Baghouses: selecting, specifying and testing industrial dust collectors The operating parameters for dust removal with filter bags determine the proper removal system. Here are practical guidelines for evaluation. ‘Milton N. Kraus® 1 To obtain a baghouse collector that will ficiently remove industrial dusts from gas streams requires: an appropriate filter system; either a pilot-plant test on the tunit to be considered, using a sample of the dust to be separated, or a pilotplant test in the field, using the actual dust-laden stream; a method for measuring the system's performance variables; and, finally, acceptance testing of the installed unit. Filter selection Selection starts with submitting a detailed specifica- tion to dust-collector vendors, prepared on the basis of information obtained from process engineers and that presented in the earlier article (Chem. Eng., Apr. 9, pp. 94-106). This information may be summarized in tabular form (Fig. 1), with details amplified in a sepa- rate procurement specification. Tabulated data should be grouped under several headings: L, Process specifications will cover all information: re- garding the dust to be handled and the dust-collection system that will deliver material to the baghouse, The dust data should include all physical and chemical properties and characteristics that may affect the filter fabric and its cleaning, or the recovery of dust from the bbaghouse. These data include density, particle-size dis- ‘uibution, temperature, moisture content, pH or corro- siveness, tackiness, toxicity, abrasiveness and clectro- static charge. Data for the dust-collection system should include: the quantity and nature of the conveying medium, its temperature and moisture content, source of the dust, to met taut Chom. Ea, AGE 9 108 anticipated dust load, allowable pressure drop across the baghouse, and the baghouse’s maximum design pressure. 2. Mechanical specifications will list the known or de- sired details of the baghouse, such as materials of con- struction, size and location of access openings, method of fabrication, fire and explosion protection, method of dust removal, need for insulation, access platforms, structural supports, painting requirements, exhauster details and performance characteristics. Spaces should be left for insertion of unknown data by the vendor. ‘3, Electrical specifications will include tabulation of available electrical services for power and control, types of enclosures required for timers, starters, relays, etc., and motor-drive data. 4, Remarks will contain information not tabulated under other headings, such as special construction de- tails, and information that would help the vendor in preparing the bid. ‘The procurement specification should accompany the tabulated data sheet (such as Fig. 1). A detailed de- scription of the desired construction for cach baghouse component should be given insofar as possible. These components should include bag enclosure, bags, dust hoppers, bag-cleaning mechanisms and lubrication re- quirements, Overall performance requirements for the baghouse would be stated, allowing the vendor to pro- pose any deviations from the detailed specifications that he believes would improve the design. Analyzing the bids Bach vendor submits a bid that may conform to the procurement specification but will be sized according to aa 133|" Nomenclature © Solids Gomten’ of hydrocyclone feed, tons roti ‘rom, alamy ‘Solids “coment of hydrocyclone o an | Binsin hydrocyclone undertow, ton Bs salds/ton slurry. ican Open rod mill followed by closed ball mill grinding systery ther grinding. The ball mill discharge then joins the rod mill discharge and is pumped to the hydrocyclone. Aless frequently encountered use of hydrocyclones in the mineral processing industry is the actual concentra- tion of the valuable mineral. This application is seen ‘most frequently in the nonmetallic minerals industry and is especially prevalent in the preparation of kaolin and other clay materials [/). In the closed-circuit grinding of ores, a universally encountered operating parameter is the circulating load. The circulating load, defined as the material returned to the grinding mill from the classifier, is expressed as a percentage of the coarse ore feed [2], In general, mill operators try to keep the circulating load in the range of 150%-300%, though exceptions are common (4). Increasing the circulating load has the effects of low- ‘ring the sharpness of the classifier separation, and pro ‘ducing a coarser grind in the mill |}. Ax the circulating load continues to increase, the mill finally overloads. When this occurs, coarse feed to the mill must be cut off| until the mill can grind through the overload. Con- versely, if the circulating load decreases, a finer grind and sharper classifier separation will result, but the coarse feedrate will be lowered. Thus the regulation of the circulating load becomes a matter of balancing the requirements for mill throughput, fineness of grind, and sharpness of separation, Hydrocyclone efficiency Hydrocyclone efficiency is almost universally defined in terms of the underflow for each size fraction. Thus 2 hydrocyclone will have a series of eficiency values rang ing from close to 1.0 at coarse sizes to approaching 0.0 at very fine sizes The split port (dy) for @ hydrocyclone is the particle size at which 50% of the material is overflow, and 50% isticle sizes at which 75% and 25 the underflow. Any consideration cf hydrocyclone operation must begin with a mass balance. If a simple closed-circuit system is considered. with the solids contents of the three streams and the coarse ore feedrate known, it is possible to determine the hydrocyclone feedrate and the Girculating load. This is done by evaluating the solids mass'balance and the overall mass balance. ‘Te must first be remembered that the grinding mill hydrocyclone system will be operating in a steady-state condition. Thus, if the coarse ore feedrate is known, the hydrocyclone overflow is also known. ‘The solids mass balance relates the tons of solids /h in the three streams: ; F=P+C a Note that the underflow is the circulating load for this system. ‘The overall mass balance relates the tons of slurry /h: oe iQ) By solving Eq. (1) and (2) simultaneously, the two un- known solids fowrates can be determined. Determining the efficiencies is a multistep proces ‘The first step isto determine the overall mass recovery for the hydrocyclone. The mass recovery can be deter mined from the solids contents of the three streams, and the specific gravity of the solids [/} = (2=4)-(2=3) 0= (S53): Ges ® ‘This gives the fraction ofthe total sli inthe feed that report to the underflow. Once the overall mass recovery has been determined, if screen analyses of the overflow and underflow solids are availabe, iis posible to eal. ulate the differential mass recovery or Tromp value. ‘The ‘Tromp value is the mast recovery for a given size frac- tion of the solid material in the feed, and is calculated ‘thus [1]: Tt peek eee (4: = RIFT OR) bw The Tromp aime is a semilogarithmic plot of the Tromp values (7) versus the median particle sie for cach tze fraction (cee Fig, 4) From the Tromp curve, the split point (i) ean be determined by locating the particle size for wkich 7, = 080; Irthe sctecs date sre Eatiactory, itis also pesble to determine dg and ake from the curve (eee Fa 3) “The Tromp ture presents a visual indication of how eficcively the hydrocyclone i doing its jo. Ideally, the carve wotld show a vertical drop af the desired sePary: tion sz, at shown in Fig. 6A curve showing a early horizontal line would indicate thatthe hydoeyclone doing almost nothing in the way of sire elaseScation (see Fig. 7) 2} I only general rests or trend ae Using the Tromp curve to determine the values of das doo, and dye Fig. | Tromp curve indicat | poor hydrocyclone exceeding! ition Fig. 7)JPMENT OATA SHEET UST COLLECTORS SERTEE CONDITIONS ENGR Toscan Rie 7 Fleswranuous Bivrenersear [RTH PURCHASE SPECIFICATION. PROCESS SPECIFICATIONS purr para [Ten rnveor enrenavone STS ARBLE PREESIRE TOF TOR FRESSRE DROP : a Hovis EGitore oh a Baits com SSRN Cerven as _ a veneers urea cower oPPEnis _ ws oneal FunwsheD sy a Geavines _OAPuineo. ee promranea TOF erecta peouinenenrs IwsuLATION [ Qreouneo O8ESuneo BPE seine! ran [OH we gaeee ramen OR, OMA Oro Recess — FoR Fon SEGAL REQUIREMENTS ee ee = _/Eee RRS FUUEIURAT] Grounee OHS ra eo ono. 0, painting | OF Zuo ‘Devorcont Donanaes aware Sod worn [oe peace como FP = ————oooaeaeeaa Purchase specification for dust collectors Fig.lasek v the vendor's experience. Bids should first be analyzed ‘ria sct by the purchaser as being desir~ able for handling the entering dust. Size may be inves- tigated later, or may be proven by pilot-plant test ‘A typical list of desirable requirements for a hot, tacky dust: ‘Absence of moving mechanical parts to be cleaned. 4% Cleaning of filter media by reverse airflow. 1 Loosened dust fiows in same direction as normal airfiow. (This prevents redeposit of dust on cleaned filter Fabric. On some designs, this directs the dust into the dust hopper.) © Minimum amount of ancillary equipment. Minimum requirements for utilities. Minimum floor space. 1 Modular, all-welded design to resist the specified ‘minimum pressure (inches of water column), “These eriteria should be revised to suit the character- istics of other types of dust. Obviously, not all bag- houses will meet the requirements, Tn addition to a tabulation of comparative data for bid analysis (Table Da separate list (Table 11) should bbe made to show how the bids conform to the selection requirements. From these tabulations, selection may be narrowed to fone or more vendors whose equipment meets most of the criteria. If more than one vendor qualifis, the submistions must be investigated more closely by con- sidering price delivery, guarantees, availability ofa test unit, etc. For hot, tacky dust, all baghouses depending on mechanical components that could be fouled if a bag were to break were eliminated. Even if a bag did not break, fine dust could build up and smear the mecha- nisms’ and affect their operation. Such baghouses can ‘only be considered where savings in first cost and main- tenance costs justify the risks, Equipment guarantees ‘A materialsand-workmanship guarantee is standard with most vendors, and lasts for a period of one year from date of shipment. Because construction of the bbaghouse and the system it serves may extend over a long period of time, the purchaser should request an extension of time for this, ‘The vendor should guarantee that baghouse emis- sions will comply with the pollution-control laws appli- cable to the baghouse location. The purchaser's accep- tance of the baghouse could be based on acceptance of the installation and issuance of an operating permit by the environmental protection agency having jurisdic- tion, or by test. The recommended test procedure is that cstablished by the Industrial Gas Cleaning Institute (/] Filter testing: vendor’s pilot plant ‘Testing ofa baghouse is time-consuming, and there is a tendency to bypass this phase of equipment selection and rely on a vendor's application experience in sizing a particular filter, However, an experienced vendor will Fequest that a quantity of the dust (about several cubic feet) be sent for test in a pilot plant. The unit would be equipped with brand-new filter tubes of the vendor's choice, based on an analysis of the dust characteristics, by the vendor's laboratory. A typical test procedure would be: 1, A volumetric feeder is calibrated by setting it for a high discharge rate, and weighing the test material discharged over a period of time. 2. A blower exhausts air from the baghouse, and a calibrated venturi tube in the suction line meters the airflow. 3, ‘The feeder is set to discharge dust into the filter's inlet connection at the desired rate. The discharged powder is dispersed by an air jet in order to simulate plant conditions 4, The cleaning-cycle programmer is set at the ven- dors recommended rate, and operating utilities for the baghouse turned on. 5. The filter is operated for at least eight hours in order to build up a filter eake on the bags. However, the differential presure across the bags were to build up to the vendor's standard operating-pressure differential prior to this time, the dust feeder would be recalibrated to a lower rate and the airflow reduced accordingly 6, The standard pressure differential is then ap- proached and maintained (by trial and error) for vari- ‘ous rate of dus loading and airflow. Each combination ‘of dust loading and airflow is maintained for several hours ‘During this test, the data are plotted as dust loading in (grains/f air}/fe cloth area. Typical curves for various materials are shown in Fig. 2. These are called Toading curves, and are only useful for sizing the unit on test when handling the tested dust. ‘The dustloading test is usually run at ambient-air temperatures. When the baghouse is operated at a higher air eemperature, the filter rate must be corrected. ‘The correction factor is obtained from Fig. 3, which is a curve for constant-volume flow. In effect, this factor increases cloth area. It compensates for the increased air viscosity at high temperatures. (The viscosity of air increases almost in direct proportion to the increase in temperature.) In general, the relationship of the correc tion factor for the filter rate is: Alt Bn where F is the filter rate at a given temperature and @ pressure drop of 3.5 in, water column,* Fy is the cor- rected filter rate, 6; isthe filter-rate correction factor for ambient temperature during test, and ¢ is the filters rate correction factor for the proposed operating tem- perature, ‘After the filter rate is corrected for temperature, a separate calculation must be made to allow for expan- sion of air at the higher temperature, so as to obtain the actual filter rate at the face of the fabric during regular operation. ‘The cloth area for a full data, is calculated from: Ao = Val FePe (dy) ize baghouse, based on test @) where dg is total area of cloth, ft%; Vp is total volume of zzas stream at operating temperature, ft/min; Fy is ‘Milropl Cor. standacd Ref 3 135_ i Moke of in water Standard Top tube. Bottom Bagecleaned _Clasning calletor _column ‘material Medium onbeg sheet’ ubesheet by controlled by_ A 120 Square Welded Stainless Yin dle, Felted Outside Camped — ——_‘Highrpresure Timet_aoes sta tocege atin vah solenoid va as. 8 17 Feet, Welded Carbon SMsinedia, Felted Qutsige Clamped ——~—_Highpressure Timer apart sea rocage tired in each solonoid valves baa ce 17 Squme Bolted Carbon Yin. dia, Felted Outside Clamped High pesture Tier gosta fee toeage in each solenoid valves ba ° 20 Rect. Welded Galvanized Bin. dia, Woven Inside Presedin Clamped Air suction Timer stat caller {no boa housing e 15° Square Welded or Carbon G:in.dia, Woven Inside Snapring Clampad High preseure Th bod" Red or ttn. ncnalne firjotin each sla Fung ean beg plus al fucuon into solenold ves housing F 40 Cyl. Bolted Carbon Ovel_«—Feltad Outside Clamped — ————Lowspresure Continuo eo by coue bast int. rotating ple ach beg anceam wack 6 12 Rect, Bol Envelope Woven Outside Hooked over ins Air suction Continuously Rear! maltions —Into'beg traveling fousing suction lena 4 16 Square Bolted Carbon 12nsdia, Felted Inside Clamped_Clampad Traveling Lowprasure ste irblom —airblowea Fingon rovers bao cochbeg —lnath. Dil presure st 1 20 Oyl.—Bolted Carbon GHndia. Woven Inside Snap ring Snap ring Traveling Lowsresue fee airblow: owen fingon” traverse bap tech bee length eon- tinvousy corrected filter rate; and F, is an application factor from Table IV. Estimating the filter rate ‘A preliminary estimate of the test-filter rate can be made by using a filter rate guide developed by one manufacturer for application to pulse-jet or blow-ring Daghouses that have continuously cleaned fellter media [2]. Such baghouses maintain a pressure differ- ential of 3.5 in. water column at design flow and load- ing conditions, The guide consists of five factors, multi plied together to arrive at a filter rate. These are: ‘Material factor, Fyy from Table IL allows for the properties and characteristics for a material that would affect its tendency to form a filter cake during normal airflow and release it during the cleaning cycle for the bags by reverse airflow. “Application factor, F,, fro Table IV allows for the type of process delivering dust to the baghouse and for upsets that cause variations in the dust load. Temperature effect, 4 from Fig. 3 compensates for in- 136 an creased air viscosity at high temperatures, as previously explained. "Particle-size factor, F,, from Table V allows for the fact that fumes and fine dusts tend to blind fabrics more rapidly than do coarse particles. ‘Dust-lead efet factor, Fa, from Fig, 4, which isa typical performance curve, is applicable only to this type of Daghouse operating at constant differential pressure ‘An actual pilot-plant test eliminates the factors for material, particle-size and dust-load effects, but the temperature and application factors must be obtained, tas described above, in order to determine the required cloth area. "The preliminary determination of filter rate can be used to roughly size a baghouse for cost-estimating purposes. For baghouses using woven fabrics that rev Quire cleaning at less-frequent intervals, the filter rate Gstablished by the above guidelines can be divided by five. The resulting rate, however, is only applicable for dust loads within the capability of such baghouses, and for cost-estimating purposes only.SRT Ts, Tobi Airflow, Motors required es path sequence _to oper eras ust hopper, ‘None tubes, tp plenum Dust hopper, None tubes, fp plenum, Dust hope None Requires removal of tubes, tp plenum fe manfoigr to replace bege ‘Rapper timers {niet plenum, None Cloth aes of 6,500 ust hopper, Fe ond fess hat tubes, bottom ‘welded constraction otbeg and indies boos” Somperiment. B Side of bog ‘Cylindrical units must compartment, be paralleled for large tubes, top Plonuot : ok Side of bag 1-Planum drive comparenent, baffle, envelopes, cleanin chembor ‘Top planum, tubes, Bottom ofa Compartment 1-Blow cose 2Blowers recycled Top plenum {bes canter otbag ‘compartmer 1.Blow core 2Blowers (vtingtical units ‘must be paralleled Fer largeeepecities Filter testing: field pilot-plant Some vendors can supply a test unit for installation ata prospective user’s plant. This method of evaluation should only be undertaken under one or more of the following circumstances: (1) when there is no precedent for filtering a specific material; (2) when a filter design is deemed novel and the user has had no previous experience with it; (3) when the filter unit must be evaluated under a variety of plantoperating conditions or when handling a variety of materials; or (4) when determination of the effect of weather conditions on fltation is desired for a proposed outdoor installation AA test unit is generally supplied on a rental basi, ‘with an option to purchase. The accrued rental charges could be applied tothe purchase price of the test unit or ofa larger unit. The agreement with the vendor should provide for: (1) freight, handling and instalation ex- Pense, usually borne by the user; (2) services of an installation supervisor, ifone is desired; (8) allocation of reconditioning expenics for damage caused beyond | normal wear during a test; (4) responsibility for loss of the unit from any cause; (5) access to the test area by the vendor's representatives and furnishing test data to them; (6) schedule of charges and method of billing; (7) statement of credits allowed upon return of the test unit; and (8) a list of auxiliary equipment furnished with the test unit Installation of a field test unit at an operating plant usually requires that a small percentage of dust-aden air be drawn from an existing exhaust system to match the test unit’s capacity. This requires insertion of a special nozzle into a straight section of the existing duct—not just a “tee” or lateral connection, The open end of the nozzle should face upstream and be centered s0 a8 to receive a representative sample of the dust load. Air drawn through the test filter should enter the sampling nozzle at the velocity that normally exists in the cross-section from which the sample is drawn. This procedure is called isokinetic sampling. The exhaust fan for drawing this sidestream through the test unit must be sized for the airflow required to obtain the isokinetic velocity at a suction pressure low enough to overcome (a) the suction resistance in the main duct plus (b) the estimated resistance of the test filter and ductwork, Some form of shutoff gate should be placed in the suction line, close to the test unit, so that the unit may be opened for inspection and modification without affecting flow in the main duct, and that the unit may be shut down without causing reverse airflow through its exhaust fan. If tess above the isokinetic velocities are planned, the exhaust-fan’s drive should be able to in- crease the fan’s speed in order to obtain higher airflows. Collected dust from the unit should discharge through an airlock into a drum for weighing. If no airlock is used, the dust may flow into a drum through an airtight drum cover, so that the drum is maintained at the same negative pressure as the test unit, However, every time the drum has to be emptied, weighed or inspected, the unit must be shut down. Provision should be made for periodic disposal of collected dust. Other- wise, the dust may back up and fill the baghouse, voiding the test data, A ypical outdoor test-setup for a unit handling hot detergent dust is shown in Fig. 5. If hot or moist dust is to be processed, provision should be made to completely insulate the unit, A heated, plywood enclosure should protect the dust hopper and collecting drum, but insu- lation and electrical tracing of the hopper may still be required. (On units supplied with compressed air for cleaning, the air lines and solenoid valves should be electrically traced. Alll electrical heat tracing should be thermostat ically controlled to prevent overheating and burning of the collected dust. A large temperature differential be- tween the air-inlet and air-outlet connections may indi cate an air leak into the collector, o the need for addi tional insulation. Measuring performance Means for measuring performance should be installed and data taken for airflow rate, dust load, temperature, pressure, and timing of cleaning cycles Airflow raie—Install a calibrated venturi meter or 2 raWhen testing, obtain the airflow rate before, during and after cleaning the bags, if possible. By measuring the pressure differential across the exhaust fan, an ap- proximate check of airflow may be made by referring to fhe fan characteristic curve of static pressure vs. air volume for the operating speed of the unit. ‘When the fan's characteristic curve rises and then drops, so that it is uncertain which of two identical Satie pressures has been read, the motor horsepower Should be measured (using clamp-on electrical measur- ing instruments for quick readings). The horsepower vs. air volume curve should then be used to determine on Which side of the characteristic curve the static-pressure reading belongs. (The horsepower-air volume curve is tusually a straight line, increasing proportionally with capacity.) During a test, a flow measurement higher than that required to obtain isokinetic velocity indicates the pres nee of an air leak in the test equipment, usually at the dust-discharge connection. Tn a filter using reverse flow of atmospheric air for cleaning, the reverse-airflow rate should be obtained by {nstalling a calibrated orifice atthe reverse-air inlet and measuring the pressure drop across it. ‘Dust load-—Install a platform scale close to the dust- Es 258 ‘Dust toad, (graing/tt®)/ft2 cloth area B08 38 oa a 8 BIO TE 4 16 18 2022 24 | Aintoslath ratio, (2/min\ 2 7 Aton ei, tein AE es | | collecting drum (Fig, 5). Weigh the load in the drum at Dust loadings for filters cleaned regular intervals and use the weights of filled, sample- __ Loyrhiahrpressute reverse sized containers of known volume to determine average powder density f ‘Tonperature-Install thermometer connections at the orifice inthe clean-air discharge duct, with the required | baghouse inlet and outlet. The thermometers can be of gtk of saight duet coming before and after the | the indicating or recording type. A glasratem indict sees Gr Shatll two connections in a straight section | ing thermometer inserted in a eork-stoppered hol in sere for making a pitottube traverse (Fig. ). These -) the duct may suffice. The sensing bulb should, be i ci ons ahculd be at right angles to each other and | cleaned after each use. Wet-bulb readings on the dust ; at the same cross-section of the duct. duet must be taken by rapidly inserting a glassstem Conformity of bids to selection requirements for dust collector Table moni : Make of Moving Ravers alt= Oust follows saulpment Foot Housing Colector parce flow only normal sfow rguired___Utiiis required per wectication A None Ye Ne Timer High-pressure sit, Minimum You ‘contol ac. Staines ates materi 3B None Yer No Timer Hidhgrenura a, Minimum Seni None No tmer ——Hahgromur a, Minimum Seco contol a. Damper Yeu High-presure sf, Multiple Special contol ac, compartment Fepper maior. © Damper Yr High-presure or, Multiple Speci ono! aca,” eompertment Tapper motors F Rotating Yes, «No Revere'sir Blower end plenum Minimum Special Bienum Blower motors Rating © Yee None Plenum motors Minimum ‘Special 4 Ye Ye freverseair | Control 2¢.,blower Minimum Special Blower ii aie i ( Teweling Yes Motor: Minimum Special Slow oarethermometer with a wetted wick over the bulb, and holding it in place until the temperature reaches its lowest level. The test opening should be sealed during this reading. The wick should be cleaned and rewetted before each reading. ‘Wet- and dry-bulb readings of the ambient air sur- rounding the test unit should be recorded at the time other data are taken, by using a sling psychrometer, Pressure—Install manometer connections in the inlet and outlet ducts, close to the filter, to measure the pressure differential across the unit. Also, install con- nections on the clean and dirty sides of the baghouse at the tubeshect to get the pressure differential across the bags If factory-installed connections are provided, make sure that they are not blocked by paint or galvanizing ‘metal, and that a cotton filter is installed in the dirty-air side. When testing, take readings of the pressure differ- tential across the bags and across the baghouse just before a cleaning period starts, and just after it ends. ‘On baghouses using compressed air for cleaning, install a pressure gage on the air-supply manifold. Ob- serve this gage regularly to ensure that proper cleaning pressure is maintained, Timing of cleaning eles —Before starting a test, operate the program timer, and check the time between closing and opening of the air-reversal dampers and the start- ing and stopping of shakers in each compartment of multicompartment units equipped with bag-shaking and reverse-airflow dampers. Check the time between the termination of the period in one compartment and the start of the cleaning cycle in the next compartment, as set up in the timer. ‘ (On units having reverse-airflow pulsing valves, time the period between successive blows or pulses, and if possible, the duration of the airflow. ‘The Cleaning-cycle time is the total elapsed time required to clean all the bags in a filter once. ~, Regular inspections of baghouses In addition to obtaining data at the baghouse, a complete record of process operations that exhaust into it should be obtained. Operating conditions, rates, downtime, upsets, etc., of the process should be corre- lated with events occurring at the baghouse, The bag- house should be opened for inspection at regular inter- vale as determined by process operations—at least once per week. The bags should be inspected for: 1, Permanent coating of dust that cannot be re moved by cleaning. 2, Plastering of moist material on one side of each filter element. 3, Hardness of the coating for evidence of condensa- tion, 4. Thickness of the coating along the length of the bag. 5. Color of the coating compared to the color of the collected dust. 6. Condition of the fabric, opening of the weave or felt, tears, wear due to rubbing or flexing, ete. Baghouse interiors should be inspected for buildup of | powder on the walls of the bag housing and dust hop- ‘per—especially at corners and ledges, at baffles opposite Tea ional Correction factors for tumparatures above standard the inlet, at valley angles, and at dust outlets, Weekend or nightly condensation can cause buildup at dust out- lets. A sample of any builtup material should be col- lected and analyzed for particle size and composition. This may reveal an unusual concentration of an ingre- dient that has been scalped off the material in the process, and should not be in the baghouse After several hundred hours of operation, a number of bags should be removed from. strategic locations within the baghouse and tested for air permeability ‘The test should be made on sections cut from the top, center and bottom of each bag. Permeability of the used. bag should be compared with that of a new one. ‘These data would be used to estimate the life of the bags or to determine the frequency of washing or dry- cleaning needed. The method of cleaning must be de- termined by trial and error to suit the filter fabric and the materials lodged in it. A cleaned bag should also be tested for air permeability: in the same manner as a dirty bag. Acceptance testing of installed filters ‘The information presented here applies mainly to field-assembled units that arrive at the jobsite knocked-down or as subassemblies. This information may be modified as necessary for factory-assembled units, Testing for acceptance can only start after com= pletion of alll mechanical and electrical work. A typical procedure for such testing is to: 1. Inspect the baghouse and all system ductwork for structural integrity 2. Check the direction of rotation of all motor-driven equipment—such as fans, conveyors, cleaning mecha~ a 139Dust load affeots filter rate Fig. 4 nisins, and airlock dischargers—by actually observing. the motor shafts. 3, Lubricate all motor-driven equipment, such as gear reducers, crank arms, and pivots. 4. Compare motor-starter overload-relay ratings with motor-nameplate ratings to ensure proper match- 3 $0 a to suit starter location, 3. Blow out dirt, water and scale from alll com- pressed-air pipelines, and make certain that airline lubricators are filled with the proper oil. This should be done before connecting the piping to air-cylinder oper- ators, solenoid valves, and measuring instruments. 6. Connect all compressed-air piping, blanking off any possible eseape openings. Check the piping for air leakage by applying full pressure, closing the air-supply_ valve, and observing a test gage for loss of pressure, Rapid loss of pressure indicates a large leak that should be sought for by following the length of the piping and listening for escaping air. Slow loss of pres- sure indicates a small leak that should be sought by applying soapy water to cach joint, after isolating the piping from the main air-supply valve. (A leaking sup- ply valve can falsely indicate a well because it replenishes the leaking air.) 7. Check the operation and sequencing of each pro- gram timer, Follow up by checking the functioning of the cleaning mechanisms controlled by the timers. For ‘example, reverse-air dampers should operate freely and seat tightly. Check timing between opening, closing and cycling of the dampers between compartments; check ‘operation of the linkage to shaking mechanisms and the duration of the shaking period; and check that solenoid valves are pulsing in proper sequence and are set for the recommended off-time, Observe the action of poppet valves on units equipped with air blast into the plenum. 8. Check carriage level, cracking of air-supply hose, and the limit-switch operation for carriage reversal, on filter units having blow-ring carriage drives. 9. Cheek operation and starting sequence of dust- discharge valves and conveyors. 10. Check settings of pressure and temperature: switches, alarms, etc. Check the scale ranges of indicat- ing instruments, including manometers so as to prevent blowout of manometer liquid. Inspection for structural integrity ‘Testing for housing tightness should be made before applying the insulation. and jacketing. This is most important on outdoor installations, where rain can enter the housing. A preliminary inspection of both the (Material characteristics affect filter rate through fabric-bag collectors Table It 18 2 10 8 ot ake mix ‘Asbestos ‘Alumine "Ammonium ‘Activated carbon CCerdboard dust Butfing dust ‘carbon black, fine ‘phosphate Carbon black, Fibrous ond fertilizer molecular) ‘calulesic Detergents materials Clay and brick dusts Fumes and other Gypsum Cost ‘isperved Lime, hydrates Fluorper produets direct Pedi Kaolin ‘rom rections salt Limestone Powered mike Sond Fock dust, ores ‘Soaps bart dst ‘and minerals Pigments Sods aah silica Plastics Tale Sugar Resins general any physically and chemically stable materials Sites ‘ncluaes solids that are unstable due to hyprossopie nature, aubiimation Starch ‘or polymerization. HetManometer 10s se Sarmping Boshouse Test. connection (biter tee) Ephaust /tan palo chamber wes wise Test \gement for baghouse collector handling a hot dust clean and dusty sides of the baghouse interior should be made to ensure removal of all construction materials and debris. ‘The bags should be checked for (1) proper assembly to tubesheets; (2) proper tension at both high and low positions of the shaker mechanisms, when furnished; (3) proper suspension, so that the bags do not touch each other or the baghouse walls. Interior walkways and bafiles should be in place. Access doors should be checked to ensure (1) that gaskets are in place and mate with the seating surfaces when the doors are closed and (2) that fastening or latching devices are in place, lubricated, and operating freely, Inspect walls for evidence of daylight from un- Table IV Factor, Procons t Nuance venting Relief of transfor Product collection ‘Airconveying vent- ing mi, ash iver, clasties, Process filtration ‘Spray dryers, kins, Reta 09 os used bolt-holes, unconnected piping, open joints or pinholes Fire detectors, sprinklers and deluge nozzles should be in place. Fusible sprinkler-head systems should be checked for tightness before closing and sealing the housing. (This should be done before bag installation, so that a leak will not damage the bags.) ‘An exterior inspection should be made after the bag- hhouse access and inspection doors are closed and latched. All bolted joints on panels, dust hoppers and connecting ductwork should be inspected for unused bolt-holes (have proper bolt installed, alignment (mis- matched holes should be redrilled, and unused holes or spaces filled in), and presence of sealing compound or ‘gasket between mating parts. On outdoor suction systems, the housing can be checked for tightness by starting the fan and pulling maximum suction on the housing, using clean air. Inlet dampers can be partially closed to obtain this condition and reduce airflow so as not to overload the motor, Size Fy ‘Over 100 mieron| 12, 150 to 100 micron 4 10 to 60 micron 19 3 t0 10 micron os Under 3 micron 08 Souree: Ref. 3 141‘While the housing is under suction, a hose should be played on all access doors, and on all bolted and field ‘welded joints exposed to the weather, taking care to avoid direct impingement into the openings on re verse-air dampers. Obviously, all outdoor electrical sys tems should be sealed before this test is done. ‘After hosing down the equipment, each compartment ‘and plenum chamber is entered and inspected for evi- dence of water, which is then traced to its source. The ‘causes of leaks, if any, are then eliminated. ‘On indoor suction systems where hosing is impracti- cal, the fan can be started and the system set to obtain maximum suction. Every housing joint or door seal is then inspected by listening or feeling for airflow into the unit. Each suspected leak is then marked for later cor- rection. Although minor leakage is tolerable, a process ‘upset or fan failure may cause positive pressure inside the baghouse, and spew dust into the room. ‘On positive-pressure systems, the fan is started and clean air directed into the baghouse. Housing joints and door seals may then be inspected for leaks. Testing for leaks on the clean-air side of positive-pressure baghouses is unnecessary, and on many large baghouses impracti- cal. The best way to check for leaks is by brushing all the joints and sealing surfaces with soapy water, espe- cially where the escaping dust is hazardous. Bag con- nections to the tubesheets must be tight because leakage goes directly to atmosphere. ‘After these checks have been made, the exhaust fan may be started. Then, the clean airflow through each equipment connection or exhaust hood is adjusted to ‘obtain the required intake velocity for dust pickup. The cexhaust-fan motor should be frequently checked during this petiod for overheating or other evidence of over loading, ‘Overloading may be caused by (a) excessive airflow due to the low resistance of new bags; (b) the blower, on a positive-pressure system, acting as a booster; or (c) an oversized fan. In the first case, the exhaust-fan discharge should be temporarily throttled until a filter cake builds up and the pressure drop in the baghouse becomes constant. A larger motor and speedup of the fan may be necessary where the positive pressure in a baghouse, bin cor receiver must be made negative. A permanent ‘damper may have to be installed on the discharge of an oversized fan. Starting up a collector The operating sequence for a typical dustcollection system is 1. Start dust-removal equipment. 2: Start bag-cleaning equipment. When starting up with brand-new filter bags delay starting the timers for the cleaning system until the anticipated pressure dif- ferential is attained. If this differential cannot be reached because of operating conditions, the timer may be started, and the system operated at a lower pressure differential 3 Start the exhaust fans. Reduce initial airflow to new bage until filter cake is formed, then increase air- iow to full load, On a unit that handles hot dust, preheat the baghouse before allowing dust to enter, in border to prevent condensation. 4. Observe the pressure differential across the bags. If this increases above a predetermined limit, make the following investigations: ‘a. Check for overloaded filter bags by manually ad- ‘vancing the cleaning cycle, and observe how this affects presure drop. If overloading is apparent, shorten the Cleaning cycle, or lengthen. the cleaning time, or do both. On a unit operating intermittently, the exhaust system may have to be shut down at regular intervals to permit cleaning, pending determination of a means to increase filter capacity or reduce dust load. b. Check the compressed-air pressure at the air- supply manifold on baghouses cleaned by high-pressure reverse-air jets, If too low, check for clogged air filters or for insufficient air capacity to accommodate the added load, . Recheck rotation of the exhaust fan, If rotation is reversed, a fan sometimes appears to be operating satis- factorily simply because it still discharges air and creates suction—especially with flat-bladed fans. d. Check the cleaning sequence between compart ments to see if they are being cleaned in too rapid succession, The hopper inay be overloaded with dust if two or more bag compartments discharge into it. 5. Observe the discharge duct for evidence of dust. A light plume is normal on new bags until the filter cake forms. The plume should cease after several hours of ‘operation. If it persists or becomes heavier, check the clean-air side or clean-air plenum for powder buildup. If buildup is present, recheck the tightness of all bag connections to the tubesheet, Tests for emissions ‘Most environmental protection codes provide a tem= porary operating period to allow for evaluation and testing of the baghouse before issuing an operating certificate, During this period, the purchaser or vendor should sample the emissions by using an established procedure acceptable to the local environmental pt tection agency (see Ref. 1). Emissions from the unit should be less than the amount originally reported in the application to construct and install the dust-control system, The inspector for the code-enforcement agencies should be invited to witness the tests in order to expe- dite approval if the emission rates are acceptable. ‘Steven Danator, Ear References 1 ge Pedr Deering Fema Pe Ben ScrAlceandrla VA 22518. catamiiam me 2, Frys RBs and Reiner, TV, New Fler Rate Guide, Air Biri, 104 2: Bilt RO, Relocnce Data and Tab, Mp or, Samii, NF 4. Knut, MN, “Pocomatie Conveying of Bulk Materia” Renal Pres, Nov Yon, Reprints Reprints of the two articles on Baghouse Dust Collectors will be available shortly. Check number 014 on the eprint order form in the back ofthis or any subsequent iatue, Please circle number 305 on the Reader Service Card for a complete free Catalog of Reprints 142 RTA RGNERING APRDESIGN GUIDE FOR DUST COLLECTORS Consider these factors when selecting, specifying and operating dust collectors to remove unwanted particulate matter from gaseous process and exhaust streams Amrit. Agarwal Pneuma:'> Conveying Consuit9 Services ust collectors are used widely in most manufacturing opera tions to remove dust and other Jairhorne particulate matter from process gases before re- leasing them to the atmosphere or to downstream processes. Sevoral differ cent types of dust collectors are widely in use today, and they vary in terms ‘ofnomenclature, design and operation or more. see the box at right) For instance, in pneumatic convey- ing svstems, dust collectors are rou tinely used for removing dust. {rom the convering gas before the gas is released to the atmosphere. They are also often installed at the end of the convering line on bins and silos, as bin vent filters or as filter-receivers, When used with systems that convey fine solid particles, dust collectors not only recover these fine particles from the convering gas, but they also clean ‘the gas before it is discharged to the atmosphere or reeycled, In most dusteollection devices, the filtration mechanism combines Doth depth filtration and surface or GETTING THE TERMINOLOGY STRAIGHT The term dus callector is very brood. Listed below are some of specific ypes of dust 6,000, posed to; otherwise, they will deform fand may rupture. Instead of earbon steel, which could rust and degrade product quality, use stainless steel or galvanized steel wire cages. Typically, the cages should have at least eight rods made from 10-gage wire with at Teast eight rings. Pleated filter elements, Pleated filter elements provide more surface ‘area than cylindrical bags; how- fever, dust can accumulate inside the pleats and not release when the fil- ters are pulsed. This is particularly true for sticky dusts. Certain pleated designs, such as those whose pleats are not folded too close to each other ‘or are not too deep, may be suitable for some dusts, Dust collectors in feed bins Dast collectors that are used in feed bins require special eare because they cean get severely damaged if discharge ‘of the material from the feed bin stops ‘and new material keeps on entering the bin from @ pneumatic conveying system or from another source. Be- cause of the rising level of the material bags ean get buried under the material, get collapsed, or restrict the flow of the feonveying gas, To prevent such inci- dents, the rotary valve, whieh is gener ally used as the feed-out device, is pro- vided with a motion switch installed ‘n its rotor such that if the rotor stops, the incoming paeumatie conveying system or any other feed stream also stops. Redundant level switehes in the bin are also provided to stop the low of incoming material into the bin. Dust collector efficiency. Felted polyester bags capture particles up to 20 micrometers. Their efficiency de- pends upon the size of the dust part cles but, in general. itis about 99.95% if the dust collector is correctly sized and operated. Bags with Gore-Tex membrane (or equivalent membrane) capture par- ustration otaouble wal NOZZLES: © Twist & Dry™ for Spray Drying We asd the capacity of cur Twist fe Dey nazz ps for operating con Bone upto 40°F, Now eeu a increase produc: ‘ity we maintaining Theconvenient locking mechanism and superior performance of out Twist Een nozles BETE’ BETE Fog Nomi, ne. coef ik 0901 ones 2.06 perenne a MON ca Dreyer neta toa To ind ot mae att is norte vis worw.beta.com/cengetd.html aalinks.che.com/4517-34 49° CHEMICALENGINEERING WWW.CHE.COM FEBRUARY 2005 ‘adlinks.che.com4517-35 ticles | ciency Guare filters filter. 1 smalle use. T times ever, t filter e mary | the saz Inst system tems a that, d vided and co +A loc feren able + Alan andentering conveying tree. Bes material material, low of the uch inc. is gener. 1, is pro. installed tor stops, onveying eam also ves in the re flow of Felted les up to eney de- 'st parti 199.959 Uy sized ane (or wwe par vot kes hi " ticles up to 2 micrometers, Their eff- ciency can be as high as 99,99: Guard filters. As noted earlier, guard filtors are often provided a2 protec- tion against the failure of 2 primary filtor Their surface area can be much smaller because of their infrequent use. The airto-cloth ratio can be five times that of the primary filter. How ‘ever, they should use the came type of filter element as that used in the pri mary filter so that they can eaptare the same size parti Instrumentation and control systems. Controls for pulse-air sys- tems are discussed above, Aside from that, dust collectors are generally pro vided with the following instruments and controls as a minimum: +A local and an optional remote di forential pressure indicator to en: able monitoring ofthe pressure drop across the filter elements ‘+ Alarms for both very high pressure and very low pressure drop: adllinks.che.com/a517.38 the filter clomonts, as indications for bag replacement +A high level ewitch interlocked to stop the incoming material flow lo- cated at least one foot below the fil- ter elements to prevent damage to the bags * For pulse;jet dust collectors, timers to control the duration of the pulse and the frequency of the pulse, and a solector switch to control the sequence of pulsing rows of bags ‘These are provided by the dust col lector vendor but ean be locally ad- Justed, + For pulse-jet dust collectors, a pres- sure control valve with a pressure gage to control and monitor the pulse gas pressure ‘The design of dust collectors is a sub- ject that is generally not included in many engineering curricula, despite the widespread use of dust collectors in many chemical process and manu- facturing operations. The information Corer ron EH&S personnel for extended on-sit See Rene knowledgeable Associates ensures an optimal i ui r & ° ° F w ry KG = 9 Zz Q fa I I () n ° 4 zk < Ps 7) provided here should give design and operating engineers sufficient. guid- ance for the successful selection, spec. ification and operation of dust collec- tors. 7 Edited by Suzanne Shelley Asi Agarwal i a on ac Coney Cona Charleston, WV 25318; Phone: SUSIE Tia Fae Sau at ese silecroing Bow Chemica Co. in'2008, Tee lecword amt apecialat and a rtident pneumatic oaneoying ‘Pesalan Adel to pre nn ear ot esgn-consrtion operating and eudleon: Sey devectene ia peotenticcSemetog and ta oahdsThandling procesven, He holden HS. in iechanieal niacin fom Une University of ‘Mc ng oan BA Fore Mar (Charleston, WVa) Hels a member of AIChESS article Technolog Forum has wetton's nan EercHaicor ail ge daaes on gnenele nveying and butkSolds handling nmental provides experienced Cer ee Deora Dee ee ect Ce SOR ae cd rote a adlinks.che.com/4517.37 (CHEMICAL ENGINEERING WWW.CHE.COM FEBRUARY 2005 49