Document v12mD6DORj2p1JJomjxxee2M8

618 CHAPTER 36 1965 Guide And Data Book procedures involve an accelerated test using a synthetic test dust. These procedures differ in the type of test dust used nnrf in the method of measuring the dust content of the air enter ing and leaving the air cleaner. The numerical value of the arrestance, or efficiency, of an air cleaner may vary for each method of test, even if the same test dust is used. However, it generally varies when the same test method is used with dif ferent test dusts. Caution must be exercised in interpreting published arrestance data, since different test procedures prevent performance comparison of two cleaners of differing designs. The test used may yield results that are considerably different from those obtained in an actual installation. The two most common methods for evaluating efficiency are: (1) the weight method, in which the weight concentration of dust in the air entering and leaving the cleaner is measured by some suitable means, and (2) the dust spot test in which the dust content of the sir entering and leaving is measured*in terms of its ability to stain filter .paper. 1. Weight Efficiency. A weight method in common use today is described in the Air Filter Institute Code* Section I, which specifies a test dust as follows: "The test dust mixture consists of 72 percent standardized air cleaner test dust, fine (Arizona road dust, fine); 25 percent K-l carbon black; and 3 percent No. 7 cotton 1inters ground in a Wiley mill with a 4-mm screen (percentages by weight)." The entering concentration is determined' by feeding a known weight of the prepared test dust at a known and con trolled rate. The concentration of dust in the air leaving .the filter, is determined by passing the entire air flow through :a high-efficiency filtering blanket, and measuring the gain .in blanket weight. Test data and results are presented on stand ardised forms made available by the Air Filter Institute. Atmospheric dust is a heterogeneous mixture of ranging in size from a small fraction of a micron up toJarger particles tens of microns in diameter. A small percentage of the largest particles in such a mixture always accounts for most of the weight; this is also true of the particle mixture in ambient outdoor atmosphere. The weight efficiency method of rating which employs an artificially generated dust cloud of heterogeneous particle size, is a description only of the ability of a filter to remove the largest dust particles in the mixture and gives little indication of the filter performance in removing the smallest particles. For thesereasons, where extremely small particles are objectionable, the weight .effi ciency method of rating would not adequately differentiate between two filters having approximately equal weight5effi ciencies on such heterogeneous dust mixtures. 2. Duet Spot Efficiency. Recognition of these facts led :to development of the dust spot method of'testing at the Na tional Bureau of Standards.* It is based onthe idea that, since one objectionable characteristic of the finer airborne dust 'particles is the capacity to soil walla, ceilings and othet in terior surfaces, filtration of air through white filter papdr would constitute an accelerated simulation of this effect and description of efficiency on the optical density basis would property reflect this operating characteristic of air-cleaning devices. The NB8 test method is composed of two parts. One (dust spot, atmospheric) employs the normally-occurring particu late matter in the outdoor ambient atmosphere of the toting laboratory as the test aerosol. The second is a fed-dust tech nique-in which a Cottrell precipitator dust-from a specified source in combination with standardized lint-ia dispersed-in a horizontal test duct containing the filter .undergoing test. In both cases, efficiency is evaluated on the basis of dust spot 'opacity comparisons, not by weight. The National Bureau of Standards ordinarily employs ambient atmospheric dust for its evaluation of electronic air cleaners and other high-efficiency filters, and the fed-dust technique for other types of filters. In evaluating tests by the dust spot method, consideration must be given to the type of contaminant used in the test, since efficiency values are con siderably higher when Cottrell dust is used as compared to ambient atmosphere. The efficiency of many panel filters tested by the dust spot (atmospheric) method is very low (10 to 15 percent or 1^)- this test method is not generally employed for a rating of this type of filter since it fails to differentiate one filter from another, even though on a weight basis the filters in question may. exhibit significant differences in performance. Con versely, while the weight efficiency of an electronic air nWner may be high, the weight method of rating is not adequate to differentiate one device from another. . While the weight method emphasizes the larger particles in a heterogeneous dust mixture, the dust spot method empha sizes the finer particles since the fines have greater staining' power than an equivalent weight of coarse particles. This is another reason that not all cleaners can be rated by a mnglo method. a. NBS Test Method. Air samples are drawn at equal flow rates from upstream and downstream of the air cleaner and passed through known areas of filter paper. The optical density of the resulting.dust spots on the two filter papers ia measured with a photometer and used to determine the arrestance of the air cleaner. Dust spots.of approximately the opacity are developed by successive selection of different upstream and downstream sampling areas,or by upstream sampling only a frac tion of the time. When unequal areas are employed. Equation 1 is used to evaluate the arrestance E and when part-time up stream sampling is employed, Equation 2 is used to evaluate the ar- .restftDce. E = X. A d Qd Aa X Qo - 0) T Qo E- I 100 Qv (2) where Au = upstream dust spot area. Ad downstream dust spot area. Qo " change in upstream opacity of stain. Qd s change in downstream opacity of stain: T -* percentage of upstream sampling time. b. AFI Bust Spot rat He Air Filter Institute has adopted a dust spot procedure 'known as the AFI Dust Spot Test Code.4 which is identical with that part of the NBS Test Method which employs ambient atmospheric particulates. The AFI procedure specifies methods and equipment for nmlring dust spot efficiency testa on atmospheric air and specifies procedures and Report Forms for Qualification Tests. Dust Holding Capacity procedures iare not included in the AFI Dost Spot Test Code as the purpose of the code is to specify an efficiency .measuring technique and procedure. Dust holding capacity procedures are specified in Section I of the AFI Code; this procedure is used to establish that value-whether efficiency is measured by the AFI Dust Spot .or the AFI Weight method. 3. Dust-Holding Capacity. . a. AFI Method. The AFI Code, Section I (Weight), specifics procedures for determining dust-holding capacity, the amount of dust which a filter can retain and have, a resistance less than ssome arbitrary value. Under the AFI procedure, dust-holding capacity is defined as the product of the weight of dust reaching the filter and the average Efficiency at a specified resistance ana air flow. Dust loading and arrestance determinations are made 'simultaneously; arrestance and resistance are determined for each increment of feed with the results platted against weight of dust fed. Fig. 1 shows the results of an Air Filter Institute Weight Code test conducted on a typical viscous impingement. type filter. Air Cleaners AIR VOLUME-HUNDREDS OF CFM 619 (fatted by NBS--Cottrefl Fad-Oat* Method) Rg. 2 .... Dust and Lint Holding Capacity of Two Rlters (Tested by AFI Code--Section I) Rg. 1 .... Typical Unit Rlter Performance Data b NBS Method. Dirt-loading measurements are made at the National Bureau of Standards* by feeding a mixture of 4 percent standardised cotton linters and 96 percent Cottrell precipitate (by weight) to the test filter. The two ingredients are separately dispersed into the air stream by two aspirating type injectors. The measurements show the relationship between the resistance and the weight-of dirt mixture received at the filter; In the NBS method, the weight of dust received by the filter is not multiplied by efficiency to obtain dust-holding capacity and, therefore, the value is frequently referred to as dust received. The curves in Fig. 2 illustrate the difference in the dirt-loading characteristics of two filters: (1) a viscous-impingement type and (2) a dry filterwith an extcnded^urface cellulose-fiber medium, with dust and with lint This chart shows the different loading effects of lint on tire two types of filters as compared to dust without lint. NBS performs loading tests on all filters except electrostatic precipi- . tators, whether the test aerosol be atmospheric air or Cottrell precipitate. Significance and Limitations of Ratings The ideal filter test method would simulate the actual per formance of the filter when ingtallpri at the use site. The desirability of having a widely accepted accelerated test method to determine the operating characteristics of filters is generally recognized, but the variety and variability of air borne particulates,*** makea the use of a synthetic test dust in any accelerated test automatically violate the principle of simulation of actual performance. Because the indicated weight efficiency of .air . filters .is highly dependent on the particle size distribution of test dust (and this includes important consideration of its state, of de agglomeration) it is universally recognized that testing of filters in this category requires a high degree of standariiization of the test dust itself, the dust dispersion apparatus,' and other elements of test equipment and of procedures. . It has been established that the concentration, size charac teristics and lint content of atmospheric dust vary greatly in different locations.* These inherent limitations of any.acceler ated test indicate that it is not possible to reproduce exact air .filter performance. Test results, therefore, are limited in'per mitting comparisons between different air cleaners, but are useful in preparing purchase and design specifications. In loading tests, the dust-holding capacity of a filter as de termined in an accelerated test will not pehnit calculation, from dust concentrations in some locality, of the actual life of a filter'. The test results must be restricted to comparisons between different filters.. " The accuracy and repeatability of filter tests performed by reputable laboratories is quite good. Sources of differences in reported values lie in the variability of the test aerosols and dusts which is very difficult to control closely. Another source of difference is the nonunifonnity of the filter medium. The current trend toward *TM*iia made of random air or waterlaid and spun fibrous materials, introduces the inherent variability of the medium into the performance data. Realiza tion of these unavoidable variations, under even the most closely controlled manufacturing conditions, will prevent TmmmdpmtAndtng and specification of impossibly close per formance tolerances. At present, figures on air cleaner efficiency can not be related quantitatively to the rate of soiling of a space or of mpphnnlral equipment, nor to the intensity of other objec tionable effects. Thus, air cleaner efficiency data do not pro vide the information necessary to design properly air cleaning installations to control wall smudging, deposits on heat trans fer surfaces, soiling of furnishings, etc. In other words, in the absence of knowledge of a quantitative relationship between air cleaner efficiency figures and the various objectionable effects, these figures have only the limited value previously indicated.. IAsking systematically developed data of this nature, knowledge developed from practical experience must be considered. Other Filter Test Methods Different dusts and techniques have been used in various laboratories to determine gravimetric and dust spot efficiency of filters, and opinions on the relative merits of each have been diverse. Questions, pertaining to this matter, however, can readily be resolved by reference to the principle, cited above, that present test results must be limited to a compari son between different filters. Any test method can only justify itself if it report test results on a wide variety of air filters under its particular test conditions. Conversely, test results on a single filter by some nonstandard method are valueless because they can not be compared with competing air cleaners.; A test method developed by Whitby et al using a pulver ized anthracite coal-carbon black mixture for a test dust, was the ha-*TM for an interesting report1* on the relative, perform ance of a wide variety of air cleaners. While that test method Viaa Dot gained official recognition, the reported test, results y