Flue Gas Cleaning Technologies and Systems
Flue Gas Cleaning Technologies and Systems
Environmental pollution is one of the biggest problems all over the world at present. From a series of global environmental problems, more and more people have now realized that environment and resources are the basic necessities for the survival and developments of human beings. Flue gases which are the product of most of the technological processes are polluted with a variety of solid particles. In order to further use the gases (if they have sufficient calorific value) or to release them into the atmosphere, it is necessary to clean the gases. However, atmospheric emissions control costs money, with seldom any financial payback to the operating organization.
The past few years have seen a complete change in the attitudes, education, responsibility, and regulations in the area of emission controls in various countries. The emission control regulations are becoming tighter with the passing time in an attempt to save the future generations against the ill-effects of the atmospheric pollution. Now several organizations are fast changing their attitudes towards the atmospheric pollution and are having pro-active engagement in the pollution control activities. Organizations now want to be seen, by the public, to be responsible organizations which produce ‘clean’ products. This is partly market driven, as markets now demand more and more ‘clean’ products. The customers in the present day situation are becoming more educated, both in their responsibility towards the environment, as well as in the benefits of having a clean environment.
The purpose of a flue gas cleaning system is to reduce atmospheric emissions of substances hazardous to the environment and health. This includes e.g. heavy metals, dioxins and substances which cause acidification and eutrophication. Because some of the substances in flue gases are toxic and carcinogenic, it is important to reduce their emissions. The acidification of forests and lakes has been reduced substantially by removing oxides of sulphur and nitrogen from flue gas.
The technological processes of metallurgical, chemical, and thermal power plants generate the waste flue gases normally laden with dust and at high temperatures. The composition and quantity of these gases depend on the nature of the technological processes and the raw materials. Emissions of waste flue gases are actually a result of the raw materials used, and the processes and reactions which takes place in these plants. The flue gases can contain carbon dioxide, carbon mono oxide, oxides of sulphur (SO2 and SO3) and nitrogen (NOx), hydrogen, hydrogen sulphide (H2S), fluorine (in the form of HF), chlorine (in the form of HCl), arsenic, mercury, volatile organic compounds (VOC), water vapour, and dust etc. Water vapour is harmless but contributes to a visible plume at the stack outlet.
There are several technological processes which take place at high temperatures. Further many of these processes handle raw materials some of them are in the form of fines. Hence all these processes are prone to emit pollutant gases and particulate matter into the atmosphere. This in turn affects the quality of air around the plant. In order to improve and protect the quality of air, different pollution control devices are used for reduction of the emissions. Earlier for many years, pollution control equipments were used only for those processes where the pollutants amounts were very high or they were toxic in nature. These equipments were also earlier used where they had some recovery value. But in the present scenario, with the environment regulations becoming more and more tighter and with the increasing concerns of the society regarding the environment, it has become necessary to look into the emissions of all the technological processes and install equipments in all the areas to reduce the emissions to minimum possible levels.
There are at least five major groupings of atmospheric sources of pollution normally attributed to the technological processes, each with specific best practice technologies for abatement. Theses grouping do not constitute a comprehensive list since where strong acid gas concentrations are encountered; alternative flue gas cleaning technologies are to be implemented such as sulphuric acid plants. These five major groupings for the gas cleaning technologies are related to (i) dust and particulate emission control, (ii) acid gases such as SO2 / HCl and HF control, (iii) NOx abatement control, (iv) acid mist and other aerosols control, and (v) mercury, dioxins/furans and VOCs control. For the acid gas fixation technologies, product disposal as always remains a challenge. In the majority of applications, waste products are simply land-filled with the necessary operating costs which come with it. The emission control equipments for these five groups of gas cleaning technologies are basically of two types (i) dust and particulate emission control equipments and (ii) gaseous emission control equipment. This article concentrates on the dust and particulate emission control systems.
The problem of high temperature gas cleaning in general is one which is perhaps the most perplexing to industry. It is difficult since the problem is normally associated with extremely fine particles dispersed in gases at temperatures which can range from 700 deg C to 1,500 deg C. In some instances even higher temperatures can be involved. Because of the fine aerosol and the high temperatures involved, the customary approaches do not normally solve the problem. Hence, the progress in this area has not been as rapid. Basic problems associated with high temperature gas cleaning are those of economics and fundamental requirements for cleaning.
In some instances the cleaning of the waste gases is necessary since it represents a material of substantial value or, if the particulate material is removed, a gas remains which is combustible and is recovered in the form of heat or energy which can be utilized in the process. In other instances economic value of the effluent, either particulate, gaseous, or the usual combination, is of such small magnitude that the cost of disposal represents a sizeable problem. In these cases the cleaning or removal necessary to prevent air pollution is one for which there is only an intangible return.
In the second category, the desire of the industry is to obtain cleaning at minimal cost without imposing a burden in the form of increased production costs. The cost of preventing community air pollution and possible injury to the property or the public is normally without any tangible recovery other than good public relations.
Effective cleaning of these gases presents serious technical problems on account of the variety of impurities. High-efficiency gas cleaning systems are vital for the reliable operation and long campaign life of high temperature metallurgical and thermal power plants, and allow the operators to meet the relevant pollution control standards. The selection of the gas cooling and cleaning plants is critical with regards to technical feasibility, economic acceptability, and environmental compatibility. Further, gas cleaning systems are to be designed to the highest levels of cleaning efficiency, safety, and reliability while providing the best possible environmental protection.
Important criteria for the design of a gas cleaning system are (i) gas volume in N cum per hour, (ii) chemical composition of the gases, (iii) moisture content of the gases, (iv) temperature of the gases, (v) dust content of the gases in kg per hour, (vi) characteristics of the dust, like corrosive, abrasive etc., (vii) particle size range of the dust, (viii) emission standards, (ix) explosion characteristic of the gas, (x) hygienic design, (xi) on line or off line systems, and (xii) materials of construction.
There are three major considerations in the design of the gas cleaning system. The first is the hood which is to be designed to capture the dust and gases emitted and prevent a smoke filled working area. The second is that the gas and dust caught by the hood is to be cleaned before release to the atmosphere. The third is that the dust collected is to be disposed of in such a way that it does not get re-entrained in the air or a stream to become a pollution problem all over again.
The flue gases coming from the metallurgical furnaces are frequently having high temperatures (700 deg C to 1,500 deg C or even higher) and high dust content. Hence, before treating these gases in the gas cleaning system, these gases are to be cooled down to a temperature below 400 deg C. Several gas cooling methods are used in practice. These are (i) waste heat boiler, (ii) indirect cooling with air, (iii) indirect cooling with water, and (iv) evaporative cooling with water.
Waste heat boilers are mainly used for cooling the flue gases from those technological processes which produce flue gases with continuous gas flow rates. This allows the application of gas cooling using waste heat boiler with good operating results.
The system for cooling the gases by indirect cooling with air is used relatively seldom in practice because of several disadvantages such as (i) cooling air has a lower temperature than the dew point of the process gas, and condensation of the acid takes place on the cooler walls which causes corrosion of the equipment, (ii) risk of accretions and blockage due to sticky dust, (iii) the retention time of the gas at high temperature (higher than 550 deg C) is long which causes the formation of additional SO3 and raises the dew point of the gas, and (iv) in case of fluctuating gas flow rate, it is difficult to control the gas outlet temperature.
The system for cooling the gases by indirect cooling with water is frequently being used. In this case, flue duct is having water tubes around its circumference through which the cooling water flows. The sizing of the tubes and the water parameters (pressure and flow) are to be such that the temperature of the heated water always remains below the levels of its evaporation. While the system avoids steam and the regulatory issues associated with the steam handling, the disadvantage of the system relates to the bulkier equipment and handling of larger volume of cooling water.
Evaporative cooling with water is a suitable technical alternative to indirect air cooling or waste heat boilers for cooling gases with fluctuating gas flow rate. Modern evaporative cooling equipments are using a special kind of spray nozzles, the so called two-component nozzles (water and pressurized air), which allow flexible operation and sensitive controlling of the gas temperature at cooler outlet. This feature is very important in order to avoid any excessive gas temperature drop which can cause a condensation of acid mist and as a consequence wetting of the dust and the creation of wet dust deposits, in the subsequent hot gas precipitators as well as corrosion. The advantages of using evaporative coolers are (i) evaporative cooling reduces the formation of additional SO3 in the gas because of a short retention time of the gas upstream the evaporative cooler at high temperature (above 550 deg C) in the presence of particulate metal compounds which are acting as catalysts. (ii) downstream of the evaporative cooler the formation of SO3 is inhibited, (iii) conditioning of the gas with water for better performance of the electro-static precipitator (ESP), and (iv) no internals like guide vanes are needed.
There are around 40 different types of gas-cleaning devices which are available today, and, based on shared characteristics , they can be grouped into five main types namely (i) mist eliminators, (ii) dust extractors (also sometimes called dust catcher) and cyclones, (iii) wet dedusters, (iv) filters, and (v) ESPs. Further, the gas cleaning systems can be based upon technologies for dry dust separation or technologies for wet dust separation. In the dry dust separation technologies, conditioning of gas with water can be needed based on the requirements of the technological process. The conditioning of the gas is carried out by injecting of water together with nitrogen in a conditioning tower to produce a water mist with droplets having a diameter of around 150 micrometers. The residence time of gas in tower is controlled in such a manner that all the droplets are fully evaporated at the outlet of the conditioning tower.
Particulate removal devices operate basically on the principle that a gas stream containing particles is passed through a region where the particles are acted on by external forces or caused to intercept obstacles, thereby separating them from the gas stream. When acted upon by external forces, the particles acquire a velocity component in a direction different from that of the gas stream. In order to design a separation device based on particulate separation by external forces, it is essential to compute the motion of a particle under such conditions.
A preliminary selection of suitable particulate emission control systems is normally based on the knowledge of four items namely (i) particulate concentration in the stream to be cleaned, (ii) the size distribution of the particles to be removed, (iii) the gas flow rate, and (iv) the final allowable particulate emission rate. Once the systems which are capable of providing the required efficiencies at the given flow rates have been chosen, the ultimate selection is normally made on the basis of the total cost of construction and operation. The size of a collector, and hence its cost, is directly proportional to the volumetric flow rate of gas which is to be cleaned. The operating factors which influence the cost of a device are the pressure drop through the unit, the power required, and the quantity of water needed (in case of a wet scrubbing system). Devices which remove particles from gas streams rely on one or more of the following physical mechanisms.
Sedimentation – The particle-containing gas stream is introduced into a device or chamber where the particles settle under gravity to the floor of the chamber. Devices of this type are called settling chambers.
Migration of charged particle in an electric field – The particle-containing gas stream is introduced into a device in which the particles are charged and then subjected to an electric field. The resulting electrostatic force on the particles causes them to migrate to one of the surfaces of the device, where they are held and collected. Devices of this type are called ESP.
Inertial deposition – When a gas stream changes direction as it flows around an object in its path, suspended particles tend to keep moving in the original direction due to their inertia. Particulate collection devices based on this principle include cyclones, scrubbers, and filters.
Brownian diffusion – Particles suspended in a gas are always in Brownian motion. Brownian motion is the random motion of particles suspended in a medium. This pattern of motion typically consists of random fluctuations in a particle’s position inside a fluid sub-domain, followed by relocation to another sub-domain. When the gas stream flows around obstacles, the natural random motion of the particles bring them into contact with the obstacles, where they adhere and are collected. Since the Brownian motion is more pronounced with the smaller particles, it is expected that devices based on diffusion as the separation mechanism are most effective for small particles.
The key parameter which influences the choice of which device is to be employed in a particular case is the particle diameter ‘Dp’. The physical mechanisms as given above vary greatly in their effectiveness depending on the size of the particle. Thus the effectiveness of particulate removing devices is a function of particle size.
The collection efficiency ‘N(Dp)’ of a gas cleaning device for particles of diameter ‘Dp’ is defined by the equation N(Dp) = 1- (number of particles of diameter Dp per cum of gas out) / ( number of particles of diameter Dp per cum of gas in). The overall efficiency ‘N’ of the device based on particle number is given by the equation N = 1 – (number of particles per cum of gas out) / (number of particles per cum of gas in). These efficiencies can be expressed in terms of the particle size distribution functions at the inlet and outlet sides of the device.
There are several different classes of particulate control equipment. The simplest particulate control device is a settling chamber, a large chamber in which the gas velocity is slowed, allowing the particles to settle out by gravity. A cyclone operates by causing the entire gas stream to flow in a spiral pattern inside a tapered tube. Because of the centrifugal force, particles migrate outward and collect on the wall of the tube. The particles slide down the wall and fall to the bottom, where they are removed. The clean gas normally reverses its flow and exits out of the top of the cyclone.
The ESP utilizes the electrostatic force on charged particles in an electric field to separate particles from the gas stream. A high voltage drop is established between two electrodes, and particles passing through the resulting electric field acquire charge. The charged particles migrate to and are collected on an oppositely charged plate while the clean gas flows on through the device. Periodically, the plates are cleaned by rapping to shake off the layer of dust which has accumulated.
A variety of filters operate on the principle that the particulate-laden gas is forced through an assemblage of collecting elements, such as a fiber or a filter mat. As the gas passes through the assemblage, particles accumulate on the collectors.
Wet collection devices called scrubbers operate on the basis of the collision of particles with droplets of water which can easily be separated from the gas because of their large size.
Mechanical collectors such as settling chambers or cyclones are typically much less expensive than the others but normally are only moderately efficient in particle removal. Since they are much better for large particles than for small ones, they are frequently being used as pre-cleaners for the more efficient final control devices, especially at high particulate loadings. ESPs can treat large volumetric flow rates of gas at relatively low pressure drops with very high removal efficiencies. However, ESPs are expensive and are relatively inflexible to changes in process operating conditions. Fabric filters (bag filters) tend to have very high efficiencies but are expensive and are normally limited to dry, low-temperature conditions. Scrubbing can also achieve high efficiencies and offers the auxiliary advantage that gaseous pollutants can be removed simultaneously with particles. However, scrubbers can be expensive to operate, owing to their high pressure drop and to the fact that they produce a wet sludge which is required to be treated or disposed of.
Gravitational settling is perhaps the most obvious means of separating particles from a flowing gas stream. A settling chamber is simply a horizontal chamber through which the particle-laden gas flows and to the floor of which the dust particles settle. It is, in principle, simply a large box through which the effluent gas stream flows and in which particles in the stream settle to the floor by gravity. Gas velocities through the settling chamber are to be kept low enough so that settling particles are not re-entrained. The gas velocity is normally reduced by expanding the ducting into a chamber large enough so that sufficiently low velocities result. Although in principle settling chambers can be used to remove even the smallest particles, practical limitations in the length of such chambers restrict their applicability to the removal of particles larger than around 50 micrometers. Thus settling chambers are normally used as pre-cleaners to remove large and possibly abrasive particles, prior to passing the gas stream through other collection devices.
Settling chambers offer the advantages of (i) simple construction and low cost, (ii) small pressure drops, and (iii) collection of particles without need for water. The main disadvantage of settling chambers is the large space which they need. In fact, the chamber can contain a number of relatively closely spaced horizontal plates so that the distance that a particle is to settle to be collected is considerably smaller than the height of the overall device. Fig 1 shows a simple gravity settling chamber design.
Fig 1 A simple gravity settling chamber design
In analyzing the performance of a settling chamber, the key feature is the nature of the gas flow through the device. In this regards, three basic idealized flow situations can be distinguished namely (i) laminar flow, (ii) plug flow (velocity uniform across the cross section) with no vertical mixing of particles, and (iii) plug flow with complete vertical mixing of particles.
Laminar flow is characterized by a parabolic-type velocity profile. Such a flow is only realized for Reynolds numbers below that for transition to turbulent flow. In a laminar flow, the time required for a particle at height ‘h’ above the floor of the chamber to settle is ‘h/V’ where V is the particle’s settling velocity. Vertical mixing of particles is absent in laminar flow. The effect of Brownian motion is normally neglected relative to the steady downward movement due to settling.
In the laminar flow settling chamber the gas velocity profile is parabolic, and as a particle below the centre streamline settles, it encounters fluid moving more slowly, and thus its residence time in the chamber increases over what it has been on the higher streamline. On the other hand, particles initially above the centre streamline encounter faster moving streamlines as they fall until they pass the centre streamline.
The second flow category is the plug flow with no vertical mixing of the particles. This type of flow is, in a sense, an approximation to laminar flow in that vertical mixing of particles is still ignored, but a flat velocity profile is assumed and the particles all settle at their settling velocities. The second type of flow situation is that of plug flow with no vertical mixing of particles. In this situation, it is assumed that the particles are distributed uniformly across the entrance to the chamber. Whether a particle is collected is determined solely by the height ‘h’ at its entrance above the collecting surface. A critical height ‘h*’ can be defined such that all particles entering with ‘h’ less than or equal to ‘h*’ are collected and those for which’ h’ is greater than ‘h*’ escape collection.
The third category, plug flow with thorough vertical mixing, is the turbulent flow. In a turbulent flow settling chamber, the gas velocity is assumed to be uniform across the chamber due to the turbulent mixing. Moreover, the turbulent mixing in the core of the chamber overwhelms the tendency of the particles to settle and maintains a uniform particle concentration vertically across the chamber. Removal by settling can be assumed to occur in a thin layer at the bottom of the chamber.
The flow in a rectangular channel can be assumed to be turbulent if the Reynolds number is greater than 4,000. In the laminar flow settling chamber, particles settle at all heights above the floor of the chamber, the key to the analysis being to calculate the overall residence time of the particles as they fall across streamlines. The mechanism of collection in a turbulent flow settling chamber is, although ultimately based on the settling of particles under gravity, rather different from that in the laminar flow chamber. The difference is due to the turbulent flow in the chamber. In the bulk flow in the chamber, turbulent mixing is vigorous enough so that particles are overwhelmed by the flow and do not settle. It is assumed that the turbulent mixing maintains a uniform particle concentration over the height of the chamber. Very near the floor of the chamber, a thin layer can be assumed to exist across which particles settle the short distance to the floor. Thus, once a particle, vigorously mixed in the core of the flow, enters this layer, it settles to the floor.
Cyclone separators are gas cleaning devices which utilize the centrifugal force created by a spinning gas stream to separate particles from a gas. A standard tangential inlet vertical reverse flow cyclone separator is shown in Fig 2. The gas flow is forced to follow the curved geometry of the cyclone while the inertia of particles in the flow causes them to move toward the outer wall, where they collide and are collected. A particle of mass ‘m’ moving in a circular path of radius ‘r’ with a tangential velocity ‘vA’ is acted on by a centrifugal force of ‘m(vA)2/r’. At a typical value of ‘vA’ = 10 m/s, ‘r’ = 0.5 m, this force is 20.4 times that of gravity on the same particle. Thus it can be seen that the substantially enhanced force on the particle over that of settling alone can be achieved for cyclone geometry.
In a cyclone, the particles in the spinning gas stream move progressively closer to the outer wall as they flow through the device. As shown in Fig 2, the gas stream can execute several complete turns as it flows from one end of the device to the other. For the design of a cyclone separator, the given gas flow rate and inner and outer radii, the length of the body of the cyclone is to ensure that a desired collection efficiency for particles of a given size be attained. Since the length of the body of a cyclone is related through the gas flow rate to the number of turns executed by the gas stream, the design frequently consists of computing the number of turns needed to achieve specified collection efficiency.
There are a variety of designs available for the cyclone separators which differ in the manner in which the rotating motion is imparted to the gas stream. Conventional cyclones can be of three categories namely (i) reverse-flow cyclones (tangential inlet and axial inlet), (ii) straight-through-flow cyclones, and (iii) impeller collectors.
Fig 2 shows a conventional reverse-flow cyclone with a tangential inlet. The dirty gas enters at the top of the cyclone and is given a spinning motion because of its tangential entry. Particles are forced to the wall by centrifugal force and then fall down the wall due to gravity. At the bottom of the cyclone the gas flow reverses to form an inner core which leaves at the top of the cyclone. In a reverse-flow axial-inlet cyclone, the inlet gas is introduced down the axis of the cyclone, with centrifugal motion being imparted by permanent vanes at the top.
Fig 2 Tangential inlet vertical reverse flow cyclone
In straight-through-flow cyclones the inner vortex of air leaves at the bottom (rather than reversing direction), with the initial centrifugal motion being imparted by the vanes at the top. This type is used frequently as a pre-cleaner to remove large particles. The main advantages of this cyclone are low pressure drop and high volumetric flow rates.
In the impeller collector, gases enter normal to a many-bladed impeller and are swept out by the impeller around its circumference while the particles are thrown into an annular slot around the periphery of the cyclone. The principal advantage of this cyclone is its compactness. The main disadvantage of the cyclone is a tendency toward plugging from solid build-up in the cyclone.
Cyclones can be constructed of any material, metal or ceramic. They are capable of withstanding high temperatures, abrasive particles, or corrosive atmospheres. It is necessary that the interior surface be smooth so that the collected particles can slide easily down the wall to the hopper. There are no moving parts in a cyclone, hence so operation is normally simple and relatively free of maintenance. Their low capital cost and maintenance-free operation make them ideal for use as pre-cleaners for more efficient final control devices, such as ESPs. Although cyclones have traditionally been regarded as relatively low efficiency collectors, some cyclones which are presently available can achieve efficiencies which are higher than 98 % for particles larger than 5 micrometers. Normally, cyclones routinely achieve efficiencies of 90 % for particles larger than 15 micrometers to 20 micrometers.
Fig 3 shows a particle entering tangentially onto a horizontal plane of a spinning gas stream at r3 is considered. Because of a centrifugal force of ‘m(vA)2/r’, the particle follows a path outward across the flow streamlines. Its velocity vector has a tangential component (vA) and a radial component (Vr). There is also an axial component (vZ).
The so-called laminar flow cyclone separator does not have laminar flow in the sense of the laminar flow settling chamber, but rather a frictionless flow in which the streamlines follow the contours of the cyclone as shown in Fig 3.
Fig 3 laminar flow and turbulent flow cyclones
The model of the turbulent flow cyclone separator is shown in Fig 3. Because of turbulent mixing the particle concentration is assumed to be uniform across the cyclone, and, as in the case of the turbulent flow settling chamber, removal occurs across a thin layer at the outer wall.
Cyclone collection efficiency increases with increasing (i) particle size, (ii) particle density, (iii) inlet gas velocity, (iv) cyclone body length, (v) number of gas revolutions, and (vi) smoothness of the cyclone wall. On the other hand, cyclone efficiency decreases with increasing (i) cyclone diameter, (ii) gas outlet duct diameter, and (iii) gas inlet area. For any specific cyclone, whose ratio of dimensions is fixed, the collection efficiency increases as the cyclone diameter is decreased. The design of a cyclone separator represents a compromise among collection efficiency, pressure drop, and size. Higher efficiencies need higher pressure drops (i.e., inlet gas velocities) and larger sizes (i.e. body length). The dimensions required to specify a tangential-entry, reverse-flow cyclone are shown in Fig 4.
Besides collection efficiency the other major consideration in cyclone specification is pressure drop. While higher efficiencies are achieved by forcing the gas through the cyclone at higher velocities, to do so results in an increased pressure drop. Since increased pressure drop needs increased energy input into the gas, there is ultimately an economic trade-off between collection efficiency and operating cost. Cyclone pressure drops normally range from 250 Pa to 4,000 Pa.
ESP is one of the most widely used particulate control device. It has wide size ranges. The ESP chamber consists of two electrodes, the discharge and the collecting electrodes. Between the electrodes, the gas contains free electrons, ions, and charged particles. The species contributing to the space charge density are ions, electrons, and charged particles. The gas molecules capture all the free electrons so that only the ions and charged particles contribute space charge density. Actually, ionic current flows in the direction of the electric field consisting of ions charged with the same polarity as the charging electrode and moving to the collecting electrode. The ions migrate to the collecting electrode with a velocity large enough to be unaffected by the turbulent flow in the chamber.
The basic principle of operation of the ESP is that the particles are charged, and then an electric field is imposed on the region through which the particle-laden gas is flowing, exerting an attractive force on the particles and causing them to migrate to the oppositely charged electrode at right angles to the direction of gas flow. ESP differs from mechanical methods of particle separation in that the external force is applied directly to the individual particles rather than indirectly through forces applied to the entire gas stream (e.g. in a cyclone separator). Particles collect on the electrode. If the particles collected are liquid, then the liquid flows down the electrode by gravity and it is removed at the bottom of the device. If the particles are solid, the collected layer on the electrode is removed periodically by rapping the electrode.
Particle charging is achieved by generating ions by means of a corona established surrounding a highly charged electrode like a wire. The electric field is applied between that electrode and the collecting electrode. If the same pair of electrodes serves for particle charging and collecting, the device is called a single-stage ESP (Fig 4). A wire serving as the discharge electrode is suspended down the axis of a tube and held in place by a weight attached at the bottom. The sides of the cylinder form the collecting electrode. The collected particles which form a layer on the collecting electrode are removed to the dust hopper by rapping the collecting electrode. In a two-stage ESP, separate electrode pairs perform the charging and collecting functions.
Fig 4 Cylindrical single-stage electrostatic precipitator
Most industrially generated particles are charged during their formation by such means as flame ionization and friction, but normally only to a low or moderate degree. These natural charges are far too low for electrostatic precipitation. The high-voltage DC (direct current) corona is the most effective means for particle charging and is universally used for electrostatic precipitation. The corona is formed between an active high voltage electrode such as a fine wire and a passive ground electrode such as a plate or pipe. The corona surrounding the discharge electrode can lead to the formation of either positive or negative ions which migrate to the collecting electrode. The ions, in migrating from the discharging to the collecting electrode, collide with the particulate matter and charge the particles.
Since the gas molecule ions are many orders of magnitude smaller than even the smallest particles and because of their great number, virtually all particles that flow through the device become charged. The charged particles are then transported to the collecting electrode, to which they are held by the electrostatic attraction. The particles build a thickening layer on the collecting electrode. The charge slowly bleeds from the particles to the electrode. As the layer grows, the charges on the most recently collected particles are to be conducted through the layer of previously collected particles. The resistance of the dust layer is called the dust resistivity.
As the particle layer grows in thickness, the particles closest to the plates lose most of their charge to the electrode. As a result, the electrical attraction between the electrode and these particles is weakened. However, the newly arrived particles on the outside layer have a full charge. Because of the insulating layer of particles, these new particles do not lose their charge immediately and thus serve to hold the entire layer against the electrode. Finally, the layer is removed by rapping, so that the layer breaks up and falls into a collecting hopper. ESPs are normally employed for gas cleaning when the volumetric throughput of gas is high.
The mechanism for particle charging in a ESP is the generation of a supply of ions which attach themselves to the particles. The corona is the mechanism for forming ions. The corona can be either positive or negative. A gas normally has a few free electrons and an equal number of positive ions, a situation which is exploited in generating a corona. When a gas is placed between two electrodes, small amount of current results as the free electrons migrate to the positive electrode and the positive ions migrate to the negative electrode.
In the positive corona discharge electrode, the wire in the cylindrical ESP (Fig 4) is at a positive potential. The few free electrons normally present in the gas migrate toward the wire. As the electrons approach the wire, the electrons’ energy is increased because of an increase in the attractive force. These free electrons collide with gas molecules, the collision leading in some cases to the ejection of an electron from the molecule, producing two free electrons and a positive ion. The two free electrons continue toward the positive electrode, gaining energy, until they collide with two more gas molecules, producing four free electrons and two positive ions. This process is referred to as an electron avalanche.
The positive ions formed migrate to the negative electrode. It is these positive ions which migrate across the entire device to the negative electrode that collide with and attach to the particles in the gas. The region immediately surrounding the wire in which the electron avalanche is established is the corona. Thus, with a positive corona the particles become positively charged. The term ‘corona’ arises from the fact that the electron avalanche is frequently accompanied by the production of light. In the negative corona the discharge electrode is maintained at a negative potential.
The electron avalanche begins at the outer surface of the wire and proceeds radially outward. Close to the wire the electrons are sufficiently energetic to form positive ions upon collision with gas molecules, thus initiating the electron avalanche. The positive ions formed migrate the short distance to the wire. As the electrons migrate outward into a region of lower electric field strength, they are slowed down by collisions with gas molecules. These electrons eventually have lower energy than those which are accelerated toward the positive electrode in the positive corona. These relatively low energy electrons, rather than ejecting an electron from the gas molecule upon collision, are absorbed by the gas molecules to produce negative ions. The formation of negative ions, which begins to occur at the outer edge of the corona, essentially absorbs all the free electrons produced in the electron avalanche at the wire surface. These negative ions then migrate to the positive electrode, in the course of which attaching to gas molecules and forming negative ions.
For a negative corona to be effective, it is necessary that the gas molecules can absorb free electrons to form negative ions. Sulphur dioxide is one of the best electron absorbing gases of those present in flue gases. Oxygen, CO2, and H2O are also effective electron absorbers. The negative corona is normally more stable than the positive corona, so it is preferred in most industrial applications. A by-product of the negative corona is the production of ozone (O3). The positive corona does not need an electron-absorbing gas.
As the ESP is operated, a layer of the collected material builds up on the collecting electrode. Particle deposits on the precipitator collection surface are to possess at least a small degree of electrical conductivity in order to conduct the ion currents from the corona to ground. The minimum conductivity required is around 10 to the power -10 per ohm-centimeter (resistivity of 10 to the power 10 ohm-centimeter). This conductivity is small compared to that of ordinary metals but is much greater than that of good insulators such as silica and most plastics. The resistivity of a material is determined by establishing a current flow through a slab of known thickness of the material.
As long as the resistivity of the collected dust layer is less than about 10 to the power 10 ohm-centimeter, the layer surrender its charge to the electrode. At the room temperature, a typical dust has a resistivity of around 10 to the power 8 ohm-centimeter. This is because of a layer of water on the surface of the particles. As the temperature is increased beyond 100 deg C, the water is evaporated and the resistivity increases to a value characteristic of the collected solids. When the resistivity of the layer exceeds around 10 to the power 10 ohm-centimeter, the potential across the layer increases so that the voltage which can be maintained across the ESP decreases and the collection efficiency decreases. The electrical resistivity of collected particulate matter depends on its chemical composition, the constituents of the gas, and the temperature.
Particle charging in ESP occurs in the gas space between the electrodes where the gas ions generated by the corona bombard and become attached to the particles. The gas ions can reach concentrations as high as 10 to the power 15 ions per cubic meter. The level of charge attained by a particle depends on the gas ion concentration, the electric field strength, the conductive properties of the particle, and the particle size. A 1 micrometer particle typically acquires the order of 300 electron charges, whereas a 10 micrometer particle can attain 30,000 electron charges. Predicting the level of charge acquired by a particle is necessary in order to predict the particle’s migration velocity, on the basis of which the collection efficiency can be calculated for a given set of operating conditions.
There are actually two mechanisms by which particles become charged in an ESP. In the first mechanism particle charging occurs when ions which are migrating toward the collecting electrode encounter particles to which they become attached. In migrating between the electrodes the ions follow the electric flux lines, which are curves everywhere tangent to the electric field vector. When the particle first enters the device and is uncharged, the electric flux lines deflect toward the particle, resulting in the capture of even a larger number of ions than are captured if the ions have followed their normal path between the electrodes. As the particle becomes charged, ions begin to be repelled by the particle, reducing the rate of charging. Eventually, the particle acquires a saturation charge and charging ceases. This mechanism is called ion bombardment or field charging.
The second mode of particle charging is diffusion charging, in which the particle acquires a charge by virtue of the random thermal motion of ions and their collision with and adherence to the particles. Diffusion charging occurs as the ions in their random thermal motion collide with a particle and surrender their charge to it. In that sense the mechanism of diffusion charging is identical to that of the diffusion of uncharged vapour molecules to the surface of a particle. However, because both the particle and the ions are charged, the random thermal motion of the ions in the vicinity of a particle is influenced by an electrostatic force. This force gives rise to a tendency of the ions to migrate away from the particle as the particle charge increases. The overall flux of ions to a particle hence is both the random diffusive motion and the electrical migration.
The theories of both field and diffusion charging, in their full generality, are quite complex and have received a great deal of attention. Strictly speaking, field and diffusion charging occur simultaneously once a particle enters an ESP, and hence to predict the overall charge acquired by a particle, one is to consider the two mechanisms together. However, since the diffusion charging is predominant for particles smaller than around 1 micrometer in diameter and field charging is predominant for particles larger than around 1 micrometer, the two mechanisms frequently are treated in ESP design as if they occur independently. In doing so, one estimates the total charge on a particle as the sum of the charges resulting from each of the two separate mechanisms.
Filtration of particles from gas streams
A major class of particulate air pollution control devices relies on the filtration of particles from gas streams. A variety of filter media is employed, including fibrous beds, packed beds, and fabrics. Fibrous beds used to collect airborne particles are typically quite sparsely packed, usually only around 10 % of the bed volume being fibers. Packed bed filters consist of solid packing normally in a tube and tend to have higher packing densities than do fibrous filters. Both fibrous and packed beds are widely used in the ventilation systems. Fabric filters are frequently used to remove solid particles from industrial gases, whereby the dusty gas flows through fabric bags and the particles accumulate on the cloth.
The physical mechanisms by which the filtration is accomplished vary depending on the mode of filtration. Conventional sparsely packed fibrous beds can be viewed as assemblages of cylinders. In such a filter, the characteristic spacing between fibers is much larger than the size of the particles being collected. Thus the mechanism of collection is not simply sieving, in which the particles are trapped in the void spaces between fibers. Rather, the removal of particles occurs by the transport of particles from the gas to the surface of a single collecting element. Because the filtration mechanisms in a fibrous bed can be analyzed in terms of a single collector, it is possible to describe them in considerable theoretical detail.
Packed-bed filters are sometimes viewed as assemblages of interacting, but essentially separate, spherical collectors, although the close proximity of individual packing elements casts doubt as to the validity of this approach. Because of the relatively closer packing in packed-bed filters, and the resulting difficulty of describing the particle collection process in clean theoretical terms, predicting collection in such systems is more empirically based than for fibrous filters. Fabric filter efficiencies must be predicted strictly empirically since the accumulated particle layer actually does the collecting.
A fibrous filter bed is viewed as a loosely packed assemblage of single cylinders. Even though the fibers are oriented in all directions in the bed, from a theoretical point of view the bed is treated as if every fiber is normal to the gas flow through the bed. The solid fraction of the filter is normally of the order of only 10 %. In addition, each fiber acts more or less independently as a collector. Thus, to compute the particle removal by a filter bed, one basically needs to determine the number of fibers per unit volume of the bed and then multiply that quantity by the efficiency of a single fiber.
The basis of predicting the collection efficiency of a filter bed is the collection efficiency of a single filter element in the bed. The filter element is taken as an isolated cylinder normal to the gas flow. Three distinct mechanisms as given below can be identified whereby particles in the gas reach the surface of the cylinder.
As per the first mechanism, the particles in a gas undergo Brownian diffusion which brings some particles in contact with the cylinder due to their random motion as they are carried past the cylinder by the flow. A concentration gradient is established after the collection of a few particles and acts as a driving force to increase the rate of deposition over that which occurs in the absence of Brownian motion. Because the Brownian diffusivity of particles increases as particle size decreases, it is normally expected that this removal mechanism is the most important for very small particles. When analyzing collection by Brownian diffusion, the particles are treated as diffusing mass-less points.
As per the second mechanism, interception takes place when a particle, following the streamlines of flow around a cylinder, is of a size sufficiently large that its surface and that of the cylinder come into contact. Thus, if the streamline on which the particle centre lies is within a distance Dp /2 of the cylinder, interception occurs. Here Dp is the particle diameter.
As per the third mechanism, inertial impaction occurs when a particle is unable to follow the rapidly curving streamlines around an obstacle and, because of its inertia, continues to move toward the obstacle along a path of less curvature than the flow streamlines. Thus, collision occurs because of the particle’s momentum. It is to be noted that the mechanism of inertial impaction is based on the premise that the particle has mass but no size, whereas interception is based on the premise that the particle has size but no mass.
Collection can also result from electrostatic attraction when either particles or fiber or both possess a static charge. These electrostatic forces can be either direct, when both particle and fiber are charged, or induced, when only one of them is charged. Such charges are normally not present unless deliberately introduced during the production of the fiber.
The size ranges in which the various mechanisms of collection are important are (i) Inertial impaction – greater than 1 micrometer, (ii) Interception – greater than 1 micrometer, (iii) diffusion – less than 0.5 micrometer, and (iv) electrostatic attraction – 0.01 micrometer to 5 micrometer. It is normal to analyze the mechanisms of collection separately and then combine the individual efficiencies to give the overall collection efficiency for the cylinder or other obstacle.
Most developments of particle collection assume, for lack of better information, that particles transported to the surface of a fiber are retained by the fiber. Experiments have shown, however, that for a variety of substances and filter media, the fraction of particles striking the collector surface which adhere is generally less than unity and can in some cases be as low as 0.5.
Industrial fabric filtration is normally accomplished in a so-called bag- house, in which the particle-laden gases are forced through filter bags. Particles are normally removed from the bags by gravity. Fig 5 shows three bag-house designs, in which cleaning is accomplished by vibration (Fig 5a), air jet [Fig 5b), or traveling ring [Fig 5c).
Fig 5 Designs of bag house filters
The fabric filtration process consists of three phases. First, particles collect on individual fibers by the above described mechanisms. Then an intermediate stage exists during which particles accumulate on previously collected particles, bridging the fibers. Finally, the collected particles form a cake in the form of a dust layer which acts as a packed bed filter for the incoming particles. As the dust layer accumulates, the pressure drop across the filter increases, and periodically the dust layer is to be dislodged into the hopper at the bottom to ‘regenerate’ the fabric bag. High efficiencies are attainable with fabric filters, particularly in treating combustion gases from the technological processes. To the extent that effective operation of an ESP depends on the presence of SO2 in the gas as an ionizable species, fabric filters can operate with no loss of efficiency with low-sulphur level.
Fabric filters consist of semi-permeable woven or felted materials which constitute a support for the particles to be removed. A brand-new woven filter cloth has fibers roughly 100 micrometers to 150 micrometers in diameter with open spaces between the fibers of 50 micrometers to 75 micrometers. Initially, the collection efficiency of such a cloth is low because most of the particles pass directly through the fabric. However, deposited particles quickly accumulate, and it is the deposited particle layer that enables the high-efficiency removal once a uniform surface layer has been established.
Although fiber mat filters are similar in some respects to fabric filters, they do not depend on the layer of accumulated particles for high efficiency. Fiber mat filters generally are not cleaned but are discarded. They are ordinarily used when particle concentrations are low, so that reasonable service life can be achieved before discarding.
In a fabric filter the particle layer performs the removal task. As the layer of collected particles grows in thickness, there is an increase in the pressure drop across the particle layer and the underlying fabric. The two major considerations in the design of a fabric filter assembly are the collection efficiency and the pressure drop as a function of time of operation (since the last cleaning). The collection efficiency depends on the local gas velocity and the particle loading on the fabric.
Fabric filters offer the several advantages such as (i) they can achieve very high collection efficiencies even for very small particles, (ii) they can be used for a wide variety of particles, (iii) they can operate over a wide range of volumetric flow rates, and (iv) they need only moderate pressure drops. The limitations of fabric filters are namely (i) operation is to be carried out at temperatures lower than that at which the fabric is destroyed, or its life is shortened to an uneconomical degree, (ii) gas or particle constituents which attack the fabric or prevent proper cleaning, such as sticky particles difficult to dislodge, are to be avoided, and (iii) bag houses need large floor areas. The advantages of fabric filter bag houses clearly outweigh their limitations.
Wet collectors, or scrubbers, employ water washing to remove particles directly from a gas stream. Scrubbers can be grouped broadly into two main classes namely (i) those in which an array of liquid drops (sprays) form the collecting medium, and (ii) those in which wetted surfaces of various types constitute the collecting medium. The first class includes spray towers and venturi scrubbers, while the second includes plate and packed towers.
Scrubbing is a very effective means of removing small particles from a gas. Removal of particles results from collisions between particles and water drops. In the humid environment of a scrubber, small, dry particles also grow in size by condensation of water and thereby become easier to remove. Re-entrainment of particles is avoided since the particles become trapped in droplets or in a liquid layer. A scrubber also provides the possibility of simultaneously removing soluble gaseous pollutants. The particle-laden scrubbing liquid is to be disposed of, a problem not encountered in dry methods of gas cleaning.
A spray scrubber is a device in which a liquid stream is broken into drops, approximately in the range 0.1 mm to 1 mm in diameter, and introduced into the particle laden gas stream. The array of moving drops becomes a set of targets for collection of the particles in the gas stream. Collection efficiency is computed by considering the efficiency of a single spherical collector and then summing over the number of drops per unit volume of gas flow. The relative motion between the drops and particles is an important factor in the collection efficiency since capture occurs by impaction and direct interception. Diffusion is also important for smaller particles.
There are two general types of spray scrubbers. The first class comprises those having a preformed spray where drops are formed by atomizer nozzles and sprayed into the gas stream. These include (i) counter-current gravity tower, where drops settle vertically against the rising gas stream, (ii) cross-current tower, where drops settle through a horizontal gas stream, and (iii) co-current tower, where spray is horizontal into a horizontal gas stream.
The second class comprises those in which the liquid is atomized by the gas stream itself. Liquid is introduced more or less in bulk into a high-velocity gas flow which shatters the liquid into drops. Devices in this class are called venturi scrubbers since the high velocity gas flow is achieved in a venturi (a contraction). Fig 6 shows four types of wet collection equipment.
Fig 6 Types of wet collection equipment
The simplest type of wet collector is a spray tower into which water is introduced by means of spray nozzles (Fig 6a). Gas flow in a spray chamber is counter-current to the liquid, the configuration leading to maximum efficiency. Collection efficiency can be improved over the simple spray chamber with the use of a cyclonic spray tower, as shown in Fig 6b. The liquid spray is directed outward from nozzles in a central pipe. An unsprayed section above the nozzles is provided so that the liquid drops with the collected particles have time to reach the walls of the chamber before exit of the gas. An impingement plate scrubber, as shown in Fig 6c, consists of a tower containing layers of baffled plates with holes (5,000 to 50,000 per square meter) through which the gas must rise and over which the water must fall. Highest collection efficiencies of wet collectors are obtained in a venturi scrubber, shown in Fig 6d, in which water is introduced at right angles to a high-velocity gas flow in a venturi tube, resulting in the formation of very small water droplets by the flow and high relative velocities of water and particles. The high gas velocity is responsible for the breakup of the liquid. Aside from the small droplet size and high impingement velocities, collection is enhanced through particle growth by condensation. Different types of particle scrubbing devices are described below.
Plate scrubber – It is a vertical tower containing one or more horizontal plates (trays). Gas enters the bottom of the tower and must pass through perforations in each plate as the gas flows counter-current to the descending water stream. Plate scrubbers are normally named for the type of plates they contain (e.g. sieve plate tower). Collection efficiency increases as the diameter of the perforations decreases. A cut diameter, that collects with 50 % efficiency, of around 1 micrometer aerodynamic diameter can be achieved with 3.2 mm diameter holes in a sieve plate.
Packed-bed scrubber – It operates similarly to packed-bed gas absorber. Collection efficiency increases as packing size decreases. A cut diameter of 1.5 micrometers aerodynamic diameter can be attained in columns packed with 2.5 cm elements.
Spray scrubber – In this scrubber, particles are collected by liquid drops which have been atomized by spray nozzles. Horizontal and vertical gas flows are used, as well as spray introduced co-current, counter-current, or cross-flow to the gas. Collection efficiency depends on droplet size, gas velocity, liquid / gas ratio, and droplet trajectories. For droplets falling at their terminal velocity, the optimum droplet diameter for fine-particle collection lays in the range 100 micrometers to 500 micrometers. Gravitational settling scrubbers can achieve cut diameters of around 2 micrometers. The liquid / gas ratio is in the range 0.001 cum to 0.01 cum per cum of gas treated.
Venturi scrubber – A moving gas stream is used to atomize liquids into droplets. High gas velocities (60 m/sec to 120 m/s) lead to high relative velocities between gas and particles and promote collection.
Cyclone scrubber – Drops can be introduced into the gas stream of a cyclone to collect particles. The spray can be directed outward from a central manifold or inward from the collector wall.
Baffle scrubber – In this scrubber, there are changes in gas flow velocity and direction induced by solid surfaces.
Impingement-entrainment scrubber – The gas is forced to impinge on a liquid surface to reach a gas exit. Some of the liquid atomizes into drops which are entrained by the gas. The gas exit is designed so as to minimize the loss of entrained droplets.
Fluidized-bed scrubber – A zone of fluidized packing is provided where gas and liquid can mix intimately. Gas passes upward through the packing, while liquid is sprayed up from the bottom and / or flows down over the top of the fluidized layer of packing.
The collection efficiency of wet collectors can be related to the total energy loss in the equipment. The higher is the scrubber power per unit volume of gas treated, the better is the collection efficiency. Almost all the energy is introduced in the gas, and thus the energy loss can be measured by the pressure drop of gas through the unit. The major advantage of wet collectors is the wide variety of types, allowing the selection of a unit suitable to the particular removal problem. As regards disadvantages, high pressure drops (and hence energy requirements) are to be maintained, and the handling and disposal of large volumes of scrubbing liquid are to be undertaken.
In case of spray scrubbing, the conceptually simplest of the devices is a gravity spray chamber. Water droplets are introduced at the top of an empty chamber through atomizing nozzles and fall freely at their terminal settling velocities counter-currently through the rising gas stream. The particle-containing liquid collects in a pool at the bottom and is to be pumped out for treatment to remove the solids, and the cleaned liquid is normally recycled to the tower. The overall efficiency of a spray tower increases as the collection efficiency of a single drop increases, as the length of the chamber increases, and as the ratio of the volumetric flow rate of water to that of gas increases. It increases as the diameter of the drops decreases.
Venturi scrubbers are used when high collection efficiencies are needed and when most of the particles are smaller than 2 micrometers in diameter. There are a number of examples, in fact, where a venturi scrubber is the only practical device for a gas-cleaning application. If the particles to be removed are sticky, flammable, or highly corrosive, for example, ESPs and fabric filters cannot be used. Venturi scrubbers are also the only high-efficiency particulate collectors which can simultaneously remove gaseous species from the effluent stream.
The distinguishing feature of a venturi scrubber is a constricted cross section or throat through which the gas is forced to flow at high velocity. A typical venturi scrubber configuration is shown in Fig 7. The configuration includes a converging conical section where the gas is accelerated to throat velocity, a cylindrical throat, and a conical expander where the gas is slowed down. Liquid can be introduced either through tangential holes in the inlet cone or in the throat itself. In the former case, the liquid enters the venturi as a film on the wall and flows down the wall to the throat, where it is atomized by the high-velocity gas stream. In the latter, the liquid is injected perpendicular to the gas flow in the throat, atomized, and then accelerated. Gas velocities in the range 60 m/sec to 120 m/sec are achieved, and the high relative velocity between the particle laden gas flow and the droplets promotes collection. The collection process is essentially complete by the end of the throat. Because they operate at much higher velocities than ESPs precipitators or bag houses, venturi scrubbers are physically smaller and can be economically made of corrosion-resistant materials. Venturis have the simplest configuration of the scrubbers and are the smallest in size. Fig 7 shows the comparison of the efficiency of venturi scrubber with the efficiencies of other gas cleaning devices.
Fig 7 Venturi scrubber configuration and comparison of efficiencies
A typical range of liquid to gas flow rate ratios for a venturi scrubber is 0.001 cum to 0.003 cum of liquid per cum of gas. At the higher liquid / gas ratios, the gas velocity at a given pressure drop is reduced, and at lower ratios, the velocity is increased. For gas flow rates exceeding about 1,000 cum / minute venturi scrubbers are normally constructed in a rectangular configuration in order to maintain an equal distribution of liquid over the throat area.