Annals of Occupational Hygiene Advance Access originally published online on August 26, 2005
Annals of Occupational Hygiene 2005 49(7):569-574; doi:10.1093/annhyg/mei034
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2005 British Occupational Hygiene Society Published by Oxford University Press
Original Article |
Evaluation of Particulate Filtering Respirators Using Inward Leakage (IL) or Total Inward Leakage (TIL) Testing—Korean Experience
1 Department of Occupational Health & Safety Engineering, Inje University, Gimhae, Gyeongnam-do, 621-749, Republic of Korea (South Korea); 2 Department of Environmental Education, Kongju National University, Gongju, Chungnam-do, 314-701, Republic of Korea (South Korea)
* Author to whom correspondence should be addressed. Tel: +82 55 320 3285; fax: +82 55 325 2471; e-mail: dhan{at}inje.ac.kr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Korean certification regulation for particulate filtering respirators requires inward leakage (IL) or total inward leakage (TIL) testing according to European Standard EN 13274-1, and the standard levels of compliance are similar to those of the European Standard. This study was conducted to evaluate particulate filtering respirators being commercially used in the Korean market using an IL or TIL test and the validity of standard level in Korea. Three half masks and 10 filtering facepieces (two top class, four 1st class and four 2nd class)—a total of 13 brand name respirators—were selected for the test with panels of 10 subjects. Each subject was classified with nine facial dimension grid squares in accordance with face length and lip length. IL or TIL testing was conducted at the laboratory of the 3M Innovation Center in which the experimental instruments and systems were established in compliance with European standards. The testing procedure followed EN 13274-1 (2001). As expected, leakages of half masks were less than those of filtering facepieces and the latter were significantly different among brands. TILs of the 1st class filtering facepieces were found to be much more than those of the 2nd class and the result may cause a wearer to get confused when selecting a mask. The main route leakage for filtering facepieces may not be the filter medium but the face seal. Therefore, it is necessary to develop well-fitting filtering facepieces for Koreans. Because leakages were significantly different for different facial dimensions, a defined test panel for IL or TIL testing according to country or race should be developed. A more precise method to demonstrate fit, for example, fit testing such as in the US regulations, will be needed before IL or TIL testing or when selecting a respirator. Another finding implies that geometric mean of five exercises for IL or TIL may be better than arithmetic mean to establish a standard individual subject mean.
Keywords: filtering facepiece • half mask • inward leakage (IL) test • Korean • respirator • test panel • total inward leakage (TIL) test
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
If the facepiece does not fit, it does not protect the wearer regardless of how effectively other components of the respirator work. Because the fit of a respirator is the crucial determinant of whether adequate respiratory protection is possible, fit testing is an important factor in protecting the wearer's health. Fit testing of respirators has been therefore required in the United States (CFR, 2003
). Presently, respirator-use standards and regulations requiring fit testing are limited to the United States, and the majority of countries do not require fit testing. Another scale related to respirator fit is inward leakage (IL) testing or total inward leakage (TIL) testing. Typically, TIL is recognized to be the sum of all sources of leakage into the facepieces (Myers, 2000):
 |
TIL of a test aerosol from all sources (e.g. face seal, filter and exhalation valve) is determined for a panel of people (e.g. 10 subjects) wearing the respirator with the filter to be used in the workplace. If neither the HEPA filter nor USA 100 series filters (broadly equivalent to European P3 filters or FFP3 filters) are used in the testing, filter penetration may not be negligible in certain cases. Therefore, TIL except for these filters may not predict how well a respirator will fit a specific individual. In many European countries, facial fit is evaluated during respirator certification using an IL or TIL test instead of fit testing (EN 136, 1998; EN 140, 1998; EN 13274-1, 2001).
There are no fit testing regulations in Korea, but IL and TIL testing regulation for particulate filtering respirators certification (Korean ministry of labor, 2002, Notice 2000-15) is similar to the European Standard. In this country IL or TIL testing has been conducted to get certification for marketing since July 2002. At this time it is necessary to evaluate particulate filtering respirators by using IL or TIL testing. This study was carried out to evaluate IL or TIL for particulate filtering respirators, to validate the standard and to know if there are any differences of leakages with facial dimensions.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
Respirators
A total of 13 brands were selected for testing: half mask A (imported), B and C (domestic); filtering facepiece top class A (imported), D (domestic); filtering facepiece 1st class A (imported), D, E and F (all domestic); and filtering facepiece 2nd class A (imported), D, E and F (all domestic). They are widely used in the Korean market and are already certified.
Test panel
Korean regulations for IL or TIL testing require a 10-member test panel (subjects) irrespective of facial dimensions. A 10-member test panel with a wide spectrum of facial dimensions was constituted for the purpose of this study. The procedure for half and quarter mask fit testing for Koreans developed by Han proposed a 16-member panel (Han, 1999); however, 10 members based on IL or TIL regulations were selected for this study. The test panel consisted of 10 male volunteers as shown in Fig. 1; face length (nasion-menton) and lip length (width of mouth) measurements were used as determinants for facial dimensions. The test panel was divided into a three by three grid and each grid square is occupied by one number subject except for grid No. 5. Grid No. 5 is occupied by two subjects since it represents the facial dimensions of most of the population.
|
Experimental method
The experiment was conducted at the 3M Korea Innovation Center in which the experimental instruments and systems for IL and TIL testing had been established in compliance with the European Standard (EN 13274-1, 2001). These instruments and systems are almost the same as those at KOSHA (Korea Occupational Safety and Health Agency) in which respirators are certified for marketing. Most instruments used in this experiment were made in the UK, including the atomizer (SFP Service, UK), photometer (Model QL30F/RFI, Electron Tubes Ltd., UK) and pulse sampling unit (SFP Service, UK).
The test procedure for IL or TIL was performed in compliance with the European Standard by using the sodium chloride (NaCl) method (EN 13274-1, 2001). Before testing each test subject was asked ‘Does the device fit?’ and if the answer was ‘Yes’ the test was continued. If not, the subject was told to adjust the respirator, and this could be repeated several times. Each subject was subjected to testing for all 13 masks one by one. Five test exercise periods were fulfilled, such as a first walking period with normal breathing (NB1), side to side (SS), up and down (UD), talking (T) and a second walking period with normal breathing (NB2); each exercise period spanned 2 min and was performed on the treadmill at the speed of6 km h–1.
Standard levels of IL and TIL in Korea
There are three categories of filtering facepieces in Korea, the top class (highest level of filtering efficiency, 
99.0%), 1st class (second highest level of filtering efficiency, 94.0%) and 2nd class (low level of filtering efficiency, 80.0%). Standard levels of IL and TIL in Korea are similar to those of European countries. For example, in the case of the half mask, at least 46 out of 50 individual results over each of the exercise periods (i.e. 10 subjects x 5 exercise periods) shall be not >5%, and in addition at least 8 out of the 10 individual wearer arithmetic means over all exercise periods shall be not >2%. Table 1 shows standard levels of IL or TIL in Korea with each classification of particulate filtering masks.
|
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
IL or TIL among brands and types
Table 2 shows 10 individual arithmetic means and standard deviations of IL or TIL by brands and types. As expected, ILs of half masks were much less than TILs of filtering facepieces. In the case of half masks there were statistically significant differences among the three brands (P < 0.05). TILs of both the top class and 1st class of filtering facepieces were significantly different (P < 0.05) among brands, but those of the 2nd class of filtering facepieces were not. Unexpectedly, the 1st class of filtering facepieces had more TILs than the 2nd class except for brand A. This result cannot be fully explained from this experiment; however, it may have resulted from the following. Filters of the 1st class may have more layers or may be thicker than those of the 2nd class, which increases the filtering efficiency. This in turn increases the pressure drop when a wearer is inhaling. This increased pressure drop of the filter media may increase face seal leakage when the mask does not match with the wearer's face. The aforementioned speculation aside, since the result may contribute to confusion when selecting filtering facepieces in Korea, it is thought that it is necessary to study more precisely the TIL of the 1st class filtering facepieces tested.
|
Although all masks tested had been certified by KOSHA, 6 out of 13 brands (46%) were found to be not in compliance with Korean standard levels. It is not easy to explain whether the differences between these measurements and KOSHA's measurements are caused by the validity of KOSHA or this study and/or the problem of the selection of the test panel or testing procedure. First, the differences may be explained by the testing procedure. The test panels were informed that if they wish to adjust the facepiece during the test due to the perception of leaking they may do so, and after adjusting the facepiece the relevant period was retested. In fact, there were small differences of leakages between the first set and the retested set. Because EN 140 does not stipulate how many set frequencies for each exercise period should be performed, one set frequency is usually the norm. The result of this study indicates that more than one set would be desirable to increase the reliability of testing. Second, the differences might be attributed to the test panel selected. If the facial dimension category of the test panel selected in this study differed from those in KOSHA's certification test, this result would not be unexpected. Accordingly it is recommended that a test panel with constant facial dimensions for each country or race should be selected for IL or TIL testing.
Penetration through filter media versus face seal leakage
It is assumed that in fit testing there is no penetration through filter media, but leakages evident from TIL testing include penetration through filter media as well as face seal leakage. It is necessary to know what proportion of the TILs occurs due to penetration through filter media and what the main leakage route is. Table 3 shows maximal and minimal leakage values as well as the ratio of maximal to minimal values by brand and type. There were huge differences between maximal and minimal leakage values. For example, there is a 292 times difference between maximal and minimal leakage values for the company A 1st class of filtering facepieces. The largest value out of minimal leakages such as 0.18 for the company B half mask, 0.39 for the company D top class, 0.60 for the company F 1st class and 0.89 for the company E 2nd class of filtering facepieces may be the most important. Compared with Korean compliance standards, these happen to be very small values. On the other hand, maximal leakages are mostly very large values, far beyond compliance with Korean standards. If penetration through filter media had been large, minimal leakage would have been large and the ratio of maximal to minimal leakage would be small. In other words, if a mask is fitted well, the wearer's face penetration through filter media will be very little. This result indicates that while penetration through filter media may be insignificant, leakages would be mostly attributed to face seal leaks. Therefore, in order to decrease IL, developing well-fitting facepieces is much more a priority than developing a high efficiency filter in Korea.
|
Differences of IL or TIL among facial dimensions
Table 4 shows, for each grid square, P-values of IL or TIL. The overall values represent all samples of the type regardless of brands, for example, analysis for 150 samples (3 brands x 50 samples each) in the case of the half mask. As shown in Table 4, except for half mask A, leakages are significantly different among the nine grid squares for each brand and each type(P < 0.05). In EN 140 and the Korean standard the faces of the 10 test subjects should be described (for information only) by four facial dimensions, (in mm) such as length of face (nasion-menton), width of face (bizygomatic diameter), depth of face and width of mouth. Facial dimensions are required only to report the test results, but there are no requirements for facial dimensions when selecting test panels, according to the standards. This implies that facial dimension grid squares or categories in accordance with country or race should be defined for IL or TIL testing.
|
In the case of filtering facepieces, the facepiece contours that contact the subject's face were usually the same within the same brand regardless of class. Differences of leakages for all samples (e.g. 150 samples for company A and D each, and 100 samples for company E and F each) according to the company were analyzed in nine grid squares. Table 5 shows the mean values of all samples of all classes of filtering facepieces in a company, mean values in each grid square and P-values between one grid square and the others. For example, the arithmetic mean of leakages for the three types of filtering facepieces of company A is 5.62 and the mean of those in grid square
is 30.72. Since leakage in grid square is significantly much more than in the other eight grid squares, filtering facepieces made by company A may be less well-fitting for wearers belonging to grid square .
|
According to the procedure for developing a test panel researched by Han (1999)
, while a proportion of the population in grid square 
is the highest out of the nine grid squares, those in grid square , 
, 
and 
are relatively low (See Fig. 1). Therefore, the Korean test panel for men for half and quarter masks consists of one subject in each of grid squares , , and , four in grid square , and two in each of grid squares 
, 
, 
and 
. For this reason, it is thought that a filtering facepiece in grid square having a lower leakage may be well-fitting for many wearers with a wide spectrum of facial dimensions. Leakages of filtering facepieces made in company A in grid square , , , and are less than the mean value of 5.62, and, additionally, significantly less in grid square compared with the other grids. This result indicates that because filtering facepieces made by company A have lower leakages in grid square , which represents a large proportion of the population, they may fit relatively many wearers.
On the other hand, it is thought that just asking about fit and perception of leaking is not enough to demonstrate adequate fit before IL or TIL testing. So a more precise method to demonstrate fit, for example, fit testing such as USA regulation (CFR, 2003), should be used. Commercially, it is the best way to manufacture a respirator that can fit as large a proportion of the population (whether national, ethnicity-based or global) as possible. If it is not possible, fit testing of individual wearers should be established to ensure the wearer's safety.
Goodness of fit for ILs
In EN 140 and the Korean standard, two determinants, such as 50 individual values over each of the exercise periods and the individual wearer arithmetic mean (10 subjects) for leakages, should be calculated. Arithmetic mean is calculated under an assumption that variables are normally distributed. It was necessary to find which distribution of leakages represents good fit. Five variables, that is, five exercises periods are too small to evaluate distribution and it is difficult to achieve a meaningful distribution. In addition, the statistical package used for testing needed at least six samples as variables (Vos and Christie, 1997). Thus, 50 individual leakage values (10 subjects x 5 exercises) were used to test goodness of fit for distribution. Testing goodness of fit was conducted using the Shapiro–Wilk's W-test in the statistical package. Table 6 shows the results of testing goodness of fit for distribution of leakages by brand. In 6 out of 13 brands lognormal distribution of leakages was found to be better fitting and in the other seven brands both lognormal and normal distributions were rejected. Although this result was produced from not five but 50 exercises, considering the characteristics of the distribution for leakages, geometric mean may be better than arithmetic mean for establishing the revised standard level in the future.
|
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This study was carried out to evaluate particulate filtering respirators, which are widely used in the Korean market, using an IL or TIL test that has been in use in Korea since July 1, 2000 and is similar to EN 13274-1. As expected, leakages of half masks were less than those of filtering facepiece masks. Because it was found that TILs of the 1st class filtering facepieces were generally much more than those of the 2nd class, and that none of TILs of the former were in compliance with the Korean standard, it seems that many individuals get confused when selecting a mask. Since leakages were significantly different among the facial dimension grid squares a test panel selected according to country or race may be needed to test IL or TIL. Because the filtering facepieces made by company A had lower leakages in the grid square, in which a large proportion of the population was represented, these masks are likely to be better masks for preventing leaks. It may be necessary to demonstrate fit, for example, fit testing, before IL or TIL testing. Otherwise, fit testing for individual wearer as required by the US regulation (CFR, 2003
) may be needed when selecting a respirator. A lognormal distribution was a better fit to the leakage data than the normal distribution in most masks. This result suggested that the geometric mean of the five exercise leakages may be better than the arithmetic mean to establish a standard individual subject mean. |
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
This work was supported by 2003 grant of the Korea Sanhak Foundation.
Received November 24, 2004; in final form June 1, 2005
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONACKNOWLEDGEMENTSREFERENCES |
|---|
CFR (Code of Federal Regulations) (2003) ‘Respiratory Protection’ Title 29, Part 1910. 134.
European Standards EN 136 (1998) Respiratory Protective Devices: Full-face masks; requirements, testing, marking. European Committee for Standardization.
European Standards EN 140 (1998) Respiratory Protective Devices: Half-masks, quarter-masks; requirements, testing, marking. European Committee for Standardization.
European Standards EN 13274-1 (2001) Respiratory protective devices—methods of test—Part I: Determination of inward leakage and total inward leakage, British Standard.
Han DH. (1999) Fit testing for respirators and development of fit test panels for Koreans. Kor Ind Hyg Assoc J; 9: 1–13(in Korean).
Korean ministry of labor (2000) Certification for personal protective equipments (particulate filtering masks). Notice 2000-15.
Myers WR. (2000) Respiratory protective equipment. In Harris RL, editor. Patty's industrial hygiene. 5th edn. New York: John Wiley & Sons, Inc. pp. 1497–8.
Vos GA, Christie JD. (1997) LogNorm 2TM—Statistics for exposure assessment. InTech Software Corp., Industrial Software Solution, Tulsa, OK 74136.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||