Emission Characteristics of Gaseous and Particulate Mercury from a Subcritical Power Plant Co-Firing Coal and Sludge

Abstract

Field tests were carried out in a subcritical coal-fired power plant co-firing coal and sludge to analyze the emission characteristics of gaseous and particulate mercury. EPA30B method was applied to determine the mercury speciation in different positions of the flue gas, including the inlet and outlet of the selective catalytic reduction DeNO_X system (SCR) and electrostatic precipitator (ESP); PM₁₀ (with aerodynamic diameter ≤10 μm) was collected using a cyclone and a Dekati low-pressure impactor (DLPI). Before accessing the SCR, Hg in flue gas from both single coal combustion and co-firing mainly existed as Hg⁰; the higher content of Hg in sludge than coal led to the much higher Hg⁰ concentration for co-firing. The total Hg concentration at not only the SCR inlet and outlet but also the ESP inlet did not change obviously. However, Hg_p concentration at the ESP inlet increased significantly, accompanied by a decrease in Hg⁰. The transformation of Hg⁰ to Hg_p appeared to be more distinct for co-firing. The higher HCl concentration of co-firing derived from the much higher Cl content of sludge than coal, and together with the higher ash content of sludge containing more minerals capable of adsorbing Hg⁰, may lead to the greater transformation from Hg⁰ to Hg²⁺ and Hg_p when co-firing. After the ESP disposal, nearly all Hg_p was removed along with PM₁₀, and most Hg⁰ was also removed. The removal efficiency of mercury after the ESP was 92.12% under coal firing and 92.83% under co-firing conditions, respectively. The slightly higher mercury removal efficiency under co-firing should be attributed to the complete removal of the higher concentration of Hg_p.

1. Introduction

Sludge is a semisolid residue produced in the process of industrial or municipal sewage treatment. Owing to rapid industrialization and urbanization, the amount of wastewater produced in China has increased dramatically during the past few decades [1]. Because of the physical and chemical processes involved in wastewater treatment, heavy metals, trace organic compounds that are difficult to degrade, and potentially pathogenic organisms in wastewater are concentrated in sludge [2]. Therefore, sludge must be carefully treated in an environmentally sound manner. However, a large proportion of sludge in China is treated through improper dumping [3]. Many sewage treatment plants dump sludge in the suburbs, causing serious pollution of soil and water resources. Sludge management, treatment, and disposal is thus a large problem in China [1].

Commonly used sludge disposal methods include combustion, agricultural utilization, sanitary landfilling, and dumping into the sea [2]. The combustion utilization of sludge can also be used as alternative energy or raw materials to reduce production costs. Sludge combustion can not only effectively use a large amount of energy carried by sludge itself and realize the recycling of energy [4], but also reduce the ecological toxicity of sludge [5]. The co-combustion of sludge in coal-fired power plants appears to be one such effective utilization method [6], and more than ten power plants in China, such as Tianjin Yangliuqing and Chongqing Luohuang power plants, have applied coal-fired coupled sludge power generation at present.

It is of great significance to study the migration and transformation law of mercury from coal-fired power plants for the removal of mercury from flue gas, as it pollutes water and soil through dry and wet deposition and forms highly toxic methylmercury through biological methylation reaction, which poses a great threat to human health and the biological environment. In China, coal utilization contributes about 70% of the total primary energy consumption [7], resulting in a large amount of Hg release [8]. There are three forms of mercury in coal-fired flue gas [9]: elemental mercury (Hg⁰), oxidized mercury (Hg²⁺), and particulate mercury (Hg_p). Hg²⁺ and Hg_p can be effectively removed by the existing air pollution control equipment of coal-fired power plants [10,11,12], especially Hg_P removal efficiency by dust removal devices, which is up to 99% [13]. Because of its high vapor pressure, volatility, and insolubility in water, Hg⁰ is difficult to capture by the air pollution control equipment in coal-fired power plants [14], and is then emitted into the atmosphere, causing harm to the environment and humans. The conversion of Hg⁰ to Hg_P is regarded as an effective way to remove mercury [13].

The Hg emission characteristics of sludge and coal co-combustion have been studied in the laboratory. Dajnak et al. [15] studied combustion of coal and coal/dried sewage sludge blend in a 0.5 MW furnace and found that the adsorption of mercury on the ash of the blend was much higher than for the coal. Lopes et al. [16] co-fired dry sewage sludges with coal in a fluidized bed and found that fly ash contributes to the retention of mercury in the particulate phase. Mercury emission and transformation in coal-fired power plants have been widely studied [17,18,19], while the on-site verification and on-site process analysis of coal-fired coupled sludge power generation have not been studied. Therefore, this study measured the emission characteristics of gaseous mercury and particulate mercury from a subcritical coal-fired power plant coupled with sludge. The analysis results are representative of the research on the migration and transformation behavior of mercury emissions of similar units.

2. Experimental

2.1. Operating Conditions and Sample Characterization

Field tests were carried out in a 600 MW subcritical coal-fired power plant. This coal-fired power plant, as shown in Figure 1, consists of an opposed firing boiler, ammonia-spraying selective catalytic reduction DeNO_X system (SCR), two-chamber four-electric-field electrostatic precipitator (ESP), and limestone–gypsum wet flue gas desulfurization system (WFGD) [20]. The boiler has a 262 t/h coal feeding rate and a 2030 t/h evaporation rate. Coal is fed through a positive-pressure direct-firing ball-mill coal pulverizing system with six double inlet and outlet coal pulverizers; each pulverizer corresponds to six cyclone low-NOx burners. The proximate and ultimate properties of coal and sludge samples are presented in

Table 1. Properties of coal and pretreated sludge.

Proximate analysis (wt%, db a)	VM	Ash	FC	Mad
coal	30.31	16.98	52.71	1.09
sludge	40.75	56.12	3.13	0.90
Ultimate analysis (wt%, ad b)	C	H	N	S	O c	Hg (μg/g)	Cl (μg/g)
coal	66.69	4.25	0.76	0.32	10.23	0.07	133
sludge	21.80	3.67	3.41	0.72	14.10	0.96	450

^a dry basis; ^b air-dried basis; ^c by difference; VM: volatile matter; FC: fixed carbon; M_ad: moisture at air-dried basis.

. Moisture content of sewage sludge at the received basis is ~35%. As can be seen in Table 1, the ash content of sludge (56.12%) is much higher than that of coal (16.98%), and the fixed carbon content of sludge (3.13%) is far less than that of the coal (52.71%), while the mercury content of sludge (0.96 μg/g) is much higher than that of coal (0.07 μg/g), along with the chlorine content (450 to 133 μg/g).

During the operating test, the boiler load was stabilized at ~55% (i.e., ~350 MW) of its full load; other operating parameters such as coal feeding rates were stable as well. Mercury was collected and tested at various sites of the APCDs (air pollution control devices) during continuous operation of the boiler [21], including the cross-sectional positions of the four sites, termed the inlet and outlet of the SCR, and the inlet and outlet of the ESP, respectively.

2.2. Co-Firing Procedures and Ultralow Emission of Boiler

The pretreated sludge sample used in the power plant was collected from local sewage disposal plant. The relatively high moisture content of sludge (~35%) was dried using waste heat from the steam turbine and then conveyed to a dry sludge storage bin. as shown in Figure 1. After the drying process, sludge was pressurized and then moved to the conveying belt and thoroughly mixed with coal in the coal pulverizers at a mass ratio of ~5% with a stable feeding rate. Several trials for each test were carried out to obtain repeatable values to avoid the nonuniformity caused when sludge was conveyed to the pulverizer by the conveying belt. Finally, the total mass of sludge combusted during each test period was recorded to make certain the mix ratio of sludge was stabilized at 5%.

The representative emission parameters of the unit during the experimental periods are recorded in

Table 2. The representative emission parameters of the unit during the experimental periods.

	Boiler Load (MW)	NOX (mg/m3)	SO2 (mg/m3)	PM (mg/m3)
coal	373.2	27.6	14	1.17
co-firing	381.2	35.7	15.6	0.95

. As shown in Table 2, the emission of SO₂, NO_x, and PM was lower than the emission limitation of the gas turbine (SO₂ < 35 mg /m³, NO_x < 50 mg /m³, PM < 5 mg /m³), which means that this coal-firing power station unit meets China’s standard for ultraclean emission technology.

2.3. Sampling Procedures and Analytical Method

2.3.1. Mercury Sampling

In coal-fired power plant, mercury in flue gas is mainly generated from the mercury in coal. During combustion, mercury in coal is released into slag and flue gas [22]. After the flue gas passes through the fly-ash removal facilities, most of the Hg_p is transferred to the fly-ash collector ash with the removal of fly ash. The flue gas sampling points are located at the inlet and outlet of the SCR and ESP (Figure 1). The contents of various forms of mercury and HCl in the flue gas were sampled and tested.

The EPA 30B method was applied for the sampling of total gaseous Hg in flue gas through extractive sampling with sorbent traps, which were subsequently analyzed [23]. When the 30B method is used for sampling in the field experiment, the flue gas from the flue enters the negative pressure sampling probe. After passing through the capture device in the probe, the mercury in the flue gas is absorbed by the adsorbent in the adsorption tube and left in the adsorption pipe [17]. The subsequent gas is dehydrated through the dehumidification device equipped with silica gel, and then discharged into the atmosphere through the flowmeter in the sampling pump. The speciation of mercury adsorbed by sorbent traps can be analyzed due to the two forms of adsorption tubes in the 30B method [24,25,26,27]: one is the total mercury tube, and the other is the valence tube. The adsorbent in the total mercury tube is activated carbon, which can adsorb total gas-phase mercury content in flue gas. The valence tube is filled with two-stage potassium chloride and two-stage activated carbon. Potassium chloride can selectively adsorb Hg²⁺ in flue gas, while Hg⁰ is adsorbed by activated carbon; the adsorbent only adsorbs gaseous mercury, thus Hg_p is calculated by the mass of mercury in particulate matter collected in the total mercury tube before adsorbance. The thermostatic sampling tube is internally a glass intubation tube, externally covered with asbestos salt, and encased in a stainless steel sheath. The temperature of the sample tube is maintained at about 120 °C to prevent the condensation of water vapor and adhesion of mercury vapor in the flue gas. Paired sorbent traps were placed in front of the sampling probe so that sorbent traps can contact directly with flue gas. The sampling process at each test site lasted 30 min and the gas flow rate was controlled at 2 L/min, with a sampling gas volume of 60 L. The main component of the thermostatic filter box is the filter used to collect mercury particles. It is mainly composed of a glass filter, quartz filter membrane, and polytetrafluoroethylene filter pad. After collection, the adsorption tube used an RA-915M analyzer manufactured by Lumex Instrument, Russia, to test the content of various forms of mercury. The mercury concentration in the gaseous phase was measured using a 30B total mercury tube, and the Hg²⁺ and Hg⁰ data were measured using a 30B valence tube.

After the flue gas passes through the wet desulfurization system, part of the gaseous mercury is transferred to gypsum and desulfurization wastewater, and the remaining gas-phase mercury is emitted into the atmosphere. The liquid and solid samples are bottom slag, fly ash, gypsum, and desulfurization wastewater. The mass production rates for liquid and solid samples are provided by the power plant.

Because chlorine content in fuels has an important influence on mercury removal [28,29,30], and to study the influence of HCl in mercury removal, HCl concentration in flue gas at each site was also sampled at the same time as mercury sampling. The gas flow rate of the HCl concentration test was controlled by a vacuum pump; HCl concentration test was divided into two steps:

Step 1: A mixed solution of Na₂CO₃ and NaHCO₃ was placed in two absorption bottles to adsorb Cl; by testing Cl in the absorption bottles, HCl + Cl₂ concentration in flue gas was obtained.

Step 2: A saturated salt aqueous solution was used before the mixed solution of Na₂CO₃ and NaHCO₃, by testing Cl in absorption bottles, the Cl₂ concentration in flue gas was determined;

The HCl concentration in the flue gas was calculated by subtracting the chlorine concentration determined in step 2 from the chlorine concentration determined in step 1.

2.3.2. PM₁₀ Sampling

The PM₁₀ sampling system mainly includes a constant velocity sampling nozzle, sampling probe, heating insulation layer, K-type thermocouple, flow controller, cyclone separator, DLPI (Dekati low-pressure impactor), and vacuum pump [31,32,33]. Some sampling pipes in the furnace were directly heated by flue gas to ensure that the flue gas entering the sampling pipe had the same state as the flue gas in the furnace (isothermal, constant velocity, and constant pressure). The sampling pipe outside the furnace adopted two-stage temperature-control heating and insulation to ensure that the flue gas temperature was higher than the dew point temperature of the flue gas after adding dilution gas. The cyclone and DLPI-adopted integral insulation were used to ensure that the internal temperature of the flue gas remained unchanged. The addition of dilution gas can reduce the concentration of particles in flue gas and prolong the sampling time, so as to improve the sampling accuracy and ensure the representativeness of samples. During the experiment, the supporting vacuum pump was connected at the tail of the DLPI, and the outlet pressure was set to 100 mbar and the gas flow was 10 L/min. PM₁₀ was segregated to 13 stages by the DLPI. At least 3 replicates were performed for each test condition. PM₁₀ was collected on aluminum foil and the collected mass was measured using a Sartorius microbalance (accuracy: 0.001 mg) to determine the particle size distributions (PSDs) of PM₁₀.

3. Results and Discussion

3.1. Mercury Concentration and Speciation at the Studied Sampling Sites

Figure 2 shows the mercury concentration at the inlet and outlet of the SCR and ESP under conditions of coal combustion and co-combustion with sludge. It can be seen that before the SCR, Hg in flue gas from coal combustion mainly exists as Hg⁰; it is consistent with the reports by Liu et al. [18] and Li et al. [17]. Hg in sludge was reported to reside as Hg⁰ mainly and in minor amounts as mercuric sulfides and residual Hg [34]. Therefore, the higher content of Hg in sludge than coal, as shown in Table 1, led to the much higher Hg⁰ concentration for co-firing than that for single coal combustion. It is noteworthy that little Hg²⁺ was formed in the SCR range, implying that hardly any Hg⁰ was oxidized by the SCR catalyst. The total Hg concentration at not only the SCR inlet and outlet but also the ESP inlet did not change obviously for both single coal firing and co-firing. However, Hg_P concentration at the ESP inlet increased significantly for both single coal firing and co-firing, accompanied by a decrease in Hg⁰. Liu et al. [18] and Li et al. [17] also found this phenomenon. After the ESP disposal, nearly all Hg_P was removed along with fly ash and most of Hg⁰ was also removed [17,18]. The remaining mercury in flue gas was mainly Hg⁰; according to previous studies, WFGD systems can remove most of the Hg²⁺ but have low treatment ability for Hg⁰ [13]. It is reasonable to assume the remaining Hg concentration at the WFGD outlet is close to the ESP outlet for both coal combustion and co-firing. The remaining Hg concentration in co-firing is higher than coal combustion at ~33%, which indicates the mercury emitted from co-firing is slightly higher than coal-combustion.

Compared with single coal firing and co-firing of coal and sludge, Figure 2 shows that the transformation of Hg⁰ to Hg_p appeared to be more distinct for co-firing in the region between the SCR outlet and ESP inlet. As shown in Figure 3, the HCl concentration in the SCR outlet was tested to be 130.86 mg/m³ for coal firing and 189.87 mg/m³ for co-firing. Meanwhile, the transformation rate for Hg⁰ indicated by the ratio of the sum of Hg²⁺ and Hg_p in the ESP inlet to Hg⁰ in the SCR outlet was 47.14% for coal firing and 70.64% for co-firing. The 45.09% higher HCl concentration of co-firing than coal firing, which should be derived from the much higher Cl content in sludge than coal, as shown in Table 1, may thus lead to the observed 49.85% greater transformation from Hg⁰ to Hg²⁺ and Hg_p. Considering that the concentration of Hg²⁺ in the flue gas did not increase and the concentration of Hg_p increased significantly (Figure 2), the Hg²⁺ oxidized by Hg⁰ assisted by HCl [35] should be adsorbed on the fly ash surface to form Hg_p, because fly ash is reported to have a strong adsorption capacity for Hg²⁺ with the decrease in flue gas temperature between the SCR outlet and ESP inlet [22,36]. Furthermore, the higher ash content of sludge compared to coal (Table 1) means it contains more minerals capable of adsorbing Hg⁰ [15,16], which may also cause more transformation from Hg⁰ to Hg_p for co-firing (Figure 3) at the relatively low temperature range of the ESP inlet [36].

The total Hg removal efficiency was calculated via the mercury concentration at the SCR inlet and ESP outlet, and the removal efficiency after the ESP was 92.12% under the coal firing condition and 92.83% under the co-firing condition. The superior removal performance higher than 92% demonstrates the appreciable synergistic removal effect of the current ultralow emission units in China [18]. The mercury removal efficiency after co-firing, although there is more disposal burden from high content of sludge, is slightly higher than that of coal combustion alone, due to the complete removal of the higher concentration of Hg_p, as shown in Figure 2. As for the ashes collected by the ESP, which contained a higher concentration of Hg, the leachability of mercury in coal fly ash from coal-fired power plants was researched. Mercury in fly ash was identified in an amorphous component [37] and demonstrated that the leaching of Hg in fly ash under natural conditions was negligible [38]; hence, the removal efficiency calculated above is reliable.

3.2. Mercury Mass Balance

In this study, mercury mass balance was used to study the migration and transformation law of mercury in coal-fired power plants. The samples were divided into gas, liquid, and solid. The solid samples include coal, sludge, bottom slag, fly ash, and gypsum. The liquid sample is desulfurization wastewater. Hg concentrations of the liquid and solid samples are shown in

Table 3. Hg concentration in various samples.

Sample (μg/g)	Desulfurization Gypsum	Desulfurization Waste Water	Bottom Slag	Fly Ash	Coal	Sludge
Coal	0.71	1.00	0.002	1.26	0.07	/
Co-firing	0.81	1.57	0.009	0.91	0.07	0.96

. It can be seen that the slag in the emissions from the coal power plant did not enrich mercury. Gypsum, desulfurization wastewater, and fly ash were the main migration directions of mercury in coal. After co-firing, the concentration of Hg²⁺ in desulfurization wastewater increased and Hg_p in fly ash decreased.

Mass balance at the ESP outlet was calculated by Hg input mass (g/h)/Hg output mass (g/h). Mass balance of coal combustion was 97.95% and 82.42% in co-firing; it was acceptable between 70% and 130% [39]. Previous studies had shown that mercury mass balance rate at the ESP and WFGD outlets was within 10% [17]; in this study, it is also discussed in Section 3.1 that the remaining Hg concentration at the WFGD outlet was close to that at the ESP outlet for both coal combustion and co-firing. It is reasonable to assume that the mass balance at the WFGD outlet was also between 70% and 130%, which is acceptable.

3.3. Emission Characteristics of PM₁₀ in the SCR Outlet and the ESP Inlet

As observed in Figure 2, much Hg_p was formed between the SCR outlet and the ESP inlet. The particle size mass distributions (PSDs) of PM₁₀ in the SCR outlet and the ESP inlet are thus presented in Figure 4. Two peaks were found in PSDs for both coal firing and co-firing and for these two sites.

In the SCR outlet, the fine-mode peak appeared near 0.3 μm, and the amount of fine modal particles decreased after co-firing with sludge, due to the condensable Ca and S vapor from coal combustion and fine Si and Al adhering to the surface of molten Fe–Al silicate or Fe–Ca–Si–Al species, forming PM₁₊ [20]. The coarse-mode peak appeared at 5 μm and showed similar concentrations for coal firing and co-firing. It appears that the co-firing of sludge and coal has a certain control effect on fine particles PM₁ before the ESP, and the addition of sludge transfers fine-mode particles to coarse-mode. In the ESP inlet, an unobvious fine-mode peak was found at 0.2–0.3 μm, and the coarse-mode peak was found at 3–5 μm. The concentration of PM₁ from co-firing was similar to that from coal firing, and the concentration of PM₁₀ from co-firing should be slightly lower. The reduction in PM₁₀ can improve the working performance and pollutant removal efficiency of the ESP, providing a superior Hg_p removal effect, as shown in Figure 2.

3.4. Relation of Emission Characteristics of PM₁₀ and Hg_P

The concentration of Hg_p and PM₁₀, including PM₁, PM_2.5, and PM₁₀ in the ESP inlet, are correlated in Figure 5. As observed in Figure 2, the concentration of Hg_p from co-firing was 2.56 times higher than that from coal firing. Figure 5 presents the ratios of Hg_p to PM₁ and shows that the ratio from co-firing was 2.46 times higher than that from coal firing, indicating that the Hg enrichment in fine particles was irrelevant to the concentration of PM₁. In contrast, the ratio of Hg_p to PM_2.5 from co-firing was 3.11 times higher than that from coal firing, and the ratio of Hg_p to PM₁₀ was 3.05 times higher. These results indicate that less PM_2.5 or PM₁₀ contained more Hg_p for co-firing. As discussed in Section 3.1, the high ash content of sludge contains abundant fine minerals capable of adsorbing Hg⁰, which may cause the more obvious transformation from Hg⁰ to Hg_p for co-firing.

4. Conclusions

The emission characteristics of gaseous mercury and particulate mercury were analyzed via field tests in a subcritical coal-fired power plant co-firing coal and sludge. Detailed conclusions follow:

(1)

Before the SCR, Hg in flue gas from both single coal combustion and co-firing mainly existed as Hg⁰, and the higher content of Hg in sludge than coal led to the much higher Hg⁰ concentration for co-firing.

(2)

Hg_p concentration at the ESP inlet increased significantly, accompanied by a decrease in Hg⁰. The higher HCl concentration from co-firing derived from the much higher Cl content of sludge than coal, and the higher ash content of sludge containing more minerals capable of adsorbing Hg⁰, may lead to more transformation from Hg⁰ to Hg²⁺ and Hg_p, when co-firing.

(3)

The removal efficiency of mercury after the ESP disposal was 92.12% under the coal firing condition and 92.83% under co-firing. The slightly higher efficiency after co-firing should be attributed to the complete removal of the higher concentration of Hg_p in the ESP inlet.