas humidity was measured by the flue gas analyzer. The flue gas temperature was measured with thermocouple. 1.2 Preparation of the oxidizing highly reactive absorbent The oxidizing highly reactive absorbent was prepared from fly ash, industrial lime and oxidizing additive M, which was described in refs. [5,6]. The preparing procedure is as follows: The mixture of oxidizing additive M and water, including fly ash and industrial lime at the ratio of 3:1 in weight, was stirred and digested at 363 K and dried after six hours. Therefore, the oxidizing highly reactive absorbent was obtained. 1.3 Definition of removal efficiencies, characterization and analysis of removal products The removal efficiencies are calculated from the measured concentrations of SO2 and NO at inlet and outlet. A SEM (KYKY-2800B type, Scientific Instrument Corporation of Academy of Science in China) was used to observe the micro-structures of fly ash, oxygen-enriched highly reactive absorbent and the spent absorbent, the composition of oxygen-enriched highly reactive absorbent and the spent absorbent was determined by an accessory X-ray energy spectrometer (Vantage DIS type,

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

Thermo NORAN Company in USA). The contents of sulfate, sulfite, nitrate and nitrite in the spent absorbent were analyzed using the methods of chemical analysis. Sulfate and sulfite were determined by barium chromate photometry. Nitrite was determined by N-(1-naphthyl)-ethylenediamine photometry, and nitrate was determined by the process of the reduction of zinc powder. 1.4 Circulating flow rate, circulating multiple and the concentration of solid materials in the reactor The availability of the absorbent in the reactor is characterized by circulating flow rate, circulating multiple and the concentration of solid materials in the reactor. Circulating flow rate (kg/(m2·s)) is obtained by measurement and calculation. The concentration of solid materials in the reactor (kg/m3), equal to the ratio of the solid materials mass to volume of the reactor, is calculated. Circulating multiple is obtained by expression of Gs·A/mds, where Gs is circulating flow rate, A is the sectional area of the reactor, and mds is the amount of absorbent. All of the three parameters can be adjusted by a screw feeder.

2 Results and discussion 2.1 Calculation methods 2.1.1 Calculation of Ca/(S+N). In flue gas desulfurization technology, an important index Ca/S is usually used to evaluate the utilization of absorbents and serves as a critical technological parameter to determine the addition of absorbent. As flue gas simultaneous desulfurization and denitrication process in this paper, Ca/(S+N) is used instead of Ca/S and expressed as (1) Ca /(S + N ) = n1 / n 2 , where n1 and n2 stand for the molar weights of calcium oxide in absorbents and the mixtures (SO2+1/2NO) in flue gas, respectively. The definition of stoichiometric relationship is based on the expected removal products mainly including CaSO3/CaSO4 and Ca(NO3)2/Ca(NO2)2 in the spent absorbent[7]. 2.1.2 Removal efficiencies for SO2 and NO. The removal efficiencies for SO2 and NO are calculated from C j ,in − C j ,out (2) Yj = × 100%, C j ,in where “j” can be SO2 or NO, “in” and “out” stand for inlet and outlet, respectively.

2.1.3 Calculation of the feed-in amount of absorbent. The feed-in amount of the absorbent is calculated from ⎛ m3 mg 10−3 g 1 mol 1h ⋅ CSO2 3 ⋅ ⋅ ⋅ ⎜Q h mg 64 g 60 min m ⎝ m3 mg 10−3 g 1 mol 1 1h ⎞ ⋅ CNO 3 ⋅ ⋅ ⋅ ⋅ ⎟ h mg 30 g 2 60 min ⎠ m 74 g 4 100 + x ⋅ ⋅ ⋅ ⋅n mol 0.9 100 100 + x = n(8.56QCSO2 + 9.13QCNO ) ⋅ ⋅ 10−5 g / min, (3) 100 where n = Ca/(S+N), x is the mass percent of the additive, Q is the gas flow rate in m3/h, CSO2 is the concen+Q

tration of SO2 in mg/m3, and CNO is the concentration of NO in mg/m3.

2.1.4 Calculation of availability of calcium. The calculating formula for availability of calcium in flue gas simultaneous desulfurization and denitrification is obtained referring to that of flue gas desulfurization: Availability of Ca = Reacted amounts of Ca in desulfurization and denitrification Added amounts of Ca in absorbent

Q ⋅ CSO2 ⋅ YSO2

Q ⋅ CNO ⋅ YNO 30 × 2 Q ⋅ CNO ⎞ + ⎟⋅n 30 × 2 ⎠ +

64 ⎛ Q ⋅ CSO2 ⎜ ⎝ 64 15CSO2 ⋅ YSO2 + 16CNO ⋅ YNO 1 ⋅ . = 15CSO2 + 16CNO n =

(4)

2.2 Effects of the running parameters on desulfurization and denitrification in flue gas CFB 2.2.1 Effect of content of M additive on removal efficiencies for SO2 and NO. In order to get the optimal additive quantity of M, the comparing experiment of desulfurization and denitrification was done. The experimental results are shown in Figure 2. The experimental results show that the content of M has no obvious effect on desulfurization, but has a significant effect on the removal of NO. The efficiencies of denitrification increased rapidly as the content M increased at lower content of M, and reached maximum at 1.6%(wt) M, and then the efficiencies increased slowly with the content of M, because M is highly dispersed on the surface of the prepared absorbent during the course of assimilation which made it contain abundant oxidizing-point. The removal reaction was primarily on the

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

137

control of the conversion from NO to NO2 at lower content of M. With the content of M, the “oxidizing-point” and the conversion from NO to NO2 also increased greatly, therefore the efficiencies of denitrification increased significantly. However, the “oxidizing-point” on the surface of the absorbent tended to saturation when the content of M reached 1.6%(wt). Correspondingly, the efficiencies of denitrification increased very slowly with the content of M. The optimal content of M was 1.6%(wt). Figure 3 Effect of the residence time on the efficiencies of simultaneous denitrification and desulfurization. The inlet temperature of flue gas in flue gas CFB is 130℃; the flue gas humidity is 6%(v). Ca/(S+N) is 1.2; the content of additive M is 1.6%(wt); the concentrations of SO2 and NO are 2965 and 963 mg/m3, respectively.

flue gas was determined to be 2.4 s in the experiment.

Figure 2 Effect of the content of additive M on the efficiencies of simultaneous denitrification and desulfurization. The inlet temperature of flue gas in flue gas CFB is 130℃; the flue gas humidity is 6%(v); the residence time is 2.4 s; the flow rate is 400 m3/h. Ca/(S+N) is 1.2; the circulation multiple is 85; the concentrations of SO2 and NO are 3087 and 1032 mg/m3, respectively.

2.2.2 Effect of residence time of flue gas on removal efficiencies for SO2 and NO. The removal efficiencies are affected by gas-solid contact and NO oxidization, which is controlled directly by the residence time of flue gas in a flue gas CFB. The optimal residence time of flue gas was achieved by the experiment. The results are shown in Figure 3. It can be seen from Figure 3 that efficiencies of desulfurization and denitrification increase with the residence time of flue gas in flue gas CFB, especially for denitrification. This could be that the contact of flue gas with absorbents is more sufficient as an increase of the residence time of flue gas, which leads to the rise of removal efficiencies. With removal of NO, due to the low water solubility of NO, it can be absorbed hardly by the absorbent, therefore, the oxidation of NO to NO2 is necessary, which can be carried out by an increase of the residence time of flue gas. However, excessive residence time of flue gas would increase the volume capacity of reactor and running cost. The optimal residence time of 138

2.2.3 Effect of Ca/(S+N) on removal efficiencies for SO2 and NO. Figure 4 shows the effect of Ca/(S+N) on efficiencies of desulfurization and denitrification. It can be seen that the efficiencies of desulfurization and denitrification increase with the value of Ca/(S+N), and the removal efficiencies appeared to be manifest trend of increase when the Ca/(S+N) lay between 0 and 1.2. However, the turning point of removal efficiencies occurred at 1.2 Ca/(S+N), thereafter, the removal efficiencies stayed on nearly constant at the higher value when the Ca/(S+N) lay between 1.2 and 2.0. The optimal Ca/(S+N) is set to be 1.2 according to Figure 4.

Figure 4 Effect of Ca/(S+N) on the efficiencies of simultaneous denitrification and desulfurization. The inlet temperature of flue gas in flue gas CFB is 130℃; the flue gas humidity is 6%(v); the residence time is 2.4 s; the flow rate is 400 m3/h; the content of additive M is 1.6%(wt); the concentrations of SO2 and NO are 3110 and 978 mg/m3, respectively.

2.2.4 Effect of circulating multiple on availability of calcium. Compared with that of circulating flow rate, circulating multiple and the concentration of solid mate-

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

rials, the effect of circulating multiple on availability of calcium is more visual. The relationship between availability of calcium and circulating multiple is shown in Figure 5, with the flue gas CFB inlet at the concentrations of SO2 3087 mg/m3 and NO 1032 mg/m3, respectively, Ca/S+N of 1.2.

Figure 5 Effect of circulating multiple on Ca availability. The inlet temperature of flue gas in flue gas CFB is 130℃; the flue gas humidity is 6%(v); the content of additive M is 1.6%(wt); the residence time is 2.4 s; the flow rate is 400 m3/h; the concentrations of SO2 and NO are 3087 and 1032 mg/m3, respectively.

From the expression of calcium availability in sec. 2.1.4, the efficiencies of desulfurization and denitrification varied with the change of calcium availability in a given concentration of SO2, NO and Ca/S+N in flue gas CFB. When circulation ratio was 1, the entering solid materiel into flue gas CFB will entirely be discharged from the reactor. Therefore the flue gas CFB experiment was regarded as the same in the duct injection system, and the flue gas CFB reactor became the duct in which reaction occurred among the absorbent, SO2 and NO. Since SO2, NO and the absorbent contacted only once, the residual calcium in absorbent was not reused to desulfurization and denitrification, and the absorbent showed a low calcium availability. The concentration of solid materials in the reactor increased with the circulating multiple. Accordingly, the quantity of materiel taken out by flue gas would also increase in unit time. Therefore, the absorbent can be diffused better in the reactor by air flow，which leads to the increase of gas-solid contact area. The contact probability between absorbent and reacting gas also increased and the availability of calcium was promoted. However, the excessive concentration of solid materials in the reactor would increase not only the pressure

drop, but also the burden of solid-gas separator. Based on expression of the ratio of Ca to (S+N) and previous work[6], the feed-in velocity of the absorbent was calculated as 2.19 g/s, while the concentrations of SO2 and NO at the flue gas CFB inlet were 3087 and 1032 mg/m3 respectively and the Ca/(S+N) was 1.2. Here, the concentration of solid materials in the reactor was 4.5 kg/m3, the circulating flow rate was 3.8 kg/(m2·s), and the circulation multiple was 85. 2.2.5 Effect of inlet flue gas temperature on removal efficiencies for SO2 and NO. The variations of removal efficiencies for SO2 and NO removal as a function of the inlet temperature of flue gas in flue gas CFB are shown in Figure 6. Satisfactory efficiencies were obtained in the inlet flue gas temperature range from 120 to 130℃. When the temperature was higher than 130℃, the removal efficiencies decreased with the temperature. The optimal inlet flue gas temperature was selected at 130℃ accordingly. In fact, removal for SO2 and NO is a semi-dry reaction process between porous solids and gases. The adsorption, absorption and diffusion of SO2 and NO on the surface of absorbent are affected greatly by the flue gas temperature. When the temperature was lower than a critical value, adsorption and absorption were predominant and the efficiencies increased with the temperature[7,8]. However, when the temperature of flue gas is higher than the critical value, the desorption of gas molecules on the surface of the absorbent enhanced and the equilibrium adsorptive capacity of gas decreased. Therefore, the removal efficiencies decreased evidently with the temperature. There is a turning point for absorption, adsorption and desorption when temperature of

Figure 6 Effect of the inlet flue gas temperature on efficiencies of denitrification and desulfurization. The flue gas humidity is 6%(v); the Ca/(S+N) is 1.2; the residence time is 2.4 s; the flow rate is 400 m3/h; the content of additive M is 1.6%(wt); the circulation multiple is 85; concentrations of SO2 and NO are 2912 and 939 mg/m3, respectively.

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

139

flue gas reaches 130℃, which closes to the flue gas temperature in coal-fired power plant after dust removal. It is convenient for flue gas simultaneous desulfurization and denitrification.

The flue gas humidity should be controlled at above 6% due to humidity of practical flue gas exceeding 6% according to Figure 7. The higher removal efficiency can be obtained.

2.2.6 Effect of flue gas humidity on removal efficiencies for SO2 and NO. Flue gas humidity is one of the important factors affecting removal efficiencies. The experimental results of denitrification and desulfurization are shown in Figure 7 in the humidity range from 0 to 8%. The results indicated that removal efficiencies increased with the humidity, and markedly increased in removal efficiencies as flue gas humidity increased at the lower humidity. However, increasing degree of the removal efficiencies is lessened little by little as humidity increased. It is considered that the removal reactions are mainly physical adsorption and slow gas-solid reaction at low humidity, the water membrane would be formed gradually on the surface of absorbent as flue gas humidity increased, especially for the formation of the monomolecular layer before, hence, the SO2 and NOx could be easily dissolved on the surface of absorbent and the rapid ionizing reactions between Ca(OH)2 and SO2 and NOx occurred, so the removal efficiencies were enhanced obviously. Figure 7 also shows that the removal efficiencies appeared constantly when the humidity exceeded 6%, at which adsorbed water on the surface of the absorbent exceeded monomolecular layer, and the diffusion of SO2 and NOx to the surface of the absorbent would be delayed. Therefore, the velocity of ionizing reactions slowed down, which is consistent with that of － Yoon and Stouffe[9 11].

2.2.7 Effect of concentrations of SO2 and NO on removal efficiencies for SO2 and NO. The emission concentrations of SO2 and NOx vary with the content of the sulfur in burned coal and the operating conditions of the boiler. Experiments of simultaneous desulfurization and denitrification were carried out in order to examine the adaptability of flue gas CFB process for coal species and conditions of the boiler (Figure 8). It can be seen that the removal efficiencies varied very little in the range of 1000－3500 mg/m3, indicating that the flue gas CFB technology is good in adapting to different coals and combustion conditions.

Figure 8 Effect of SO2 and NO concentrations on the efficiencies of simultaneous denitrification and desulfurization. The inlet temperature in flue gas CFB is 130℃; the flue gas humidity is 6.0%(v); the residence time is 2.4 s; the flow rate is 400 m3/h; Ca/(S+N) is 1.2; the content of additive M is 1.6%(wt); the circulation multiple is 85.

2.3 Parallel experiments of simultaneous desulfurization and denitrification in flue gas CFB

Figure 7 Effect of flue gas humidity on the efficiencies of simultaneous denitrification and desulfurization. The inlet temperature of flue gas in flue gas CFB is 130℃; the residence time is 2.4 s; the flow rate is 400 m3/h; Ca/(S+N) is 1.2; the content of additive M is 1.6%(wt); the circulation multiple is 85; the concentrations of SO2 and NO are 2487 and 980 mg/m3, respectively.

140

The optimal conditions of flue gas CFB were achieved through the above experiments, with the inlet temperature of flue gas of 130℃, the flue gas humidity of 6%(v), the residence time of 2.4 s, the flow rate of 400 m3/h, Ca/(S+N) of 1.2, the content of additive M of 1.6%(wt), and the circulation multiple of 85. Five parallel experiments of simultaneous desulfurization and denitrification in flue gas CFB were carried out using oxidizing highly reactive absorbent when the inlet concentrations of SO2 and NO were 3087 and 1032 mg/m3 respectively. The results are showed in Table 1. It can be seen that higher efficiencies of simultaneous denitrification and desulfurization were achieved, which was valuable

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

Table 1 Results of simultaneous flue gas desulfurization and denitrification with a flue gas CFB Items 1 2 3 4 Efficiencies SO2 (%) 95.1 95.4 95.8 95.3 NO (%) 64.9 64.5 65.1 64.8

for industrial application. 2.4 Chemical analysis of removal products The removal products of SO2 and NO in the spent absorbent were analyzed with chemical analysis methods described in sec. 1.3[12]. The theoretical contents of sulfur and nitrogen species in the spent absorbent were calculated according to the inlet concentrations of SO2 and NO and removal efficiencies. The contents of sulfur and nitrogen species in the spent absorbent resulting from theoretical and measured values are listed in Table 2. It is indicated that the complicated reactions of SO2, NOx and calcium species contained in the absorbent have been carried out. The molar ratio of sulfate and sulfite in the spent absorbent was 1.85, and that of nitrite and nitrate was 2.98, which shows that sulfate was the main desulfurization product and nitrite was the main denitrification one. The suggestion can be provided by the above results for the assessment and utilization of the removal products. 2.5 Desulfurization and denitrification mechanism for oxidizing highly reactive absorbent

Average

Sample variance S2

95.7 64.7

95.5 64.8

0.08 0.05

grid of Si－O, Al－O, and finally makes activated SiO2 and Al2O3 react with calcium hydroxide. Hence, the achieved highly active absorbent showed a highly specific surface area and adsorption activation. From the SEM images shown in Figures 9, 10 and 11, the surfacial micro-structure for material particles changed significantly before and after digesting. Figure 9 shows that distinct spherical bodies with glossy and tight surface can be seen in the un-preparative fly ash, indicating that the main crystal should be SiO2 and Al2O3[13]. Figure 10 reveals that the surface of oxidizing highly reactive absorbent appears coarse and destroyed seriously. Some gelling amorphous materials and plentiful loopholes are formed on the surface, demonstrating that oxidizing highly reactive absorbent has a highly specific surface area. Figure 11 shows the SEM image for the spent oxidizing highly reactive absorbent. Compared with Figure 10, the micro-structures for the spent particles are different, demonstrating that after the reaction there are plentiful sediments and the surface turns to be glossy, because the surface of absorbent is

Calcium absorbent is widely used in flue gas desulfurization. However, the availability of the absorbent is low in semi-dry or dry process. To overcome this disadvantage and realize simultaneous desulfurization and denitrification especially, the oxidizing highly active absorbent was prepared[4,5] by fly ash, lime and additive. As the key material for preparation of the oxidizing highly active absorbent, fly ash is chiefly composed of SiO2, Al2O3, Fe2O3 and CaO, which is regarded as volcanic ash because it contains a great amount of non-crystal SiO2 and Al2O3. Plentiful silicon and aluminum oxides are gradually activated by the effect of alkali excitant. Calcium hydroxide diffuses to the spherical glass body surface on fly ash to cause chemical adsorption and erosion, which dissolves the glass body and destroys the Table 2

5

Figure 9 Surface of fly ash particle.

Theoretical and measured values of sulfur and nitrogen species in spent absorbent (mmol/g) Theoretical valuesa)

Content of S

Measured valuesb)

Content of N

S[SO2− ] 4

S[SO2− ] 3

N[NO− ] 2

Relative error

N[NO− ] 3

S species

N species

1.597 0.778 0.997 0.538 0.555 0.186 3.88% 4.76% a) The calculated values resulting from the contents of S and N multiply the removal efficiencies; b) the measured values of S and N in spent absorbent. ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

141

Figure 10

Surface of oxidizing highly reactive absorbent.

Figure 11

Surface of the spent absorbent.

covered by removal products including calcium sulfate, calcium sulfite, calcium nitrate, and calcium nitrite and so on, after the resorption and adsorption. Figure 12 shows the energy spectrum on the surface of fly ash, providing the main elementary compositions and the relative contents.

Figure 13 shows the energy spectrum on the surface of one single particle of the spent absorbent. Compared with Figure 12, it is obvious that sulfur species appears in the image. Unfortunately, the Vantage DIS X-ray energy spectrometer can only analyze the elements between Na and U, as the nitrogen element could not be analyzed. The absorption of nitrogen species was verified by chemical analytical methods (Table 2). Because the oxidizing highly reactive absorbent is rich in calcium, from Table 2, it is clear that the chemical absorptions of SO2 and NO2 and calcium species contained in the absorbent have been carried out. Masanori Sakai’s research on simultaneous denitrification and desulfurization by hydrated lime[14] has demonstrated that the coexistence of SO2 and NO could promote the removal with each other. The infrared band for the spent sample showed that without NO in flue gas, the main product of SO2 absorbed by calcium hydroxide was SO32−. While with the coexistence of NO, the main product was SO42−, showing that the existent NO enhances the oxidation of SO32− to SO42−. The analysis of desulfurization product in the experiment showed that the molar ratio of sulfate radical to sulfite radical in the spent absorbent was 1.85, which was similar to ours. Tomohiro Ishizuka et al.[15] investigated the reaction between SO2/NO and O2 on the surface of CaO by means of infrared band and pre-organization temperature desorption technology. The experimental results further demonstrated that surface species formed from NO adsorption was relevant to the conversion from SO2 to SO42−. However, NO2/O2 was not effective to such conversion, namely that the adsorbed SO2 on the surface which had already pre-adsorbed NO2 could not be converted to SO42−. The coexistence and pre-adsorbed oxygen were of no use in this conversion, whereas useful in the presence of NO. O2 became useful.

Figure 12 Energy spectrum at 1# point of Figure10. 142

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

Figure 13

Average energy spectrum on the surface of one single particle for the spent absorbent.

The predominant form of NOx in flue gas is presented as NO, whose solubility in water is much lower than NO2, HNO2 and HNO3 and so on[16], which is the main reason why NOx cannot be removed together with SO2 from flue gas in the conventional desulfurization techniques. Therefore, the key for simultaneously removing NOx and SO2 from flue gas is to oxidize NO to NO2 rapidly. It was found in our experiments that the absorbent containing no additive M showed scarce removal for NOx. However, with the absorbent containing additive M, many oxygen-rich points were formed on the surface of it where NOx can be oxidized to NO2 within the residence time 2.4 s of flue gas in the reactor. Firstly, NO2 could be adsorbed physically together with SO2 on the surface of absorbent, when atomizing water was sprayed into the reactor, water film formed on the surface of absorbent. SO2 and NO2 were dissolved into the water film and reacted rapidly with Ca(OH)2 in the absorbent. According to the results of Masanori Sakai[14] et al. and our results in this paper, the final products of the desulfurization and denitrification process were calcium sulfate, calcium sulfite, calcium nitrate and calcium nitrite. There are two functions of additive M for removing NOx. As a strong oxidant, it can oxidize NO to NO2 rapidly. On the other hand, its reduced products may have catalytic activity. Furthermore, the effect of the oxygen content in flue gas should not be neglected, which should be investigated in the future. Combined with the results of the X-ray energy spectrometer and those of the chemical analysis, the reaction mechanism in flue gas CFB can be deduced as follows: SO2+H2O ⎯⎯ → H2SO3

(5)

→ CaSO3+2H2O Ca(OH)2+H2SO3 ⎯⎯

(6)

→ reactive complex CaSO3+O2+NO ⎯⎯ compounds ⎯⎯ → CaSO4+NO2

(7)

→ NO2+M(reductive products) (8) NO+M (oxidant) ⎯⎯ → 2HNO3 + NO 3NO2 +H2O ⎯⎯

(9)

→ 2HNO2 NO2+NO+H2O ⎯⎯

(10)

Ca(OH)2+2HNO3 ⎯⎯ → Ca(NO3)2+ 2H2O

(11)

→ Ca(NO2)2+2H2O Ca(OH)2+2HNO2 ⎯⎯

(12)

The previous work of Chironna et al.[17] showed that the ratio of NO/NO2 is extremely important in the conventional scrubbing technology, and the highest removal efficiency occurs when the molar ratio of NO to NO2 is 1. In this work, the molar ratio of nitrite to nitrate in denitrification products was measured to be 2.98. Research results[18] show that the heavy metals, such as Hg, Cd, Ni, Cr, V, Se, Fe, Mn, Cu, Zn, Pb and As, are usually attended in environment. Quite a few heavy metals are essential trace elements for organisms, while some are environmental pollutions, such as Cd, Pb, As, Cr and Hg, excluding additive M. Moreover, additive M applied in this experiment is in trace and will cause little environmental influence.

3 Conclusion (1) Oxidizing highly reactive absorbent based on lime, fly ash and additive M was prepared. The analysis results of SEM and X-ray energy spectrometer showed that volcanic ash reaction occurred between fly ash and lime during the digesting. Rough and plentiful pores were observed on the surface of the oxidizing highly reactive absorbent, which indicated that oxidizing highly reactive absorbent had a highly specific surface area. Additive M dispersed uniformly on the absorbent surface.

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144

143

(2) The experiments of simultaneous desulfurization and denitrification in flue gas CFB were carried out by using oxidizing highly reactive absorbent. The results showed that flue gas humidity, flue gas temperature, flue gas residence time, content of additive M, Ca/(S+N) and circulating multiple are principal influencing factors for removal of SO2 and NOx from flue gas. The achieved optimal parameters of simultaneous desulfurization and denitrification using flue gas CFB are as follows: the inlet temperature of flue gas in flue gas CFB is 130℃; the flue gas humidity is 6%(v); the residence time is 2.4 s; the flow rate is 400 m3/h; Ca/(S+N) is 1.2; the content of additive M is 1.6%(wt) and the circulation multiple is 85. Removal efficiencies for SO2 and NO under the optimal conditions are 95.5% and 64.8% respectively. The 1

2

3

4

5

6

7

8

144

Zhong Q. Techniques of Flue Gas Desulfuriztion and Denitrification for Coal-fired and Examples (in Chinese). Beijing: Chemical Industry Press, 2002. 182－203 Chu H, Chien T W, Li S Y. Simultaneous absorption of SO2 and NO from flue gas with KMnO4/NaOH solutions. Sci Total Environ, 2001, 275(3): 127－135 Hori M, Matsunga N, Malte P C, et al. The effect of low concentration-fuels on the conversion of nitric oxide to nitrogen dioxide. In: Twenty-Four Symposium (International) on Combustion. The Combustion Institute, Sydney, Australia, 1992. 909－916 Zhao Y, Ma S C, Li Y Z, et al. Experimental investigation of desulfurization and denitrification from flue gas by absorbents based on fly ash. Zhongguo Dianji Gongcheng Xuebao/Proceedings of the Chinese Society of Electrical Engineering (in Chinese), 2002, 22(3): 108－112 Zhao Y, Ma S C, Huang J J, et al. Experimental study on SO2 and NOx removal and mechanism by highly reactive sorbent. Proc Chin Soc Electr Eng (in Chinese), 2003, 23(10): 236－240 Fan B G, Qi H Y, Yu C F, et al. Mass balance and chemical change of bed materials in circulating fluidized bed during desulfuriztion. Therm Energy Power Eng(in Chinese), 2001, 23(10): 236－240 Gao X, Luo Z Y, Chen Y F, et al. Study on effect of moisture on desulfurization characteristic of calcium-based sorbent. Combust Sci Technol (in Chinese), 1999, 5(1): 39－45 Gao X, Luo Z Y, Liu N. Desulfurization characteristic of calciumbased sorbent during activation process. J Chem Eng JPN, 2001, 34(9): 1114－1119

mass percent of additive M significantly affects the denitrification, while weakly affects the desulfurization. (3) The experimental results in flue gas CFB also showed that the technology of simultaneous denitrification and desulfurization appears stable and simple, which is valuable in industry application. (4) According to the analysis of SEM and X-ray energy spectrometer and the chemical analysis of the removal products, the mechanism of desulfurization and denitrification using oxidizing highly reactive absorbent can be described as follows: NO is oxidized to NO2 firstly, then reacted with Ca(OH)2 contained in the absorbent together with SO2. Therefore, calcium sulfate, calcium sulfite, calcium nitrate and calcium nitrite are formed. 9 10

11

12

13 14

15

16 17 18

Yoon H, Stouffer M R, Rosenhoover W A, et al. Pilot process variable study of coolside desulfurization. Environ Prog, 1988, 7(2): 101－111 Stouffer M R, Yoon H F, Burker P. An investigation of the mechanisms of flue gas desulfurization by in-duct dry sorbent injection. Ind Eng Chem Res, 1989, 28: 20－27 Cook J L, Khang S J, Lee S K, et al. Attrition and changes in particle size of lime sorbents in a circulating fluidized bed absorber. Power Technol, 1996, 89: 1－8 Christopher H N, Gary T R. Simultaneous sulfur dioxide and nitrogen dioxide removal by calcium hydroxide and calcium silicate solids. J Air Waste Manage, 1998, 48(9): 819 －828 Zhao P G. Synthetic Utilization of Fly Ash (in Chinese). Shenyang: Liaoning Science and Technology Publishing House, 1993. 68－99 Sakai M, Su C L, Sasaoka E J. Simultaneous removal of SOx and NOx using slaked lime at low temperature. Ind Eng Chem Res, 2002, 41(20): 5029－5033 Tomohiro I, Hajime K, Tsutomu Y, et al. Initial step of flue gas desulfurization——An IR study of the reaction of SO2 with NOx on CaO. Environ Sci Technol, 2000, 34(13): 2799－2803 Ervin B M J, Thomas J O. Hydrogen peroxide scrubber for the control of nitrogen oxides. Environ Eng Sci, 2002, 19(5): 321－327 Chironna R J, Altshuler B. Chemical aspects of NOx scrubbing. Pollut Eng, 1998, 31: 32－37 Gu J G, Zhou, Q X, Wang X. Reused path of heavy metal pollution in soils and its research advance. J Basic Sci Eng (in Chinese), 2003, 12(2): 143－151

ZHAO Yi et al. Sci China Ser B-Chem | February 2007 | vol. 50 | no. 1 | 135-144