Effects of Temperature on Electrostatic Precipitators of Fine Particles and SO3

The fine particles and SO3 at the outlets of electrostatic precipitators (ESPs) were measured when the temperature of the flue gas decreased to 90°C on a pilot-scale platform with 50,000 m h of real flue gas. Based on the test results and scanning electron microscopes (SEMs) of the particulate matter, the mechanism of fine particle change and SO3 removal was analyzed. The results show that the efficiency of SO3 removal reaches 73% and the penetration concentration of PM1 decreases from 0.5 to 0.17 mg m when the gas temperature decreases from 120–130°C to 90–95°C in a coal-fired power plant.


INTRODUCTION
As air pollution has been a serious problem, the ultralow emission in China requires that the particles emissions must be less than 10 mg m -3 (Saeki et al., 2003;Ando et al., 2011;Nicol et al., 2013;Preston et al., 2013;China et al., 2015).
The collection efficiency of electrostatic precipitator (ESP) is over 99.9%. However, the difficulty of charging is one of the major limiting factors for particles collection of ESP (Kato et al., 1994;Yi et al., 2006;Roudier et al., 2013;Afshar-Mohajer et al., 2014;Xu et al., 2015), and the mass concentration at the outlet of ESP was about 50 mg m -3 with high penetration efficiency of dust particles in the size range of 0.1-1 µm (Tamaru et al., 1998;Misaka et al., 2009). Chang et al. (2015) had enhanced particles agglomeration by using a pre-charger to increase mean particle size but without reducing the dust resistivity and taken more power.
There are some ESPs operated at different gas temperatures in coal-fired power plants (Xia et al., 2011;Li et al., 2017), Hitachi Plant Technologies developed a gas treatment system with a low low temperature electrostatic precipitator (LLT-ESP), the moving electrode type ESP with gas temperature of 90°C was described (Misaka et al., 2008). In 2016, the industrial standard of LLT-ESP was implemented in China (JB/T 12591-2016), but there is few study focused on its mechanism. Thus, this paper will describe the major performances of LLT-ESP and the changes to ESP at different gas temperatures by tests.

Testing Apparatus
(1) Pilot-scale Platform The pilot-scale platform ( Fig. 1) was built beside a real coal-fired power plant. The flue gas of the pilot-scale platform was pumped from a 300 MW heat-supply unit in Heibei plant after the economizer, thus ensured that the characters of flue gas in the pilot-scale platform. The flue gas volume is 50,000 m 3 h -1 in 350°C while 32,221 m 3 h -1 in 90°C.
The main parameters of LLT-ESP in the pilot-scale platform is shown in Table 1, which includes of the main components of coal and ash, the inlet flue gas parameters during the texts. The LLT-ESP pilot-plan is shown in Fig. 2. (2) Test ESPs in 300-1000 MW Coal-fired Power Plants The three ESPs were set in Changxing coal-fired plant (Power plant name, the unit capacity is 660 MW), Shangan coal-fired plant (the unit capacity is 660 MW) and Ninghai coal-fired power plant (the unit capacity is 1000 MW) for in-depth study of the performances of LLT-ESP in large units. The main parameters of ESPs and the entrance flue gas parameters are shown in Table 1.

Experiment Devices and Methods
In this paper, to test the dust concentrations, SO3 concentrations and dust practical resistivity of LLT-ESP, GGH means gas-gas heater, SCR means selective catalytic reduction   Table 2. The measurement of dust concentration was on the basis of ISO 12141-2002. The sampling points were set at the inlet and outlet of ESP. The PM was sampled by a normal sample tube, a glass fiber cartridge at the inlet of ESP and by a low concentration sample tube, quartz membranes at the outlet of ESP, respectively. Especially, the gas was sampled with a heating sample tube by isokinetic sampling.
The fiber cartridges or membranes were weighed before and after by an electronic balance in tens of millions.
The measurement method of SO 3 was on the basis of GBT 21508-2008 as shown in Fig. 3. The gaseous SO 3 would become mist SO 3 when the flue gas temperature was reduced under acid dew point by the 60-65°C temperaturecontrolled water bat, and be collected by the snake-like glass tube, depended on the inertial forces. During the test, the constant temperature should not be lower than 60°C to  prevent SO 2 from being condensed. The heating sample tube was set with a quartz filter or a glass fiber to filter the particulate matter and the heating temperature is not lower than 260°C. The content of SO 4 2was tested by titration method with an ultraviolet spectrophotometer.

Dust Removal Efficiency with Different Diameters
The result of the tests done on the LLT-ESP pilot-plan is shown in Fig. 4. It showed that the dust concentration via gas temperature and diameter. The sizes of dust particles between GGH and LLT-ESP were growing from 130 to 90°C and the particles ratio of ≤ 1 µm, ≤ 2.5 µm, ≤ 10 µm decreased by 65%, 50% and 35% respectively. In general, the dusts whose size was in a range of 0.1-1 µm were difficult to remove (Zhao et al., 2018). When the temperature reduced, the penetration concentrations of PM1 decreased from 0.5 to 0.17 mg m -3 . As the penetration concentration of PM 1 decreases about 65% when the temperature of flue gas decreases to 90°C according to the tests, about 60% of PM 1 has grown up to big particle which was easy to collect by ESP.
As the same trend, the tests of LLT-ESP in Changxing plant (600 MW) showed that the penetration efficiency of PM 1 decreased 65% as the gas temperature decreased from 120 to 90°C as shown in Fig. 5. And the total particles concentration (TPS) were less than 11.5 mg m -3 at the outlet of LLT-ESP, contrasted with 19 mg m -3 at the outlet of normal ESP, accorded by the test of Shangan plant.
The LLT-ESP which reduced the gas temperature to near 95°C made a good effect on the performance of ESP and effectively solved the problem of low removal efficiency of fine particles. Moreover, the WFGD would show an  apparent increase of dust removal efficiency as the particles diameters increased at the inlet of WFGD when took a LLT-ESP (Ondov et al., 1979).

SO3 Removal Efficiency
The tests done on the LLT-ESP pilot-plan showed that the removal efficiency of SO 3 was 22.84%, 96.15% and 96.61% when the gas temperature was 120°C, 90°C and 80°C, respectively. Furthermore, the tests done on the LLT-ESP in Changxing plant showed that the removal efficiency of SO 3 achieved 73% by GGH and LLT-ESP, and the SO 3 concentration was less than 6 mg m -3 , as shown in Fig. 6, while the penetration concentration of PM decreased from 28 to 18 mg m -3 . SO 3 removal efficiency is increased with LLT-ESP.

Mechanism Analysis of Fine Particles Change and SO3 Removal
To analyze the mechanism of fine particles change and SO 3 removal, membranes was set at the inlet of ESP in the pilot-scale platform, and particulate matter was observed by scanning electron microscope (SEM) as shown in Fig. 7.
Contrast of 120°C, the particulate matters in the gas temperature of 90°C got together, and the fine particulate adhered to coarse particulate.
As the flue gas before ESP includes relatively much PM than SO 3 , the gas SO 3 condensed into SO 3 mist and was condensed on PM when the flue gas temperature was reduced below acid dew point. According the analysis of those SEM images, the rule of condensation, adsorption, the model of fine particles change and SO 3 removal was put forward as shown in Fig. 8, which supports the test results of dust particles grow when the flue gas decreases to 90°C in Fig. 4. At the same time, the SO 3 would be collected together with PM and solved the high penetration rate of SO 3 of normal ESP in 120-130°C.

CONCLUSIONS
When the gas temperature in a coal-fired power plant decreased from 120-130°C to 90-95°C, the particulate matter aggregated and adhered to coarse particles, and the H2O SO3 H2SO4 mist

Fine particle
Coarse particle Gas temperature is reduced SO3 mist is condensed on particle Water film forms around coarse particle.
Water film forms around fine particle. gaseous SO 3 condensed into SO 3 mist and on the PM. The problem of charging fine dust particles was negated by reducing the flue gas temperature below the acid dew point, as most of the PM 1 developed into large particles, which were easily collected by the ESP. More than 70% of the SO 3 was removed with the PM, and the penetration concentration