%4? -0* o"J1V. % * 4r ^* < * A v w ^Kil^ "*W~ -^^iv- "«-** \/ v^y . \*-> h v^v \-?SV* v^v 1 .0 >■ A o V ^ A^ ^ ,0^ ^ «?Y.s* A oK *«k ^ # **m£a°o X c* • jffev ***** .'i$S&> X/ 0* Vis. * q , ,^ t . — % % ,0* .o^° % *o, ,*> \r» ^ *%. ' .: v 77, '> '.- ^.v •"• • v^ . • • .; v ' ' ' y u ^;iA " ■" ^;. ri-?\* " ">° • • •-• - # - ^ V * 1 °° ^o< ^^ oV SAcr °0. ^^^ -0° r *^T* .A v5. s * A ^ ...» / ^vrr^V ^ % ^\/ %^^-V .. V-— o ^ .^>\ b5^ ^-o 5 ;«w 8?"V %^ ^^fe* P ° X4 ^ ^ ' w 6?> ^ .Oii«, aV *^> ^ . I ' * *- & U .** SjsStiii: \..

.-' ^ *+*• XC 8967 Bureau of Mines Information Circular/1984 Analysis of Data on Respirable Quartz Dust Samples Collected in Metal and Nonmetal Mines and Mills By W. F. Watts, Jr., D. R. Parker, R. L. Johnson, and K. L. Jensen UNITED STATES DEPARTMENT OF THE INTERIOR Information Circular 8967 Analysis of Data on Respirable Quartz Dust Samples Collected in Metal and Nonmetal Mines and Mills By W. F. Watts, Jr., D. R. Parker, R. L. Johnson, and K. L. Jensen UNITED STATES DEPARTMENT OF THE INTERIOR William P. Clark, Secretary BUREAU OF MINES Robert C. Horton, Director IA A rfl ,\0 % Library of Congress Cataloging in Publication Data: Analysis of data on respirable quartz dust samples collected in metal and nonmetal mines and mills. (Bureau of Mines information circu ar ; 8967) Bibliography: p. 21-22. Supt. of Docs, no .: I 28.27:8967. 1. Mine dusts. 2. Quartz dust. I. Watts, W F. (Winthrc P F.). II. United States. Bureau of Mines. HI. Series: Information circu- lar (United States. Bureau of Mines) ; 8967. J TNB9WJ4 [TN312] 622s [622'. 8] 83-600327 CONTENTS Page Abstract. 1 Introduction 2 MSHA coding system 2 Mine Inspection Data Analysis System (MIDAS) 3 File structure 3 Data editing 3 Sampling strategy 4 Results 6 Data distribution 6 Variables analysis 9 Year and mine type. 9 Commodity and occupation 9 Commodity, location, and time.. 14 Commodity and occupation group 16 Summary and conclusions 20 References 21 Appendix A. — MSHA sampling codes and descriptors and selected GM C/TLV data.... 23 Appendix B. — MSHA mine health ranking criteria 27 Appendix C . — Abbreviations used in this report 28 ILLUSTRATIONS 1. Log-normal probability plot of C/TLV's for RQ 8 2. Cumulative frequency plot of quartz concentrations grouped by location... 15 3. Cumulative frequency plot of quartz concentrations grouped by commodity.. 18 TABLES 1. Contaminants most frequently sampled in metal and nonmetal mines 6 2 . Number of mines and employment by mine type 7 3. Yearly statistics for RQ, respirable and total nuisance dust, and total silica dust 10 4. RQ exposures by mine type for 1974-81 and 1978-81 11 5. Ranking of 24 commodities by 1978-81 RQ GM C/TLV 12 6. Ranking of 28 occupations by 1978-81 RQ GM C/TLV 13 7. RQ GM C/TLV by commodity, location, and time period 14 8. Occupation groups with highest RQ exposures 17 9. Other occupation groups with high RQ exposures 19 A-l. MSHA contaminant codes for personal samples 23 A-2 . MSHA contaminant codes for area samples 24 A-3. RQ samples collected 1974-81 for each MSHA occupation code.. 24 A-4. RQ samples collected 1974-81 for each commodity 25 A-5. RQ samples from metal and nonmetal mines, by location 25 A-6. GM C/TLV's for 9 occupation groups 26 A-7. GM C/TLV's for 8 commodity groups 26 UNIT OF MEASURE ABBREVIATIONS USED IN THIS REPORT dBa decibel (A scale) ppm parts per million cm 3 cubic centimeter yg/m 3 microgram per cubic meter h hour Mm micrometer mg/m 3 milligram per cubic meter WL working level mppcf million particles per cubic foot yr year pet percent ANALYSIS OF DATA ON RESPIRABLE QUARTZ DUST SAMPLES COLLECTED IN METAL AND NONMETAL MINES AND MILLS By W. F. Watts, Jr., ' D. R. Parker, 2 R. L. Johnson, 3 and K. L. Jensen 4 ABSTRACT This report describes a statistical analysis of 41,502 respirable quartz dust samples collected from 1974 through 1981 in metal and non- metal mines and mills. The goal of this Bureau of Mines study was to identify commodities and occupations associated with high quartz dust exposure so that future enforcement and research activity can be focused on the high-exposure areas. For this analysis, the Bureau used its computerized Mine Inspection Data Analysis System (MIDAS), which contains data on nearly 350,000 air samples, including data for 61 contaminants in 45 commodities. The sam- ples analyzed were collected by inspectors of the Mine Safety and Health Administration (MSHA). The data were entered into the MIDAS computer and coded according to commodity, occupation, location, and other cate- gorical information. Occupations and commodities with high quartz dust exposure were identified by grouping the data by code categories. Workers at sandstone, clay and shale, and miscellaneous nonmetallic mineral mills had the highest quartz dust exposures. Within these mills, occupations with high exposures were baggers, general laborers, and persons involved in crushing, grinding, and sizing operations. Dust concentrations in these mills were not necessarily the highest, but be- cause the dust contained a high percentage of quartz, the health risk from dust exposure was the greatest. 1 1ndustrial hygienist, Twin Cities Research Center, Bureau of Mines, Twin Cities, MN. ^Program analyst, Mine Safety and Health Administration, U.S. Department of Labor, Arlington, VA. -'Operations research analyst, Division of Automatic Data Processing, Bureau of Mines, Denver, CO. ^Math aide, Twin Cities Research Center (now graduate student, Iowa State Univer- sity, Ames, IA) . INTRODUCTION The 1980 National Academy of Science report on respirable dust in mines (_O s concluded that there was a critical need to assess the magnitude of respirable dust exposures in noncoal mines. Expo- sure to respirable quartz dust (RQ) 6 has been linked to silicosis, a form of pul- monary fibrosis. Pulmonary fibrosis is disabling, progressive, and sometimes fatal. Silicosis tends to occur after years of exposure, but occasionally high exposures of only a year or more lead to acute silicosis (2). To assist research- ers and enforcement personnel in their efforts to protect workers from RQ haz- ards, the Bureau of Mines undertook a statistical analysis designed to identify mine and mill commodities and occupations associated with high quartz dust expo- sure. The results of this analysis are presented in this report. The Federal Government agency respon- sible for establishing and enforcing mine health and safety standards is MSHA. MSHA inspectors determine compliance with metal and nonmetal mine and mill Federal health standards through the collection and analysis of environmental air sam- ples. They collect many types of samples for a variety of contaminants in noncoal mines and mills , and these samples con- stitute a large body of historical indus- trial hygiene data. Since 1974 MSHA has collected nearly 350,000 samples in metal and nonmetal mines and mills, and records of these exposures are stored in MIDAS, a computerized information system operated by the Bureau (3). MIDAS is a data base and software system designed to analyze the industrial hygiene data collected by MSHA, and to associate these data with other mine information collected from other sources. MIDAS is accessible from remote terminals around the country via the Bureau's telecommunications network. Certain aspects of the system are dis- cussed in this report, and a description of the software used in MIDAS has been previously reported (4). MSHA CODING SYSTEM MSHA uses a coding system to describe the occupational environments where per- sonal samples (obtained from monitors worn by personnel) are collected. The codes are substituted for a compre- hensive written description of the work- place in order to facilitate automated data processing. They provide a descrip- tion of the industry (commodity produced at the mine site), operation (occupation of the person monitored for sampling), and a location (mine type) for personal sample collected. Personal samples are collected over an entire workshift, and ^Underlined numbers in parentheses re- fer to items in the list of references preceding the appendixes. "Except for unit of measure abbrevia- tions, all abbreviations are identified the first time they are used and are also listed in appendix C. (Unit of measure abbreviations are listed after the table of contents. ) for each sample an individual computer record is maintained. Other valuable in- formation in the computer record includes mine identification number, date, contam- inant measured, concentration, threshold limit value (TLV), and social security number of the miner. A typical set of codes might describe a respirable dust sample collected for a load-haul -dump diesel operator in a limestone quarry in Illinois. The approach used in this analysis was to calculate exposures for different pop- ulations of workers. The codes were used individually or in combinations to define population groups. Grouping similar codes together helped compensate for the recurring problem of inadequate sample size. A complete discussion of the groups used in the analysis is included in the "Results" section. Complete lists of the descriptors used for contaminant, commodity, location, and occupation are in appendix tables A-l through A-5. MINE INSPECTION DATA ANALYSIS SYSTEM (MIDAS) FILE STRUCTURE MIDAS includes 5 master files that con- tain over 400,000 records. The master files are used to create smaller files for analysis. The master files are known as the personal exposure file, the area sample file, the Minerals Availability System (MAS) data base, the 22-mine sur- vey file, and the mines file; and each has a different format. The RQ data are contained in the edited personal exposure file, which is described in the next section. The mines file is an index of all metal and nonmetal mine properties in the United States. Each property is listed with a unique mine identification number, its location, the property name, company name, approximate number of employees (if any), year-round or other status, and other data. Information from this file is stripped off and added to records of personal and area sam- ples, thus allowing records from similar commodities and mine types to be sorted and grouped together. This sorting and grouping is possible because every record in MIDAS, regardless of its format, has a mine identification number of seven digits. Thus all data are cross-referenced by mine identification numbers. Worldwide commodity information for about 200,000 mine properties is avail- able in the MAS data base. Data from this file contribute to the ability of MIDAS to group producers of similar com- modities together for analysis, particu- larly when mine properties that produce both primary and secondary products are involved. Often the first clue point- ing to a potentially hazardous exposure comes from knowledge of what product or products are mined. The classification system used in the MAS is more extensive than that available in the mines file. Among the sources of data gathered for the MAS are mine oper- ators, government agencies, and published literature. In 1976-77 MSHA conducted an environ- mental health survey of 22 underground metal and nonmetal mines with assistance from the National Institute of Occupa- tional Safety and Health (NIOSH) and the Bureau. The results of this survey are in the 22-mine survey file. (The Public Health Service and the Bureau had previ- ously surveyed some of the same mines in the late 1950' s.) About 17,000 personal and area samples have been collected at the 22 mines. These samples are the only industrial hygiene data in MIDAS that were not collected during mine inspections. The area record file contains 136,174 records of area samples. An area sample is a grab sample collected by a sampling device (detector tube, bistable, vacuum bottle, instantaneous monitor, glass impinger, or charcoal tube) located near the worker. The area samples are for toxic and asphyxiant gases, mists and vapors, and radon daughters. The RQ data are contained in the per- sonal exposure file. This file also con- tains data on occupational exposures to other airborne contaminants regulated by MSHA. The data were collected by moni- toring miners over the course of an entire workshift. The file, as received from MSHA, contained 206,888 records of personal exposure, of which 183,995 (89 pet) passed an editing process. Editing was necessary because of errors in the records. These errors included duplicate entries, decimal place errors, coding errors, erroneous TLV's, and misclassifi- cation of samples. DATA EDITING MSHA regulates health and safety condi- tions in mines under the authority of the Federal Mine Safety and Health Act of 1977 (5). The specific regulations are found in the Code of Federal Regulations, Title 30, (£), but for RQ and other air- borne contaminants, MSHA adopted the 1973 recommended TLV's of the American Conference of Governmental Industrial Hygienists (ACGIH) (7). The TLV for RQ is determined by collecting a respirable dust sample, analyzing for quartz con- tent, 7 and calculating the TLV using the formula 10 mg/m 3 percent respirable quartz + 2 when the quartz content (percent respir- able quartz) is XL pet. (The resultant TLV is expressed in milligrams per cubic meter. ) The TLV for respirable dust con- taining quartz is therefore inversely proportional to the quartz content of the sample. Thus, for a given exposure lev- el, the magnitude of toxicity is propor- tional to the quartz content (8). Respirable dust is capable of deep lung penetration and is generally defined as that fraction of an aerosol which in- cludes particles with an aerodynamic diameter of less than 5 um. However, the size classifiers used with personal sam- plers do not have a sharp size cutoff; they reject some dust particles with di- ameters less than 5 pm and allow small numbers of larger particles to pass (8). The factor 2 in the denominator of the TLV formula ensures that dust exposures will not be excessively high when the quartz content is less than 5 pet and effectively limits the dust concentration to 5 mg/m 3 when no quartz is identified in the sample. The MIDAS records of RQ exposure were edited using the TLV formula. By substi- tuting 1.0 and 100 for percent respirable quartz in the formula, it was determined that the TLV for RQ-bearing dusts must be between 0.10 and 3.33 mg/m 3 . Therefore, when the TLV recorded in a record fell within this range, the record was con- sidered valid; otherwise, the data were recoded based upon alternative rules (3). Editing the RQ records resulted in re- tention of 75 pet of the original rec- ords, the creation of 11,582 respirable nuisance dust records (21 pet of the original records), the recoding of 937 records (1.7 pet), and the rejection of 1,047 records (1.9 pet). The number of respirable nuisance dust records created during the editing process was large be- cause MSHA inspectors have only one code for respirable dust regardless of quartz content. MIDAS was used to create a code for respirable nuisance dust when the TLV exceeded 3.33 mg/m 3 , which indicated the sample contained less than 1.0 pet quartz. Other important features of the editing process were additions made to each rec- ord to simplify sorting and analysis. MIDAS calculated the concentration-to-TLV ratio (C/TLV) for every record of per- sonal exposure and added it to each rec- ord. The percentage of silica in every dust sample was also calculated and added to the records. The following informa- tion from the mines file was added to each record: commodity code, mine type, mine status, size group, and number of employees at the sampled mine. For com- puting geometric statistics (statistics based on a log transformation) , concen- trations reported as zero were considered equal to 0.001 mg/m 3 . SAMPLING STRATEGY MSHA inspectors collect samples to en- sure compliance with the 1973 ACGIH TLV's, which have been adopted as stan- dards by MSHA. The samples are not collected as they would be in a scientif- ically designed survey; rather, sampling 'Quartz content is determined by X-ray diffraction after the filter has been weighed. is judgmental. This creates a statisti- cal problem: These data may not be representative, so the classical sta- tistical assumptions of randomness, homo- scedasticity (homogeneous variance), and normal distribution may not apply. The factors discussed below (in order of rel- ative importance) are believed to influ- ence MSHA's sampling strategy. The 1977 Mine Safety and Health Act mandates that each underground mine be inspected four times per year and each surface mine or mill must be inspected two times per year. This requirement reflects the general opinion of Congress and the miners * unions that a Federal presence is important at every mine, and that underground mines are more hazardous than surface mines and mills. MSHA establishes mine health ranking criteria which are used as sampling guidelines (9). (The criteria are listed in part in appendix B.) In fiscal year 1982, these guidelines included three ranks: A, B, and C. Rank A lists mine categories known to have the greatest health risks. Mines with a history of noncompliance receive more frequent inspections. However, resampling of an area previously out of compliance is not done unless the mine operator has insti- tuted environmental controls or a per- sonal protection program. The mine oper- ator may request additional sampling to demonstrate the effectiveness of the con- trol measures, which may result in a cluster of samples being collected at certain properties. Inspectors are in- structed to sample employees having the greatest potential for exposure, and this determination depends upon the judgment of the inspector. In addition, the num- ber of employees at the mine site affects the proportion of employees sampled and to some extent the number of inspectors required per visit. An illustration of the effect commodity and employment have on sample distribution was provided in a previous report (3) where it was shown, for instance, that 0.13 samples were col- lected per employee from molybdenum prop- erties and 0.41 samples were collected per employee from limestone properties. From the same properties, 7 samples were collected per limestone property in con- trast to more than 50 samples per molyb- denum property. In this case there were many small limestone mines with few employees, but only a few large molyb- denum mines with many employees . Market forces affect the number of mines of any one type which are in operation, and their production and em- ployment levels. Mines inspected one year may not be inspected the next, be- cause of strikes, mine shutdowns or clo- sures which result from a drop in com- modity price or declining ore quality, or other similar circumstances. MSHA is affected by changes in budget appropriations and administrative poli- cies. These changes affect the number and location of MSHA district offices, the number of inspectors, and the number and types of inspections conducted each year. Since much of MSHA's sampling strategy is "worst case," it is nonrandom and may be biased. However, the degree and direction of all possible biases are ob- scure. Several factors suggest that MSHA's data are representative. Compan- ies are not given prior notice of inspec- tions, so operators cannot make special preparations. No operator can deny an MSHA inspector access to its property, thus every mine can be inspected. This differs from a research survey in which mines are asked to participate and know well in advance when samples will be col- lected. Although inspectors are in- structed to select workers with the greatest potential for overexposure, most of the samples reported had low dust con- centrations. In attempting to pick out "high-risk" workers, the inspectors can- not know if individual workers will ac- tually encounter high dust exposures. However, the results were typical of environmental sampling. Environmental samples are typically log-normally dis- tributed, meaning that many low concen- trations are reported. The authors therefore believe the MSHA data are typi- cal of the results that would be expected from a survey using sampling strategies designed to assure randomness. The re- sults suggest that the MSHA inspectors, using common sense and judgment, col- lected "representative" samples within each occupation, location, and commodity group. Finally, certain commodities, oc- cupations, and locations are sampled so frequently that MSHA may approach a cen- sus of those mining subpopulations. RESULTS DATA DISTRIBUTION The first part of the analysis was aimed at determining (1) the distribution of the number of samples by contaminant, commodity, location, and occupation and (2) the best measures of exposure. Table 1 shows that 12 contaminants accounted for 94 pet of all the samples collected, and that RQ samples accounted for 13 pet, ranking it second to noise (28 pet). However, if the samples for RQ, total nuisance dust, 8 respirable nuisance 8 Total nuisance dust has a TLV of 1 mg/m-*. This value is. used when samples collected for total dust contain less than 1 .0 pet quartz. Total dust samples are collected on filters without a cy- clone preclassifier to remove nonrespir- able particulate. ^Respirable nuisance dust has a TLV of 5.0 mg/m-*. This value is used when sam- ples collected for respirable dust contain less than 1.0 pet quartz. Res- pirable dust samples are collected on filters with a cyclone preseparator ahead of the filter to remove nonrespirable- size particulate. dust, 9 and total silica dust 10 are com- bined, they acount for about 25 pet of all the industrial hygiene samples col- lected. Tables A-l and A-2 in appendix A list all airborne contaminants for which personal or area samples were collected, and tables A-3 through A-5 show the num- ber of RQ samples distributed by com- modity, location, and occupation. It can be seen from these tables that the number of RQ samples collected varied consider- ably from code to code. From 1974 through 1981 MSHA collected 41,502 RQ samples at 4,815 mine and mill properties on 32,188 miners. From table 2 it can be calculated that these samples were collected at 30 pet of the properties on 12 pet of the workforce. 10 Total silica dust has a TLV deter- mined by the formula 30 mg/m-* percent quartz + 3 This value is used when a dust sample contains 1.0 pet or more quartz. TABLE 1. - Contaminants most frequently sampled in metal and nonmetal mines Contaminant Type of sample Number of samples Percentage of total samples Noise 1 RQ Methane Carbon monoxide Radon daughter Carbon dioxide Oxygen Total nuisance dust Total silica dust Respirable nuisance dust 2 Nitrogen dioxide Hydrogen sulfide Subtotal, Others , Total..., P P A A A A A P P P A A 89,830 41,566 28,217 28,170 27,311 24,292 15,663 13,415 12,515 11,582 9,261 2,450 304,272 17,897 322,169 27.9 12.9 8.7 8.7 8.5 7.5 4.9 4.2 3.9 3.6 2.9 .8 94.4 5.5 100.0 A Area. P Personal. RQ Respirable quartz dust. includes both noise dosimeter and sound level meter measurements. 2 Category created during editing process. TABLE 2. - Number of mines and employment by mine type Mine type Underground j Surface Mill Total NUMBER OF MINES 829 213 133 507 1,069 6,554 3,759 585 757 1,863 1,921 2,039 6,554 5,755 1,175 11,889 3,205 16,269 EMPLOYMENT Metal , Nonmetal. , Sand and gravel. Stone Total , 28,547 9,987 2,081 39,915 30,152 12,184 35,725 34,557 112,618 37,397 25,157 43,792 106,346 96,096 46,628 35,725 80,430 258,879 NOTE. — Based on final 1981 figures from MSHA, Health and Safety Analysis Center, which show 1,182 uniquely identified mills and 2,023 associated mills. These are crude estimates because the number of properties and employment fluc- tuate from year to year, but it is clear that a large percentage of the mines, mills, and workforce was sampled. Over- all, the average number of samples per mine sampled was 8.6, and the average number of samples per employee sampled was 1.3. Table 2 provides information on employ- ment and mine type which is essential for understanding the numerical distribution of the RQ samples. A majority of the mines were either sand and gravel pits or stone quarries with relatively few em- ployees, while the largest segment of the workforce was employed at very large metal mines and mills. The average num- ber of employees at an underground metal mine is 50, whereas the average number of people employed at a sand and gravel pit is 5. Nearly every worker can be sampled at a sand and gravel pit during the course of an inspection; but because of the large number of pits, samples are not collected at every site visit. In con- trast, the relatively few large under- ground metal mines are visited fre- quently, but only a tiny percentage of the workforce is sampled each time. Thus, the number of RQ samples collected was affected by the number of inspections required by the 1977 Mine Health and Safety Act, the number and types of mines and mine employment, and the degree of hazard perceived by individual inspectors. In an array of data such as the RQ data, frequency distributions and proba- bility ploys are useful in selecting the appropriate geometric or arithmetic sta- tistics to estimate central tendency and dispersion. A previous report (3) used these techniques to show that the RQ con- centrations and C/TLV's reported from 1974 through 1980 approximated a log- normal distribution. This finding sug- gested that the geometric mean (GM) and geometric standard deviation (GSD) were better indicators of central tendency and dispersion than the corresponding arith- metic values. Figure 1 is a log-normal probability plot of the RQ C/TLV's from 1974 through 1981. The figure suggests that the distribution of C/TLV's approxi- mates a log-normal distribution, although extreme values are present at both tails. (A description of the computer program used to generate this plot can be found in reference 10.) A straight-line normal probability plot indicates that a population is normally distributed. The abscissa of a point is an observed value, and the ordinate is its expected standard normal value, which is computed by ordering the observed val- ues, assigning a cumulative probability _l < > O O LU h- O LU Q_ X LU 4.5 3.6 2.7 .8 -.9 -1.8 -2.7 -3.6 4.5 -2. ■A A A A A rA A A A A A A AA AAAA A AAAA AAAA AAAA AAAA AAA AAA AAAA AA AAA AAA AAA AAA AA AAA AA AA AAA AA AA AAA AA AA AA AA AA AA AA AA A A A AA A A A A A I -1.40 -700 .700 1.40 LOG C/TLV 2.10 2.80 3.50 FIGURE 1. - Log-normal probability plot of C/TLV's for RQ. value to each based upon its rank order within the sample, and then determining the standard normal value corresponding to that cumulative probability value. A standard normal value is a value on the normal distribution (or log-normal dis- tribution in this case) with a mean equal to zero and a standard deviation equal to one. The absolute value of the ex- pected standard normal value of a par- ticular observed value can be understood as the number of standard deviations away from the mean that an observed value of that rank would be expected to lie. Thus, an observed value is related to a cumulative probability value, which is then related to an expected standard nor- mal value. If the plot representing these relationships is close to a straight line, the population may be nor- mally distributed (11). For log-normally distributed data, the GM C/TLV is an excellent index of exposure because it relates the concen- tration to the TLV, which is determined by the quartz content of the sample. This allows an immediate assessment of exposure. A C/TLV greater than 1 indi- cates that an overexposure has occurred. Other measures of exposure are the geo- metric mean concentration (GM CONC), the median concentration, and the percentage of samples greater than the TLV. The concentration of quartz dust can also be calculated (percent quartz x concentra- tion) and compared to the NIOSH recom- mended permissible exposure limit (PEL) of 50 ug/m 3 crystalline silica ( 13 ) or the maximum allowable level of 100 ug/m 3 under the current TLV (12). The GM C/TLV, the percentage of samples greater than the TLV, and quartz concentration are used in the next section to assess exposure. VARIABLES ANALYSIS Year and Mine Type The second part of the RQ analysis was based on partitioning of the RQ data by year, commodity, occupation, location, and combinations of variables. Such an analysis was possible because each record of exposure includes supplementary coded information. However, as shown in the previous section, the frequency with which each code was used varied consider- ably, and this was the limiting factor in grouping the data. Table 3 shows the yearly statistics for RQ; and for comparison, similar statis- tics are given for respirable nuisance, total nuisance, and total silica dust. It appears that the RQ dust levels were on a downward trend until 1980 and 1981, when slight increases occurred; but in those same years there was a sharp de- crease in the number of samples col- lected, which suggests a shift in dust sampling strategy. Each year, 15 to 20 pet of the RQ dust samples equaled or exceeded the TLV, which demonstrates the continuing problem mines and mills have with RQ. The average TLV and the average quartz content can be calculated from the GM RQ CONC and the GM C/TLV shown in table 3. These two averages provide a crude esti- mate of the permissible level of dust in metal and nonmetal mine environments under existing standards. For all years the average TLV was 1.16 mg/m 3 and the average quartz content was 6.6 pet. In comparison, the GM TLV for total silica dust was 3.91 mg/m 3 and the average quartz content was 4.7 pet. Table 4 shows RQ exposure by mine type. The data were grouped according to the codes shown in tables A-4 and A-5. Table 4 also shows a third variable, time; the RQ data for all years (1974-81) are shown on the left side of the table, and data for only the last 4 yr (1978-81) are shown on the right. Most of the samples were collected in surface stone mines, mills, and quarries, which would be ex- pected given the large number of these facilities. The highest concentrations of dust were found in samples from under- ground stone mines, but the highest GM C/TLV *s were found in samples from non- metal mills. Commodity and Occupation Many of the samples were collected at clay and shale, barite, and miscellaneous nonmetal mills, and a high proportion of the samples from these mills exceeded the TLV (table 5). In this analysis, the category "miscellaneous nonmetals" was used to clarify MSHA's commodity code number 59, which is termed "other non- metal. " This commodity group includes three or four producers of tripoli. Tripoli is typically ground to a fine particle size and contains a high per- centage of quartz; as a result, the TLV is low. Samples collected at these two mills severly skew the data for the en- tire commodity group. Higher concentra- tions of dust are tolerable at most other facilities because the dust contains a lower percentage of quartz, which in- creases the TLV. 10 TABLE 3. - Yearly statistics for RQ, respirable and total nuisance dust, and total silica dust Year Total samples > TLV, pet Concentration, mg/m c GM Median GSD GM C/TLV RQ 1974, 1975, 1976. 1977, 1978. 1979. 1980. 1981. Total or average, 284 2,652 6,232 7,579 7,854 7,806 4,750 4,345 41,502 46.13 28.39 25.69 19.44 16.16 14.16 15.24 17.35 18.82 .58 .49 .51 .44 .40 .38 .40 .43 .43 0.54 .47 .48 .43 .38 .35 .37 .40 .40 2.62 2.95 3.12 3.02 3.04 3.07 2.71 2.61 2.98 0.86 .53 .45 .37 .34 .32 .33 .36 .37 RESPIRABLE NUISANCE DUST 1974 8 479 1,309 1,840 2,312 1,803 1,624 2,149 8.35 3.82 2.77 3.16 2.50 .99 1.26 0.18 .52 .36 .29 .23 .23 .14 .11 0.12 .56 .39 .45 .36 .27 .13 .12 2.87 5.70 5.65 8.12 9.76 7.43 6.49 8.06 0.04 1975 .10 1976 .07 1977 .06 1978 .05 1979 .05 1980 .03 .02 11,524 2.62 .21 .27 7.99 .04 TOTAL 1 NUISANCE DUST 1974 67 436 1,194 1,600 2,061 3,213 3,035 1,761 25.37 40.37 34.17 34.00 25.52 20.67 14.53 16.47 3.95 6.14 4.59 4.65 3.09 2.22 1.59 1.86 4.19 7.47 5.10 5.62 4.17 3.00 2.16 2.70 3.72 6.36 5.97 6.08 7.40 7.69 7.19 8.24 0.04 1975 .61 1976 .46 1977 .46 1978 .31 1979 .22 1980 .16 .19 13,367 22.94 2.55 3.43 7.49 .26 TOTAL SILICA DUST 1974 34 196 781 1,622 1,693 3,823 3,291 1,071 41.18 66.84 42.25 39.27 22.15 19.88 21.67 23.06 1.92 6.50 3.29 2.78 1.43 1.41 1.66 1.62 1.65 8.28 3.13 2.71 1.46 1.30 1.50 1.51 4.29 4.20 3.68 4.80 4.90 3.69 3.35 4.29 0.69 1975 1.59 1976 .76 1977 .69 1978 .35 1979 .37 1980 .43 .42 12,511 25.63 1.76 1.61 4.10 .45 GM GM C/TLV GSD RQ TLV Geometric mean. Geometric mean concentration-to-TLV ratio, Geometric standard deviation. Respirable quartz dust. Threshold limit value. 11 TABLE 4. - RQ exposures by mine type for 1974-1981 and 1978-1981 1 Industry group 1974-81 1978-81 UG mine Surface mine Mill Total UG mine Surface mine Mill Total Stone: N GM CONC mg/m 3 .. GM C/TLV Metal: N GM CONC mg/m 3 .. GM C/TLV Nonmetal: N GM CONC mg/m 3 .. GM C/TLV Sand and gravel: N GM CONC mg/m 3 .. GM C/TLV Total: N GM CONC mg/m 3 .. GM C/TLV 797 0.79 0.40 3,301 0.56 0.45 631 0.76 0.57 NAp NAp 4,730 0.62 0.45 12,745 0.36 0.28 1,474 0.27 0.26 1,471 0.39 0.34 5,180 0.27 0.29 20,869 0.33 0.29 7,071 0.55 0.45 2,942 0.43 0.40 3,314 0.77 0.70 1,545 0.37 0.59 14,872 0.54 0.50 20,613 0.43 0.34 7,717 0.44 0.39 5,416 0.64 0.56 6,725 0.29 0.35 40,471 0.43 0.37 602 0.73 0.36 1,516 0.51 0.40 299 0.67 0.47 NAp NAp 2,418 0.58 0.40 8,505 0.36 0.27 1,007 0.26 0.24 729 0.34 0.28 3,701 0.27 0.29 13,941 0.32 0.28 3,525 0.51 0.45 1,611 0.39 0.35 1,718 0.72 0.64 942 0.39 0.59 7,796 0.50 0.46 12,612 0.41 0.31 4,134 0.39 0.34 2,746 0.56 0.50 4,643 0.29 0.34 24,155 0.40 0.34 GM CONC Geometric mean concentration. GM C/TLV Geometric mean concent rat ion-to-TLV ratio. N Sample size (number of samples). NAp Not applicable. UG Underground. Excluded are 1,031 samples which could not be class the mines from which they were taken are permanently leted from the master mine file. ified by industry closed and were group because therefore de- The 24 commodities listed in table 5 account for 97 pet of all the samples. The commodities are ordered according to their 1978-81 GM C/TLV s, and the per- centage of samples that exceeded the TLV is shown. (Miscellaneous nonmetals en- compasses nonmetals not elsewhere classi- fied, such as aplite, vermiculite, ky- anite, and tripoli; and miscellaneous stone includes a variety of products such as basalt, diabase, gabbro, and others.) Each mine was assigned a commodity code, but some mines produce more than one product, so the classification system was not perfect. The periods 1974-77 and 1978-81 were chosen because the 1977 Mine Health and Safety Act resulted in changes in the MSHA inspection program, which in turn affected the sampling strategy. With the exception of traprock, each GM C/TLV was lower in 1978-81 than it was in the preceding 4 yr, which suggests there was improvement in environmental condi- tions. The GM C/TLV ranking of the first seven commodities in table 5 listed would be the same for 1974-77 as for 1978-81, except that feldspar would be rated sec- ond rather than twentieth. 12 TABLE 5. - Ranking of 24 commodities 1 by 1978-81 RQ GM C/TLV (In order of decreasing GM C/TLV for 1978-81) Commodity 1974-77 N > TLV, pet GM C/TLV 1978-81 N > TLV, pet GM C/TLV Barite Misc. nonmetals Molybdenum Clay and shale Sands tone Misc. stone Gold and silver Granite Copper Talc Lead and zinc Slate Traprock Fluorspar Sand and gravel Lime Cement Phosphate. Uranium Feldspar Mica Iron Limestone Gypsum Total Total metal and nonmetal. 84 242 348 1,723 1,363 161 259 1,236 1,214 54 566 137 307 56 2,082 193 1,264 341 487 65 18 650 3,308 34 16,212 16,747 61.9 43.4 34.2 34.9 37.4 38.5 28.2 22.9 21.7 29.6 18.2 18.2 11.1 30.4 21.8 20.2 18.3 19.3 18.5 40.0 16.7 20.6 14.4 11.1 ND 23.2 1.44 .83 .71 .65 .67 .63 .60 .47 .44 .52 .42 .51 .28 .59 .35 .31 .36 .46 .41 .88 .56 .39 .29 .28 ND .43 56 499 178 1,490 1,767 307 615 2,273 785 88 339 129 404 66 4,643 200 1,083 121 610 128 116 1,527 6,444 134 24,002 24,755 26.8 40.9 30.9 28.3 29.9 25.7 16.1 19.2 18.2 13.6 11.8 7.0 17.3 12.1 15.1 14.5 10.8 9.1 6.2 12.5 6.0 10.6 6.6 2.2 ND 15.1 0.76 .68 .65 .58 .55 .53 .43 .42 .41 .40 .40 .40 .38 .37 .34 .30 .30 .30 .28 .27 .27 .26 .23 .20 ND .34 GM C/TLV Geometric mean concent rat ion-to-TLV ratio. N Sample size (number of samples). ND Not determined. TLV Threshold limit value. 1 Not included are marble, antimony, bauxite, beryl, chromite, manganese, tungsten, mercury, other metals, asbestos, boron, magnesite, potash, pumi sodium compounds, sulfur, gilsonite, and oil shale. titanium, ce, salt, Table 5 also shows the sample size (N) for each commodity group. For limestone the total N (1974-81) was large, 8,752 samples; but for fluorspar total N was only 122 samples. The uneven distribu- tion of samples became a problem when further subgrouping by occupation, loca- tion, and time was attempted, because N sometimes became too small — or it became zero. This was especially true of the minor commodities. The problem was re- duced somewhat by grouping similar com- modities together when there was an insufficient number of samples. (This is discussed in more detail later. ) Table 6 lists 28 occupations which account for 97 percent of all the RQ sam- ples. The occupations are ranked accord- ing to their 1978-81 GM C/TLV's; and the sample size is included; and the per- centage of samples that exceeded the TLV is shown. The 1978-81 GM C/TLV for bag- gers, the first-ranked occupation, was 35 pet higher than that of the second-ranked occupation, slushing; and nearly 40 pet of the bagger samples exceeded the TLV. A closer examination of the data for bag- gers showed that the product most often bagged is some type of industrial sand such as glass sand, frac sand, or silica 13 TABLE 6. - Ranking of 28 occupations 1 by 1978-81 RQ GM C/TLV (In order of decreasing GM C/TLV for 1978-81) Occupation 1974-77 N > TLV, pet GM C/TLV 1978-81 N > TLV, pet GM C/TLV Bagger Slushing Supply Blasting Grinding Welding Percussive drilling General labor. Drying, filtering, and thickening Crushing Rotary drilling Administration. Sizing General shop Complete cycle Roasting and retoring. . . . Concentrating Concrete operations Mechanic Load, haul, dump electric Rock sawing Bulldozing Pelletizing Hoisting Mining machine Load, haul, dump gas Technical services Load, haul, dump diesel. . Total Total metal and nonmetal. 786 103 152 57 762 89 808 2,101 561 2,215 374 232 640 148 613 136 189 150 651 400 142 442 98 43 22 194 134 4,116 16,358 16,747 46.2 55.3 25.7 24.6 37.0 27.0 25.9 31.3 30.5 30.1 28.6 18.5 29.1 16.9 14.7 22.8 23.8 27.3 15.0 15.2 15.5 16.5 15.3 7.0 18.2 20.1 14.9 9.9 ND 23.2 0.90 1.09 .47 .46 .71 .51 .48 .58 .54 .57 .51 .32 .48 .35 .38 .45 .45 .56 .34 .34 .43 .35 .32 .21 .46 .35 .29 .25 ND .43 1,232 111 198 56 622 61 1,047 2,450 753 4,402 552 133 936 196 479 173 268 67 559 574 117 519 204 89 39 306 112 7,999 24,254 24,755 39.8 27.0 27.3 17.9 22.7 18.0 22.2 21.9 21.0 18.5 19.4 15.8 18.1 16.8 11.5 13.9 15.7 11.9 10.7 10.6 6.8 10.2 13.7 10.1 5.1 7.5 5.4 6.2 ND 15.1 0.83 .54 .49 .48 .47 .47 .46 .44 .44 .40 .39 .39 .37 .37 .37 .36 .31 .31 .30 .30 .29 .28 .26 .26 .25 .24 .24 .23 ND .34 GM C/TLV Geometric mean concentration-to-TLV ratio. N Sample size (number of samples). ND Not determined. TLV Threshold limit value. *Not included are machine mucking, hand mucking, timbering, rock ing, diamond drilling, load, haul, dump compressed air, track crew, operations, dredging, and jet piercing. bolting, slurry backfill- , chemical flour, which all have high quartz con- tents. The high quartz content results in a low TLV, meaning that dust controls must be very effective in order for the baggers to avoid overexposure. Of the 28 occupations listed in table 6, 23 showed a reduction in the GM C/TLV from the first to the second time period, although 5 showed slight GM C/TLV increases (supply, blasting, admini- stration, general shop, and hoisting). Blasting and administration showed de- creases in the percentage of samples exceeding the TLV, and the other three either showed slight increases or re- mained the same. There were extreme differences in N among the occupational categories. The load, haul, dump diesel operator code was used 12,115 times, but the code for min- ing machine was used only 61 times. The explanation for the extreme difference 14 is that most metal diesel equipment, continuous miners, of rock breaking mines is drilling for a few industr trona, potash, and ing methods used adapted (1). and nonmetal mines use but relatively few use The principal method in underground noncoal and blasting, except ial minerals such as gypsum, for which min- in coal mines can be Commodity, Location, and Time It was possible to further subdivide the RQ data using location (i.e., under- ground mines, surface mines, or mills) as a variable, but in doing this the problem of inadequate sample size became very evident in specific subsets. However, despite this limitation, problem areas can be identified with such an analysis. Table 7 utilizes the GM C/TLV to show a commodity, location, and time analysis for the same commodity groups as are listed in table 5. Each of the mining locations includes commodities that pose potential problems . Underground molyb- denum and clay and shale mines have GM C/TLV s that were considerably higher than the mean for all underground loca- tions. The same was true for surface TABLE 7. - RQ GM C/TLV by commodity, location, and time period Commodity 1 Underground Surface Mill 1974-77 1978-81 1974-77 1978-81 1974-77 1978-81 Barite 1.74 + 0.51 + .68 * .83 * ND ND .55 * ND .45 * ND .07 + .46 * ND .78 + ND .67 + ND .30 + .47 * ND ND ND .50 * .27 + ND 0.51 ++ .63 * .64 * ND ND .40 * ND .41 * ND ND .42 * ND .38 + ND .38 -H- ND .31 + .32 * ND ND .58 + .36 * .19 + ND 0.55 + ND .35 * .35 + .27 ++ .58 + .45 * .31 * .22 ++ .42 + ND .24 * ND .30 ** .44 * .33 * .51 * .30 + .71 + ND .20 * .24 ** .39 + ND 0.24 * ND .36 * .37 ++ .49 * .30 + .41 ** .36 * .22 -H- .37 + ND .32 * ND .30 ** .44 * .23 * .28 + .19 ++ .18 + .22 -H- .22 * .21 ** .20 ++ 1.32 + 0.96 * 1.27 + .78 ** .64 ++ 1.07 ++ .99 + .53 * .54 * .38 * 1.29 + .31 * .35 * .21 + .59 * .83 * .38 * .40 ++ .26 -H- 1.00 + .56 + .45 * .39 ** .18 + 0.90 + 1.09 * .94 + .68 ** .45 + .62 * .61 * .50 * .45 * .37 * Talc .41 ++ .37 ++ .53 * .34 + .59 * .80 * .35 * .31 + .26 * .33 ++ Mica .32 ++ .30 * .28 ** .23 + .52 .40 .31 .28 .55 .46 ND + * ** Not determined 10-49 samples. 50-99 samples. 100-999 sample: More than 999 because there were less than 10 samples samples. Listed in same order as in table 5 to allow comparison. 15 miscellaneous stone, sandstone, and gran- ite mines, although the magnitude of the GM C/TLV's was lower. Of the three loca- tion categories, mills had the highest GM C/TLV's. Within the mill category, mis- cellaneous nonmetal mills had the highest GM C/TLV's, and barite, molybdenum, and sandstone mills followed close behind. One problem with this analysis is that samples taken at some mills were grouped together with the surface and reported with surface mine location codes. Sam- ples collected indicated that exposures in mills were generally higher than those in surface mines. Therefore, the effect of grouping some mill samples with the surface mine samples would be to raise the surface mine exposures and decrease the difference between mill and surface mine exposures. Unfortunately, there was no way to separate these data. The data shown in table 7 are useful for identifying locations with higher RQ exposures. Listed below are the eight locations with the highest GM C/TLV's for 1978-1981. At least 50 RQ samples were collected at each of these locations dur- ing this time period. The first value listed for each location is the GM C/TLV, and the percentage of samples greater than or equal to the TLV is shown in parentheses. Misc. nonmetal milling 1.09 (56.0) Sandstone milling 80 (41.7) Clay and shale milling 68 (33.7) Clay and shale underground mining 64 (27.3) Molybdenum underground mining .63 (31.5) Misc. stone milling 62 (29.1) Gold and silver milling 61 (30.0) Sand and gravel milling 59 (30.1) The bulk of the samples collected from miscellaneous nonmetal mills came from two tripoli mills with histories of dust problems. These and similar mills have received special attention from MSHA and NIOSH because they produce a finely ground product with a high quartz con- tent, and because acute cases of sil- icosis in young workers have been identified at these mills (14). An occu- pational breakdown of data collected from these mills showed that general laborers, baggers, and workers involved with crush- ing and other mill activities encountered exposures for which the GM C/TLV exceeded 1.00. The GM RQ CONC for these facili- ties was 0.65 mg/m 3 , and the average quartz content of each sample was 60 pet, resulting in an average TLV of 0.17 mg/ m 3 . Dust control at these plants would have to be excellent in order for them to meet the RQ TLV. Workers in metal and nonmetal mills are generally exposed to more quartz dust than their counterparts in mining activ- ities. This is illustrated in figure 2, 1,000 100 — 10 1 1 1 1 1 1 1 I i - < n ' w - '11 - fit 1 4 / / / / / // / 1 / - : / A V <' /// ~ '' /// '' // * / / _ * / / * / / * * / ' -V / / y / / ." / / ." / ~ / " / - //// - // - - .//// - KEY Other surface facilities • Underground mining - i i i i i i i i 1 10 20 30 40 50 60 70 CUMULATIVE, pet 80 90 100 FIGURE 2. - Cumulative frequency plot of quartz concentrations grouped by location. 16 a cumulative frequency plot of quartz concentrations grouped by location for 1978-81. Quartz concentrations are derived by multiplying the quartz content by concentration. Samples from mills contained more quartz than samples from other locations. A sample containing 100 ug/m 3 or more of quartz always exceeds the RQ TLV when the respirable dust con- centration is less than 10 mg/m 3 , and 26 pet (838) of the samples from mills exceeded this level. This compares to 8 pet of the underground and surface mines samples and 14 pet of the other surface facilities samples. Commodity and Occupation Group As previously mentioned, the sample size N becomes smaller every time the RQ data are partitioned into smaller sub- groups. One way to deal with this prob- lem is to group similar codes together, thus making fewer subgroups. An example of this has already been shown in table 4, in which all the commodities were grouped into four categories and all the location codes were grouped into three categories. However, table 4 does not show occupation, which is also a variable of interest. In order to group the codes in a meaningful manner, several param- eters must be considered. These include N for each code, the similarity of codes, and the distributions of dust concentra- tion and C/TLV. For some commodity and occupation codes, N is very large; for example, limestone and load, haul, dump diesel can stand alone as individual groups. Using the parameters listed above, the codes for location, commodity, and occu- pation were grouped for further analysis. The location codes were logically grouped by similar mine activities defined as (1) underground or surface mining and (2) surface mineral processing including milling. Since 24 commodities and 28 occupations accounted for 97 pet of the 1974-81 data, only these commodities and occupations were included in the consoli- dated commodity and occupation groups. This eliminated data from 20 commodities and 12 occupations, which are listed below tables 5 and 6. The remaining codes were grouped into 8 commodity groups and 9 occupation groups. Both sets of groups are listed below. The commodity groups included — 1. Limestone. 2. Sandstone. 3. Stone; cement, granite, lime, slate, traprock, and miscellaneous stone. 4. Metal; copper, gold and silver, iron, lead and zinc, molybdenum, and uranium. 5. Clay and shale. 6. Miscellaneous nonmetals. 7. Nonmetals; barite, feldspar, fluor- spar, gypsum, mica, phosphate rock, and talc. 8. Sand and gravel. The occupation groups included — 1. Bagger. 2. Crushing. 3. Load, haul, dump; electric, diesel, gasoline, and bulldozer. 4. Drilling; rotary and percussive. 5. Grinding and sizing. 6. Finishing; roasting and retorting; drying, filtering, and thickening; con- centrating; and pelletizing. 7. General labor; including complete cycle and general shop work. 8. Welder and mechanic. 9. Other; slushing, blasting, rock sawing, mining machine operator, concrete operations, hoisting, supply handling, technical services, and administration. 17 The GM C/TLV s and percentages of samples greater than the TLV for these commodity and occupation groups are shown in tables A-6 and A-7. The data were further broken down into the two time periods used in the previous analyses, the periods before and after passage of the Mine Health and Safety Act of 1977 (1974-77 and 1978-81). Combining the 8 commodity groups, 9 occupation groups, 3 location groups, and 2 time periods creates a 8x9x3 x2 matrix with 432 cells for which the GM C/TLV, GM CONC, GSD, and other statistics can be calculated by a computer program in MIDAS. The table produced by this pro- gram is too extensive to be reproduced in this report, but the highlights are shown in table 8. The computer program that calculated the figures in table 8 is also capable of calculating the same values for different commodity, occupation, location, and time groups (and in fact this was done before the final grouping was adopted). There are 28 occupation, commodity, and location combinations listed in table 8. For each, N was at TABLE 8. - Occupation groups with highest RQ exposures 1 (In order of decreasing GM C/TLV for 1978-81) Occupation group Commodity group Location group N > TLV, pet GM CONC, mg/m 3 GM C/TLV General labor Bagger Do Do General labor Crushing Grinding, sizing. . . Bagger Grinding, sizing. . . Bagger Other Crushing Do Drilling Grinding, sizing. . . Finishing General labor Finishing Grinding, sizing. . . Drilling Bagger Load, haul, dump... Drilling Crushing Finishing Grinding, sizing. . . Crushing Finishing Misc. nonmetal... Sandstone Misc. nonmetal... Sandstone ...do .do, • • •do •••••••••••• Clay and shale... Sand and gravel.. ...do Clay and shale. . . ...do Metal Sandstone Clay and shale... Sand and gravel.. Clay and shale... ...do Sandstone Limestone Stone Clay and shale. . . Stone Metal Sandstone Stone ...do Sand and gravel.. M S M M M M M M M M M M UG S M M M M S UG M UG S M M M M S 64 60 100 191 88 135 85 300 98 196 81 62 53 57 143 183 114 150 93 112 93 54 620 409 137 134 350 100 75.0 73.3 62.0 52.9 53.4 40.7 45.9 41.7 42.9 38.3 39.5 37.1 34.0 36.8 32.9 32.8 29.8 29.3 35.5 26.8 23.7 16.7 29.8 25.4 24.8 21.6 21.4 17.0 0.69 .40 .82 .44 .52 .42 .37 1.16 .48 .45 .91 .72 .69 .44 .83 .43 .75 .71 .26 1.43 .68 .79 .56 .45 .45 .66 .60 .32 1.89 1.58 1.54 1.25 1.05 .90 .90 .85 .84 .84 .77 .75 .73 .73 .72 .66 .64 .62 .62 .60 .58 .56 .56 .56 .53 .52 .52 .52 GM CONC Geometric mean concentration. GM C/TLV Geometric mean concentration-to-TLV ratio. M Mill. N Sample size (number of samples). S Surface. TLV Threshold limit value. UG Underground. Based on the 1978-81 data with a minimum N of 50 and a minimum GM C/TLV of 0.50. 18 least 50 and the GM C/TLV was at least 0.50; in all but two groups, more than 20 pet of the samples exceeded the TLV. Occupations in sandstone, clay and shale, and miscellaneous nonmetal mines and mills account for 17 of the occupa- tion, commodity, and location groups listed in table 8. Eight of the ten highest ranked groups fell into one of these categories. The four sand and gravel combinations in table 8 (grinding, sizing — mill, bagger — mill, finishing — mill, and finishing — surface) primarily represent data from industrial sand pro- cessing plants. A mine-by-mine investi- gation of the ten highest ranked groups revealed that the results for these occu- pational groups were greatly influenced by samples collected at industrial sand facilities. The industrial sand pro- ducers, including the producers of silica flour, are classified by MSHA in the sandstone, sand and gravel, and mis- cellaneous nonmetal categories. The extremely high quartz content, small par- ticle size, and abrasive nature of sand make sand dust difficult to control and result in high GM C/TLV s at these estab- lishments. An anomaly in the table is that certain finishing occupations occur twice, once with a mill location and once with a surface mining location. One rea- son for this is that some mines do not have a separate identifiable mill, and thus it was necessary to use the surface mine location code; however, in some cases this anomaly resulted from failure to use the correct code. The occupational groups most frequently listed in table 8 are bagger, crushing, grinding, and sizing. They account for 16 of the 28 high exposure groups. The data shown in table 8 indicate that work- ers in these areas are frequently exposed to levels exceeding the TLV. The occupa- tion bagger occurs 6 times in table 8, accounting for 940 out of a total of 1,232 (or 76 pet of the total) samples, for the period 1978-81. Baggers had the highest risk of overexposure of any occupation; 40 pet of all bagger samples exceeded the TLV. Three underground mine groups appear in table 8: Drilling — limestone, load- haul-dump — clay and shale, and crushing — metal. Drilling at surface sandstone and other stone quarries also had an above- average risk of excessive exposure. Another way of estimating the risk from exposure is to calculate the quartz con- tent in each sample and plot the fre- quency distributions, as was done in fig- ure 2 for location groups. Figure 3 shows the 1978-81 frequency distributions for seven of the eight commodity groups shown in table 8. The other group, stone, is not included because the plotted line for stone closely resembles the line for sand and gravel. Quartz 1 w 90 1 III 7 Ft 1 1 1 1 ^^J^^ r- / / E/ ' 80 r c/ / / / / A / 70 - ^_ o Q. uj" 60 _ > 1- < _l Z> 2 50 - Z> o KEY 40 1 // / // A 1/ / / B /// / / c Misc nonmetal Sandstone Clay and shale 30 r// / D // / E V / F // G Sand and gravel Metal Nonmetal Limestone 20 l l 1 1 I Mil 1 1 10 50 100 400 QUARTZ, /xg/m 3 FIGURE 3. - Cumulative frequency plot of quartz concentrations grouped by commodity. content is important in determining TLV's and in assessing the impact of changes in regulations. NIOSH has proposed an occu- pational standard of 50 pg/m 3 of crys- talline silica as a time-^weighted average to protect workers for up to a 10-h work day and 40-h work week over a working lifetime (13). This contrasts with the 100-pg/m 3 standard which is used by the Occupational Safety and Health Admini- stration (OSHA) and is now being proposed by MSHA. Figure 3 clearly shows the effect these two levels would have had on the rate of past overexposure. If the NIOSH proposed standard had been in effect, many more samples would have been considered to represent overexposures. Miscellaneous nonmetals, sandstone, and clay and shale are commodities with high quartz contents , and therefore they would have had a greater percentage of samples exceeding a lower standard than would limestone producers. 19 High concentrations of respirable dust with less than 1 pet quartz content could constitute overexposures if the standard were changed to 100 pg/m 3 of quartz. Table 9 lists 15 occupation groups that are not listed in table 8 but have GM dust concentrations greater than 0.50 mg/m 3 . Again, mill locations predom- inate, but table 9 also includes four underground occupational groups for metal and limestone mines. Other commodities with high dust levels were identified from total nuisance dust data not shown; they include underground salt, trona, and potash mines where quartz is not a con- stituent of the host rock. Many of the workers sampled by MSHA inspectors were wearing respirators at the time the sample was collected, so the actual exposure of the individual was somewhat lower than the reported expo- sure. In 1981, 4,345 RQ samples were TABLE 9. - Other occupation groups with high RQ exposures 1 (In order of decreasing GM CONC) Occupation group Commodity group Location N > TLV, GM GM CONC, group pet C/TLV mg/m 3 M 98 15.3 0.43 0.77 M 342 21.9 .43 .68 M 182 11.5 .32 .68 M 120 20.0 .41 .67 UG 231 8.2 .31 .67 M 70 7.1 .35 .58 M 532 10.3 .30 .58 UG 173 9.2 .37 .56 M 162 11.1 .32 .56 UG 204 18.6 .46 .55 UG 466 11.4 .40 .54 M 50 14.0 .33 .50 M 122 22.1 .45 .52 S 510 22.6 .47 .50 S 920 9.0 .26 .50 Bagger . General labor..., Grinding, sizing. Finishing Load, haul, dump. Welder, mechanic. Crushing Drilling General labor..., Other , Load , haul , dump . Finishing , Load, haul, dump, Crushing , Do , Nonmetal. < Stone...., Limestone, Stone. . . . , Limestone, Stone...., Limestone, Metal...., Limestone, Metal...., ...do...., Nonmetal Clay and shale, Stone Limestone GM CONC GM C/TLV M N S TLV UG ^ased mg/m 3 . Geometric mean concentration. Geometric mean concentration-to-TLV ratio. Mill. Sample size (number of samples). Surface. Threshold limit value. Underground, on the 1978-81 data with a minimum N of 50 and a minimum GM CONC of 0.50 20 collected, and 17.35 pet of these samples equaled or exceeded the TLV (table 3). In instances where the C/TLV was 1.0 or more, 68 pet of the workers were wearing respirators. The percentage wearing res- pirators decreased as exposure declined below the TLV, but 52 pet of the 261 workers exposed where C/TLV ranged be- tween 0.80 and 0.99 were wearing respir- ators. Data from table 8 showed that general laborers and baggers employed in sandstone and miscellaneous nonmetal mills had the greatest risk of RQ overexposure. MSHA records show that of the 89 samples collected on baggers at these properties in 1981, 64 pet had C/TLV ratios of 1.0 or greater, and of these, 52 workers (91 pet) wore respir- ators. Of general laborers with similar exposures, 82 pet wore respirators. MSHA regulations require the use of respira- tory protection on an interim basis in situations where overexposures occur and the implementation of engineering con- trols to permanently protect workers. SUMMARY AND CONCLUSIONS The objective of this report was to statistically analyze records of RQ expo- sures collected by MSHA from 1974 through 1981 to identify commodities and occupa- tions associated with high quartz dust exposure. The analysis was conducted using the Mine Inspection Data Analysis System (MIDAS), a computerized system designed by the Bureau with the collabor- ation of MSHA. Editing the RQ data left 41,502 records for analysis, and these were unequally distributed among the occupation, commod- ity, and location codes used by MSHA. The primary reason for the unequal dis- tribution is that MSHA is mandated by law to inspect all properties within its jurisdiction, and this type of sampling strategy would tend to assess dust levels in all types of mines rather than in high-risk areas only. The C/TLV was shown to approximate a log-normal distribution, and the GM of this distribution was shown to be a good index of exposure because it relates dust concentration to the TLV, which is deter- mined by a formula dependent upon the percentage of quartz found in the sample. The greater the quartz content the lower the TLV. Use of the C/TLV as an index of exposure places an equal weight upon both the quartz content and the dust concen- tration of the sample; hence the ranking of high-risk subpopulations reflects both. Only samples that contained more than 1.0 pet quartz were considered in this analysis; samples with less than 1.0 pet quartz were reclassified as respir- able nuisance dust. The GM CONC for all samples was calcu- lated to be 0.43 mg/m 3 with a GSD of 2.98. The median concentration was 0.40 mg/m 3 , and 18.37 pet of all samples (7,623 samples) exceeded the TLV. The GM C/TLV was 0.37 with a GSD of 3.40. The TLV calculated from these data was 1.16 mg/m 3 with an average quartz content of 6.6 pet. The range of quartz content was between 1.0 and 98.0 pet. The highest GM C/TLV s and the highest percentage of overexposure were in mills, for workers involved in bagging, crush- ing, grinding, sizing, and general labor activities. Exposures were highest in industrial sand processing plants. The respirable dust concentrations measured at these plants were not the highest, but the allowable level of exposure was the lowest because the samples contained high percentages of quartz. These mills are located at mines classified as sandstone, sand and gravel, miscellaneous nonmetal, and miscellaneous stone producers. Other mills with high exposures were molyb- denum, clay and shale, barite, and gold and silver producers. Crushing, grind- ing, sizing, and general labor activities at limestone mills were also dusty activ- ities, but the quartz content of the rock was low, making these exposures less serious. More than 100,000 workers were employed in noncoal mills in 1981, which was more than 40 pet of the entire metal and nonmetal mine workforce, so the 21 problem was significant in terms of the number of potentially exposed workers. Fewer samples were collected in under- ground mines, and exposures were gener- ally lower than in mills ; but several commodities mined underground had rela- tively high levels of exposure. Under- ground molybdenum, clay and shale, iron, and miscellaneous nonmetal mines stood out as having higher GM C/TLV's and a greater percentage of overexposures. Underground mine activities with higher than average exposures were crushing, drilling, and diesel vehicle haulage. Slushing was also a dusty underground mine occupation, but very few samples were collected. Surface mines had the lowest overall exposures, and in these mines drillers in stone quarries had the highest exposures. Crushing at limestone mines was also a dusty occupation, but their relativly low quartz contents resulted in higher allow- able dust levels. The analysis showed that quartz concen- trations varied when data were grouped by location and commodity. About 3,200 more samples would have exceeded the exposure limit had the NIOSH recommended PEL of 50 ug/m 3 been in effect. Quartz content was highest in samples from mills producing products such as industrial sand and tripoli, and in underground mines pro- ducing molybdenum and clay and shale. Specific dust control research projects should focus on these high-risk areas with the intention of developing controls that can be transferred to other seg- ments of the industry with minimum modification. REFERENCES 1. National Materials Advisory Board. Measurement and Control of Respirable Dust in Mines. Natl. Acad. Sci., Wash- ington, DC, NMAB-363, 1980, 405 pp. 2. National Institute for Occupational Safety and Health. Occupational Health Guidelines for Crystalline Silica. DHHS (NIOSH) 81-123, Jan. 1981, 121 pp. 3. Watts, W. F., R. L. Johnson, D. J. Donaven, and D. R. Parker. An Introduc- tion to the Mine Inspection Data Analysis System (MIDAS). BuMines IC 8859, 1981, 41 pp. 4. Jahsman, W. E., R. L. Johnson, D. J. Donaven, and W. F. Watts. MIDAS User's Manual. Internal BuMines manual, 1981, 210 pp.; available from the Divi- sion of Automatic Data Processing, Bu- Mines, Denver, CO. 5. U.S. Congress. The Federal Mine Safety and Health Act of 1977. Public Law 91-173, as amended by Public Law 95- 164, Nov. 9, 1977, 83 Stat. 6. U.S. Code of Federal Regulations. Title 30 — Mineral Resources; Chapter 1 — Mine Safety and Health Administration; July 1, 1981. 7. American Conference of Govern- mental Industrial Hygienists. TLV's — Threshold Limit Values for Chemical Sub- stances in Workroom Air Adopted by the ACGIH in 1973. Cincinnati, OH, 1973, 54 pp. 8. . Documentation of the Threshold Limit Values. 4th ed. , 1980, Cincinnati, OH, pp. 364-465. 9. Mine Safety and Health Administra- tion. Fiscal Year 1982 Health Mine Rank- ing and Inspection Guideline. Available from Health Div. for Metal and Nonmetal Safety and Health, Mine Safety and Health Administration, Arlington, VA, 1982, 3 pp. 10. Dixon, W. J. (ed.). BMDP Sta- tistical Software. Univ. of CA Press, 1981 ed., 1981, 715 pp. 22 11. Freund, J. E., and R. E. Walpole. Mathematical Statistics. Prentice-Hall, 1980, pp. 206-211. 12. JRB Associates. Technological Control of Asbestos and Silica at Mines and Mills, Task 1 Progress Report Expo- sure Profiles. (NIOSH contract 210- 81-4104), Oct. 30, 1981; for inf., contact Roy Flemming, TPO, NIOSH, Wash- ington, DC. 13. National Institute for Occupa- tional Safety and Health. Criteria for a Recommended Standard. . .Occupational Expo- sure to Crystalline Silica. HEW (now HHS), 1974, HEW 75-120. 14. Center for Silicosis-Illinois. tality Wkly. Rep. , pp. 205-206. Disease Control. Morbidity and Mor- v. 29, May 9, 1980, 23 APPENDIX A.--MSHA SAMPLING CODES AND DESCRIPTORS AND SELECTED GM C/TLV DATA TABLE A-l. - MSHA contaminant codes for personal samples and number of edited records for each contaminant Code Type of measurement RQ Respirable nuisance dust Talc, nonasbestiform Nuisance dust 1 Cristobalite, respirable Tridymite, respirable Mercury vapor Lead 1 Cadmium 1 > Arsenic and compounds 1 Manganese 1 Beryllium 1 Iron oxide 1 Asbestos , fibers >5ym. ....... Cobalt 1 Copper fume 1 Molybdenum 1 Nickel 1 Vanadium fume ^ Zinc oxide fume 1 Chromium 1 Oil mist 1 Airborne silica dust, total.. Welding fume 1 Noise, dosimeter measurement. , Sound level meter measurement, Magnesium oxides 1 Aluminum oxides 1 Titanium oxides 1 pirable quartz dust, particulate. Unit of measure used mg/nr* , mg/m 3 mppcf , mg/m 3 , mg/m 3 mg/m 3 mg/m 3 mg/m 3 mg/m 3 yg/m 3 yg/m 3 ug/m 3 mg/m 3 fibers/cm 3 mg/m 3 mg/m 3 mg/m 3 mg/m 3 Pg/m 3 mg/m 3 mg/m 3 mg/m 3 mg/m 3 mg/m 3 pet dBA mg/m 3 mg/m 3 mg/m 3 Number of records 1... 2... 3... 4... 11.. 12.. 13.. 14.. 15.. 16.. 17.. 18.. 19.. 20.. 21.. 22.. 23.. 24.. 27.. 28.. 29.. 31.. 34.. 35.. 40.. 41.. 81.. 82.. 83.. 41,566 11,582 283 13,415 168 3 70 1,357 379 457 806 332 1,430 1,365 527 714 490 609 508 860 830 56 12,515 1,030 86,229 3,601 898 945 717 RQ Res 1 Total 24 TABLE A-2. - MSHA contaminant codes for area samples and number of records for each contaminant Code Type of measurement Mercury vapor Radon daughter measurement Nitrogen oxide Nitrogen dioxide Nitrogen oxides Carbon monoxide Carbon dioxide Aldehydes Ammonia Hydrogen sulfide Sulfur dioxide Chlorine Sulfuric acid mist Hydrogen cyanide Carbon disulfide Perchlorethylene. Phosgene. Oxygen Hydrocarbons, total Methane Unit of measure used mg/nr* WL PPm PPm PPm PPm PPm Ppm PPm PPm PPm PPm mg/m 3 PPm PPm PPm PPm pet PPm pet Number of records 13.. 50.. 70.. 71.. 72.. 73.. 74.. 75.. 76.. 77.. 78.. 79.. 80.. 87.. 88.. 89.. 90.. 91.. 92.. 93.. 810 27,311 180 9,261 431 28,170 24,292 132 259 2,450 577 22 1 248 4 17 20 15,663 109 28,217 TABLE A-3. - RQ samples collected 1974-81 for each MSHA occupation code Code Occupation Number of personal samples Code Occupation Number of personal samples 1... 2... -J • • • 4... 5... 6. . • 7... 8. . . 9... 10.. 11.. 12.. 13.. 14.. 15.. 16.. 17.. 18.. 19.. Slushing Machine mucking Hand mucking Timbering Rock bolting Backfilling Blasting. Rock sawing Drilling, percussive... Drilling, rotary Drilling, diamond Loading, hauling, dump- ing electric. Loading, hauling, dump- ing diesel. Loading, hauling, dump- ing gasoline. Loading, hauling, dump- ing compressed air. Mining machine operator Track crew Complete mining cycle. . Concrete operations.... 214 79 58 83 78 4 113 259 1,855 926 66 974 12,115 500 158 61 36 1,092 217 20... 21... 23... 24... 25... 26... 27... 28. • . 29... 30... 31... 32... 34... 35... 36... 37... 38. . • 39... 40... Hoisting Bulldozing Slurry General labor and cleanup. General shopwork Welding Mechanic Crushing Grinding Roasting, retorting.... Drying, filtering, and thickening. Sizing Concentrating Chemical operations.... Supply handling Technical services..... Administration Bagger Pelletizing Dredging Jet piercing 132 967 79 4,551 344 150 1,210 6,617 1,384 309 1,314 1,576 457 92 350 246 365 2,018 302 58 99 25 TABLE A-4. - RQ samples collected 1974-81 for each commodity Commodity N Commodity N STONE Cement. . . , Granite. . , Lime , Limestone, Marble. . . , 2,347 3,509 393 9,752 37 Sandstone, Slate...., Traprock. , Misc. . ... , 3,130 266 711 468 METAL MINES AND MILLS Antimony Bauxite (including alumina mills). Beryl Chromite Copper Gold and silver, lode and placer. . Iron 33 12 2 1,999 874 2,177 Lead and zinc, Manganese. . . . , Molybdenum. . . . Titanium , Tungsten , Uranium , Mercury , Other metals. . 905 13 526 20 38 1,097 6 15 NONMETAL MINES AND MILLS Asbestos , Barite , Boron minerals. Clay and shale, Feldspar , Fluorspar , Gypsum. ........ Magnesite , Mica Phosphate rock, 3 140 7 3,213 193 122 180 12 134 462 Potash , Pumice , Salt , Sodium compounds , Sulfur , Talc, soapstone, and pyrophyllite, Gilsonite , Oil and/or shale. , Misc. nonmetals 1 , 4 31 7 11 142 2 4 741 MISC. NONFUEL MINES AND MILLS Unspecified 1,031 Sand and gravel | 6,725 N Sample size (number of samples). 1 0ther metals and miscellaneous nonmetals are defined in the Standard Industrial Classification (SIC) codes published in 1979 by the National Bureau of Standards. TABLE A-5. - RQ samples from metal and nonmetal mines, by location Category N Underground mine locations: Mill 77 Shop 22 Metal mine 3, 245 Nonmetal mine 1,134 Stone mine 455 Open pit mine locations: Crushed stone mine 11,339 Sand and gravel quarry 5,492 Nonmetal mine 3, 050 Metal mine 1,445 N Sample size (number of samples). Category N Surface mill locations: Mill (bagging, screening, etc.) 5,575 Crushing 4,210 Grinding 1, 553 Flotation and reagents 516 Shop 313 Misc 3, 068 26 TABLE A-6. - GM C/TLV's for 9 occupation groups used in table 8 (In order of decreasing GM C/TLV for 1978-81) Occupation group 1974-77 N > TLV, pet GM C/TLV 1978-81 N > TLV, pet GM C/TLV Bagger , Drilling , General labor. . . < Crushing Grinding, sizing. Finishing Other Welder, mechanic. Load, haul, dump. 786 1,182 2,862 2,215 1,402 984 1,035 740 5,152 46.2 26.7 27.0 30.1 33.4 26.6 23.5 16.5 11.3 0.91 .49 .52 .57 .60 .48 .43 .36 .27 1,232 1,559 3,125 4,402 1,558 1,398 922 620 9,398 39.7 21.3 20.0 18.5 19.8 17.9 16.0 11.4 6.7 0.83 .44 .43 .41 .41 .37 .37 .32 .23 GM C/TLV Geometric mean concent rat ion-to-TLV ratio. N Sample size (number of samples). TLV Threshold limit value. TABLE A-7. - GM C/TLV's for 8 commodity groups used in table 8 (In order of decreasing GM C/TLV for 1978-81) Commodity group 1974-77 N > TLV, pet GM C/TLV 1978-81 N > TLV, pet GM C/TLV Misc. nonmetal. . Clay and shale. . Sandstone Stone. Sand and gravel. Metal , Nonmetal • . Limestone , 242 1,723 1,363 3,298 2,082 3,524 672 3,308 43.4 34.9 37.4 20.5 21.8 22.2 27.7 14.4 0.83 .65 .67 .40 .37 .46 .56 .29 499 1,490 1,767 4,396 4,643 4,054 709 6,444 40.9 28.3 29.9 16.8 15.1 13.2 10.2 6.6 0.69 .58 .56 .39 .34 .34 .31 .23 GM C/TLV Geometric mean concent rat ion-to-TLV ratio. N Sample size (number of samples). TLV Threshold limit value. 27 APPENDIX B.--MSHA MINE HEALTH RANKING CRITERIA Rank A a. All underground uranium mines. b. All nonuranium underground mines where employees are exposed to radon daughters in excess of 0.10 WL. c. All mines and mills where employees are exposed to mineral fibers classified as asbestos or talc. d. All mines and mills where employees are overexposed to toxic or asphyxiant gases. e. All mines and mills where employees are overexposed to contaminants which have ceiling values recommended by the ACGIH. f . All mines and mills where employees are overexposed to particulates (except silica) and whose TLV's are less than 10 mg/m 3 or 30 mppcf . their TLV: antimony, arsenic, beryllium, cadmium, chromium, cobalt, manganese, mercury, nickel, lead, uranium, and vanadium. j . Any other mine or mill where the inspector and supervisor believe that a serious health hazard exists. Rank B a. All mines and mills where employees are overexposed to particulates whose TLV's are 10 mg/m 3 or 30 mppcf. b. All mines and mills where employees are occasionally overexposed to dust con- taining 1 pet or more free silica. c. All mines and mills where employees are overexposed to fumes containing the following substances at or above half their TLV: antimony, arsenic, beryllium, cadmium, chromium, cobalt, manganese, mercury, nickel, lead, or vanadium. g. All producers of silica sand, glass sand, industrial sand, and silica flour; and all mines and mills where cristo- balite or tridymite occur. h. All mines and mills where employees are repeatedly overexposed to dust con- taining 1 pet or more of free silica. d. All mines and mills where employees are overexposed to noise. e. Any other mine or mill where the inspector and supervisor believe that a health hazard exists. Rank C i. All mines and mills where employees are exposed to dust containing the fol- lowing substances at or above one-half a. All mines and mills where cernible health hazard exists. no dis- 28 APPENDIX C. —ABBREVIATIONS USED IN THIS REPORT NOTE. — This listing does not include unit of measure abbreviations, which are listed after the table of contents, or abbreviations that are used only in the tables and identified below the tables. ACGIH American Conference of Governmental Industrial Hygienists C/TLV concentration-to-threshold limit value ratio GM geometric mean GM CONC geometric mean concentration GSD geometric standard deviation MAS Minerals Availability System MIDAS Mine Inspection Data Analysis System MSHA Mine Safety and Health Administration N sample size NIOSH National Institute for Occupational Safety and Health PEL permissable exposure limit RQ respirable quartz dust TLV threshold limit value INT.-BU.OF MIN ES,PGH.,P A. 27324 HW160 84 * «►, 'oho* <$ * * v ^A&" Wvfc> ^ <£*" •* »'. <^> ^bv" .A . >■ ' « * ^ A* iV . o " ° ■» 'O "oF ■a? ^ •• ^\ *v a v <$> •? i * A 'Cj. ' •» ^ A* *. vV«* '^ A A°* ^i^L'*_ ^ . "^ Xs* ,-*- v ^ - -^0* o, >V „ « o . 'Pi. A> .0 « o * O * ^ • ay o *■ > v » " • °* ^v A> _,c. ^ ^ *^"T.T*' .^A <^. *^. , ;'» ^G v "b, **"^T*' A ,/\ » ay cv * V \s 4> ^* 4 „ i ' * \/ .-^fe*' V/ ## ^/ -ill*- V #fc ^/ •»■, V V /,!^>>o A^*-V y.-Sa^V ./\v^X *° $► * M 1>«** £ % 4- v V> ■ ,0* *o. '♦TJTT.' A . • cS^K^'.. O '■< *W A «. L ( s V./ .**j£SKSfc\ V,** ..SaSfcfc \.J :&£: %u? MM'. V°;--\ 4? 'titer** * v ^°;-o\ / 0*- *o, -*7«7«' A )* .••^-••-% A v .>•'», ^^ HECKMAN BINDERY INC. ^ JUL 84 'mJR^ N.MANCHESTER, ^sSr INDIANA 46962 *> .. • ' ^.v : , « **"V •' «'«.'•*'**!■, sS-XEVwi V-