Commercial test of a slurry jet FGD system

Commercial Test of a SlurryJet FGD System Suhao He,Guangming Xiang, Dingkai li,Yon Li, Qiang Yao, and Xuchang Xu State Key Laboratory of Clean Coal Combustion, Department of Thermal Engineering, Beijing 100084, People’s Republic of China

A flue gas desulfurization (FGD) system that treats dis-

chargesfrom three industrial boilers using a slurry jet technique, was tested at the Shenyang Fertilizer Factory in Shenyang in Northeast China. At the design conditions, the FGD system had a desulfurization efficiency greater than 93%,and a calcium utilization efficiency of approximately 93%with S(IV)fraction oxidized to S(W) in excess of 98%. The influence of the main operating parameters are also analyzed in thispaper. The system has a high desulfurization efficiency at varying inlet SO, concentrations and gas flow rates, and a low cost f o r industrial use, especiallyfor small- and medium-size industrial boilers in China. INTRODUCTION

Air pollution and the related acid rain are serious problems in China. It is estimated that, if n o control strategies are implemented, the amount of SO, discharged in China will reach 33 million tons in 2010, about one third of which will come from small- and medium-size industrial boilers. Several flue gas desulfurization (FGD) techniques have been introduced in China, but most are not suitable because they d o not satisfy the criteria of low equipment cost, low water consumption, and low waste discharge [ll. In addition, technology is urgently needed to simultaneously provide desulfurization and dust removal [21. Wet FGD is the most appropriate technique for SO, pollution control [31. It can achieve a high desulfurization efficiency of 95% and stable operation relatively easily. Depending on the gas-liquid contact method, an FGD scrubber may be classified as a packed tower, a spray tower, a bubbling tower, or the so-called slurry jet tower. Each type of scrubber has its own characteristics. Currently. the spray tower is the most widely used scrubber, since its sprayers generate an even spatial distribution of droplets to absorb the SO, from the gas [41, while a packed tower uses a liquid film 151, and a bubble tower uses bubbles as the gas-liquid contact Environmental Progress (V01.21, No.2)

interface, such as in the Chiyoda Thoroughbred 121 FGD system (CT-121) [61. However, all these devices have shortcomings. To g e n e ra te a n a p p r o p r iate droplet distribution, sprayers in a spray tower have complex designs and can easily be fouled. Plugging can occur in the packing layers of packed towers if the slurry pH is not carefully controlled. Bubbling towers have similar plugging and control problems. All these problems increase the operating risk and the final cost of FGD systems. A slurry jet tower avoids these problems since the jet nozzle is bigger and there are no packing materials. The operating problems and the final cost are also reduced since the control system demands and plugging potential are reduced. Construction of the FGD system shown in Figure 1 was started in August 1999, and completed in March 2000. Commercial tests were then conducted. The FGD system treats flue gas discharged by three 10 t/h industrial boilers by removing not only SO,, but also particulate matter, HCl, HF, SO,, and trace components in the flue gas. Analysis of the system showed that the circulating slurry pH, the gas flow rate, the circulating slurry flow rate, and the inlet SO, concentration are the main operating parameters. For a pH of 5.5 to 6.5, a gas flow rate of 25,000 m3/h (all gas flow rates are given at 273 K and 1 atm), a circulating slurry flow rate of 195 to 240 m3/h, and inlet SO, concentrations from 500 ppm to 2,000 ppm, the absorption tower desulfurization efficiency varied from 90 to 96%, and the dust removal efficiency was greater than 95%. For a slurry pH in the oxidation tank of 3.5 to 6.5, the calcium utilization efficiency was 90 to 95%, and the fraction of S(IV) oxidized to S(VI) was 97% to 99%. Therefore, the system had high desulfurization efficiencies with variable inlet SO, concentrations and gas flow rates, a high level of automation, high reliability, and simple operation and maintenance. This makes it widely applicable in industry, especially for small- and medium-size industrial boilers in China. July2002 131

SO,,- + 1/2 0, -+ so4,-(partial)

Ca(OH), -+ Ca2++ 2 OH- (partial) Ca2++ SO,*- -+ CaSO, (partial) Ca2++ SO4*-+ 2 H,O -+ CaS04.2H,0 (partial) H+ + OH- + H,O In the lower part of the absorption tower, the slurry coming down from the upper section reacts with fresh slurry, and the S(IV) is oxidized to S(V1). The main reactions taking place in the circulating oxidation region are: HSO, + 1/2 0, -+ HSO, -+ H+ + SO,*-

so32-+ 1/2 0, -+ so4,Ca2++ so4,-+ 2 H,O

Figure 1. Slurry jet FGD project in Shenyang, China. OPERATION OF THE SLURRY JET FGD

As shown in Figure 2, the absorption tower can be divided into two regions: the upper section, which is the desulfurization region where the circulating slurry mixes with the flue gas to absorb the SO,, and the lower section, which is the circulating oxidation region. In the lower section, fresh slurry mixes and reacts with the reacted slurry so that the CaSO,, produced in the desulfurization reaction, is oxidized to C a S 0 4 . This c o m p o u n d t h e n crystallizes t o form CaS04.2H,O. Flue gas enters the absorption tower tangentially at the bottom of the upper section of the tower, producing a swirling flow in the reaction region. Specially designed jet nozzles (Figure 3) are arranged beneath the flue gas inlet, since the slurry jet pattern significantly affects the removal efficiency [71. The slurry is sprayed upward to form slurry columns that reach a peak, and then disperse a n d fall down. During the process, the flue gas mixes well with the circulating slurry so that SO, in t h e flue g a s reacts with t h e absorbent, and is then removed. The wet chemistry in all FGD plants is similar, irrespective of the liquid-gas contact mode [Sl. The main reactions taking place in the desulfurization region are: SO, (8) + H,O -+' SO, (1) + H,O SO, (1) + H 2 0 + H+ + HSOj+ 2 H+ + SO:-

HSOj t 1/2 0,-+ HSOi (partial) 132 July 2002

+ CdSO4.2H,O

In addition, because the flue gas enters the absorption tower tangentially, part of the particulates move towards the wall due to the centrifugal force and are then removed by the water film. Mist eliminators are also used to remove some tiny dust particles, which further increase the dust removal efficiency. PROCESS

The FGD system receives exhaust gas from the boilers as shown in Figure 2 . The main design a n d economic parameters for the system are listed in Table 1 . Carbide sludge, a waste residue from the nearby Shenyang Chemical Factory, was used to replace lime as the desulfurization absorbent as a way of recycling the waste, and also to reduce operating costs. The carbide sludge analysis is given in Table 2. Fresh slurry was prepared in a pulping tank and then pumped to the FGD scrubber. Dewatered gypsum, the byproduct of the desulfurization process, was used as a soil additive in Kangping County near Shenyang. Reliable controls and instrumentation were critical f o r satisfactory operation. The project used a PCbased data control system, with local PLC nodes collecting and controlling signals. Some of the major control loops are shown in Figure 2. TEST RESULTS AND ANALYSIS

The circulating slurry pH, gas flow rate, circulating slurry flow rate, a n d inlet SO, concentration were selected as experimental parameters for the tests. These parameters could be easily adjusted in the system to achieve the desired operating conditions, and all were recorded once the system reached steady state.

Influence of the pH Absorption at low pH eliminates sulfite ions that produce scaling and plugging, resulting in lower operating a n d maintenance costs [91. T h e system w a s Environmental Progress (V01.21, No.2)

Boiler

++= Legend:

Valve

a

PH

-@

Gas flowrate

-@

slurq flow rate

a

+)

Pressure SO^ concentration

-0 Temperature Slurry supply

Water supply

Figure 2. Sketch of the slurry jet FGD system. Table 1. System design and economic parameters. ._.

__._

_ _

.

-

___ -

-

___.

_

.

_ - .

.-

__

Liquid-gas ratio .. PH of the circu.&in_g slurry __ Inlet flue gas temperature __ Outlet fluegaskmpsrature . Absorbent Absorbent purity __ Quantity or lime supply Multi-cyclone dust removal effj@enqDust removal efficiency Capital investment. . . Water consumption. Electricity consu-mEtion Operating cost . _ - - _. . ._ Unit cost -

-

--_

lm __ ..6m - 1.7% - __ 1,120 ppm

.

-

.

__

__

_

-

-

--

-

-

--

_

160°C_

. -.

. .

--

>93%_ -

-

-

.

_-

_ .

.

.

-

-

- 80%

-

.

__

Carbide sludge Approx. 90% is Ca(OH), 400 E&li 00 % %-slujc_@)

__ __

.

_.

_

.

- - ...-

.

_

.

10_um3_ _ _ _ _

_-5.5 . -

. -_

GO0 C

. _

._ -

. -.

-

_

_ .

_ _

_

-

_ _

-

-

23 400 m3/h --_ > 80% _ - ____

--

-.

-

. --

-

.

-

_ _ ____--

._. ._

Gas flow ?ate- . SO, removal efficiency

_ - _

_-

-

lot!”

-

Steam output of each of thethree chain stokers __ Radius of &scrubber - -___ _._ -. . ._ .__Height of the absoqtion-section - ___ - ___ - Sulfur content-on-the basis of-dyair - __ . - SO, concentration of flue gas

-

.

-

.

.

-

-

-

-

.

.

2,8!q,OO9_wB: 14,000 t& -525 000 Kw.h/yr . 300 _-L-000 RMB*/y_r 0.59 RMB*/Kg SO, absorbed 7

__ __

-

__

_.

* 1 US Dollar = 8.28 RMB

Environmental Progress (V01.21, No.2)

July2002 133

J RoW rate: 16500'M

-+Gas

,

65!

,

2

, 3

. , . , 4

SiO, 2.01 Na,O 0.49

1

1 I

A1,03 1.78 TiO, 0.01

I

1

I

I I

I I

Fe,03 0.26 MnO 0.01

1

CaO

,

MgO K,O 0.14 0.15 P,05 , Ignition loss I 0.10 j 28.88

1 65.79 '

, . ,

6

7

.

, 6

Figure 4. Influence of pH and gas flow rate on desulfurization efficiency. a. Circulating slurry flow rate: 240 mj/h; inlet SO, concentration: 500 ppm.

I

,

designed to operate at a pH of 5.5. During the tests, desulfurization efficiency variation with the circulating slurry pH had good repeatability for different liquidgas ratios and inlet SO, concentrations as shown in Figures 4a, b, and c. For a gas flow rate of 24,000 m3/h, a n inlet SO, .'. .concentration at 1,000 ppm, and pH values from 2 to 8,' the desulfurization efficiency increased from 78% to 97%, b e c a u s e a higher pH m e a n s m o r e alkali absorbent in the slurry and, hence, enhanced desulfurization ability. When the pH was too high, the SO, absorption was enhanced, but the concentrations of both SO3,- and C03,- increased, so that CaCOj and CaS03 easily crystallized o n the absorbent surface, which prevented further absorption and reaction with the absorbent [lo]. However, when the pH was too low, the calcium absorbent utilization efficiency was very high so that if CaCO, is used as the absorbent, absorbent a n d energy d e m a n d s a r e reduced, but desulfurization efficiency is low. At normal operating conditions, the pH is controlled at about 5.5 and the desulfurization efficiency remains above 95% for lower inlet SO, concentrations (500 ppm), and is 90% to 95% for relatively high inlet SO, concentrations (1,000 ppm and 2,000 ppm).

J --t Gas flow rate 16500 m'h

-L Gas flow rate 24000 m 'h

b. Circulating slurry flow rate: 240 m3/h; inlet SO, concentration: 1,000 ppm.

100,

Influence of the Gas Flow Rate and the Circulating Slurry Flow Rate The gas flow rate determined the gas velocity in the scrubber while the circulating slurry flow rate determined the height of the absorption section. Thus these two parameters determined the liquid-gas ratio and the gas residence time in the FGD process. In a spray tower, the residence time and the liquidgas ratio positively influence the desulfurization. However, when the gas velocity increases, although the residence time decreases, the liquid-gas ratio in the 134 July 2002

.

PH

Figure 3. Nozzle used in slurry jet FGD tower. Table 2. Carbide sludge composition.

5

& Gas flow rate' 240004h 65!

, . , . , .

2

3

4

5

6

7

8

PH

Circulating slurry flow rate: 240 m3/h; inlet SO, concentration: 2,000 ppm.

C.

Environmental Progress (V01.21, No.2)

100,

-

Circulabng slurry flow rate: 2 4 h tCirculabng stuq flow rate: 2 i h A Circulating slurry flow rate: i&n I

2

.

,

3

.

I

4

.

,

5

.

,

6

v

1

7

.

1

-

8

80 250

500

750

1000

1250

1500

1750

2000

2250

Inlet Sq concentration (ppm)

PH

Figure 5. Influence of the circulating slurry flow rate on desulfurization efficiency. (Gas flow rate: 16,400 m3/h; inlet SO, concentration: 1,000 ppm.)

Figure 6 . Influence of inlet SO, concentration on desulfurization efficiency. (Gas flow rate: 16,400 m3/h; circulating slurry flow rate: 240 m3/h.)

absorber can be reduced to achieve a constant SO, removal efficiency ill].Thus, the gas velocity and residence time both positively influence desulfurization and there is an optimal set of conditions. In the ABB design of the LS-2 FGD system, the relative flue gas velocity was designed to be as high as 4.5 m / s 141, but the high velocity increases the risk of re-entrainment. With the slurry jet technique, breakup and coalescence of droplets occurs throughout the falling process, which enhances the absorption enough to offset the negative influence of the lower velocity, and increases the positive influence of the longer residence time. Moreover, a lower velocity also decreases the load on the mist eliminator, because the reduced gas velocity minimizes plugging [4].During the tests, the gas flow rate was set to -12,000 m3/h, -16,400 m3/h and -24,000 m3/h, which correspond to flue gas velocities of 1.5 m/s, 2.0 m/s, and 3.0 m / s in the FGD scrubber. When the circulating slurry flow rate was 240 m3/h with a 6 m high slurry column, the gas-liquid contact times were 4 s, 3 s, and 2 s respectively. As shown in Figure 4, with the other operating conditions holding constant, the desulfurization efficiency decreased with increasing gas flow rate. The shorter gas-liquid contact time means that less SO, is absorbed by the droplets formed by the slurry jet in the FGD scrubber, which reduces the desulfurization efficiency. Figure 4 also shows that, when the pH of the circulating slurry was lower and the inlet SO, concentration was not high, desulfurization efficiency was quite sensitive to flue gas velocity, because the chemical absorption was not as great as the physical absorption. Longer gas-liquid contact times would then effectively increase the desulfurization efficiency.When the pH was high enough and the inlet SO, concentration increased, the chemical absorption was great enough to eliminate most of the SO, during the initial contact time. Therefore, there would be no obvious effect as the gas-liquid contact time increased. The circulating slurry flow rate is also an important operating parameter of the FGD system which significantly influences the capital investment and the operating cost. In a slurry jet scrubber, when the circulating slurry flow rate increases, the heights of the slurry

columns increase and the gas-liquid contact section becomes longer. Therefore, the desulfurization efficiency rises, but the energy consumption also increases, which increases the capital investment and operating cost, and the burden on the mist eliminator and the induced draft fan. The circulating slurry flow rate was set at 240 m3/h, 220 m3/h and 200 m3/h during the tests, which corresponded to slurry column heights of 6 m, 4.9 m, and 4 m. When the gas flow rate was 16,400 m3/h, the liquid-gas ratios were 14.7 L/m3, 13.3 L/m3, and 11.9 L/m3, respectively. As shown in Figure 5 , a higher liquid-gas ratio leads to higher desulfurization efficiencies when the other operating parameters remained constant.

Environmental Progress (V01.21, No.2)

Infiuence of the Inlet SO, Concentration During the tests, the inlet SO, concentration was adjusted to -500 ppm, -1,000 ppm, and -2,000 ppm. When the inlet SO, concentration was increased and other conditions were held constant, including the pH, liquid-gas ratio and gas flow rate, the desulfurization efficiency decreased linearly, with steeper slopes as the pH was lowered. (Figure 6 ) When the SO, concentration was increased, the SO, partial pressure increased, which led to lower mass transfer resistance in the gas phase and facilitates the chemical reaction. But, total desulfurization efficiency decreased, because the slurry with the same SO, absorption ability reached its absorption limit more quickly at high SO, concentrations. Because the slurry with lower pH has less absorption capability, it is more sensitive to the increased inlet SO, concentration, so desulfurization efficiency decreased much more rapidly. Oxidationof NIV) and Calcium Absorbent Utilization Efficiency Thickened slurry was discharged from the absorption tower to a receiving tank where samples were collected. The slurry contained CaS03, CaS04 and CaC03, formed by the reaction between Ca(OH), and CO,, unreacted Ca(OH),, dust removed from flue gas by the slurry jet, and other substances contained in the calcium absorbents themselves.

July 2002 135

.-

J

tem. The scrubber itself has a dust removal efficiency greater than 95%. Due to the success of the pilot operation, the slurry jet FGD system has been chosen for use in two other plants by Tsinghua Tongfang Energy & Environment Co., Ltd.

.

ACKNOWLEDGMENTS

s o d . , 2

3

.

, .

4

, 5

.

, 6

. 7,

.

I80

e

PH

Figure 7. Influence of pH on oxidation ratio and calcium utilization efficiency. (Gas flow rate: 24,000 m3/h; circulating slurry flow rate: 240 m3/h; inlet SO, concentration: 1,000 ppm.)

Forced oxidation is widely used since it generates larger crystals that are more easily dewatered and less thixotropic than naturally oxidized sludge 1141. Lowering the slurry pH increases the dissolved HSO, concentration and improves the HSO, oxidation rate. When the oxidation air flow rate was 500 m3/h and the other conditions were held constant in these tests, the fraction of oxidized S(IV) stabilized at 97 to 99%, regardless of the pH, as shown in Figure 7. The other conditions had a minor influence on the oxidation of S(1V). The utilization efficiency of calcium absorbent refers to the ratio of the molar content of CaSO, and CaS03 to the total calcium molar content. As shown in Figure 7, increasing the pH from 3 to 7 by adding more calcium absorbent into the desulfurization system resulted in more un-reacted calcium dioxide being discharged, and utilization efficiency decreasing from 95% to 90%. CONCLUSIONS

The desulfurization efficiency was always very high during the tests. At the design condition, when the gas flow rate was 24,000 m3/h, the inlet SO, concentration was 1,000 ppm, the circulating slurry flow rate was 240 m3/h, and the pH was 5.5, the desulfurization efficiency was over 93%. Test results show that the pH of the circulating slurry, the inlet SO, concentration, the liquid-gas ratio and the gas-liquid contact time are the main process parameters. The last two parameters are functions of the gas flow rate and the circulating slurry flow rate. The desulfurization efficiency increases as the pH of the circulating slurry increases, th e liquid-gas ratio increases, the gas-liquid contact time increases, or the inlet SO, concentration decreases. When the slurry pH in the oxidation tank was 3.5 to 6.5, the calcium utilization efficiency was 90% to 95%, and the fraction of oxidized S(IV) was 97% to 99%. In addition to removing SO.,,, the scrub b e r increased the total dust removal efficlency of the sys-

136 July 2002

This research o n the slurry jet FGD system was funded by the Major State Basic Research Program of the People’s Republic of China - Project G19990222. Japan Science & Technology Corporation UST) also funded part of the investigatior . LITERATURE CITED

1. Xu, X., et al.,“Development of Coal Combustion Pollution Control for SO, and NO, in China,” Fuel

Processing Technology, 62, 2-3, p p 153-160, February 2000. 2. Song, F. and H. Ji, “Application Research of Integrated Desulfurization and Dust Removal Technology,” FACT 23, 1, p p 501-504, American Society of Mechanical Engineers, Fuels a n d Combustion Technologies Division, July 1999. 3. “Scrubber Myths and Realities,” Power Eng., 99, 1, p p 35-38, Institute of Clean Air Companies, Inc., January 1995. 4. Klingspor, J.S. and G.E. Bresowar, “Advanced Limestone-Based Wet Flue Gas Desulfurization,” ABB Review, 8, p p 23-27, August 1995. 5. Muramatsu, K., et al., “Development of Mitsubishi Wet Flue Gas Desulfurization System,” Chem. Econ. Eng. Rev., 16, 11, p p 15-22, November 1984. 6. I d e m u r a , H . , “New T h o r o u g h b r e d Flue Gas Desulfurization Plant (Chiyoda Thoroughbred 121 Plant),” Chem. Econ. Eng. Rev., 16, 11, p p 23-27, November 1984. 7. Henry, D.W., et aL , “Scrubber Upgrade Achieves 95% Removal Efficiency,” Power Eng., 95, 3, p p 25-28, March 1991. 8. Kiil, S., et al., “Experimental Investigation and Modeling of a Wet Flue Gas Desulfurization Pilot Plant,” Ind. Eng. Chem. Res., 37, 7, p p 2792-2806, July 1998. 9. Baron, E.S., “Clean Coal Project Nears Commercial O p e ra tio n ,” PowerEng., 99, 2 , p p 31-36, February 1995. 10. “LimeLkestone Scrubbing for SO, Removal,”Air Pollution Control andDesignforIndzcZry, P.N. Cheremisinoff, Editor, Marcel Dekker, New York, W,1993. 11. Lani, B.W. and M. Babu, “Phase 11: The Age of High Velocity Scrubbing,” Proc. Int. Tech. Conf. Coal Util.Fuel Syst., 21, p p 145-155, 1996. 12. S p a r m a n n , A., et al., “Operating Experience with the FGD Wet Scrubbing Process and Gypsum Processing at VEAG’s Janschwalde Power Station,” VGB PowerTech, 79, 1, pp 59-64, January

1999.

Environmental Progress (V01.21, No.2)