14.2.1 Control of suspended particulate matter (SPM) using ESP

Electrostatic precipitators depend on static electricity to do their work. The main components of an ESP are two sets of electrodes (Figure 14.5). The first is comprised of rows of electrically grounded vertical parallel plates, called collecting electrodes, between which the gas to be cleaned flows. The second set of electrodes is wires, called discharge electrodes, are centrally located between each pair of parallel plates. These electrodes may be of different shape and design. Figure 14.6 shows different types of collecting electrodes, while Figure 14.7 presents various types of discharge electrodes.

The wires carry a unidirectional, negatively charged, high-voltage DC current from an external source that generates unidirectional, non-uniform electric field. In a non-uniform electric field, the field intensity near the conductor is considerably higher than that on the periphery or on the plate. Therefore, as the voltage increases, the ionization takes place about the conductor. No ionization takes place near the periphery or on the plate. When the voltage is high enough, a blue luminous glow, called a corona, is produced around them. The corona is an indication of generation of negatively charged gas ions that travel from wires to the grounded collecting electrodes as a result of strong electric field between them. The field intensity that causes the corona to start may be calculated using the following formula [6]:

$E_C=3.04\left(\beta +0.0311\sqrt{\frac{\beta }{r}}\right)*{10}^6$ (14.1)

(14.1)

where

$=\frac{(P_B-P_G)*298}{101.3*T_G}$ (14.2)

(14.2)

where

14.2.2 Fundamentals

It is known from basic physics that all bodies fall into two categories, i.e., conductors and dielectrics. A body in which electric charges may move freely throughout its volume is called a conductor. While a body that does not possess any such property and acts as an insulator is called dielectric. Gases under normal condition are insulators (dielectric), since they comprise electrically neutral atoms and molecules. However, when subjected to the influence of various artificial methods, i.e., X-rays, radioactive emissions, cosmic rays, strong heating, bombardment of gas molecules by rapidly moving electrons or ions, etc., these gases get ionized, i.e., split into electrons (negative ions) and positive ions, and become conductors. A combination of some of the free electrons of gases with their neutral gas molecules also results in the formation of negative ions.

Out of all artificial methods of ionization just described, the high-voltage direct-current corona is by far the most effective to ionize gas particles. In this method electrodes are connected to a voltage source and produce a non-uniform electric field. When gases are passed between these electrodes, the free charges within gases move along lines of forces of the non-uniform electric field. Thus under the influence of strong electric field suspended particles in gas stream migrate toward collecting electrodes. On reaching the electrodes, particles release part of their charge and get collected. The collected dust is then discharged periodically by rapping and gets collected in storage hoppers. The whole process includes the following steps:

The separation force acting on a particle is given by Coulomb’s law, which states that force is proportional to the product of particle charge and the intensity of the collecting field. Hence,

$F_C\infty E*q$ (14.3)

(14.3)

where

Although the Coulomb force steers particles to accelerate toward the collecting electrodes, inertial and viscous forces counteract the Coulomb force and oppose the motion of particles. For suspended particles, inertial forces are negligibly small and may be ignored. The viscous force of the gas is governed by Stokes’ law and is expressed by Eq. 14.4. The resultant of the Coulomb force and Stokes (viscous) force determines velocities to be attained by a particle in the precipitation field. The velocity at which the particles move in the gas streams toward the collecting electrodes is called the migration velocity of the particle.

Thus, the viscous force,

$F_S=6\pi \mu rw$ (14.4)

(14.4)

where

$w=\frac{E_o*E_p*r}{\mu }(\mathrm{for\; particle\; size}>1\mathrm{micron})$ (14.5)

(14.5)

$w=\frac{E_p}{\mu }(\text{for}\text{sub}-\mu \mathrm{particles})$ (14.6)

(14.6)

where

For E_o=E_p=E, and if w is expressed in cm/s, E in V/m, r in m, and μ in N.s/m², Eq. 14.5 and Eq. 14.6 change as follows [6]:

$w=\frac{{10}^{-9}*E^2*r}{\mu }(\mathrm{for\; particle\; size}>1\mathrm{micron})$ (14.7)

(14.7)

$w=\frac{0.17*{10}^{-9}*E}{\mu }(\text{for}\text{sub}-\mu \mathrm{particle})$ (14.8)

(14.8)

Equation 14.7 shows that the migration velocity of particles greater than 1 μ in diameter is determined by the particle size as well as by the product of field intensities and inversely by the gas viscosity. While from Eq. 14.8 it follows that sub-micron particles travel at velocities directly proportional to the field intensity and inversely proportional to the gas viscosity, but independent of their size. Since the migration velocity is dictated by gas viscosity, which in turn is directly proportional to gas temperature, it may be concluded that the migration velocity is sensitive to changes in gas temperature.

14.2.3 Factors to consider for sizing an ESP

Sizing of an ESP is determined by various factors or parameters, discussed in the following sections. The more accurate the values of these parameters can be predicted during the design stage the better performance of an ESP is expected in the field. Hence, the more the field feedback data available on the performance of a particular design of ESP, the more accurate these parameters can be predicted at the design stage that will ensure better performance of similar design of ESPs in future. Eventually the design of an ESP is more of an art than a science.

A single parameter, which is to be judiciously assumed prior to sizing an ESP, is effective migration velocity (EMV), which influences the collection efficiency of an ESP. Deutsch–Anderson in 1922 developed the following formula for predicting the collection efficiency of an ESP,

$\eta =\left\{1-e^{-\frac{wf}{100}}\right\}*100$ (14.9)

(14.9)

where

From Eq. 14.9 it is evident that the factors that affect the performance of an ESP are EMV and SCA.

In the 1970s, a modified version of the Deutsch–Anderson equation, known as the Matts–Ohnfeldt equation, was developed to predicting better performance of an ESP. The modified equation is

$\eta =\left\{1-e^{-{\left(\frac{w_{k}f}{100}\right)}^k}\right\}*100,$ (14.10)

(14.10)

where

14.2.4 Effective migration velocity (EMV), w

From field experience it has been established that it is easier to collect larger particles than smaller particles, and the collection efficiency of an ESP decreases with a drop in particle size. Therefore, based on experimental studies the collection efficiency of an ESP at various operating conditions as well as for particle sizes ranging larger than 1 μ to sub-micron an overall effective migration velocity is estimated. The value of effective migration velocity is a measure of ease or difficulty of collection.

As explained earlier, the migration velocity is determined by the magnitude of particle size, the strength of electric field and the viscous force or drag of the particle. This theoretical value may become excessively high compared to actual value due to uneven gas flow, particle diffusion, electric wind, particle charging time, and re-entrainment.

Hence, during design stage, some characteristics/values of these factors are assumed to arrive at an acceptable effective migration velocity, which may get reduced significantly during actual operation. In practice actual migration velocity may deviate considerably from that estimated at design stage, which is usually based on empirical observations. In view of this, it is safe to select a low value of effective migration velocity, and even more so, for handling high resistivity fly ash (Table 14.6), since the migration velocity declines as resistivity increases (Figure 14.8).

14.2.5 Specific collection area (SCA), f

The SCA is determined by dividing the total effective collecting electrode surface area by the actual volume flow rate of wet gas to be treated. The value of the SCA depends on physical, chemical, and electrical characteristics of the dust. Table 14.7 summarizes some typical values of SCA for different resistivity of fly ash.

In addition to the EMV and SCA other factors that influence ESP sizing are:

14.2.6 Aspect ratio

This is the ratio of total collecting length to the collecting height of an electrode. If the length of collecting zone is too short compared with its height, some of the falling dust will be carried out of the precipitator before it reaches the hopper, resulting in re-entrainment of dust particles. As a result dust carryover to stack will increase. For good design, a designer should be comfortable with an aspect ratio of ESP of 2.5 or more.

14.2.7 Number of fields in series

To enhance the reliability of ESP performance, the longitudinal electrical section of gas flow path is divided into a number of independent electric fields. In practice a number of electric fields in series should be no less than 7, in case ash resistivity is too high number of fields in series can be even 9. With this arrangement, the ESP performance would not be impaired much, even with the loss of one electric field, particularly in the case of handling high resistivity ash.

The more the number of fields within a given longitudinal section, the smaller the size of each section, which in turn will provide the following advantages:

14.2.8 Length of a field

The time required by each particle to migrate to collecting electrode from discharge electrode determines the length of a field of precipitator passage in the direction of gas flow. This migration time has to be less than the passage time of each particle while passing through a field of precipitator with the same velocity as gas for ensuring effective removal of PM in a field vis-à-vis and ESP. The length of a field of precipitator thus is ensured by [7]:

$\frac{s}{w}\le \frac{L}{V_g}$ (14.10)

(14.10)

or

$L\ge s\frac{V_g}{w}$ (14.11)

(14.11)

where

14.2.9 Treatment time

The time taken by a particle to pass through the total length of the precipitator is known as the treatment time. For a good design the value of treatment time varies between 25 and 35 s for low-sulfur, low-sodium coals and between 7 and 20 s for high-sulfur, high-sodium coals [8].

14.2.10 Electrode spacing

Electrode spacing is double the gap between discharge electrodes and collecting electrodes, which means the gap between two parallel collecting electrodes. The electrode spacing considerably influences the migration velocity. For constant current density and field strength, an increase in electrode spacing facilitates choosing a higher migration velocity as is clear from Table 14.6. As a result, the total collection area is reduced substantially to achieve the desired collection efficiency, which is obvious from Eq. 14.9; thus the total capital expenditure gets reduced. A decrease in electrode spacing would result in lower migration velocity, resulting in higher capital expenditure.

To meet today’s stringent pollution control regulations, the efficiency of an ESP has to be very high, about 99.98% or more, for handling high resistivity ash. Thus, from an economical perspective industry practice is to provide electrode spacing of the order of 300/350/400/450/500 mm [9].

14.2.11 Automatic voltage control

The collection efficiency of an ESP is determined by the precipitation voltage potential between the discharge electrode and collecting plate. The maximum precipitator efficiency may be achieved by maintaining the operating voltage in the individual field of the precipitator below, but close to the flashover limit (the point of sparking between discharge electrodes and collecting plates that switches off the plant). Further, as the flashover voltage varies with operating conditions of the precipitator, such as gas volume, dust content, gas composition, gas temperature, humidity, etc., continuous adjustment of precipitation voltage according to prevailing operating conditions is needed. Therefore, in order to avoid the frequent readjustment and possible trip-outs on manually operated sets, operating personnel often reduce the voltage and current to much lower than the flashover limit.

Automatic voltage control works in two steps: first, the voltage is gradually raised at a pre-set rate until flashover takes place and switches off the system. Then after a pre-set time delay, the system switches on automatically and the voltage is raised to slightly reduced level than the flashover limit. Thereafter, this mode of cycle gets repeated continuously.

14.2.12 Transformer rectifier (T/R) set

The high-voltage power needed to discharge the electrode system is supplied by a transformer rectifier set along with a controller. Two transformer rectifier sets are normally needed to power each precipitator field. In practice a medium voltage AC supply is regulated and transformed into a nominal voltage of 50–120 kV, rectified to a negative DC output then supplied to the discharge electrode system.

The current density usually is maintained over a range of 0.25–0.45 mA/m², from which the capacity of transformer rectifier set can be determined, since the collecting area of each field is already known.

14.2.13 Hopper heating and storage volume

In practice, dust collected near hopper walls will agglomerate and plug the dust-removing system. Hopper heating arrangement prevents formation of dust cakes and gets rid of the hopper plugging problem. Additionally, the bottom part of the hopper is sometimes provided with a stainless steel liner that ensures satisfactory flow of dust from it. Fluidizing pads are also fitted inside hoppers to ensure smooth flow of dust.

The dust hoppers must have adequate storage capacity to hold all the collected dust at a boiler maximum generating condition of the unit while firing the worst coal. An overfill hopper may result in loss of an electrical section. Therefore, the hopper storage volume should be adequate enough such that during the interval between two consecutive steps of removal of dust from the hopper usually (6–8 hrs) as per the industry practice, dust build-up will not bridge the high voltage frame.

14.2.14 Rapping mechanism

When dust builds up on plates, it deposits in a layer of increasing thickness with possible re-entry into the gas flow path, hence build-up ash needs to be removed periodically by hitting/rapping collecting electrodes with hammers. Based on the changing characteristics and collection rates of dust the intensity and frequency of rapping of individual collection plates varies from inlet to outlet of precipitator. Rapping should be carried out sequentially so that only a fraction of accumulated dust is disturbed at any one time.

There are two types of hammers: drop hammer and tumbling hammer. Drop hammers are either pneumatically driven or electrically driven. Each drop hammer is connected with a number of electrodes and is used to rap from the top of electrodes. Tumbling hammers are basically side hammers where a set of hammers is mounted on a horizontal shaft in a staggered fashion. As the shaft rotates one hammer at a time tumbles and hit a shock bar.

Excessive rapping of collecting electrodes is detrimental to the performance of an ESP as it causes re-entrainment of collected dust and reduces the life of internal components. In contrast, inadequate rapping results in build-up of a dielectric dust layer on the collecting plates, which is penetrated by corona current. As a result the dust layer reaches a level of resistivity that generates ionization or back-corona and causes dust particles to become positively charged and move toward negative high voltage discharge electrodes. Thus, localized excessive sparking occurs, reducing power input to the ESP, which in turn lowers the collection efficiency.

14.2.15 Factors that affect performance of ESP

The factors that determine the type, size, and efficiency of the electrostatic precipitator are discussed in following sections. During normal running if any of these factors varies substantially from the design, the performance of the ESP will be jeopardized, which may lead to violating a local SPM emission regulation.

14.2.16 Volume of gas flow

The volume of gas to be treated is usually determined at the design stage of any project. Subsequently, if it is found that the actual volume of gas to be handled is much higher than the specified volume of gas, the following detrimental effects on ESP performance could be encountered:

14.2.17 Velocity of gas flow

This is a critical design parameter and is calculated by dividing the actual volume flow rate of gas by the cross-sectional area of the precipitator. For low-sulfur, low-sodium coals, the gas velocity may be in the range of 0.5–1 m/s, and for high-sulfur, high-sodium coals, the gas velocity could be as high as 1.5 m/s [8]. At velocities greater than that, problems of re-entrainment of suspended particles are aggravated.

14.2.18 Temperature of flue gas

Temperature of flue gases passing through precipitator will affect the collectability of ash and removal of collected ash from collecting plates. When this temperature reaches the dew-point temperature moisture will form and cause dust particles to solidify into cakes. Further build up of these cakes may result in short-circuits or wire failures. In contrast, if the flue-gas temperature exceeds a high limit, then the gas volume will increase, leading to a reduction in collection efficiency. At higher temperature, the viscosity of the gas also increases, resulting in high viscous drag with an additional reduction in collection efficiency.

14.2.19 Gas distribution

While flowing through the precipitator the distribution of gas must be uniform all along its path, from the inlet up to its outlet. In sections where velocity is high, ash may be swept off collector plates and can show up in the stack as a visible plume, while at low velocity sections, although gas is more efficiently cleaned, the effect will be a reduction in overall collection efficiency. Hence, proper devices, e.g., perforated distribution plates, turning vanes, and baffles in the precipitator inlet and duct work, should be installed to prevent uneven gas distribution.

14.2.20 Dust resistivity

The electrostatic precipitator is very sensitive to the resistivity of ash or dust deposited on collecting plates. Experience has shown that coal ashes with resistivity between 10⁸ and 2×10¹⁰ ohm-cm is normal for satisfactory and predictable precipitator design. Resistivity of ash that exceeds this limit becomes highly resistive to dust collection. The presence of constituents, such as SiO₂, Al₂O₃, CaO, etc., in fly ash causes high resistivity of ash that does not readily lose its charge when collected on plates, and the agglomerated ash is very difficult to get removed. When deep-enough deposits collect on the plate, back-corona (local sparking on dust layer) develops, causing the collection rate to reduce (Figure 14.10). A corona discharge occurs only in an inhomogeneous electric field with the electrodes definitely shaped and located [6].

The resistivity of ash is lowered by cooling the inlet gas to the ESP while simultaneously increasing the water vapor content or by introducing SO₃, NH₃, etc., into the flue gas. These additives are adsorbed onto fly ash particles and produce an electrically conductive surface condition, improving precipitation. The resistivity may be lowered by “hot” precipitation as well. Experimental results reveal that resistivity of ash is maximum at around 423 K (Figure 14.11).

14.2.21 Inlet dust concentration

At the time of design certain dust concentration at inlet to the ESP is determined. If this concentration changes during operation, there will be consequent change in dust collection efficiency (Figure 14.12).

14.2.22 Weight of dust per unit volume of gas

During the design stage the expected weight of fly ash per unit time may be calculated, considering analysis of a worst coal. If at a later stage it is found that the as-received coal has much higher ash content than a precipitator is designed to handle, then the ESP inlet dust concentration would increase, causing a fall in collection efficiency as shown in Figure 14.12.

14.2.23 Basic properties of the gas and dust particles

A precipitator is designed specifically for an ash produced by coal containing a certain amount of sulfur; if coal with different sulfur content is burned, the efficiency will vary. A coal with lower sulfur content may give satisfactory separation when the flue-gas temperature is increased. For corrosive gases and particles, the shell interior must be made of corrosion-resistant materials, such as lead for sulfuric acid mist and tile block for paper mill salt cake.

The particle size also affects the settling rate. Particles larger than about 1 μ will settle too rapidly. Sub-micron particles (within the range of 0.1–0.5 μ) will settle very slowly and exhibit Brownian motions. The settling rate again becomes rapid for particle size less than 0.1 μ [8].

The optical properties of particles are of great importance in connection with gas cleaning and air pollution, because the degree of pollution is commonly judged by visual appearance of stack discharge.

14.2.24 Disposal of collected gas

An ESP is enclosed and insulated. If the hoppers are not insulated, ash near hopper walls will agglomerate as it cools. There is a tendency for the agglomerated mass to form cakes, which may plug the ash conveying system. If ash is not removed regularly from hoppers, it can build up to a point where it will bridge high voltage frame, causing a short-circuit and loss of an electrical section. Prolonged accumulation of ash may endanger ESP structure and cause catastrophic failures.

14.2.25 Control of SPM using a bag filter

A bag filter, also known as a fabric filter or baghouse, comprises multiple compartments (similar to ESP fields), each housing several vertical (2–10 m high) small diameter (120–400 mm dia) cloth or fabric bags (in place of electrodes in ESP) [11]. Depending on the temperature of the entering flue-gas the type of fabric material may be any one of Polypropylene (PP), Polyester (PES), Draylon T (PAC), Ryton (PPS), Nomex (APA), P84 (PI), Teflon (PTFE), Fiber glass (GLS), etc [4]. Of these materials, from an economical perspective, the most suitable one for handling fossil fuel-fired flue gas is Ryton, which can withstand a flue-gas temperature up to 453 K.

14.2.26 Factors to consider for sizing a bag filter

Three factors that affect the sizing of bag filters are:

14.2.27 Factors that affect the performance of the bag filter

In addition to the air/cloth ratio other factors that may affect the efficiency of a bag filter are as follows:

14.2.28 Comparison between ESP and bag filter

Table 14.8 summarizes comparative analysis of ESP and bag filter.

14.4.1 Wet scrubber

A wet scrubber may be used for boiler applications with high- to low-sulfur coals. This is the most widely used FGD technology for SO₂ control. In wet scrubbers flue gas enters at the bottom of an absorber tower and exits as clean gas from the top. Both absorption and oxidation take place within the absorber tower. Water is sprayed at the top of the tower through a series of spray nozzles and air is injected through the tower from the bottom at a point lower than the entry point of flue gas. A slurry mixture of calcium-, sodium-, and ammonium-based sorbents, prepared in an external container, is injected into the tower to react with the SO₂ in the flue gas. From a techno-economic consideration sorbent that is widely used in operating wet scrubbers is limestone (CaCO₃). Next preferred sorbent is lime (CaO). Figure 14.13 shows the schematic arrangement of a typical wet scrubber system.

Limestone units generally operate in a forced-oxidation mode producing gypsum as a by-product. In this mode limestone reacts with SO₂ in flue gas producing calcium sulphite (CaSO₃). The injected air along with air contained in flue gas first oxidizes calcium sulphite producing calcium sulphate (CaSO₄), which in the next step reacts with incoming sprayed water to produce gypsum (CaSO₄·2H₂O). The corresponding chemical reactions are:

$\begin{array}{l}\mathrm{C}\mathrm{a}\mathrm{C}{\mathrm{O}}_3+\mathrm{S}{\mathrm{O}}_2=\mathrm{C}\mathrm{a}\mathrm{S}{\mathrm{O}}_3+\mathrm{C}{\mathrm{O}}_2\hfill \\ \mathrm{C}\mathrm{a}\mathrm{S}{\mathrm{O}}_3+½{\mathrm{O}}_2+2{\mathrm{H}}_2\mathrm{O}=\mathrm{C}\mathrm{a}\mathrm{S}{\mathrm{O}}_4·2{\mathrm{H}}_2\mathrm{O}\hfill \end{array}$

In a magnesium-enhanced wet lime process, about 4–7% (by weight) of magnesium oxide (MgO) is added to quick lime (CaO). This mixture on reacting with incoming spray water results in a slurry of calcium hydroxide (Ca(OH)₂) and magnesium hydroxide (Mg(OH)₂). The magnesium-enhanced lime-scrubbing reagent reacts with the SO₂ in the flue gas producing hydrated calcium sulfite and hydrated magnesium sulfite. Thus this process would not yield gypsum as the end product as is evident from the following chemical reactions:

$\begin{array}{l}\mathrm{C}\mathrm{a}{(\mathrm{O}\mathrm{H})}_2+\mathrm{S}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}=\mathrm{C}\mathrm{a}\mathrm{S}{\mathrm{O}}_3·2{\mathrm{H}}_2\mathrm{O}\hfill \\ \mathrm{M}\mathrm{g}{(\mathrm{O}\mathrm{H})}_2+\mathrm{S}{\mathrm{O}}_2+{\mathrm{H}}_2\mathrm{O}=\mathrm{M}\mathrm{g}\mathrm{S}{\mathrm{O}}_3·2{\mathrm{H}}_2\mathrm{O}\hfill \end{array}$

All limestone units are capable of removing 90–95% of SO₂ from flue gas. Magnesium-enhanced lime systems are considered as the most efficient SO₂ removal systems, where up to 98% removal can be ensured.

The advantages of wet scrubbers are:

The disadvantages of this method are:

14.4.2 Spray dry scrubber

Lime (CaO) is normally used as a sorbent in a spray dry scrubber. Prior to spraying into the absorber CaO is hydrated to Ca(OH)₂. Hydrated lime is directly injected as spray into the absorber tower, where moisture is evaporated by the heat of flue gas. SO₂ and SO₃ in flue gas react with hydrated lime within the absorber tower to form a dry mixture of calcium sulphate/sulphite. SO₂ removal efficiency of a spray dry scrubber falls within 90 and 95%. It is the second most widely used process to arrest SO₂.

The advantages of this process are:

The disadvantages are:

14.4.3 Sorbent injection system

A sorbent injection system is the simplest method of SO₂ removal from flue gas. Sorbent generally used in this system are limestone (CaCO₃), dolomite (CaCO₃·MgCO₃), lime (CaO), or hydrated lime (Ca(OH)₂). Based on location of injection of sorbent, this system is classified into furnace sorbent injection (FSI), economizer sorbent injection (ESI), duct sorbent injection (DSI) (Figure 14.14), and hybrid sorbent injection (HSI) process. In all these types the effectiveness of SO₂ removal depends on how evenly and quickly the sorbent is sprayed across the cross-section of flue-gas stream. The solid waste obtained downstream of each process is removed in the dust-collecting equipment.

In a furnace/post furnace sorbent injection system (FSI) powdered dry sorbent (usually limestone or hydrated lime) is sprayed into the entire cross-section in the upper part of the furnace, where the temperature is maintained above 1023 K but not more than 1523 K. When flue gas passes through this zone the SO₂ and O₂ content of the flue gas reacts with the sorbent to form CaSO₄.

The SO₂ removal efficiency that may be attained in FSI is 50% or higher based on even distribution of sorbent, fastness of application, ensuring optimum residence time (1 to 2 seconds), maintaining proper sorbent particle size (not more than 5 μ, but preferably less than 3 μ) and adopting Ca:S ratio between 2 and 3.

Sorbent used in an economizer sorbent injection process (ESI) is hydrated lime. When flue gas flows through the economizer zone its temperature remains between 573 K and 923 K. At this temperature range when sorbent is sprayed into the flue-gas stream Ca(OH)₂ reacts with SO₂ to produce CaSO₃ instead of CaSO₄.

In a duct sorbent injection (DSI) process sorbent is sprayed into the flue-gas stream at a temperature around 423 K, located downstream of air pre-heater. Along with the sorbent some water may also be sprayed to humidify the flue gas with the precaution that humidity always remains above the dew point. SO₂ removal efficiency of this process could be as high as 80%.

Adopting FSI in combination with DSI is called a hybrid sorbent injection (HSI) process. This combination results in higher sorbent utilization and greater SO₂ removal. With humidification SO₂ removal efficiency of HSI can be boosted up to 90%.

14.4.4 Dry scrubber

A circulating fluidized bed (CFB) involves a scrubbing technology known as a dry scrubber process. In this process, hydrated lime, limestone, or dolomite and water are intimately mixed with flue gas in a dedicated reaction chamber, which is the fuel-firing bed. Reacted solids get dried completely by the sensible heat of the flue gas. The process is easy to maintain and operate since it does not include high-maintenance mechanical equipment. The process can achieve about 95% SO₂ removal efficiency. Because of high SO₂ removal efficiency the Ca:S ratio (1.2–1.5) of the dry scrubber process is less than that of the sorbent injection system. (For more information, refer to Section 5.4, in Chapter 5, Fluidized Bed Combustion Boilers.)

14.4.5 Regenerable process

In a regenerable process SO₂ removal efficiency is very high (>95%), but the process suffers from high cost and high power consumption. In this process, once SO₂ is removed, the sorbent is regenerated for further use. As a result, this process provides an advantage of low sorbent make-up requirement. Another advantage of this process is it produces virtually no residual solid or liquid wastes. The sorbent may be regenerated either thermally by heating it to a temperature of about 923 K or chemically by bringing the sorbent in intimate contact with steam and a reducing gas, i.e., H₂S, H₂, and natural gas.

The process involves stripping of SO₂ from flue gas by scrubbing action of sodium sulfite solution or organic-amine liquid absorbent. Use of other chemical solutions is also possible, but currently this approach is at the research stage.

14.4.6 Combined SO₂/NO_x removal process

This process involves injection of ammonia in the flue gas upstream of a selective catalytic reduction (SCR) unit for NO_X control and injection of sorbent for SO₂ control. As a result this process is relatively complex and capital intensive and also requires lot of maintenance. This process still is in the development stage. Nevertheless, some field results show more than 70% SO₂ removal and about 90% NO_X removal by the combined process.

14.4.7 Seawater FGD

Seawater FGD is attractive to plants that are located near coastlines. Seawater is a source of natural alkalinity that can be used economically to arrest SO₂ from flue gas without the use of any other supplemental reagents. Seawater contains significant concentrations of alkaline ions including sodium, magnesium, potassium, calcium, carbonates, and bicarbonates. It also contains significant concentrations of chloride and sulphite ions. Desulfurization is accomplished by seawater scrubbing, and 90% SO₂ removal efficiency could be achieved by adopting this process.

Before entering into the seawater scrubber flue gas is cooled to its adiabatic saturation temperature of typically 365 K. The scrubber is packed with certain material proprietary in nature through which flue gas flows from the bottom and seawater is sprayed counter-current to the flow of flue gas in order to achieve effective SO₂ mass transfer and chemical absorption into the liquid phase. Thus, SO₂ in the flue gas is absorbed and converted to sulphite first. Sulphite is further oxidized to produce sulphate for safe disposal to the sea without jeopardizing the marine environment. For disposal of liquid effluent to sea, it should be ensured that liquid effluent is alkaline with a pH no less than 8.

14.5.1 Burner replacement

This method is normally adopted for retrofitting purposes. Old burners, which were not designed to meet currents standards for maximum NO_X emission, are replaced with low NO_x burners (LNBs). Figure 14.15 and Figure 14.16 show the cross-sectional details of typical LNBs from different manufacturers. These burners facilitate mixing of fuel and air in a controlled way to provide larger and more branched flames. As a result, the peak flame temperature is reduced, resulting in less formation of NO_x.

Within the flame four different processes take place as follows (Figure 14.17):

Use of LNBs, however, suffers from generating higher carbon loss in ash. The burner replacement method in isolation can not meet present-day emission level norm. To make this method effective LNBs are combined with combustion modification as discussed in the following.

14.5.2 Combustion modification

In order to ensure complete combustion with reduced NO_x emission, control of fuel and air supply is very important. This is accomplished at the burner as well as inside the furnace. Supply of secondary air and over-fire air facilitates vertical staged combustion. While supply of air at different points along the flame or introducing tertiary air at the burner ensures horizontal staging of combustion.

In addition to the above replacement of burners, adjustable angle over-fire air (OFA) dampers are also installed to quench flue-gas temperature, minimizing formation of NO_x. Furnace over-fire air helps in separating combustion air into primary and secondary air to achieve complete combustion. The primary air combines with fuel that produces a relatively low-temperature fuel-rich zone. Then the secondary air is introduced above the burners to complete the combustion also at a relatively low temperature. Thus, N₂ is produced in lieu of NO_x. The effectiveness of controlling NO_x formation by providing OFA dampers has already been “proven.”

Flue-gas recirculation, like over-fire air, quenches the flame and at the same time reduces oxygen level, thereby NO_x formation is minimized.

Use of low excess air in the combustion region and then re-burning fuel-rich mixture above this region provides fuel staging. Fuel staging can also be achieved by taking the burner out of service (BOOS). This technique, however, is seldom practiced in pulverized coal-fired boilers. Fuel biasing is another method of fuel staging. In this method, the fuel supply is shifted from the upper-level burners to the lower levels or from central burners to burners located in the periphery. These types of arrangements create a fuel-rich zone in the lower level and peripheral burners and fuel-lean zone in the upper and central burners. As a result, the flame temperature is reduced, minimizing the formation of NO_x, and at the same time, improving the balance of oxygen concentration in the furnace.

NO_x formation is inherently minimized in furnaces with large cross-sections. Furnaces that are designed for high heat release rates are susceptible to high NO_x emissions, since an air-in leakage can quickly counter weigh low NO_x gain. The fundamental concept of low-temperature fluidized bed firing yields low NO_x emissions. (For additional information refer to Chapter 5, Fluidized Bed Combustion Boilers.)

The disadvantages of combustion modification method are:

14.5.3 Steam or water injection

In oil/gas-fired combustors flame quenching is performed through the use of water or steam injection, resulting in lower peak flame temperatures. This technique is mainly adopted in gas turbines. The added benefit of adopting this technique is a boost in power output as a result of greater mass flow. This technique, however, demands ultra-pure water, since even a minute presence of alkali in the combustor may destruct the gas turbine operating at an inlet temperature of 1423 K or above.

14.5.4 Selective catalytic reduction (SCR)

In the SCR process ammonia vapor is used as a reducing agent. This vapor is injected into the flue-gas stream in the presence of a catalyst, thereby converting NO_x into H₂O and N. In the absence of a catalyst these reactions take place at temperatures between 1143 and 1478 K, but in the presence of a suitable catalyst these reactions may take place at much lower temperatures of 573–673 K, normally prevalent at the economizer outlet. Common catalytic materials are aluminium oxide (Al₂O₃), vanadium pentoxide (V₂O₅), titanium dioxide (TiO₂), tungsten trioxide (WO₃), silicon dioxide (SiO₂), zeolites (e.g., alumina silicates), iron oxides, activated carbon, etc. In coal-fired power plants vanadium pentoxide in combination with titanium dioxide is mostly used as a catalyst. SCR system can accomplish over an 80% reduction of NO_X from flue gases [14].

In coal-fired power plants, the SCR system is placed in three typical locations: between ESP and FGD in low dust configuration (Figure 14.18) or between economizer and air heater in a high dust system, especially with dry bottom boilers (Figure 14.19), or after the FGD (Figure 14.20), primarily with wet bottom boilers with ash recirculation to avoid catalyst degradation.

14.5.5 Selective non-catalytic reduction (SNCR)

In the SNCR process a reagent, i.e., urea, ammonium hydroxide, anhydrous ammonia, or aqueous ammonia, is injected into flue gases in the furnace within the appropriate temperature zone, typically in the range of 1173–1373 K [14]. The NO_x and the reagent (urea, etc.) react to form N₂ and H₂O and do not require a catalyst. The process is relatively simple, but highly sensitive to ammonia slip. In the event of any ammonia slip it will react with SO₃ in flue gas to form ammonium bisulfate, which would tend to precipitate at a temperature generally prevailing at air heater and lead to fouling and plugging of air heater. This process can continuously achieve 40–70% NO_x reduction. Generally, it is difficult to use SNCR in larger fossil-fired furnaces and in gas turbines, but it is widely applied to fluidized-bed-boilers in cases where further reduction of NO_x from flue gases is needed.

SNCR is an ideal retrofit technology and is compatible with other techniques (i.e., low- NO_x burners, OFA, FGR, gas reburning, etc.). Along with combustion modification, SNCR facilitates even higher reduction of NO_x.

The major disadvantage of the SNCR process is that it has a tendency to produce nitrous oxide (N₂O), which propagates the greenhouse effect.