A

i Lime Slaking

Particulate Collection Device

Particulate Collection Device

Water Alkali

Recycle Regeneration j

Required Optional

Cleaned Flue Gas

Water

Sorbent

Stack

Disposal

Recycle

Cleaned Flue Gas

Stack

Disposal

Cleaned Flue Gas

Stack

Cleaned Flue Gas

Stack

Cleaned Flue Gas

Heat Recovery

Stack

Disposal

Disposal

FIGURE 6-6. Simplified process schematics for dry-injection SO2 control technologies. (Adapted from Rhudy et al. [13].)

Sorbent Injection Temperature, °F

FIGURE 6-7. SO2 capture regimes for hydrated calcitic lime at a Ca/S molar ratio of 2.0. (From Rhudy, R. et al., Status of Calcium-Based Dry Sorbent Injection SO2 Control, in Proc. of the Tenth Symposium on Flue Gas Desulfurization, November 17-21, 1986, pp. 9-69-9-84.)

Sorbent Injection Temperature, °F

FIGURE 6-7. SO2 capture regimes for hydrated calcitic lime at a Ca/S molar ratio of 2.0. (From Rhudy, R. et al., Status of Calcium-Based Dry Sorbent Injection SO2 Control, in Proc. of the Tenth Symposium on Flue Gas Desulfurization, November 17-21, 1986, pp. 9-69-9-84.)

surface area. At temperatures higher than ~2300°F, dead-burning or sintering is experienced.

When limestone is used as the sorbent, it is rapidly calcined to quicklime when it enters the furnace:

Sulfur dioxide diffuses to the particle surface and heterogeneously reacts with the CaO to form calcium sulfate:

Sulfur trioxide, although present at a significantly lower concentration than SO2, is also captured using calcium-based sorbents:

Approximately 15 to 40% SO2 removal can be achieved using Ca/S in the flue gas molar ratio of 2.0. The optimum temperature for injecting limestone is -1900 to 2100°F.

The calcium sulfate that is formed travels through the rest of the boiler flue gas system and is ultimately collected in the existing particulate control device with the fly ash and unreacted sorbent. Some concerns exist regarding increased tube deposits as a result of injecting solids into the boiler, and the extent of calcium deposition is influenced by overall ash chemistry, ash loading, and boiler system design.

The following overall reactions occur when using hydrated lime as the sorbent:

Approximately 50 to 80% SO2 removal can be achieved using hydrated lime at a Ca/S molar ratio of 2.0. The hydrate is injected at very nearly the same temperature window as limestone, and the optimum range is 2100 to

The FSI process can be applied to boilers burning low- to high-sulfur coals. The factors that affect the efficiency of the FSI system are flue gas humidification (to condition the flue gas to counter degradation that may occur in ESP performance from the addition of significant quantities of fine, high-resistivity sorbent particles), type of sorbent, efficiency of ESP, and temperature and location of the sorbent injection. The process is better suited for large furnaces with lower heat release rates [12]. Systems that use hydrated calcium salts sometimes have problems with scaling; however, this can be prevented by keeping the approach to adiabatic saturation temperature above a minimum threshold.

The FSI system has several advantages [12]. One advantage is simplicity of the process; the dry reagent is injected directly into the flow path of the flue gas in the furnace, and a separate absorption vessel is not required. The injection of lime in a dry form allows for a less complex reagent handling system, which lowers operating labor and maintenance costs and eliminates the problems of plugging, scaling, and corrosion found in slurry handling. Power requirements are lower because less equipment is needed. Steam is not required for reheat, whereas most LSFO systems require some form of reheat to prevent corrosion of downstream equipment. The sludge dewater-ing system is eliminated because the FSI process produces a dry solid, which can be removed by conventional fly ash removal systems.

The FSI process has a few disadvantages when compared to the LSFO process [12]. One major disadvantage is that the process only removes up to 40 and 80% SO2 when using limestone and hydrated lime, respectively, at a Ca/S molar ratio of 2.0, whereas the LSFO process can remove more than 90% SO2 using 1.05 to 1.1 mol CaO per mol SO2 removed. This is further illustrated in Figure 6-8, which shows calcium utilization (defined as the percent SO2 removed divided by the Ca/S ratio) of hydrated lime and limestone at various injection temperatures [13]; hence, more sorbent is needed in the FSI process, and lime, which works better than limestone, is more expensive than limestone. There is a potential for solids deposition and boiler convec-tive pass fouling, which occurs during the humidification step due to the impact of solid droplets on surfaces. Also, there is a potential for corrosion at the point of humidification and in the ESP, downstream ductwork, and

Ca(OH)2(s) + heat CaO(s) + H2O(v) CaO(s) + SO2(g) + 1O2(g) CaSO4(s) CaO(s) + SO3(g) CaSO4(s)

FIGURE 6-8. Calcium utilization as a function of sorbent injection temperature for furnace sorbent injection. (Adapted from Rhudy et al. [13].)

stack. The corrosion at the point of humidification is caused by operating below the acid dewpoint, whereas downstream corrosion is caused by the humidified gas temperature being close to the water saturation temperature. Plugging can also occur, thereby affecting system pressures. The efficiency of an ESP can be reduced by increased particulate loading and changes in the ash resistivity. This can, in turn, lead to the installation of additional partic-ulate collection devices. Sintering of the sorbent is a concern if it is injected at too high of a temperature (e.g., >2300°F for hydrated lime). Multiple injection ports in the furnace wall may be needed to ensure proper mixing and follow boiler load swings and hence shifting temperature zones. Hydration of the free lime in the product may be required. Lime is very reactive when exposed to water and can pose a safety hazard for disposal areas.

Economizer Injection In an economizer injection process (shown in Figure 6-6b), hydrated lime is injected into the flue gas stream near the economizer inlet where the temperature is between 950 and 1050°F. This process is not commercially used at this time but was extensively studied because it was found that the reaction rate and extent of sulfur capture (see Figure 6-7) are comparable to FSI. However, the economizer temperatures are too low for dehydration of the hydrated lime (only about 10% of the hydrated lime forms quicklime), and the hydrate reacts directly with the SO2 to form calcium sulfite:

This process is best suited for older units in need of a retrofit process and can be used for low- to high-sulfur coals. The advantages and disadvantages of this system are similar to the FSI process (but will not be discussed in detail here as this process is not currently being used in the power industry) with the notable exception that no reactive CaO is contained in the waste.

Duct Sorbent Injection: Duct Spray Drying Spray dry scrubbers are the second most widely used method for controlling SO2 emissions in utility coal-fired power plants. Lime is usually the sorbent used in this technology, but sodium carbonate is also used, specifically in the western United States. Spray dryer FGD systems have been installed on over 12,000 MW of total FGD capacity, as shown in Table 6-1, as well as numerous industrial boilers.

The first commercial dry scrubbing system on a coal-fired boiler in the United States was installed in mid-1981 at the Coyote station (jointly owned by Montana-Dakota Utilities, Northern Municipal Power Agency, Northwestern Public Service Company, and Ottertail Power Company) near Beulah, North Dakota. The 425 MW unit burns lignite from a mine-mouth plant and initially used soda ash (Na2CO3) as the sulfur removal reagent. The spray dryer was modified about 10 years later, and the unit currently uses lime as the reagent. The second dry scrubbing system on a coal-fired utility boiler was installed on two 440 MW units; it became operational in 1982 and 1983 at the Basin Electric Power Cooperative's Antelope Valley station, also located near Beulah, North Dakota. These units fire minemouth lignite and use a slaked lime slurry to remove SO2 in the spray dryer.

A slaked-lime slurry is sprayed directly into the ductwork to remove SO2 (see Figure 6-6c). The reaction products and fly ash are captured downstream in the particulate removal device. A portion of these solids is recycled and reinjected with the fresh sorbent. Dry spray drying (DSD) is a relatively simple retrofit process capable of 50% SO2 removal at a Ca/S ratio of 1.5. The concept is the same as conventional spray drying except that the existing ductwork provides the residence time for drying instead of a reaction vessel. The main difference is that the residence time in the duct is much shorter (i.e., 1-2 sec, compared to 10-12 sec in a spray drying vessel).

The slaked lime is produced by hydrating raw lime to form calcium hydroxide. This slaked lime is atomized and absorbs the SO2 in the flue gas. The SO2 reacts with the slurry droplets as they dry to form equimolar amounts of calcium sulfite and calcium sulfate. The water in the lime slurry improves SO2 absorption by humidifying the gas. The reaction products, unreacted sorbent, and fly ash are collected in the particulate control device located downstream. Some of the unreacted sorbent may react with a portion of the CO2 in the flue gas to form calcium carbonate. Also, a little more SO2

removal is achieved in the particulate control device. The reactions occurring in the process are:

Ca(OH)2(s) + SO2(g) CaSOs ■ 2^O(s) + iH2O(y) (6-36)

Ca(OH)2(s) + SO2(g) + lO2(g) + H2O(v) CaSO4 ■ 2H2O(s) (6-43) Ca(OH)2(s) + CO2(g) CaCOs(s) + H2O(v) (6-44)

There are two different methods for atomizing the slurry. One method is the use of rotary atomizers, with the ductwork providing the short gas residence time of 1 to 2 seconds. When using this atomizer, the ductwork must be sufficiently long to allow for drying of the slurry droplets. There must also be no obstructions in the duct. The second method for atomizing the slurry is the use of dual-fluid atomizers, where compressed air and water are used to atomize the slurry. This process is referred to as the confined zone dispersion (CZD) process. The dual-fluid atomizer has been shown to be more controllable due to the adjustable water flow rate. This atomizer is also relatively inexpensive and has a long and reliable operating life with little maintenance. The spray is confined in the duct, which allows better mixing with the flue gas rather than impinging on the walls.

The DSD process has several advantages compared to wet processes. The DSD process is less complex because the reagent is injected directly into the flow path of the flue gas, and a separate absorption vessel is not needed. Less equipment is needed so power requirements are lower. The waste from this process does not contain reactive lime, as the FSI process does, and therefore does not require special handling.

Some of the problems encountered by the DSD system are also common to other dry processes. A main disadvantage of the system includes limited SO2 removal efficiency (i.e., ~50%) and low calcium utilization compared to wet processes. Quicklime is more expensive than limestone. If an ESP is used, there is the potential for reduced efficiency due to changes in fly ash resistivity and the increased dust loading in the flue gas. Additional collection devices may be needed as well as humidification to improve ESP collection efficiency. There must be sufficient length (i.e., residence time of the ductwork) to ensure complete droplet vaporization prior to the particu-late collection device. This is necessary for good sulfur capture and to avoid plugging and deposition, which in turn results in an increased pressure drop that the induced draft fans must overcome.

Duct Sorbent Injection: Dry Sorbent Injection Dry sorbent injection (DSI), also referred to as in-duct dry injection, is illustrated in Figure 6-6d. Hydrated lime is the sorbent typically used in this process, especially for power generation facilities; however, sodium-based sorbents have been tested extensively, including full-scale utility demonstrations, and are used in industrial systems such as municipal and medical waste incinerators for acid gas control.

When hydrated lime is used in this process, it is injected either upstream or downstream of a flue gas humidification zone. In this zone, the flue gas is humidified to within 20°F of the adiabiatic saturation temperature by injecting water into the duct downstream of the air preheater [12]. The SO2 in the flue gas reacts with the calcium hydroxide to form calcium sulfate and calcium sulfite:

Ca(OH)i(s) + SOi(g) + ¿O2te) + H2O(v) CaSO4 ■ 2HiO(s) (6-43)

Ca(OH)i(s) + SO2(g) CaSO3 ■ iHiO(s) + ±H2O(v) (6-36)

The water droplets are vaporized before they strike the surface of the wall or enter the particulate control device. The unused sorbent, products, and fly ash are all collected in the particulate control device. About half of the collected material is shipped to a landfill, while the other half is recycled for injection with the fresh sorbent into the ducts [12].

The DSI system offers many of the same advantages and disadvantages that other dry systems offer [12]. The process is less complex (i.e., no slurry recycle and handling, no dewatering system, fewer pumps, and no reactor vessel) than a wet system, specifically LSFO. The humidification water and hydrated lime are injected directly into the existing flue gas path. No separate SO2 absorption vessel is necessary. The handling of the reagent is simpler than in wet systems. DSI systems have less equipment to install so operating and maintenance costs are reduced. The waste product is free of reactive lime so no special handling is required.

Some of the problems encountered by the DSI system and its disadvantages, as compared to the LSFO system, are common to other dry processes. Sulfur dioxide removal efficiencies are lower (as is calcium utilization) than wet systems and range from 30 to 70% for a Ca/S ratio of 2.0. Quicklime is more expensive than limestone. When an ESP is used for particulate control, there is the potential for reduced efficiency due to increased fly ash resistivity and dust loading in the flue gas. Additional collection devices may be required. A sufficient length of ductwork is necessary to ensure a residence time of 1 to 2 seconds in a straight, unrestricted path. Plugging of the duct can occur if the residence time is insufficient for droplet vaporization, leading to increased system pressure drop.

In the dry sodium desulfurization process, a variety of sodium-containing crystalline compounds may be injected directly into the flue gas. The main compounds of interest include [14]:

• Sodium carbonate (Na2CO3), a refined product of ~98% purity;

• Sodium bicarbonate (NaHCO3), a refined product of ~98% purity;

• Nacholite (NaHCO3), a natural material of ~76% purity containing high levels of insolubles;

• Sodium sesquicarbonate (NaHCO3 ■ Na2CO3 ■ 2H2O), a refined product of -98% purity;

• Trona (NaHCO3 ■ Na2CO3 ■ 2H2O), a natural material of -88% purity containing high levels of insolubles.

Sodium bicarbonate and sodium sesquicarbonate have been the most extensively tested in pilot-, demonstration-, and full-scale utility applications due to proven success and commercial availability. In addition, sodium bicarbonate is extensively used in industrial applications for acid gas control. Of the compounds listed above, sodium bicarbonate has demonstrated the best sulfur capture in coal-fired boiler applications, as shown in Figure 6-9, which illustrates SO2 removal as a function of normalized stoichiometric ratio (NSR) for sodium bicarbonate injection into a coal-fired, pilot-scale test facility and industrial boiler equipped with fabric filter baghouses. Note that the NSR represents the molar ratio between the injected sodium compound and the initial SO2 concentration in the flue gas, considering that it takes 2 mol of sodium to react with only 1 mol of SO2.

The flue gas stream must be above 240° F for rapid decomposition of the sodium bicarbonate when it is injected or little SO2 capture will occur. While SO2 will react directly with the sodium bicarbonate, in the presence of nitric oxide (NO) this reaction is inhibited and does not result in significant sulfur capture; therefore, for acceptable SO2 capture to progress rapidly and attain acceptable levels of utilization, the bicarbonate component must begin to decompose. As flue gas temperatures increase into the optimum range of 240 to 320°F, the carbonate is decomposed, and the subsequent sulfation reaction occurs [14]. When the bicarbonate component decomposes, carbon dioxide and water vapor are evolved from the particle interior, creating a network of void spaces. Sulfur dioxide and NO can diffuse to the fresh sorbent surfaces, where the heterogeneous reactions to capture SO2 (and to a lesser extent NO) take place. The decomposition and sulfation reactions are:

2NaHCO3(s) + heat Na2CO3(s) + H2O(v) + CO2(g) (6-45)

SO2(g) + Na2CO3(s) + lO2(g) Na2SO4(s) + CO2(g) (6-46)

Lower NOZ emissions also result from injection of dry sodium compounds [4]. The mechanism is not well understood, but reductions up to 30% have been demonstrated which are a function of SO2 concentration and NSR ratio. There is a side effect of this reduction, though. Nitric oxide is oxidized to NO2 (a reddish-brown gas), and not all of the NO2 is reacted with the sorbent. As the NO2 concentration increases in the stack, an undesirable coloration in the plume can be created.

Hybrid Systems Hybrid sorbent injection processes are typically a combination of FSI and DSI systems with the goal of achieving greater SO2 removal and sorbent utilization [4]. Various types of configurations have

0 0

Post a comment