Influence of SO3 on mercury removal with activated carbon: Full-scale results

Activated carbon injection is considered one of the most cost-effective options for mercury control at PRB-fired power plants. However, roughly 30% of sites firing PRB coal use SO₃ for flue gas conditioning. The presence of SO₃ in flue gas can decrease mercury capture by activated carbon, sometimes dramatically. Overcoming activated carbon performance limitations caused by SO₃ conditioning for units with this configuration is essential to enable these plants to cost-effectively meet pending mercury emission regulations. Ameren's Labadie Unit 2 fires PRB coal and uses SO₃ to enhance particulate capture in the electrostatic precipitator (ESP). Full-scale sorbent injection tests at Labadie were conducted from 2005–2007. Six sorbents were tested at SO₃ injection concentrations ranging from 0 to 10.7 ppm. Sorbent performance was evaluated at two injection locations (the air preheater (APH) inlet and outlet). Native mercury capture on fly ash was typically less than 15%. When the mercury sorbents were injected downstream of the air preheater, the SO₃ concentration resulted in a decrease in mercury capture from 85% (no SO₃ injection) to 17% (SO₃ injection set at 10.7 ppm). Mercury sorbents were more effective when injected upstream of the air preheater. With the SO₃ system off, mercury removal increased from 75% when injecting 5.1 lb/MMacf of brominated carbon at the APH outlet, compared to 95% when injecting at the inlet. With the SO₃ system on, test results indicated an increase from about 30% injecting at the outlet to 58% injecting at the inlet. Tests evaluating the ADA-ES patented onsite milling process showed that 85% mercury capture was achieved injecting 4 lb/MMacf of milled activated carbon compared to a requirement of 10 lb/MMacf to achieve the same removal using as-received carbon, representing a 60% reduction in activated carbon consumption. No changes in opacity, APH and ESP performance, or other balance-of-plant effects were observed in these tests.     

SO2 and NO selective adsorption properties of coal-based activated carbons

The aim of this paper is to study binary gas adsorption on the activated carbon in the fixed-bed reactor. Coal-based granular activated carbons can selectively adsorb SO₂ and NO. Physically adsorbed NO is replaced and desorbed by SO_2. Chemically adsorbed NO can promote the absorption of SO₂. The presence of SO₂ and NO can enhance the chemical adsorption of NO and SO₂, respectively. When the diameter of granular activated carbon decreases and the specific surface area increases, both the penetration time of the activated carbon bed and SO₂ removal efficiency increase. The whole removal efficiency of SO₂ is more than 99% in the penetration time, but the whole removal efficiency of NO is only 55% in the coexistence of SO₂ and NO. SO₂ adsorption capacity of HNO₃ dipped granular activated carbon is higher than that of non-treated one. The two experimental results are agree with each other.     

SO2 retention on CaO/activated carbon sorbents. Part II: Effect of the activated carbon support

The present paper analyses the role of the activated carbon (AC) properties on the SO₂ uptake capacity of CaO/AC sorbents prepared by AC impregnation or ionic exchange with calcium acetate water solutions. Gas adsorption and mercury porosimetry have been used for textural characterization of the AC and surface oxygen groups have been characterized by temperature programmed desorption (TPD). Thermogravimetry has been used for SO₂ retention tests and CO₂ chemisorption at 300 °C for CaO dispersion (d) determinations. The results show that the surface calcium on CaO/AC samples, determined as “Ca loading · CaO dispersion” (parameter Ca(%) · d), governs the SO₂ uptake. The surface oxygen content is the AC property that mainly controls both the calcium loading and surface calcium on CaO/AC samples, which could be explained by the fact that the surface oxygen lowers the hydrophobic character of the AC supports therefore favouring the interaction with the calcium acetate water solutions. The combination of high calcium loading and dispersion leads to SO₂ uptakes up to 123 mg SO₂/g. The textural properties of the supports have some influence in the calcium loading. However, the effect is masked by the blockage of AC porosity by the calcium loaded.     

Impact of calcium chloride addition on mercury transformations and control in coal flue gas

Pilot-scale experiments were conducted to investigate mercury transformations in coal flue gas when firing subbituminous coal with a CaCl₂ additive. Cofiring the CaCl₂ additive with the subbituminous coal resulted in approximately 50% oxidized mercury, as a result of reactive chlorine species formed in coal flue gas, compared to the dominance of elemental mercury in the baseline flue gas. The mercury data indicate that mercury-flue gas chemistry reactions may occur at fairly high temperatures (>400 °C) in chlorine-enriched flue gas. Field tests were conducted to further demonstrate the impact of cofiring CaCl₂ on the eventual fate of mercury. These tests were completed on a 650-MW subbituminous coal-fired power plant equipped with selective catalytic reduction (SCR), a fabric filter (FF), and a wet scrubber. Overall mercury removals across the SCR-FF-wet scrubber system ranged from 75% to 96% with 200-800 ppm (coal basis) chlorine addition compared to 18-32% during baseline operations. Field data indicate that the SCR enhanced mercury oxidation, possibly as a result of the supplemental formation of reactive chlorine species and the aid of the SCR catalyst. As a result, most of the mercury in the flue gas was in an oxidized state and was removed in the downstream wet scrubber, indicating that cofiring CaCl₂ is an effective mercury control approach for a subbituminous coal-fired plant equipped with an SCR and wet scrubber.     

Simultaneous reduction of NOx and SO2 emissions from coal combustion by calcium magnesium acetate

The potential of calcium magnesium acetate (CMA) as a medium for the simultaneous control of NO_x and SO_x emissions has been investigated using a pulverized coal combustion rig operating at 80 kW. A US and a UK coal of significantly different sulphur contents were used as primary fuel and CMA was injected in solution form into the combustion gases by horizontally opposed twin-fluid atomisers at temperatures of 1100-1200 °C. SO₂ reductions typically greater than 80 and 70% were found for initial SO₂ levels of 1000 and 1500 ppm, respectively, at Ca/S ratios greater than 2.5. There did not appear to be significant limitation on sulphation by pore blockage using CMA due to the open structure formed during calcination and there is clear potential for zero SO₂ emissions at higher Ca/S ratios. The Ca content of the CMA in the form of CaO, via a droplet drying/particle calcination process, absorbs SO₂ by sulphation processes by penetration into the open pore structure of these particles. The effect of primary zone stoichiometry (λ₁=1.05, 1.15 and 1.4) on NO_x reduction was investigated for a range of CMA feed rates up to a coal equivalent of 24% of the total thermal input. NO_x reductions of 80, 50 and 30% were achieved at a primary zone stoichiometry of λ₁=1.05, 1.15 and 1.4, respectively, for a reburn zone residence time of 0.8 s. At lower equivalent reburn fuel fractions, coal gave greater NO_x reductions than CMA but similar levels were achieved above Rff=18%. The mechanism for NO_x reduction involves the organic fraction of CMA which pyrolyses into hydrocarbon fragments (CH_i), but to a lesser degree than coal, which may then react with NO_x in a manner similar to a conventional ‘reburn’ mechanism where NO_x is partly converted to N₂ depending on the availability of oxygen.     

Adsorption of CO2 and SO2 on activated carbons with a wide range of micropore size distribution

The adsorption isotherms of N₂ at 77K, CO₂ at 251, 273 and 298K, and SO₂ at 262 and 273K have been determined on a series of physically activated carbons with a wide range of micropore size distributions. Since the series includes carbons with very high burn-off, it shows the problems involved in the characterization of microporsity in superactivated carbons. On the other hand, the results show that the carbon surface-adsorbate interactions for SO₂ at low relative pressures are weaker than for N₂ and CO₂, as a result of the strong adsorptive-adsorptive interactions in the bulk gas phase.     

Full-scale measurements of SO2 gas phase concentrations and slurry compositions in a wet flue gas desulphurisation spray absorber

Two measurement campaigns were carried out at ENERGI E2's Asnæs Power plant, unit 5. The unit has a capacity of 620 MW_e and is equipped with a wet flue gas desulphurisation (FGD) plant employing a counter-current spray absorber with five spray levels. In the first campaign, the power plant was firing Orimulsion^® with 2.85 wt% S resulting in a flue gas concentration of SO₂ exceeding 2000 ppmv. In the second campaign, the fuel applied was a low-S blended coal and the SO₂ concentration in the raw gas was around 400 ppmv. A novel probe for in situ sampling of gas phase concentrations in wet FGD spray absorbers was developed and applied for measuring axial profiles of the SO₂ gas phase concentrations in the absorber. The expected decrease in SO₂ concentrations along the height of the absorber was found in the spray section (from height 26.5 to 36.2 m) whereas the SO₂ concentration above the holding tank and below the gas inlet was quite low probably due to long local residence times in the region. Horizontal variations, due to somewhat different flow conditions near the column wall were investigated and the SO₂ concentrations were found to be higher near the wall. Measurements at different gross loads showed that the SO₂ gas phase concentration at a given position inside the absorber was roughly linearly related to the L/G ratio in the measuring interval. Turning off one of the lower spray levels, while burning coal with low S content, did not lower the overall removal efficiency of the absorber. However, the SO₂ gas phase concentration inside the lower part of the absorber was increased by a factor of 2-3. Measurements of slurry pH at different positions showed a decrease of approximately 0.5 units from the upper to the lower part of the absorber. The full-scale measurements provide a detailed set of experimental data for validation of mathematical models of a wet FGD spray absorber.     

SO2 and NO removal from flue gas over V2O5/AC at lower temperatures — role of V2O5 on SO2 removal

Supporting V₂O₅ onto an activated coke (AC) has been reported to significantly increase the AC's activity in simultaneous SO₂ and NO removal from flue gas. To understand the role of V₂O₅ on SO₂ removal, V₂O₅/AC is studied through SO₂ removal reaction, surface analysis, X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS) and Fourier transform infrared (FTIR) techniques. It is found that the main role of V₂O₅ in SO₂ removal over V₂O₅/AC is to catalyze SO₂ oxidation through a VOSO₄-like intermediate species, which reacts with O₂ to form SO₃ and V₂O₅. The SO₃ formed transfers from the V sites to AC sites and then reacts with H₂O to form H₂SO₄. At low V₂O₅ loadings, a V atom is able to catalyze as many as 8 SO₂ molecules to SO₃. At high V₂O₅ loadings, however, the number of SO₂ molecules catalyzed by a V atom is much less, due possibly to excessive amounts of V₂O₅ sites in comparison to the pores available for SO₃ and H₂SO₄ storage.     

Elemental mercury vapor capture by powdered activated carbon in a fluidized bed reactor

A bubbling fluidized bed of inert material was used to increase the activated carbon residence time in the reaction zone and to improve its performance for mercury vapor capture. Elemental mercury capture experiments were conducted at 100 °C in a purposely designed 65 mm ID lab-scale pyrex reactor, that could be operated both in the fluidized bed and in the entrained bed configurations. Commercial powdered activated carbon was pneumatically injected in the reactor and mercury concentration at the outlet was monitored continuously. Experiments were carried out at different inert particle sizes, bed masses, fluidization velocities and carbon feed rates. Experimental results showed that the presence of a bubbling fluidized bed led to an increase of the mercury capture efficiency and, in turn, of the activated carbon utilization. This was explained by the enhanced activated carbon loading and gas-solid contact time that establishes in the reaction zone, because of the large surface area available for activated carbon adhesion/deposition in the fluidized bed. Transient mercury concentration profiles at the bed outlet during the runs were used to discriminate between the controlling phenomena in the process. Experimental data have been analyzed in the light of a phenomenological framework that takes into account the presence of both free and adhered carbon in the reactor as well as mercury saturation of the adsorbent.     

Novel promoting effect of SO2 on the selective catalytic reduction of NOx by ammonia over Co3O4 catalyst

Spinel nano-Co₃O₄ was prepared by solid-state reaction at room temperature and investigated for selective catalytic reduction of NO_x by NH₃ (NH₃-SCR). Although suffering from pore filling and plugging, treatment of this catalyst by SO₂ showed novel promoting effect on NH₃-SCR above 250 °C. Bulk cobalt sulfate was observed over the sulfated Co₃O₄ with XRD, which would be an active component for NH₃-SCR. The sulphated Co₃O₄ catalyst exhibited good resistance to SO₂ (500 ppm, 100 ppm) and 10% H₂O at a space velocity of about 25 000 h⁻¹ at 300 °C, as tested for 12 h.     

Influence of low concentration of SO2 for selective reduction of NO by C3H8 in lean-exhaust conditions on the activity of Ag/Al2O3 catalyst

The effect of SO₂ for the selective reduction of NO by C₃H₈ on Ag/Al₂O₃ was investigated in the presence of excess oxygen and water vapor. The NO_x conversion decreased permanently even in the presence of a low concentration of SO₂ (0.5–10 ppm) at <773 K. The increase in SO₂ concentration resulted in a large decrease in NO_x conversion at 773 K. However, when the reaction temperature was more than 823 K, the activity of Ag/Al₂O₃ remained constant even in the presence of 10 ppm of SO₂. The sulfate species formed on the used Ag/Al₂O₃ were characterized by a temperature programmed desorption method. The sulfated species formed on silver should mainly decrease the deNO_x activity on the Ag/Al₂O₃. The sulfated Ag/Al₂O₃ was appreciably regenerated by thermal treatment in the deNO_x feed at 873 K. The moderate activity remains at 773 K in the presence of 1 ppm SO₂ for long time by the heat treatment at every 20 h intervals.     

Removal of elemental mercury from coal combustion flue gas by chloride-impregnated activated carbon

In this work, adsorption of vapour-phase elemental mercury (Hg⁰) from pulverised-coal combustion flue gas by commercially available granular activated carbons treated with zinc chloride (ZnCl₂) impregnation was investigated. The experiment results showed that ZnCl₂ impregnation significantly enhanced the adsorptive capacity for mercury vapour, but decreased the specific surface area of the activated carbon. This could be explained by the occurrence of chemisorption, which was confirmed by adsorption tests over a wide range of temperatures. The influence of ZnCl₂ solution concentration on the mercury removal performance was also studied. Mechanisms of mercury adsorption onto the Cl-impregnated activated carbon were proposed.     

High rate performance of highly dispersed C60 on activated carbon capacitor

Fullerene-activated carbon composite electrodes were prepared and their charge/discharge characteristics were studied for use in a high power electric double-layer capacitor. The capacitance of the C₆₀-loaded activated carbon fiber (ACF) electrodes became greater than that of the unloaded ACF at charge/discharge current densities above 50 mA/cm². In order to obtain a highly dispersed C₆₀-loaded electrode, an ultrasonic treatment was performed. The size of the C₆₀ agglomerate decreased from 1-2 to 0.1 μm or less, and the capacitance of the C₆₀-loaded ACF electrodes increased with an increase in the ultrasonic treatment time. A higher capacitance of 172 F/g was obtained at 50 mA/cm² on a 1 wt% C₆₀-loaded electrode with ultrasonic treatment, and the C₆₀-loaded ACF electrode also showed a higher cycle performance.     

Influence of preparation conditions on porous structures of olive stone activated by H3PO4

An olive factory residue was used as a precursor in the preparation of granular activated carbon by chemical activation with H₃PO₄. Effects of final activation temperature, time, and H₃PO₄ concentration used in the impregnation stage on the porous development were investigated. SO₂ adsorption experiments were also performed for some of the activated carbon samples to represent their adsorption performance. Activation at low temperature represented that micropores were developed first at early stages of the temperatures. Mesoporosity developed at around 250 °C, enhanced up to 400 °C, and then started to decrease due to possibly shrinking of pores. The optimum temperature for olive stone was found to be around 400 °C on the basis of total pore volume and BET surface area. It was clearly demonstrated that H₃PO₄ concentration used in the impregnation stage was not only effective for development of surface area and pore volumes but also an effective tool for tailoring the pore structure and size distribution.     

Impact of SO2 concentration on the corrosion rate of X70 steel and iron in water-saturated supercritical CO2 mixed with SO2

The corrosion behavior of X70 steel and iron in water-saturated supercritical CO₂ mixed with SO₂ was investigated using weight-loss measurements. As a comparison, the instantaneous corrosion rate in the early stages for iron in the same corrosion environment was measured by resistance relaxation method. Surface analyzes using SEM/EDS, XRD and XPS were applied to study the morphology and chemical composition of the corroded sample surface. Weight-loss method results showed that the corrosion rate of X70 steel samples increased with SO₂ concentration, while the corrosion rate increased before decreasing with SO₂ concentration for iron sample. Comparing resistance relaxation method results with weight-loss method results, it is found that the instantaneous corrosion rate of iron is much higher than the uniform corrosion rate of the iron tablet specimens which are covered with thick corrosion product films after a long period of corrosion. The corrosion product films were mainly composed of FeSO₄ and FeSO₃ hydrates. The possible reaction mechanism under such environment was also analyzed, and the electrochemical reaction between the dissolved SO₂ in the condensed water film with iron is the critical reaction step.     

Influence of the clinker SO3 on the cement characteristics

This paper aims to clarify the influence of the clinker SO₃ on the cement characteristics. The impact on the strength development rate and the level of sulfate resistance were studied .The results show that increasing the amount of clinker SO₃ at low alkali level reduces the percentages of the tricalcium aluminate (C₃A) and alite as well as the alite/belite ratio, leading to a modification in the cement quality.For these reasons cements produced from a clinker containing high sulfate and low alkali, have slower strength development and higher sulfate resisting level than that produced with low sulfate clinker.     

Low cost coal-based carbons for combined SO2 and NO removal from exhaust gas

The aim of this paper is to show how a cheap carbonaceous material such as low rank coal-based carbon (or char) can be used in the combined SO₂/NO removal from exhaust gas at the linear gas velocity used in commercial systems (0.12 m s⁻¹). Char is produced from carbonization and optionally activated with steam. This char is used in a first step to abate the SO₂ concentration at the following conditions: 100 °C, space velocity of 3600 h⁻¹, 6% O₂, 10% H₂O, 1000 ppmv SO₂, 1000 ppmv NO and N₂ as remainder. In a second step, when the SO₂ concentration in the flue gas is low, NO is reduced to N₂ and steam at the following experimental conditions: 150 °C, space velocity of 900 h⁻¹, 6% O₂, 10% H₂O, 0-500 ppmv SO₂, 1000 ppmv NO, 1000 ppmv NH₃ and N₂ as remainder.It has been shown that the presence of NO has no effect on SO₂ abatement during the first step of combined SO₂/NO removal system and that low SO₂ inlet concentration has a negligible effect on NO reduction in the second step. Moreover, this char can be thermally regenerated after use for various cycles without loss of activity. On the other hand, this regenerated char shows the highest NO removal activity (compared to parent chars, either carbonized or steam activated) which can be attributed to the activating effect of the sulfuric acid formed during the first step of the combined SO₂/NO removal system.     

Adsorption of gaseous SO2 and structural changes of montmorillonite

Several montmorillonite samples after adsorption of gaseous SO₂ were analyzed to evaluate structural and textural changes. The equilibrium adsorption of the SO₂ gas was measured at 25 °C and 0.1 MPa. The samples were characterized by X-ray diffraction (XRD), infrared spectroscopy (IR), swelling index (SI), pH measurements, and N₂ adsorption–desorption isotherms. SO₂ adsorption increased with the specific surface area of montmorillonite. SO₂ retention decreased pH of the dispersed samples from 6 to 1 and released interlayer and octahedral cations from the structure, which increased the specific BET surface area and specific micropore surface similar to that of acid-activated montmorillonite.     

Effect of SO2 on the catalytic activity of Ga2O3–Al2O3 for the selective reduction of NO with propene in the presence of oxygen

The effect of coexisting SO₂ on the catalytic activity of Ga₂O₃–Al₂O₃ prepared by impregnation, coprecipitation and sol–gel method for NO reduction by propene in the presence of oxygen was studied. Although the activity of Al₂O₃ and Ga₂O₃–Al₂O₃ prepared by impregnation (Ga₂O₃/Al₂O₃(I)) and coprecipitation (Ga₂O₃–Al₂O₃(CP)) was depressed considerably by the presence of SO₂, NO conversion on Ga₂O₃–Al₂O₃ prepared by sol–gel method (Ga₂O₃–Al₂O₃(S)) was not decreased but increased slightly by SO₂ at temperatures below 723 K. From catalyst characterization, SO₂ treatment was found to cause two important effects on the surface properties: one is the creation of Brønsted acid sites on which propene activation is promoted (positive effect), and the other is the poisoning of NO_x adsorption sites on which NO reduction proceeds (negative effect). It was presumed that the influence of SO₂ treatment on the catalytic activity is strongly related to the balance between the negative and positive. The activity enhancement of Ga₂O₃–Al₂O₃(S) by SO₂ was accounted for by the following consideration: (1) increase of the propene activation ability by SO₂, (2) incomplete inhibition of NO_x adsorption sites by SO₂.