Electrostatic Precipitators, Better Wet Than Dry - Technical Paper

CHEMICAL ENGINEERING/NOVEMBER 23, 1987 159 lectrostatic precipitation is made possible .by the corona .discharge. Through an effect known as the avalanche process, the corona discharge provides a simple and stable means of generating the ions to electrically charge and collect suspended particles. In the avalanche process, gases in the vicinity of a negatively charged surface break down to form a plasma (glow) region when the imposed voltage reaches a critical level (Fig. 1). Free electrons in this region are then repulsed toward the positive (ground) surface, and finally collide with gas molecules to form negative ions. These ions, being of lower mobility, form a space-charge cloud of the same polarity as the emitting surface. By restricting further emis-sion of high-speed electrons, the space charge tends to stabilize the corona. With a corona established, dust parti-cles in the area become charged by the ions present, and are driven to the posi-tive electrode by the electric field. Of course, for the foregoing to be success-ful, the proper electrode geometry, gas composition and voltage must be pre-sent. (Fortunately, the corona process works very well for most gas streams.) Particle charging is only the first step in the precipitation process. Once charged, the particles must be collect-ed. As explained, this happens as a matter of course because the same forces that cause a particle to acquire a charge also drive the like-polarity parti-cle to the ground surface. The next step is particle removal. It is here that the dry and wet precipita-tion processes differ. In the dry process, the collected material is removed by rapping the collecting surface to slough off the particles, which fall into a hopper. In the wet process, the material is rinsed from the collecting surface with an irrigating liquid. Analyzing performance The Deutsch equation describes precipitator efficiency under conditions of turbulent flow: E = 1- exp (-AW/Q) (1) Here, E = efficiency, 1- (outlet particle concentration/inlet particle concentration); A = area of the collecting surface; W = velocity of particle migration to the collecting surface; and Q = upward gas flowrate (gas velocity X cross-sectional area of the passage). The derivation of Eq. (1) depends on simplifying assumptions, the most important being: all particles are the same size; the gas-velocity profile is uniform; a captured particle stays captured; the electric field is uniform; no zones are untreated (i.e., no particles sneak by the electric field). Eq. (1) cannot be rigorously applied to describe the performance of dry precipitators because of nonidealities inherent in conventional designs. Among these: rapping and reentrainment losses, gasSteven A. Jaasund, Jaasund AirTech, Inc. Figure 1 –- In the avalanche process, the corona discharge generates ions that Electrically charge and collection suspended particles. 160 CHEMICAL ENGINEERING/NOVEMBER 23, 1987 sneakby, and resistivity effects of the dry-dust layer. In- stead, a modified Deutsch equation must be used, in which the term W (particle migration velocity) is replaced by anoth- er known as EMV (effective migration velocity). Empirically determined, EMV is a characterizing parameter that ac- counts for all the nonidealities mentioned, as well as for the true particle-migration velocity. Values for EMV used in the modified equation are considerably lower than true particle velocities calculated or measured in the laboratory. Wet precipitators do not suffer from, many of the nonidealities encountered by the dry type. Rapping and reentrain- ment losses and troublesome dust-layer resistivity effects, particularly, are eliminated by irrigating the collecting electrodes. Also, because the wet type is frequently configured for vertical gas flow, sneakby is avoided. Therefore, EMV values for wet precipitators are usually higher than for dry ones. This means that, for a specific application, a wet unit can be smaller than an equivalent dry one. This is additional- ly true because a wet precipitator operates on a cooled, lower-volume gas stream. A full-scale wet precipitator in an iron-ore sintering plant operating in parallel with a dry precipitator was only one-seventh the size of the latter in collecting area per unit of gas volume treated. Yet the wet precipitator consistently outperformed the dry unit [G]. When to consider a wet precipitator Because its collecting surface is cleaned by a liquid, the wet precipitator can be used for virtually any particle emission. However, most emission-control goals can be met with other equipment. For instance, a wet precipitator would provide excellent control of emissions from a coal-fired boiler; but, if flyash resistivity were in a reasonable range, dry precipita- tion would be the technology of choice. Similarly, fabric filtration is preferred for electric arc furnaces and scrubbing is customarily used on lime kilns, even though wet precipita- tion would also work. Nevertheless, there are many services for which a wet precipitator should be carefully considered, and even some for which wet precipitation should be the technology of choice. Some such conditions occur when: the gas stream has already been treated in a wet scrubber; the temperature of the gas stream is low and its moisture content is high; gas and particles must be simultaneously removed; the loading of submicron particles is high and removal must be very efficient; liquid particles are to be collected; and the dust to be collected is best handled in liquid. Unlike with other gas cleaning methods, the applicability of wet electrostatic precipitation strongly depends on the particular design. Thus, the pulse-jet mechanism of a specific design of fabric filter usually will not affect the viability of gas cleaning. In the case of the wet precipitator, however, some designs will not be suitable for certain services. For instance, a precipitator for gas streams containing adherent particles must be continuously, not intermittently, irrigated. Types of wet precipitators The design of wet precipitators can be characterized by configuration, arrangement, irrigating method, and materi- als of construction. Configuration - There are two basic precipitator config-urations: plate and tube. The first consists of parallel plates with discharge elements assembled between each plate (Fig. 2). The second consists of an array of tubes, round or multisided, with a discharge electrode located in the center of each (Fig. 3). Arrangement - Gas flow can be arranged in parallel or series, and horizontally or vertically. This feature also distin-guishes a wet from a dry precipitator; because particles are removed from the latter through rapping, it is always ar- ranged horizontally. Irrigation method - This has a greater impact on the operation of a wet precipitator than any other factor. There are many irrigation methods. . In self-irrigation, the most common method, captured liquid droplets wet the collecting surface. Of course, this method only works when the particles are mostly liquid. In a specialized variation, condensation from the gas stream wets the collecting surfaces. A cold fluid, usually air, is circulated on the outside of the collecting tube to promote condensa- tion. As with mist collectors, irrigation by condensation works best with a gas stream high in moisture content and low in particle loading. In spray irrigation, spray nozzles assisted hydraulically or by air, or both, continuously irrigate the collecting surfaces. The spray droplets and the particles form the irrigating film. In intermittently flushed irrigation, the precipitator operates cyclically. During collection, it operates as a dry precipi- tator without rapping. It is periodically deenergized and Figure 2 – Electrode-discharge elements are placed between theplates of the plate-type electrostatic precipitator CHEMICAL ENGINEERING/NOVEMBER 23, 1987 161 flushed by overhead spray nozzles. This method only works well if the particles are easily removed (i.e., are not sticky). In film irrigation, a continuous liquid film flushes the collecting surface. Because the film also acts as the collect- ing surface, the plate or tube does nothing more than support the film. Therefore, the electrical conductivity of the irrigating fluid becomes an important factor. Nonconductive irrigants will not work. Also important are the physical properties of the film and the liquid-distribution network. The film must be smooth and well distributed, to avoid high- voltage arcing, which can damage the unit. Additionally, the distribution piping, plenums and weirs must be designed to avoid dead zones that promote settling or plugging. Materials of construction The second most important factor in design is materials of construction, because wet precipitators operate at, or below, tJ;1e adiabatic saturation temperature of the irrigating fluid (usually water), and corrosion is a constant concern. Unlike dry precipitators, wet ones are rarely made of carbon steel. Steel construction is only feasible when the stream is high in pH and low in oxygen. Even under this condition, care must be taken to avoid even temporary excursions into the low-pH range. At a pH below 3.0, the working life of carbon steel is measured in days, even hours. Ordinarily, wet precipitators are constructed. of one or more corrosion-resistant materials. These can include simple stainless steels, exotic high-nickel alloys, reinforced thermo-setting materials, and thermoplastics. From a materials standpoint, the housing of the wet precipitator is the least critical element; the outside need not even be corrosion resistant, only capable of withstanding ambient conditions. If the housing is of reinforced thermosetting plastic, its ex-terior surfaces should be fire retar-dant. In, addition, corrosion-prone points of fabrication on internal sur-faces, such as welds, joints and patch-es, must be done properly, and the means provided for making repairs. The collecting surfaces should afford the maximum resistance to chemical attack. Also, fabrication points subject to corrosion should be minimized, be-cause failures in the collecting surfaces can disturb the electric field and cause arcing, lowering performance. Because the discharge electrodes are usually not irrigated, there is a concen-trating effect on their surfaces that does not occur on wetted areas. For example, if the gas stream contains 200-500 ppm S02, 10-20 ppm HCI and 0--5 ppm HF, the pH on the moist surface of the discharge electrodes will be about 1.0, even if the irrigant is kept at a pH of 3.0, or higher. The galvanic effects of operation in the range of 40,000 volts compounds the corrosion potential of the concentrating effect. For these reasons, the discharge elec-trodes should always be fabricated of a material of significantly greater corro-sion resistance than that of any other part of the precipitator [3]. Costs of wet precipitators The factors that most influence the cost of wet precipitators are efficiency re-quirements, materials of construction, and size. Efficiency can be eliminated as a factor if comparisons are based on cost per unit of collecting surface area. Some costs for recently installed wet precipitators: Size, treated Type of Cost, equipm,ent gas, ft3/min service only, $/ft2 500,000 Corrosive 50-60 40,000 Corrosive 120-150 > 500,000 Noncorrosive 20-30 30,000 Noncorrosive. 40-50 These must be considered only order-of-magnitude costs, Figure 3 – Discharge electrodes are located inside the tubes of the Vertical-flow, tube-type wet electrostatic precipitator 162 CHEMICAL ENGINEERING/NOVEMBER 23, 1987 Figure 4 – Wet electrostatic precipitators capture fine particles moreefficiently than even the highest-energy wet scrubbers Figure 4 – Wet electrostatic precipitators capture fine particles more efficiently than even the highest-energy wet scrubbers not suitable for estimating the cost of a facility. Wet precipitator operating costs are among the lowest for gas cleaning equipment, because. they operate at lower pressure drops than scrubbers or fabric filters, and general1y have less collecting area (hence, require less high-voltage power) than do dry precipitators. For estimating purposes, high-voltage power consumption wiH usually range between 0.1 and 0.5 W factual ft3/min, depending on efficiency requirements. Auxiliary equipment, such as purge air blowers, heaters and pumps are highly site specific, so estimates of their power consumption should be done on a case-by-case basis. Operating characteristics of wet precipitators Wet precipitators capture fine particles without high energy consumption. Their capture efficiency of submicron particles is greater than even that of the highest-energy wet scrubber (Fig. 4). The size of the wet precipitator strongly affects its performance in col1ecting fine particles (Fig. 5). Wet precipitators are particularly effective in capturing large particles. Although most gas cleaners do a good job of this, 30-40% of the emissions from a dry precipitator consist of'large particles, mainly because of emissions due to rap- ping and reentrainment. Similarly, a considerable portion of the emissions from a wet scrubber is caused by mist car- ryover (another form of large particles). Wet precipitators are relatively insensitive to the chemical and physical characteristics of the gas stream or the parti- cles. Gas streams at almost any temperature or of any composition can be treated. With proper quenching, wet precipitators can handle fluegases at over 2,000°F, because the adiabatic saturation temperature wiH always be less than about 180°F. Because of the wide variety of materials from which they can be constructed, wet precipitators can also treat the most aggressive gas streams. Generally not an important factor in the design of wet precipitators are the physical and chemical properties of the particles, as well as factors that are normally of concern in the design of dry precipitators, such as resistivity, surface adhesion and flammability. A possible exception is the dielectric constant of the particles. It has a weak effect on the maximum charge that can be achieved, according to the theoretical relationship for predicting particle saturation charge [6]: N = { 1 + 2 [(k-1)/(k+2)] }(Eoa2/e) (2) Here, N = saturation charge; k = dielectric constant; Eo = charging field; a = particle diameter; and e = electron charge. The effect of dielectric constant on performance is not normally considered in the design of precipitators (wet or dry), because the dielectric constant of most particles is high, and (as the equation shows) has, little effect on the charge. However, the constant may be important in oil mist col1ection by a wet precipitator. Some oils tend to have very low constants, and this can markedly lower, efficiency [2]. Particle Efficiency, % dia., µ k = 100 k = 2 0.2 97.05 96.55 0.5 98.86 98.12 0.7 99.62 99.18 1.3 99.99 99. Although the efficiencies in both columns are very high, the uncollected fractions differ significantly. For example, in the case of 0.5 I-L particles, the uncollected fraction for the low dielectric constant is 165% of that for the high one. Applications for wet precipitators Although wet precipitators continue to comprise a smaIl share of the market for electrostatic precipitators, they are routinely selected for certain applications: Collecting sulfuric acid mist has long been a leading application. The typical unit is self-irrigating, tube-type, and CHEMICAL ENGINEERING/NOVEMBER 23, 1987 163 lead-fabricated, although construction of reinforced thermo setting plastic has gained increased acceptance. Efficiencies in excess of 99%, and outlet loadings of less than 0.01 grains/ft3, are routine. As mentioned, however, care must be taken to ensure that the dry dust loading does not exceed the unit's self-irrigating ability. Wet precipitators are increasingly being used to treat emissions from chemical-waste incinerators when significant quantities of fine particles are present. Typically, HCl or S02, or both, must be scrubbed from emissions, so that further wet treatment for particle abatement is logica1. Sometimes, a major concern is the corrosion potential from the very high levels of halogen acids, as is the potential for high concentrations of very fine particles that result from firing waste materials containing fume-forming inorganics, such as sodium or potassium. Wind-box emissions from iron-ore sintering plants have for years been controlled by dry precipitators. However, the advent of high-basicity sinter and the increasing use of oily revert materials have diminished the effectiveness of dry units. Wet units were tested and found successful in several major steelmaking plants. Although the ability of the wet precipitator to capture the particles in this service has been thoroughly demonstrated, fouling problems in the treatment of recycled water remain if the oil level in the water becomes too high. Wet precipitators are well, suited for treating emissions from wood and veneer dryers, because the process gases are already high - in moisture and' wood oil content, and are generally water miscible and not corrosive. . Also, wet units have for many years separated tar and oil from producer fuel gas. Applications include both commer-cial single- or double-stage gasifiers, and treating coke oven gas, as well as collecting liquid oil from oil-shale retorts. For any detarring service, special care must be taken in design to avoid the possibility of explosive mixtures from the leaking in of air. Because all precipitators spark nearly continuously, the presence of an explosive mixture, however transitory, may very well lead to a catastrophe. Common wet-precipitator problems Corrosion has been a major stumbling block in the applica-tion of wet precipitators. Many units have failed because of the lack of suitable cost-effective corrosion-resistant materi-als. Recent advances in materials and fabrication techniques have lowered costs and improved performance to the point that corrosion is no longer a problem. Nevertheless, avoiding corrosion remains important. The smallest oversight in de-sign or fabrication can lead to a corrosion-induced failure. An unsuitable irrigant can cause scaling or solids deposi-tion, with a resulting loss in performance. Scaling usually results from excessive concentrations of one or more of the following salts: CaC03, CaS03, CaS04 and CaF2. The first two are easily controlled by pH adjustment. Less predictable and less easily controlled than CaC03 or CaS03 is gypsum (CaS04). A concentration of CaS04 approaching 2,000 mg/L presents a potential for scaling. Its content should be kept below 1,000 mg/L., Least predictable and most troublesome of the common scales is that of CaF 2' The theoretical solubility limit for CaF2 can be as low as 15 mg/L, but the scale actually begins to form at a much higher concentration. Nevertheless, CaF2 scaling should be avoided, because the scale, being an adher- ent, is very difficult to remove. Wet precipitators are frequently irrigated with recirculat- ed water. Poor-quality water can hamper their operation. Whether problems arise from the accumulation of suspend- ed solids depends on the service. In some cases, concentra-tions of suspended solids in the percent range can be recy-cled through a wet precipitator without trouble. In others, such as when the gas contains free oil, not even the smallest amounts can be tolerated. The only remedy is to carefully design and operate the water treatment system. - Wet plumes are a problem. However, as in the case of wet scrubbers, the condensing "steam" plume can be eliminated through reheating, or rendered harmless by means of tall stacks. Although mist entrainment is inherent with wet scrubbers, this is not the case with wet precipitators. Be-cause water droplets are larger than 5}-t, they are easily eliminated by a precipitator. Also, the low gas velocities (3-15 ft/s) in wet precipitators do not promote the formation of droplets. The electrical-ground network must be adequate for pre-cipitators using film irrigation and having nonconductive collecting surfaces. The irrigating film should have an elec-. trical conductivity of at least 1,000 }-tmhos/ cm, and the liquid- contact surfaces should be extensive and evenly distributed. Contacts of dissimilar metals should be avoided, because galvanic corrosion will rapidly destroy the joint. Finally, a wet precipitator must be operated at, or below, the adiabatic saturation temperature of the irrigating fluid. Operation at as little as 10-deg. higher can dry the solids and cause fouling. A still higher operating temperature can damage the unit. J. Matley, Editor References 1. Gooch, J. P., and Dean, A. H., "Wet Electrostatic Precipitator System Study, prepared for U. S. EPA, May 1976, Publication No. EPA-600/2-76142, pp. 15, 17. 2. Gooch, J. P., and McCain, J. D., "Particulate Collection Efficiency Measurements on a Wet Electrostatic Precipitator," prepared for U. S. EPA, Publication No. EPA-650/2-75-033, March 1975, p. 24. 3. Haaland, H. A., "Materials Selection for Air Pollution Control Equip-ment," Proc. from NACE Symp. "Solving Corrosion Problems in Air Pollution Control Equipment," 1979. 4. Jaasund, S. A., Control of Fine Particle Emissions with Wet Electrostatic Precipitation, Environment International, Vol. 6, 1981, pp. 233-238. 5. Ramsey, G. H., Sparks, L. E., and Daniel, B. E., "Experimental Study of Particle Collection by a Venturi Scrubber Downstream from an Electro-static Precipitator," Proc. Symp. on Transfer and Utilization of Particu-late Control Technology, Vol. 3, p. 166, pub. by U. S. EPA, Publication No. EPA-600/7-79-044c, February 1979. 6. White, H. J., "Industrial Electrostatic Precipitation," Addison-Wesley Publishing Co., 1963, Chap. 1, p. 135. The author Steven A. Jaasund is president of Jaasund AirTech, Inc. (2659 W. Guadalupe Rd., D-1O3, Mesa, AZ 85202; tel.: 602-838-8337), an engineering consulting firm spe-cializing in industrial gas cleaning and air, pollution control. Formerly manager of engineering and manu-facturing with Fluid Ionic Systems of Dresser Indus--tries' Environmental Products Div., he also had been with Bethlehem Steel Corp.'s research department and with Betz, Converse and Murdoch, an environment testing firm. He holds a B.S. in chemical engineering from Lafayette College (Easton, Pa,) and an M.S. in engineering (air resources) from the University of Washington.