Theory

Air filtration involves the separation of "particles" from airstreams. Understanding separation techniques requires an exact definition of what particles are. The size of filterable particles is normally defined in terms of microns. One micron equals 1/25,400 of an inch. Even with a microscope, particles less than 1 micron are difficult to see. Tobacco smoke, for example, typically ranges in size from 0.01 to 1.0 micron. Particulate generated in grinding plastic might be as small as a few microns and as large as several hundred microns. When the naked eye "sees" tobacco smoke, it is actually seeing the light which is being reflected or scattered from millions of small particles. The wavelength of such light is in the ultraviolet range, hence the blue haze normally associated with a smoke-filled room. Plastic dust from grinding, on the other hand, can be seen as individual particles floating in the air and rather quickly settling to the floor.

As particles become very small, they cease to behave so much like particles as they do gas phase molecules. It is difficult to tell whether such small particles are actually suspended in air (particles) or diffused throughout it (gas or vapor). The bottom boundary where particles act as true particles is about 0.01 micron. The normal theory of separation does not apply to particles below this size and removing them from air requires techniques reserved for gaseous materials. Particles above 0.01 micron are usually considered to be filterable. Removal equipment is almost as diverse as the size ranges of the particulates generated.

Filtration of particles relies on four main principles:

1 Impaction
2 Diffusion
3 Straining
4 Electrostatics

Particles can be influenced by any one of these principles, or all of them simultaneously.

Impaction

As large particles move along with an airstream, their inertias prevent them from making abrupt changes in direction. If an obstruction such as a series of water droplets or fibers (glass, foam rubber, cloth, etc.) is placed randomly across the airstream path, there is a certain probability that a given particle will collide with the obstruction. As particle size and the number of particles increase, so does the probability of collision. Thus the efficiency of removing particles from an airstream by impaction is a function of particle size, fiber size and the number of fibers. The greater the number of fibers (thus the deeper the bed, the higher the pressure drop), the higher the efficiency. In turn, as the dust particles collect, they themselves become part of the filter media, thereby increasing efficiency by adding to the number of possible collisions for other suspended particles. As the collected particles build up on the filter, system pressure drop increases, usually a good indication that the path through the filter, the equivalent filter depth, is increasing. This principle of filtration is most commonly found in fiber filters and certain wet collectors.

Diffusion

When particles become very small, their mass is so low that, should they collide with any air molecules, just the random motion of the air molecules will cause them to rebound randomly. This motion is commonly referred to as Brownian Movement. If the velocity of the airstream is low, this diffusion movement will in turn cause random collisions with fiber or droplets in the way of the airflow. Hence, much like impaction, probabilities can be developed for collisions due to this diffusion. Key factors are fiber size, fiber quantity and airstream velocity. As with impaction, as more and more particles collect, the probability of collision (efficiency) for other particles is enhanced, but with an associated increase in pressure drop.

Straining

if the width of a passage is smaller than that of the particle suspended in the airstream, then the particle will be stopped and held. However, as each particle plugs a hole, air resistance increases. Standard house screens are typical of this filter type. Small particles pass through it but bugs cannot pass through. Very small particles are seldom collected using this method which is primarily used only for specialized laboratory experiments. This is ahnost never the primary filter mechanism in industrial dust collection.

Electrostatics

If a charged particle passes through an electrostatic field it is attracted to an oppositely charged body. Such charges can be generated and imparted to particles in an airstream in much the same way as static charges develop during the combing of one's hair or just walking across a rug. Electrons are stripped from large quantities of molecules with the net effect that particles of dirt not otherwise collected might be charged by friction as they pass through, then collected as they attach themselves to oppositely charged bodies. This effect can occur inside such filtering devices as fiber beds which operate primarily on the principles of impaction and diffusion but have their efficiencies enhanced by electrostatic effects.

Charges may be purposely induced onto airstream particles by applying energy to a special configuration of wires and plates stretched across the airstream. These devices, called electrostatic precipitators, form a special category of air filtration. Whether particle charges are induced by applying energy to a dirty airstream or occur naturally, they can be valuable tools in increasing air cleaning effectiveness.

Types of Filter Systems

Particle control technology can be roughly divided into two broad categories, namely,

1 Air Filtration
2 Dust Control

The first is a means for controlling the contamination of ventilation air inside of office and other occupied building space. The second usually pertains to heavier industrial emissions, either inside or outside of a plant. In this latter case, particle concentration often dictates the control technology employed. For normal ventilation purposes, contaminants are usually defined as particles less than 10 microns in size and concentrations less than 2 Mg/M3 (milligrams per cubic meter). In contrast, dust control devices must be able to handle concentrations of 20 to 40,000 Mg/M3 and particle sizes ranging from 0.5 to 200 microns.

Control equipment for nuisance industrial dust can be divided into the following types:

1 Media Collectors
2 Electrostatic Precipitators.

Media Collectors

Filter media for dust collectors can be anything from cotton bags to glass fiber sheets. While temperatures of operating airstreams or of their entrained particulates usually influence media selection, basic operating principles are the same for most filter types. They operate on the principles of impaction and diffusion.

Where dust concentrations are very heavy, devices called 'baghouses' collect dust particles and are usually equipped with 'auto-renewal' hardware such as shaking or vibratory mechanisms. These shake collected materials from the media, thus extending media life and allowing the device to stay on line for long periods of time even where loading is heavy. The rapping, or shaking, action can be activated while the system is on line and operating to its airflow capacity. Normally only one section of bags is cleaned during each rapping cycle.

The initial buildup of dust on the filter media is called the 'cake'. It aids collection efficiency by extending the equivalent filter collecting surface. Buildup of cake on the filter media is watched closely, since completely cleaning it off results in a severe decrease in collection efficiency.

Such equipment normally requires a large amount of floor space for installation. With proper fabric selection, this type of system can achieve efficiencies of 99.9% by weight, but very poor performance is the rule for particles in the 0.1 to 2.0 micron range. Pressure drops are usually very high in this type of collector.

Other types of filters, used in lighter applications more closely approximating ventilation, include panel, vee bag and pleated filters.

Electrostatics

Electrostatic precipitators are divided into two categories, namely, the Single-Stage and Two-Stage types. The single-stage version is used mostly where heavy concentrations of industrial combustion by-products are present. Concentrations as high as 200 Mg/M3 can be handled by such a device. Efficiencies of 99% are possible, with low pressure drops. Space requirements are large per CFM of airflow, although usually not as large as with the fabric baghouse. Voltages range from 20 to 65 KV, velocities from 60 to 300 FPM. Components can handle temperatures up to 850 degrees F. Efficiencies on particles 2 microns and smaller are extremely low.

Two-stage electrostatic precipitators operate at much lower voltages and temperatures. They are designed to handle much smaller particles (down to 0.01 microns) associated with industrial oil smoke, metal oxides such as welding, etc. Normal concentrations are much lower.

How all this works

The self-contained multiple-stage media type collector operates by means of a system fan, usually built into the filter cabinet, it moves contaminated air through its filters, cleans the air and returns it back to the work space. The number of filters placed in series varies with the application, but the two most con-tmon filter configurations are called two-stage and three stage, respectively.

Two stage filtration consists of a prefilter and final filter and is used in most standard dust removal situations. For dust applications, a 30% pleated prefilter is followed by a bag filter with efficiency up to 95%. For oil mist applications, a special impingement-type channel prefilter removes large oil and grease droplets, allowing a 95% bag filter to take care of remaining particulates.

Three-stage filtration consists of prefilter, intermediate filter and final filter. For dust applications, a 300/o prefilter is followed by an intermediate bag filter with efficiency to 95% and anchored by a final box-cell filter rated up to 99.97% efficient (Hot DOP Test Method). Where odors are a problem, the final filter takes the form of a vee-bank carbon cell.

Where the application dictates, further filtration may be necessary. Accessory filter configurations such as 'wraparound' prefilters are available as the application dictates. The wraparound prefilter is a bolt-on accessory designed for applications where much of the dust to be collected is . very large and removable from the air by large-pore, inexpensive media. This type of filter significantly extends the life of other filters operating downstream.

Cleaned air passes through a built-in unit blower and is exhausted, usually back to the plant. Airflow and duct static pressure losses at the site dictate blower/motor drive factory settings of the Model DB unit. Another unit, the Model DA, is designed only for hanging directly in the plant environment (not for ducted installations) and has a multiple speed, direct drive motor with no blower/motor adjustments.

Rating Filter Performance

To understand filters and their application, it is necessary to rate them for pressure drop, dust holding capacity and efficiency. These factors are often abused in industry when only data favorable to a particular filter is presented in sales literature. To fairly evaluate filter effectiveness, several test procedures have been adopted by filter manufacturers and users.

For the kind of dust concentrations encountered in most industrial work areas, the standard most often used for measuring filter efficiency is ASHRAE (American Society of Heating, Refrigerating and Air Conditioning Engineers) Standard 52-76. This standard addresses multiple methods of testing, the most important filter tests being:

1 Weight Arrestance
2 Dust Spot Efficiency
3 Hot DOP (Di-octyl Phthalate) Test, is U. S. MIL STD 282 (1958).

The Weight Arrestance Test is a simple test which involves feeding a synthetic dust to a filter and ratioing the weight of dust exiting the filter to the weight of dust originally fed into the filter. Since small particles have little mass, this method offers almost no way of factoring small particle collection efficiency. The method is used for low and medium efficiency media filters. During filter testing, pressure drop is observed until it reaches a final value (usually 11' w.g.). The number of grams of dust, less that logged on a special high efficiency filter behind the test filter, yields the filter's final dust holding capacity at 1" w.g. pressure drop. Collection efficiency is likewise a function of the weight of dust collected on the test filter and the final filter. Since dust weight is not logged until the test filter becomes loaded, this method yields only an average efficiency, averaged over the entire test run. It gives no indication how long it took for the filter to build to the final rated efficiency or give any clue about starting efficiency.

Where small-particle efficiency is critical (most industrial oil smokes generate particles in the 0.01 to 2.0 micron size range), the Dust Spot Test is often used. Here standard ambient air is passed through the test filter and the airstream has special test filters in front of and behind the test filter to monitor the presence of airborne particulate. Over time, both filters become soiled and are measured optically for relative soiling. These results are then translated into a filter efficiency rating.

A very common method for measuring the efficiency of high-efficiency filter media is the Hot DOP Test. Here DOP is boiled and the vapor injected into the airstream in front of the test filter. As the vapor condenses back to ambient temperature, it forms very uniform droplets about 0.3 micron in diameter. By the use of light scattering instrumentation, upstream and downstream particle concentrations can be measured. Filters called HEPA (High Efficiency Particulate Air), usually rated for efficiencies in excess of 99.9% on these 0.3 micron size particles, are tested using this method.

To rate the performance of filter media or a filter assembly, a wide range of test methods may be used, one which might be ideal for industrial use and another for commercial use. Several EPA test methods could be applied but tend to be oriented toward emissions with high, continuous concentrations of particulate as generated by continuous processes.

To understand how deceptive efficiency tests can be, Table 9-1 shows how certain filters would respond to testing by the three test types. For example, one of the standard filters used is rated 65% as tested under ASHRAE 52-76 (Dust Spot). Going down the Dust Spot Method column in Table 9-1, this filter would fall into the 60-80 range in that column. Looking left and right, note that this same filter could be rated 95% efficient if measured by the weight arrestance method or 35-40% efficient if measured by the Hot DOP Method. The only types of filters which will show high efficiency on the sub-micron particles generated in the Hot DOP Test are HEPA filters and electrostatic precipitators.

TABLE A

ASHRAE WEIGHT	ASHRAE DUST	MIL-STD 282
 ARRESTANCE	   SPOT		  HOT DOP
   METHOD	  METHOD 	  METHOD

   70-80	  15-30		    0
   80-90	  20-35		    0
   90-95	  40-60		  15-25
    95		  60-80		  35-40
    NA		  80-90		  50-55
    NA		  90-98		  75-90
    NA		   NA		95-99.999*

Most units can be supplied with a number of filter combinations, each arranged to suit the application. Every filter configuration can be tested individually for efficiency and pressure drop. Standard filters have even been tested for efficiency by particle size . Most standard filter combinations have been tested for pressure drops at nominal airflow ratings. When units are specified as being ducted for source capture, blower speeds are preset at the factory to match specified field conditions. Detailed specifications for all standard unit configurations are listed from their manufacturer.

APPLICATION CONSIDERATIONS

Selection Guidelines

Some filter units use multiple filters in series in the direction of airflow. The object of such a configuration is to achieve high efficiency of particle removal at the lowest possible operating cost. Choosing the best combination from the many filter sets available depends on detailed knowledge of the generating process, the problem material and its probable interaction with the filter media. Some of the more important considerations in filter selection are described below.

Respirable Fraction

Particles which float in air are inhalable into human lungs. Particles 10 microns and below fall into a category known as the "inhalable fraction", those particles which can by-pass the body's filtration mechanisms and deposit. Particles 5 microns down to 0.1 microns have a high probability of being trapped in human lungs and are referred to as the "respirable fraction".

If particles larger than 10 microns are present in any quantity in plant air, they can make breathing very difficult for workers and should be removed frombreathing zones. They are relatively easy to filter in mechanical filtration devices.

Particles in the inhalable and respirable fraction ranges are dangerous, but in a different way. They cannot be individually seen, they are easily trapped in the lungs and they are difficult to expel. Filters to remove such small particles must have extremely small pores.

Efficiency

In general, there is little advantage in selecting filters of higher efficiency than necessary. Besides having higher pressure drop and shorter life, they cost more. Using filters with higher than required efficiency can cost the user on the front-end (high-efficiency filter media is more expensive) and during operation (higher pressure drop, therefore higher fan horsepower cost, and shorter filter life). While choosing too low of an efficiency can result in ineffective particulate removal, excessive efficiency for the task can make filtration expensive, annoying and sometimes downright impractical.

Loading

This type of filter is designed for light dust loading, no more than about 2 Mg/M3. Heavy loading will quickly blind the filter network, reducing airflow. Yes, filter efficiency does increase as filters load, but so does the pressure drop across the filters. Since air to most filters of this type comes from fixed speed fans, and since most fixed speed fans move less air as pressure drop increases, airflow to most filter units drops with loading Excessive loading increases filter changeout frequency, sometimes unacceptably so. Thus, while this type of filter gives great effectiveness for the investment, there are heavy loading applications where a less desirable filter alternative must be selected.

Filter Combinations

Using filters in combination can extend the life of the final (primary) filter by as much as 150%. If a higher than normal percentage of large particles is present, the extended life of the final-filter would naturally be even greater. Conversely, where only sub-micron particles are present, they tend to pass through low-efficiency prefilters with little helping effect for the final filters.

As an example, assume there is a light but troublesome concentration of fume in a manufacturing area. Particles are thermally generated oxides in the 0.1 to 5.0 micron range. To collect a high percentage of this material would require a 4 inch deep pleated prefilter, with a nominal 30% efficiency, followed by a 95% efficient bag filter. From TABLE B below, the approximate efficiency of the first filter is 58% on 5 micron particles. Referring to TABLE C for the efficiency of the 95% bag filter the efficiencies shown in the next table can be derived.

TABLE B

Size(micron)   4"(30%*)Filter   95%*Bag   Total

    5		   58		  99	   99
    1		   5		  97	   97
   0.5		   1		  55	   53
   0.1		   6		  52	  54.9

If it is known that particle sizes are those for oil smoke, where particles mostly fall between 0.1 and 1.0 micron, then another filter would have to be added to the system to collect a high percentage of the expected sub-micron material. This filter might be necessary to keep a haze from forming in the work area. Efficiency might appear as shown in TABLE C.

TABLE C

Size(micron)   4"(30%*)Filter   95%*Bag   HEPA**   Total

    5		   58		  99	  99-99+   99-99+
    1		    5		  97	  99-95    99-99+
   0.5		    1		  55	  99.92    99.99+
   0.1		    6		  52	  99-91    99-99+

* Data per TABLE A

Sample Unit Efficiency On Small Particles Using HEPA Final Filter

The reason for keeping the 95% filter in place when adding the HEPA filter is to reduce the sub-micron particle loading to the HEPA filter. Note the 55% efficiency of the cheaper 95% bag on 0.5 micron particles.

Selecting bag filters for new applications is no easy task for the inexperienced. Choosing between 65 and 95% efficient bags as final filter in a two-stage system can be a very subtle exercise. One helpful rule of thumb is that, if individual dust particles are visible, particles would be bigger than 5 micron (probably bigger than 25 micron since that is probably the smallest particle visible to the human eye) and a 65% filter would be the logical choice for final filter. Whenever a blue "haze" is visible in the smoky area, that is a tell-tale sign that sub-micron particles are present and, at a minimum, a 95% bag filter should be used. Prefilter material is usually a standard pleated media unless the incoming airstream is wet with oil or water droplets. Then the metal impingement filter should be chosen since it it holds up in the presence of moisture and can be spray or dip cleaned of greasy buildup and reinstalled into the unit.

Gases, Vapors and Odors

Control for gases and vapors is somewhat more complicated than for particulate. Particles occupy points in space whereas gases try to occupy the whole space. Gases are assemblages of evenly spaced molecules. Using the same principles to collect them as to collect actual particles is possible but, in most cases, far from economical. Before deciding how to remove a gaseous material from an airstream, one must confirm that a problem exists, and exactly what it's all about. Like no other form of emission control, control of the gas phase requires accurate definition of the problem.

Gases present a variety of environmental problems. They may be irritating to the eyes (ammonia), have a cumulative effect on the human body or the environment (phenols), have carcenogenic potential (benzene), be toxic (hydrogen sulfide) or simply make the body feel uncomfortable and 'stuffy'. They may simply have an acutely unsavory odor. Safe concentrations of these disagreeable materials vary widely and there may be significant differences in control solutions meeting legal compliance, making employees comfortable and adequately protecting those employees.

Perhaps the toughest gas phase problem to solve is that of odor. The presence of a particular compound may be detectable from its pronounced, 'signature' aroma. Its concentration may be magnitudes less than that recognized as a 'problem', but employees might define it as totally unacceptable. The human nose is a truly remarkable and sensitive mechanism and is often responsible for the installation of control equipment. There are ways of removing troublesome materials from the air, but reducing concentration and eliminating odor do not necessarily go hand in hand. The table on page 21 of the Appendix gives an indication of the threshold odor levels for some common materials. The two most common vehicles for removing such materials from air are "adsorbent" materials and "reactive-oxidant" materials. The majority of nuisance vapors encountered in industry are most easily controlled through the use of carbon as the removal agent. Activated carbon is an adsorbent. Carbon can be found in nature in a variety of forms, from lampblack to diamonds. The form used in adsorption is called "activated carbon" or "activated charcoal".

This type of carbon is 'manufactured', in that specially selected materials such as nut shells are 'burned' at high temperatures in furnaces where controlled low quantities of oxygen are present. The resulting carbonaceous material is highly porous, resembling coke, and has a special tendency to trap molecules of various materials inside its pores (but still on its surface) as they try to pass, while allowing air molecules to pass through. This bonding of foreign materials to itself is called 'adsorption' and carbon is one of only a handful of materials which can do this effectively across a wide range of materials.

Carbon is an active element but as such tends to interact differently with different gaseous materials, depending on their physical and chemical properties. It is particularly useful in industry where materials with medium to high molecular weight and high boiling points are generated. These materials are very responsive to adsorption by carbon, however, in addition to its affinity to adsorb these troublesome hydrocarbons and other gaseous materials, it unfortunately can also hold a high weight percentage of water. Some good news is that carbon has the redeeming characteristic that heavier molecular weight materials tend to displace the lighter water from the carbon surface and that water can normally occupy no more than about 25% of carbon's nominal adsorption capacity. It is true, though, that carbon's effectiveness is significantly reduced in high humidity applications.

Besides activated carbon, other dry adsorbents used in industry include silica gel, zeolite and various activated alumina. Each material has its own special application range. Since carbon is the most conunon, it is used as standard by United Air Specialists. These specialty adsorbents will not be discussed in detail in this text, but can be made to handle such materials as sulfur dioxide, annnonia, ethylene and formaldehyde which are not adsorbed effectively by carbon.

Carbon filters in units consist of a series of filter "leaves", filled with fine carbon granules, and having screens fore and aft in the direction of airflow. These leaves are arranged in a "vee" configuration, for maximum surface area facing the airstream. Each filter leaf forms a small 'bed' of carbon one inch deep in the direction of airflow. Each cell can handle up to 2000 CFM of airflow. Depending on the material to be adsorbed, this filter can hold up to 30 or more percent of its own weight in object gas. Retentivity of carbon for various materials is published in various journals and manufacturers' literature.

Handling production quantities (pounds or gallons of solvent per hour) requires a different configuration of carbon system, one with many inches of carbon thickness facing the airstream and with some in-place means of "regenerating" the carbon, (i.e., removing the solvent). Thus, connecting a carbon system directly to a solvent holding tank with high solvent concentration is less than desirable.

Periodically, special needs arise which dictate the use of materials referred to as "reactive oxidants" or "absorbents". These materials interact chemically with offending vapors, converting them into simple new compounds which are retained in or on the media. Materials like this are used up over time and eventually require disposal and replacement with fresh material. In certain circumstances, air cleaners can be supplied with such material, either in bulk form or as part of filter frame assemblies.

Air Change Guidelines

Where filter units are to be unducted, for area capture of airborne contaminants, it is important to determine what percentage of the workspace volume will be cleaned per unit time. This subject is covered in detail in the Interior Air Cleaning Systems section devoted to Unducted Units. The procedure for calculating air changes in an application is similar except that slightly different safety factors must be brought into play.

Most importantly, keep in mind that, unlike most precipitators, filters are intended to load up so much with collected material that a pressure gage or pressure switch wams of decreasing performance and impending

TABLE D

	MATERIAL		  PARTICLE SIZE

	Offset Powder			M
	Paper Dust			M
	Carbon Black			M
	Paint Overspray			M
	Fugitive Dust			M
	Portland Cement			M
	Pottery Dust			Mm
	Grinding			M/L
	Cut Off Saw			M/L
	Welding				S
	Blue Smoke			S

	where S = small, M = medium, L	large

filter changeout. The accompanying scenario is that unit fans, running at fixed speeds, move less air as the filters become dirty. Therefore, to maintain a good air turnover rate even when the filters are soiled to a point of changeout, an additional factor needs to be assigned when sizing an area. Another differing factor for determining air changes, and therefore number of filter units, is the particulate size to be collected and the quantity generated. TABLE D shows some typical media filter applications and the expected particle size range for each. TABLE E shows a combination of air change guidelines. While there are no hard and fast rules, note that air changes for filters always exceed those for electrostatic precipitators.

TABLE E
A. Media Filter Air Change Rates

    FILTER               NUMBER OF AIR CHANGES
  EFFICIENCY
(ASHRAE 52-76		 MEDIUM		   SMALL
  DUST SPOT)		PARTICLES	 PARTICLES

  90 to 95%	       Same Number    1.2 Times Number
   Filter	     of Air Changes    of Air Changes

  60 to 65%	     1.5 Times Number
   Filter	      of Air Changes	   N/A

       AIR CHANGES		  CONTAMINANT
	PER HOUR		GENERATION RATE

	4 to 6			    Light
	6 to 8			    Medium
	8 to 12			    Heavy

In many cases, particulate loading at the inlet is difficult to predict. Selection of filter media combinations is then based on experience with similar materials, with a goal of final filter life being between 3 and 6 months or longer if possible. On-line time will, of course, vary from one application to another and is what makes filter media selection an inexact science.

If final filter life is measured in months, prefilter life might be measured more in terms of weeks, or even days. All filter life, after all, depends, to a great degree, on the generation rate of particulate in the work area, and in particular on the concentration at the filter unit.

Concentrations of 1 Mg/M3 are considered relatively light, while those of 5 Mg/M3 are usually heavy and the cause of rather frequent filter changeout.