Document 69ZO3N536G5VRpxber8VMvJ3

Strategiesfor Mine Dust Measurement MEASUREMENT STRATEGIES IN (I.S. UNDERGROUND COAL MINES THOMAS F. TOMB Mine Safety and Health Administration, Pittsburgh, Pennsylvania, USA ABSTRACT The 1969 Federal Coal Mine Health and Safety Act (ACT) mandated standards for occupational exposures to respirable coal mine dust. For mine environments where die respirable dust contains less than 5 percent quartz die standard is 2.0 milligrams of dust per cubic meter of air (mg/m3), where the respirable dust con tains more than 5 percent quartz die standard is adjusted according to die quartz percentage. The Act also required mine operators to carry out a dust sampling program. This paper presents an overview ofdie current methods used in die United States ofAmerica to assess exposures to respirable dust in coal mines, the sampl ing strategies used to enforce the mandatory dust standard, die sampling requirements ofcoal mine operators and a description of the laboratory used to process the more than 100,000 samples per year collected by the coal mine operators. ENFORCEMENT PROGRAMS (STRATEGIES) Since December of 1969, die United States of America has had a Federally mandated respirable dust standard of 2.0 mg/m3 for its underground coal mine environments. Respirable dust, for die purpose of this standard, is defined as die fraction of dust recommended by the British Medical Research Council (BMRQ and adopted by the Johannesburg Pneumoconiosis Conference in 1959. The sampling efficiency curve representative ofdie respirable dust criteria adopted at that conference is shown in Figure 1. Particle diameters in this figure refer to equivalent spherical diameters, which are de fined as the diameter ofspherical particles ofunit density hav ing the same falling velocity as die particles in question. Because of die recognized increased health risk associated with exposure to quartz (crystalline silicon dioxide), the man dated exposure standard is to be adjusted (reduced) when die quartz content in the respirable dust exceeds 5 percent. The adjusted standard is determined by dividing die percent quartz in the respirable dust into the number 10 (i.e., 10/% Si02>. In the United States there are two programs to enforce the mandatory respirable dust standard, a program conducted by the mine operators in accordance with mandated regulatory requirements and a program conducted by the Federal govern ment. Under the operator's program each operator is required to collect five respirable dust samples from a "designated oc cupation," the occupation on a coal getting operation that previous sampling has shown to have die highest dust ex posure, in each coal getting operation every two months. The samples must be collected on consecutive production shifts or on production shifts on consecutive calendar days. The collected samples are sent by mail, within 24 hours after collection, to a central laboratory in Pittsburgh, Pennsylvania, where the amount ofdust collected is determined by weighing. 1124 A data card, shown in Figure 2, is submitted with each sam ple. The dust concentration is determined for each sample us ing the weight ofdust collected, the time over which the sam ple was collected and the flow rate ofthe sampling device (in all cases this is 2.0 liters ofair per minute). All samples are required to be collected for a full production shift (portal-to-portal). The dust concentrations determined from these five samples are averaged. The average concentration is then compared to the 2.0 mg/m3 dust standard (or adjusted standard) to establish compliance or noncompliance with regulatory re quirements. In addition to the five samples collected bi monthly on die designated occupation, die mine operator is SmSSM Figure 1. Comparison ofBMRC respirable size criteria with pulmonary deposition curve. also required to collect an additional sample bimonthly at specified locations throughout die mine. These locations are strategically selected so that the environment where miners normally work or travel is monitored for compliance with the respirable dust standard. If at any time it is determined from any of these samples that the respirable dust standard is ex ceeded, five additional samples are collected (either on con secutive days or consecutive production shifts) at the site where it was determined that the applicable standard may be exceeded. The dust concentrations determined from these samples are averaged and compliance is determined using die applicable standard for the area where the samples were col lected. In accordance with regulatory requirements, die mine operator also submits to die Federal government a ventilation system and a methane and dust control plan which are to in clude: sources of dust generation in the outby areas of the mine, methods being used to control dust at these sources of dust generation and the specific location of places where samples will be collected to monitor the levels ofdust in areas where miners normally work or travel. Also specified in the plan are the parameters characterizing the measures that are being used to control dust at the coal mining (getting) opera tion. The typical parameters specified include the quantity and velocity of air used to ventilate die face, the quantity and pressure ofwater and the number, type and location ofnozzles used in the water spray system. The Federal government's program to enforce the legislated respirable dust standaid(s) consists ofa mine inspector visiting each coal mining operation to approve, or check for com pliance, that portion of the ventilation and dust control plan that describes the measures to be used by the mine operator to control respirable dust levels in the mine environment. To approve the "dust control" portion ofthe plan, an inspector will collect a personal sample on at least five miners work ing in the immediate area ofthe coal mining operation where die parameters described in the plan are being used to con trol the dust. If the type of mining is "room and pillar" employing continuous mining equipment, one sample must be collected from the environment of the continuous miner operator, one from die environment of die roofbolter operator and three from other occupations working in the immediate area. Typically these other three samples are representative ofthe environments of shuttie car operators, continuous miner operator helpers and laborers. If the mining operation is a longwall mining operation, the samples are representative of the shearer operators and shield (jack) setters. The sampling equipment is normally mounted on the miners (referred to as personal sampling) prior to die start ofthe shift and removed after die shift is finished. After the samplers are removed from die miners, a mine data card is completed and the sample and data card taken to a local Federal enforcement laboratory for processing. The respirable dust samples col lected are weighed to a tolerance of 0.1 mg which is the same as for those samples collected by the mine operators. After the samples are weighed and the net weight of the col lected dust determined, die concentration of dust, in mg/m3, is calculated using die weight of die dust collected and die volume of air sampled. To determine ifthe parameters being used to control dust are Strategies for Mine Dust Measurement effective in reducing the respirable dust level in the environ ment to die applicable standard, the dust concentrations deter mined from the five samples are averaged. For the plan to be considered adequate, die average dust concentration must be below 2.0 mg/m3 and the concentration of no individual sam ple can be greater than 2.0 mg/m3. If the average concentra tion determined from the five samples exceeds 2.0 mg/m3, the work area is found to be in noncompliance and the mine operator must improve the practices being used to control dust and specify these changes in his dust control plan. Ifthe average concentration determined from the five samples is below 2.0 mg/m3, but one or more of the individual samples is greater than 2.0 mg/m3, then sampling continues on all five occupations on subsequent production shifts. Sampling is continued until the average concentration deter mined from the individual occupation samples collected on Dust Oata Card 1. Cassette Number 2. Min* ID Number 3. Contractor Coda 4. Mina Hama 5. Company Mama 4. Data Sampled 7. Sampling Tima Mo. Oa.______ Yf._______________(min) 6. Ton* This Shift ATTACH CASSETTE HERE 9. Type of Sample (select one) <1} designated occ (ug) (2) nondesignated occ (ug) <31 designated area (ug) (4) designated work position (sur) (5) part 90 miner 10. MMUOA/SA 11- Occ Code RETURN THIS COPY TO MSHA cm mkt i WITH CASSETTE. Figure 2. Mine data card. 1125 Strategies for Mine Dust Measurement consecutive production shifts and the average concentration determined from samples collected on the same shift are both equal to or less than 2.0 mg/m3. No more than five produc tion shifts are sampled. As previously discussed, die 2.0 mg/m3 respirable dust stan dard is reduced whenever it is determined that the quartz con tent ofthe respirable dust exceeds 5 percent. Determination ofthe quartz percentage ofdie respirable dust is based on die analysis ofa selected number ofsamples collected during the plan approval process. Those samples typically selected for analysis are die designated occupation sample, all roofbolter samples and any other sample that may be suspected ofhav ing a high quartz percentage. After sampling has demonstrated that the procedures specified in the plan for controlling dust are adequate, subsequent in spections (up to three) during die year are limited to check ing on conformance with die dust control plan; i.e., no dust samples are collected, only dust control procedures are evaluated. RESPIRABLE EXIST SAMPLING INSTRUMENTATION To measure the respirable dust concentration ofcoal mine en vironments in the United States, a two-stage sampling instru ment is used. The instrument, commonly referred to as a per sonal respirable coal mine dust sampler, is shown in Figure 3. The sampler was designed to be an instrument that was capable of sampling die environment to which a miner is ex posed during his full work shift Therefore, die instrument has the flexibility of either being mounted on a person (as shown in Figure 4) to obtain his exposure or of stationary mounting to obtain measurements ofany general environment where it is located. The sampler consists of a 10 mm diameter nylon cyclone, a filter and a pump. The 10 mm nylon cyclone, the first stage ofthe sampling system, separates the sampled aerosol into two fractions: a respirable fraction and a nonrespirable fraction. The particle selectivity curve that defines the separated frac tions is shown on Figure 3. The nonrespirable fraction is collected and retained in die cyclone (Figure 6) while die respirable fraction passes through the cyclone and is collected on a 37 mm diameter, 5 micrometer pore size, vinyl metricel membrane filter. The filter is preweighed by its manufacturer to a precision of 0.1 milligram. The cyclone and filter assembly, commonly re ferred to as die ``sampling head," is designed to be mounted on the miner at his "breathing zone.** The pump, used to induce air into the sampling system, is 1126 Figure 3. Personal respirable dust sampler. Strategies for Mine Dust Measurement Figure 4. Personal sampler worn by miner. battery powered and can be easily worn by a miner during die performance of his duties. It weighs less than one kilogram and has overall dimensions of approximately 5 cm x 10 cm x 13 cm. Air is sampled at the rate of2.0 liters per minute (0.1 liters per minute). Because the 2.0 mg/m3 dust standard is based onmeasurementdata obtained with an instrumentthat sampled with respect to the BMRC selectivity curve shown in Figure 1, respirable dust concentrations determined from measurements obtained with die personal coal mine dust sampler must be multiplied by a factor of 1.38 before die measurements can be used to determine compliance with the mandatory dust standard. PROCESSING COAL MINE OPERATOR DUST SAMPLES As a result ofdie Federally mandated regulatory program, ap proximately 110,000 dust samples are collected by mine operators each year. These samples and associated data are mailed to the Federal government's central processing laboratory located at Pittsburgh, Pennsylvania. At die central processing laboratory, samples are processed in a "clean room" environment. The laboratory is maintained 1127 Strategies for Mine Dust Measurement at a slight positive pressure to limit the entry of extraneous dust from surrounding work areas. The environment in die room where samples are weighed is maintained at 23 1C and 50 percent 5 percent relative humidity. Prior to weighing, samples are vacuum desiccated to remove moisture that may be present on die sample. The internal pressure ofthe desiccator chamber is reduced to 5 mm Hg and held at that pressure for 15 minutes. Since January, 1985, respirable dust samples have been pro cessed using die Automated Weighing System (AWS) shown in Figure 7. The AWS is a robotic system which has been designed for unattended weighing offilter capsules on a Mettler AE163 analytical balance. The robotic arm (Figure 8) has die ability to rotate 360 around its central vertical axis, move up and down its vertical axis as well as in and out from the horizontal axis. At one end of the robotic arm is a * `hand' ' with a pair offingers which may be made to open and close as well as rotate 180 in wrist-like movements around die arm's horizontal axis. The system is designed so that the robot can sequentially process up to 200 samples from five trays without manual intervention. Process ing time for 200 samples is approximately four hours. Performed tasks are programmed into a power and event con troller. The power and event controller zero's the balances before weighing each sample, switches a relay to select either of two balances, activates a solenoid to open and close a balance door and to sound an alarm buzzer when manual in tervention with the AWS is required. Upon completion of a weighing, die controller activates a printer which prints the weight of each filter capsule and a sequence number on a 1 cm x 5 cm pressure sensitive label. The label is subsequently affixed to the data card. The Mettler Model AE163 analytical balance used with the AWS is shown in Figure 9. This state-of-the-art analytical balance has a weighing precision of 0.02 mg. Each balance is calibrated twice daily and checked with a Class M certified calibration weight. A radioactive deionizing unit is used to eliminate die presence of static charge on filter capsules. To isolate vibrations, the balances are positioned on a marble table weighing approximately 320 kg. The AWS has been programmed to systematically weigh a sample twice on two different Mettler AE163 balances. One in eight ofeach filter capsule weighed is reweigbed on the sec ond balance. Ifthe weight difference obtained between the two balances is within 0.1 mg, the weighings are considered to be within tolerance and weighings are continued. If the weights are out of tolerance, an alarm sounds and both balances are recalibrated. The system then reweighs die last seven filters, performs another quality control check weighing and continues processing additional samples if die check weights are within the established tolerance. As previously discussed, each respirable dust sample is Figure 5. Comparison ofthe 10 mm diameter cyclone selec tivity curve with pulmonary deposition curve. 1128 Figure 6. 10 mm cyclone with filter. Strategies far Mine Dust Measurement Figure 7. Automated weighing system. accompanied by a mine data card (Figure 2). The data on each card is manually transcribed (Figure 10), in numeric notation, onto magnetic discs. Each card contains 62 keystrokes or digits. Data transcription is verified using a double entry system. Data retranscribed by a second operator is compared to that originally transcribed. The verifying operator is alerted to resolve errors or mismatched data. All disks generated dur ing the day are then machine edited for completeness and ac curacy. After editing, all data is accumulated and telecom municated to an Information Systems Center in Denver, Colorado. The information telecommunicated to the Information Systems Center is compiled and the respirable dust concen tration for each sample calculated. A copy ofall the data and sample results are mailed directly to the mine operators. The results are also telecommunicated to local enforcement offices which have interactive access to all dust data file information. SUMMARY The promulgation of a respirable dust standard for underground coal mine environments and the programs in stituted to enforce that standard have resulted in a more healthful working environment for U.S. coal miners. As shown in Figure 11, occupational exposures have steadily decreased since promulgation ofthe respirable dust standard. However, as the data on this graph also depicts, the reduc tion ofdust levels on longwall mining operations has not been as great as on the other types ofmining operations. Work still needs to be done to develop methods to control dust on longwall mining operations. The program requiring coal mine operators to sample their mine environments and to submit the samples to the Federal government for analysis has been effective in reducing underground respirable dust levels, and has provided the im petus for them to institute procedures to control dust. 1129 Strategiesfor Mute Dust Measurement Figure 8. Robotic arm. 1130 Measurement Strategies for Mine Dust automated weighs system. ^^calbalaueeusedwMb Figure 9 Strategies for Mine Dust Measurement Figure 10. Data processing station. 1132 DUST LEVEL, m g /m ' Strategies for Mine Dust Measurement LEGEND {= CONVENTIONAL m CONTINUOUS Q LONGWALL AUGER 1970 / y y y / / y y / y / / > / y y / /I A / y z1 1971 / / / / y y > / / y A } A / / y } / / / / / / / 1972 1973 1974 1975 YEAR A 1976 1977 1982 1987 Figure 11. A yearly comparison of dust levels for four types of mining. 1133 Strategies for Mine Dust Measurement THE THRESHOLD LIMIT VALUE FOR VARIOUS FORMS OF AMORPHOUS SILICA RONALD S. RATNEY, Ph.D., CIH ACGIH Threshold Limit Value Committee (Chemical Agents), Bedford, MA, USA Silica is the common name for silicon dioxide (SiC>2). In silica each silicon atom is covalently bound to four oxygen atoms which are arranged tetrahedrally around it. Each oxygen atom is bound to two silicon atoms. In the crystalline forms ofsilica die silicon and oxygen atoms are arranged in a highly ordered lattice which extends infinitely in all directions. Of course, the lattice is not truly infinite since it must end at the surface of the solid. The surfaces of crystalline silica particles or macroscopic pieces are bounded by flat surfaces joined at sharp straight edges. In some forms ofcrystalline silica such as tripoli or quartzite, die surfaces and edges may have been worn away to produce what appears to be amorphous particles. All naturally occurring crystalline silica was formed by crystallization from aqueous solution or from molten magma. Depending on the temperature and pressure at which die crystallization takes place, one ofthree different geometrical arrangements ofthe silicon and oxygen atoms will be formed. The most common crystalline form of silica is quartz, which occurs as solid crystals from several inches in size down to microscopic dimensions. Other forms ofcrystalline silica are cristobalite and tridymite. Under several natural and artificial conditions, silicon diox ide will form solids with no overall spatial ordering of the atoms. These products are amorphous silicas. Solid objects and particles ofamorphous silica do not display flat faces and sharp edges. More importantly, amorphous silicas do not display X-Ray diffraction patterns as do die crystalline forms. The several forms of amorphous silica display different physical and chemical properties and substantially different toxicological characteristics. The only naturally occurring form of amorphous silica is diatomaceous earth whose particles are the fossil skeletons of microscopic marine plants known as diatoms. While alive these organisms extract silica from the sea water and deposit it in complex regular forms with numerous voids. In Califor nia and other parts ofdie world there are very large deposits of the mineral diatomite or diatomaceous earth consisting almost entirely of fossilized diatoms, and which is a highly porous substance with a very low bulk density. The overly ing and surrounding rock firequendy contains quartz which contaminates die final product. Some deposits contain traces of cristobalite, apparently formed by metamorphism. Fused quartz, or more properly, fused silica, is formed by die relatively slow solidification of molten quartz. In the melt 1134 there is no long range order. As it cools, die molten material becomes highly viscous and then solidifies so that die atoms become immobilized in their random positions. Fused silica is produced as lumps of glassy material but during crushing and grinding, respirable particles can be produced. Glass or silica dissolves in sodium hydroxide to form a solu tion of sodium silicate, also known as water-glass. On acidification, this forms die insoluble flocculent precipitate of silicic acid (H4Si04 or Si(OH)4). As water is eliminated between nearby SiOH groups Si-O-Si bridges are formed. Depending on the dehydration process, precipitated silica or silica gel is produced. These both can be considered to be par tially hydrated silicon dioxide. Silica gel can be dried to a very low moisture content to form a granular product which ab sorbs water and polar organic substances with great avidity. Fumed silica is produced synthetically by a vapor phase hydrolysis of silicon tetrachloride in a flame ofhydrogen and oxygen. It is a widely used filler in paints, plastics and rub ber and as an antiskid and antislip agent. Elemental silicon is produced by reacting coke and silica sand (crystalline) in an electric arc furnace. If iron is included in the charge, the product is ferrosilicon. In both cases, silicon monoxide is apparently produced as a byproduct which escapes from the furnace and is oxidized by ambient oxygen to produce what can be called silica fume. Although it is not a deliberately manufactured product, baghouse dust from silicon and ferrosilicon furnaces has been used in the same way as fumed silica. Although both fumed silica and silica fume are fumes in die usual industrial hygiene sense (they are finely divided solids produced by condensation from the gas phase) their mode of formation and worker exposure are different. As is discussed below, the toxic effects are also quite different. Precipitated silica and silica gel could be considered to be the prototypical nuisance dusts. The ACGIH considers a material to be a nuisance dust ifit causes no adverse health effects when exposures are kept under reasonable control (e.g. near or below 10 mg/m3) and further does not alter the lung air spaces, does not form collagen to a significant extent and whose tissue reactions are potentially reversible.3 Klosterkotter showed that silica gel injected intratracheally in rats did not cause fibrosis.6 Schepers et al observed no fibrosis in guinea pigs and rabbits exposed by inhalation at 126 mg/m3 for two years.7 There were macrophage ac cumulations and mild proliferation of reticulin fibers. In a group of 165 workers exposed to precipitated silica estimated to be near or below 10 mg/m3 for an average of 8.6 years, Wilson et al observed no serial changes in pulmonary func tion or chest radiographs.9 The TLV or 10 mg/m3 (total dust) assigned to precipitated silica and silica gel was not chosen to avoid any known adverse health effect.3 Rather it represents a recommendation for good industrial hygiene practice. Airborne exposure above this level may reduce visibility, may cause unpleasant deposits in the eyes and nasal passages and may cause injury to the skin and mucous membranes by purely mechanical action. Fumed silica and silica fume display entirely different tox icides. This contrast illustrates the confusion created by misidentification of the toxic substance in epidemiological studies and the risk of predicting toxicity on the basis of chemical similarity. As noted above, both products are true fumes, ultrafine solid particulates formed in gas phase reac tions. However, fumed silica appears to be only slightly more toxic than precipitated silica and silica gel. ASTM standard El 156-87 reviewed three studies involving a total of 353 workers exposed for up to 32 years to fumed silica concen trations from 1.6 to 53 mg/m3.1 No pulmonary dysfunction was observed except in smokers. Schepers exposed rats, rab bits and guinea pigs to filmed silica at 53 mg/m3 for a year causing emphysema which reversed after exposure ceased and fibrosis which partially reversed.7 Groth observed significant interstitial hyperplasia and collagen deposition in monkeys ex posed to 15 mg/m3 of fumed silica for 13 months.4 However, the monkeys' lungs showed the presence ofmineral dust which had apparently been inhaled in die wild or in captivity prior to purchase ofthe animals. No changes were observed in rats and guinea pigs similarly exposed. On the other hand, several studies in the elemental silicon and ferrosilicon industries show that silica fume produces a unique complex ofacute and chronic effects which are reversible after exposure ceases. The observations ofBowie at an African fer rosilicon plant are typical.2 Brief high exposures to silica fume produce the symptoms of metal fume fever, which can persist for up to three months. Chronic exposure produces X-Ray and pulmonary function evidence of silicosis which regresses or disappears after cessation of exposure. The TLV for fumed silica has been set at 10 mg/m3; the value assigned to nuisance dusts.3 No value has been established for silica fume but a TLV of 0.2 mg/m3 (twice the TLV for quartz) seems reasonable. It is tempting to speculate on the causes of die radical dif ference between the two silica fume materials. In the case of silica fume, the effects may be produced bv repeated high ex posures but no airborne measurements are available to sup Strategies for Mine Dust Measurement port this hypothesis. It is also possible that silicon and fer rosilicon workers are exposed to a much more freshly formed fume since they work at die tapping ports ofthe furnaces while the synthetic fumed silica may have aged for a few minutes before reaching the workers' breathing zones. Again there is no evidence to support this. In contrast to the other forms of amorphous silica, the TLV for fused silica is based on very litde actual data, animal or human. Hie Documentation references only two studies both published in the early 1950s; one an acute intraperitoneal in jection in rabbits, die other an intratracheal instillation in rats.3,8,5 No inhalation experiments in animals or epidemiological studies in exposed workers have been published since then. Both references indicated that fused silica was less active in inducing a tissue reaction than crystalline quartz but no comparisons with nonfibrogenic forms of amorphous silica were performed. On die basis of the fact that there was a tissue reaction at all, a TLV of 0.1 mg/m3 was established; die same as for quartz.3 Fused quartz is now used in several advanced technological products such as ablative surfaces for rocket reentry vehicles and in fiber optics. It is anticipated that more workers will be exposed to this hitherto exotic material and it is unfortunate that more solid toxicological data is not available. REFERENCES 1. ASTMStandard El156-87. Health Requirements for Occupational Ex posure to Synthetic Amorphous Silica. American Society for Testing and Materials, Philadelphia (1987). 2. Bowie, D. St. J., Ferro Alloy Workers' Disease. Cent. Afr. J. Med. 24:81-86 (1978). 3. Documentation ofthe Threshold Limit Values andBiological Exposure Indices, 5th Ed. American Conference of Governmental Industrial Hygienists, Cincinnati (1987). 4. Groth, D.H., Moorman, W.J., Lynch, D.W., Stettler, L.E., Wagner, W.D., Horaung, R.W.: Chronic Effects ofInhaled Amorphous Silicas in Animals. In Health Effects ofSynthetic Particulates. ASTM STP 732. Dunnom, D.D., Ed. American Society for Testing and Materials, Philadelphia (1981). 5. King, E.J., Mohanty, G.P., Harrison, C.V., Nagelschmidt, G.: The Action of Different Forms of Pure Silica on the Lungs ofRats. Brit. J. Ind. Med. 10:9-17 (1953). 6. Klosterkotter, W.: Zur Frage der silikogenen Wirkung des amorphen Siliciumdioxide in Jotten, K.W. Klosterkotter, W.: Die Staublungenergrankungen, Vol. 2., pp. 73-84. Steinkopf, Darmstadt (1954). 7. Schepers, G.W.,Durkan,T.M.,Delahant,A.B.,Crecdon,F.T.,Redlin, A.J.: The Biological Action ofDegussa Submicron Amorphous Silica Dust. AMA Arch. bid. Health 16:125-46, 203-244, 280-301 (1957). 8. Silverman, L., Moritz, A.R.: Arch. bid. Hyg. Occup. Med. 1:499 (1950). 9. Wilson, R.K., Stevens, P.M., Lovejoy, H.B., Bell, Z.G., Richie, R.C.: Effects ofChronic Amorphous Silica Exposure on Sequential Pulmonary Function. J. Occup. Med. 21:399-402 (1979). 1135 Strategies for Mine Dust Measurement COMPARISON OF THE SAMPLING STRATEGIES RECOMMENDED BY THE EUROPEAN COMMUNITIES FOR THE PROTECTION OF WORKERS FROM THE RISKS RELATED TO CHEMICAL AGENTS AT WORK, ASBESTOS, LEAD AND MINE DUST B. PREAT IREA, Havennarkt 22, B-3500 Hasselt, Belgium INTRODUCTION The purpose of this paper is to evaluate die monitoring strategies implemented by the various European Directives on die exposures to airborne workplace contaminants. Monitoring strategies to determine compliance with occupa tional health standards entail a number of requirements that are usually determined by consensus rather than through the scientific process. Statistical models will be used to compare die proposed sam pling strategies. Past research has shown that the concentra tion distribution of most air pollutants can be described as lognormal. Therefore, the analysis will be based on die lognormal distribution. Such distributions are completely defined by die geometric mean (GM, a measure of central tendency) and by die geometric standard deviation (GSD, a measure of die variability of exposures). When a monitoring strategy provides consecutive shift or daily samples to determine compliance, the autocorrelation ofdie exposures should be taken into account. DIRECTIVES OF THE EUROPEAN COMMUNITIES The first Directive laid down by the Council ofdie European Communities is the Directive 80/1107/EEC of 27th November 1980 on die protection of workers from die risks related to exposure to chemical, physical and biological agents at work.1 This is a global Directive providing for the laying down of individual Directives for specific agents. Directive on Lead The first of these individual Directives has been the Council Directive 82/605/EEC of28th July 1982 regarding exposure to metallic lead and its ionic coumpounds at work.2 Taking into account the biological half-life ofthe agent, the limit value for lead is based on the time-weighted average concentration over one week (40 hours). The strategy provides die follow ing stages: initial designation if the sample exceeds 1/2 of die limit value; a quarterly sampling cycle in the first instance; die frequency ofmonitoring may be reduced to once a year iftwo consecutive measurements are below 2/3 ofthe limit value. 1136 Directive on Asbestos The Council Directive 83/477/EEC of 19th September 1983 relating to exposures to asbestos at work was the second in dividual Directive.3 Here the limit values are measured or calculated in relation to an eight-hour reference period. The general rule is to measure die level ofasbestos at least every quarter. This frequency may be reduced to once per year when die results ofthe two preceding measurements have not exceed ed 1/2 ofthe limit value. As die time-weighted average over 8 hours has a greater standard deviation than the average over 40 hours,4 a lower action level has been chosen: namely 1/2 ofthe limit value instead of2/3. Ifdie concentration is lower than 1/4 of die limit value, monitoring is terminated. Proposal for Modification of the Frame Directive 80/1107/EEC On 6 June 1986, the Commission of die European Com munities presented a proposal for modification ofthe Frame Directive 80/1107/EEC.5 The strategy is similar to the one for asbestos, except that die action level has been lowered to 1/3 ofthe limit value and the decision to end sampling is taken when the concentration does not exceed 1/5 ofthe limit. This proposal has not been approved and Technical Committee TC 137 of die CEN (European Committee for Standardization) has been invited to draw up its own sampling scheme for die determination of airborne hazardous substances at die workplace. Draft Proposal ofthe Safety and Health Commissionfar the Mining and other Extractive Industries of the E.C. to the Governments ofdie Member States to reduce die risk to health associated with die exposure to fibrogenic mineral dust in die non-coal mining and quarrying industries. This draft proposal, in its last version (Doc. 5761/10/85) of 19 May 1988, provides for die same measuring strategy as the Directive on asbestos. Moreover, it provides for an alter native approach towards dealing with the problem ofexposure fluctuation by interpreting the limit in terms ofthe mean ex posure over one year. Daily levels are allowed to exceed the limit ifthey are compensated by die days of low exposure so that the one-year time-weighted average remains below the limit. Intuitively, it would appear difficult to accurately characterize Ae exposure over one year with only one, two or three isolated estimates of daily exposures. For example, in West German underground coal mines, the exposure over one year is estimated from averaging the results of 12 monthly measurements. STRATEGY EVALUATION An evaluation ofthe sampling strategies has been conducted using a lognormal model. For this model, a value of 1.7 has been chosen for GSD, a typical value for die distributions ofone-shift respirable dust concentrations in European underground coal mines (although it can vary from 1.2 to 3.0 in other work environments). When GSD = 1.7, it means that 5% ofthe shifts have a concentra tion exceeding 2.8 times the geometric mean or 2 times die arithmetic mean. Low values ofGSD indicate good dust con trol. When GSD > 2.5, it is likely that there are no function ing engineering controls.6 Thus dust measurements carried out for industrial hygiene purposes, even performed with the same instruments at the same place, can provide very different results, implying capricious decisions concerning both compliance and con trols. It is clear that this can lead to all forms and kinds of injustice and may have litde to do with the degree of chronic hazard. Strategies for Mine Dust Measurement For the study of the autocorrelation, we have used the time series made up by a string of consecutive measurements recorded by a recently developed respirable dust continuous measuring instrument [HUND, Wetzlar, West Germany] set up in the tailgate of a German longwall coal face.7 Figure 1 shows the time series plot ofdie shift averages at the sampling point at 50 m from die face. It is seen that the range ofthe shift averages observed during a period of 138 consecutive shifts (4 shifts a day) exceeds a factor of 13. Consider two concentrations, c(t) and c(t + h), at two shifts t and t + h, separated by the interval h. The autocorrelation between these two quantities is characterized by die variogram function Gamma (h) which in turn is defined as the expecta tion of the random variable [c(t) - c(t + h)]2/2.(4) Figure 2 shows the experimental corresponding semivariogram, to which the following theoretical model has been fitted: Gamma (h) = 0.15 + 0.11 (1.5 h/40 - 0.5 (h/40)3), where h is the time lag (i.e. the number of shifts between the sequences being compared). Using this model, allowing for the autocorrelation between the shift averages, GSD of the distribution of the five-shift average concentration over 5 days (one measurement every 4 shifts) can be estimated by using the following formula (Equation 12,(4)): Tailgate of a Coal Face Shift No Figure 1. Respirable dust concentration. 1137 Strategies for Mine Dust Measurement (InGSD)2 = flnl.7)2 - 1/10 (4Gamma(4) + 3Gamma(8) + 2Gamma(12) + Gamma(l6)). This leads to GSD = 1.4, instead of GSD = exp(((ln 1.7)2/5)05) = 1.27 for the averages of five independent shift concentrations. Figure 2 shows that, for this example, the time interval over which there is autocorrelation between the concentrations, covers 40 shifts (i.e., 10 days). Note the pseudo-periodicity of the experimental semivariogram, with lower values at h = 8 and 12 shifts, shew ing a stronger autocorrelation between concentrations measured during die same shift at two or three days intervals. In this case, random sampling would be more appropriate. Other work situations do not present any autocorrelation. Figures 3 and 4 show, by way of illustration, die respirable dust concentration in a mechanized road heading area. The strategies were simulated for various values of die geometric mean with 1000 runs in each case. The results are illustrated in Figures 5 and 6 showing die operating characteristics ofdie monitoring strategies after a maximum monitoring duration of5 years. From Figure 5, it is seen that die workplaces where the limit is exceeded more than 40% of die time are detected very quickly, usually in less than 1138 5 years operation. Conversely, Figure 6 shows that die prob ability is very low for finding die workplace in compliance ifthe exposure exceeds die limit value during at least 40 % of the shifts (i.e., with a mean greater than the health standard). All this means that the compared strategies are extremely con servative. The probability of noncompliance if die mean is lower than die limit ("operator's risk' ' with all its associated engineering and administrative consequences) is much higher than die chance of finding die workplace in compliance ifdie mean is above the limit ("worker's risk"). This probability of escaping a citation being less than 5%. The worker's risk with the Directive on Lead is higher than with die Directive on Asbestos and it is near-zero with die draft modification ofdie Frame Directive which appears to be the safest strategy but with the highest operator's risk, of un justifiable expenses and labor problems. CONCLUSION As die design of monitoring strategies aims at keeping die "operator's" and "worker's" risks as low as possible, (0,05 being the goal for the last one) it may be concluded that die strategy provided in the Directive on Asbestos is die most ef ficient. Whilst this strategy allows for a reasonable level of risk for die worker's health, it imposes a lower sampling burden than die others. However the compliance outcome of these strategies is related to the fraction ofdays above the limit Road Heading Strategies for Mine Dust Measurement Figure 3. Respirable dust concentration. Road Heading Gamma (h) 1139 Strategiesfar Mine Dust Measurement PROBABILITY OF NONCOMPLIANCE (%) DAYS ABOVE THE LIMIT VALUE (%) Figure 5. Operating characteristics. O ao o Q Z Ld o ffl a o a: o_ --D-- LEAD 1140 DAYS ABOVE THE LIMIT VALUE (%) ----------1-- ASBESTOS ----- o -- CHEMICAL AGENTS Figure 6. Operating characteristics. value, whereas die chronic hazard is related to the average level ofexposure. Therefore, die Safety and Health Commis sion for the Mining and Other Extractive Industries of die European Community in its draft proposal on die worker's protection from the risk due to die fibrogenic mineral dust in die non-coal mines and quarries, allows die operator to choose between die asbestos monitoring scheme and die compensa tion method, this last one being more expensive but with lower operator risk. REFERENCES 1. OfficialJourr%alctftheEuropeanCommunities,L3Z7:Z-13((tt.l2.8)). 2. OfficialJovmaltftheEuropean Communities, L247:l2-2\(23.0%.%2). 3. OfficialJournalofdie European Communities, L 263:25-32 (24.09.83). Strategies for Mine Dust Measurement 4. Preat, B.: Application of Geostatistical Methods for Estimation ofthe Dispersion Variance ofOccupational Exposures. Am. bid. Hyg. Assoc. J. 48:877-884 (1987). 3. Official Journal ofthe European Communities, C 164:4-8 (02.07.86). 6. Rock, J.C.: A comparison between OSHA-compliance criteria and action-level decision criteria. Am. bid. Hyg. Assoc. J. 43:297-313 (1982). 7. Annbruster, L., Robock, K., Neulinger, G., Fleckner, K.-D., Freisewinkel, U., Heiderich, R.: Weiterentwicklung der Feinstaubmessanlage und Detiicbserfahnmgen im Steinkohlenbergbau. Ergebnisse von Untersuchungen aufdem Gebiet der Staub- und Silikosebekanpfimg im Steinkohlenbergbau, Band 16, pp. 67-78. Stemkohleabergbauverein, Essen, West Germany (1987). ACKNOWLEDGEMENTS: The author wishes to thank K. ROBOCK and L. ARMBRUSTER ofthe Bergbau-Forsdumg GmbH, HauptabteQung Staubbekampfung und Pneumokonioseverhiitung (Essen, West Germany) for the provision of the data used for the study of die autocorrelation. 1141 Strategiesfor Mine Dust Measurement ANALYSIS OF RESPIRABLE COAL MINE DOST SAMPLES BY INFRARED SPECTROSCOPY P. PAROBECK S. Ainsworth T. Tomb Mine Safety and Health Administration, Pittsburgh, Pennsylvania, USA ABSTRACT To control the health hazard associated with quartz in the United States coal mining industry. Federal regula tion requires that whenever the quartz content of die respirable dust in die coal mine environment exceeds five percent, the applicable respirable dust standard be reduced. This regulation, which is applicable for both surface and underground mining operations, has been in force since the promulgation ofthe Federal Coal Mine Health and Safety Act in 1969. To enforce this regulation, the Mine Safety and Health Administration (MSHA) analyzes approximately 6,000 respirable coal mine dust samples per year for quartz content. The quartz con tent of these samples is determined using an infrared spectroscopic method. This paper presents an overview ofMSHA's quartz enforcement program, the analytical method used for quartz determination and efforts underway to enhance the sensitivity of the method. INTRODUCTION At the time ofpromulgation ofthe Coal Mine Health and Safe ty Act of 1969, the Congress ofthe United States ofAmerica stipulated that a limit be placed on die allowable quantity of quartz to which miners would be exposed. The requirement for this limit was based on work performed in die 1930's and 1940's1*2 which showed that the presence of quartz increased the health hazard associated with exposure to coal dust. Based on this data, the United States Bureau of Mines in 1948 established a dust exposure limit when the (hist in the environ ment was found to contain more than five percent quartz. The limit at that time, determined by multiplying the dust parti cle concentration by the percent quartz, was not to exceed five million particles per cubic foot of air. When the United States Congress promulgated foe Coal Mine Health and Safety Act in 1969, they directed that a formula be developed for lowering die applicable respirable dust stan dard when the quartz content ofthe dust to which miners are exposed is greater than five percent. Such a formula was developed on March 10, 1971, and was included in Parts 70.101,71.101 and 90.101 ofTitle 30 ofdie Code ofFederal Regulations. According to the formula, the applicable dust standard (mg/m3) is determined by dividing the percent quartz into the number 10 (i.e., standard = 10/% quartz). This formula was continued under die Federal Mine Safety and Health Act of 1977 which amended the 1969 Act. To enforce this standard, die quartz content ofrespirable coal mine dust samples is determined by the use ofinfrared absorp tion spectrophotometry. Three common methods available for the analysis of crystalline silica are die use ofX-ray diffrac tion, visible absorption spectrophotometry and infrared ab sorption spectrophotometry. The advantage of using die in 1142 frared method over the others for die analysis of coal mine dust samples is that die sensitivity is greater than either that of the X-ray Diffraction Method3 or the visible absorption spectrophotometric method, also referred to as die Talvitie method.4'5 In addition, the X-ray method, although able to differentiate between different forms of free silica (quartz, cristobalite and tridymite), is affected by several compounds which have diffraction peaks that interfere with die majorpeak for quartz. The Talvitie method, which cannot distinguish be tween the crystalline forms, requires extensive sample preparation using various corrosive adds and is a time con suming procedure. Since cristobalite, tridymite and amorphous silica, all ofwhich would cause an interference in the infrared analysis for quartz, have not been detected in coal mine dust,3 the infrared method is ideal for die determination of quartz in coal mine dust samples. From 1970 through 1980, quartz analysis was conducted by MSHA using a high temperature ashing (800C) technique and die subsequent pelletizing of the ash with potassium bromide (KBr). This procedure required a sample mass ofone to four milligrams, thus requiring the compositing (combining) ofa number of samples from various coal mine operations to obtain a sample of suffident weight for analysis.6 In 1981, the method was upgraded to the current method which is known as die Low Temperature Ashing (LTA) Method. This LTA method allows for the analysis of individual coal mine dust samples containing from 0.5 to 2.5 mg ofdust. The method, developed by die Bureau ofMines, has been ruggedized and evaluated.3 From 1970 through December of 1985, enforcement of die quartz standard in die United States was determined solely from die analysis of a single sample or composited samples collected by Federal mine inspectors. In December of 1985, MSHA's enforcement policy was revised so that die quartz content of die dust in the environment is based on a number of samples (up to three) collected over a period of several months and includes samples collected by the coal mine operators. QUARTZ ENFORCEMENT PROCEDURE The rudiments of MSHA's current quartz enforcement pro gram require the analysis ofselected respirable dust samples collected by mine inspectors during die approval or verifica tion ofmine operators' dust control plans. Samples typically selected for analysis are those collected on the designated oc cupation (DO), that occupation in an underground mining operation that has die highest respirable dust exposure, die roof bolting (RB) operation in underground mining and designated work positions (DWP) in die surface coal mining industry. If die analysis of any of these samples shows that die quartz content is in excess of five percent, the mine operator is notified of die option of collecting a sample for analysis on the mine entity representative ofthe original sam ple which was in excess offivepercent quartz. Ifthe difference between the quartz percentage of the operator's sample and die quartz percentage of die MSHA sample is within plus or minus two percent, (e.g., MSHA value seven percent, operator value five to nine percent) die results of the two analyses are averaged and the respirable dust standard is set accordingly. Ifthe quartz determination ofthe operator's op tional sample differs from die MSHA sample by more than plus or minus two percent, the operator is given the option of collecting a second sample on die mine entity. Following analysis ofdie operator's second sample, die average quartz percentage is determined from the three samples (MSHA sam ple plus die two samples submitted by the operator). If die operator elects not to collect a sample or if die samples sub mitted have insufficient dust for analysis (less than 0.5 milligrams), the standard is adjusted based on die analysis of die MSHA sample. At six month intervals, any entity on a reduced standard is automatically reevaluated by analyzing for quartz one of die mine operator's samples submitted for dust compliance, provided there is sufficient weight gain on die sample. Analysis of MSHA inspector samples, mine operator optional samples and six month operator samples ac counts for the analysis of approximately 6,000 samples per year. ANALYTICAL METHOD Analysis of respirable coal mine dust samples for quartz is conducted in a central laboratory located in Pittsburgh, Penn sylvania. The operation of this laboratory is a function of MSHA's Pittsburgh Health Technology Center. The analytical method used for die analysis employs die princi ple ofinfrared spectrophotometry. The current LTA method allows for die analysis ofdie quartz content ofa sample with a mass of0.5 milligrams or greater. The method has a detec tion limit of 10 micrograms of quartz and a precision of 13 to 22 percent for quartz masses ranging from 25 to 160 micrograms.3 Samples are collected with approved respirable coal mine dust sampling assemblies equipped with a quartz-free, ashable Strategies for Mine Dust Measurement filter medium. Following weighing of the filter to determine sample mass, die filter medium is ashed in a low-temperature ashing system. This ashing system, shown in Figure 1, operates at a temperature ofapproximately 120C and utilizes radio frequency energy to generate an oxygen plasma which destroys die filter matrix and die carbonaceous material pres ent in the sample. Following ashing, isopropyl alcohol is added to the residue. The residue is dispersed in the alcohol using an ultrasonic generator. The suspension is filtered onto one halfofa Gelman DM-450 vinyl metricel filter. The filtering is accomplished by washing foe sample through a specially constructed, glass filter funnel on a vacuum manifold system, shown on Figure 2. The funnel is designed to produce a 10 millimeter diameter deposit. Once filtration is complete, foe filters containing foe ashed deposits are dried on a slide warmer for approximate ly 20 minutes at a temperature of approximately 42C. Analysis for quartz is then conducted using a dispersive in frared spectrophotometer. The DM-450 filter halfcontaining the ashed residue is mounted in a sample holder and placed in the sample beam ofthe infrared spectrophotometer. A blank DM-450 filter half which has been treated with alcohol and dried is similarly mounted in foe reference beam offoe instru ment. Following appropriate parameter setting offoe infrared instrument, foe sample is scanned in the absorbance mode from 1,000 to 710 cm-1. Quartz absorbs infrared energy in foe 800 cm-1 region. The clay mineral kaolinite, which is also found in coal mine dust, also absorbs infrared energy in this region.3 Its presence causes a slight overestimation offoe quartz content. To correct for this overestimation, the absorp tion for kaolinite is measured at 915 cm-1. Thus, measuring the absorbance ofinfrared energy by the sample from 1,000 to 710 cm-1 allows for foe quantification of kaolinite at 915 cm-1 and foe correction for its interference with quartz ab sorbance at 800 cm-1. Figure 3 shows a sample of an infrared scan of a typical coal mine dust sample. As illustrated on foe figure, foe peak in tensities at 915 and 800 cm'1 are determined by measuring the height from established baselines to the peak maximums. The baselines are drawn from 950 to 890 cm-1 for foe 915 cm*1 kaolinite band and from 810 to 760 cm'1 for the 800 cm-1 quartz-kaolinite band. The measured net peak heights are converted into absorbance units and the interference due to kaolinite is determined from a calibration curve ofkaolinite absorbance at 915 cm'1 versus kaolinite absorbance at 800 cm'1. The calculated absorbance for kaolinite at 800 cm'1 is subtracted from the measured absorbance at 800 cm-1 (quartz-kaolinite) to give the absorbance due to quartz. The amount of quartz is determined from a calibration curve of absorbance of quartz at 800 cm'1 versus mass of quartz. For MSHA's quartz enforcement program, -5 pm Minusil, a commercial product offoe Pennsylvania Glass Sand Com pany, Berkeley Springs, West Virginia, is used as foe quartz standard.7 Kaolinite used for foe standard is Hydrite UF, supplied by the Georgia Kaolin Company, Elizabeth, New Jersey. Once foe analysis and calculations are completed, foe percent quartz in foe coal mine dust sample is computed by using the 1143 Strategies for Mine Dust Measurement following equation: quartz (,percent,,) =-------------m--a--s-s--o--f- q-u--a--r-t-z--(-o-g-)------------- x 100 mass of coal mine respirable dust (og) The percent quartz determined for a coal mine dust sample is truncated to die whole percent value which is subsequent ly used in die formula for die determination of die reduced coal mine dust standard when die quartz percentage is in ex cess of five percent (reduced standard -- 10/% quartz). To insure the integrity of die quartz analyses performed, MSHA conducts an internal quality assurance program. This program consists of die analysis of three specially prepared samples, containing varying known quantities ofquartz, with each group of20 to 25 samples analyzed. The mass ofquartz on each ofthe three quality control samples is unknown to the analysts. These samples undergo die same processing as die coal mine dust samples; i.e., ashing, deposition by filtration and IR scanning. The analysis of these samples is used to verify dial die process is controlled, assuring the reliability of analytical results. IMPROVEMENTS TO ANALYTICAL CAPABILITY In an effort to improve the sensitivity ofthe present analytical technique, MSHA recently acquired a Fourier transform in frared spectrophotometer (FTIR). A KllK operates different ly than a dispersive infrared spectrophotometer. A FTIR employs an interferometer to obtain information about the transmission of infrared energy of all wavelengths (simultaneously) emitted by die source and passing through die sample, whereas a dispersive spectrophotometer uses a monochromotor and slit system to divide die infrared radia tion into frequency elements. The interferometer ofthe FI'lK contains a fixed mirror and a moving mirror, die position of which is determined by a helium-neon laser. The information obtained from a sample is digitally stored as signal intensify versus mirror displacement as shown in Figure 4. This is known as an interferogram. The instrument's computer then performs a Fourier transform ofthe interferogram to produce the desired absorbance versus frequency (in wave numbers) spectrum as shown in Figure 5. The b ilk has many advantages over die dispersive instru ment. Since there are no entrance or exit slits in die FTIR, a greater amount ofenergy reaches the detector, resulting in increased sensitivity. The laser tracking of die moving mir ror results in greater precision of the wavelength measure ment, permitting multiple scans to be averaged and thereby increase the signal to noise ratio ofthe absorbance spectrum. Precise duplication ofthe analytical frequencies and computer control of the calculations with the FTIR reduce die errors associated with die electromechanical components ofthe dis persive instruments and the necessary manual measurements of frequencies and peak intensities. 1144 Figure 1. Low-Temperature ashing system. Strategies for Mine Dust Measurement Figure 2. Vacuum filtration of ashed samples. 1145 Strategiesfor Mine Dust Measurement The current quartz analysis procedure employing dispersive IR is used to detect from 25 to 250 micrograms ofquartz for coal dust sample masses ranging from 0.5 to 2.5 milligrams. With the Fl'IK, it is anticipatedthat 10 micrograms ofquartz will be detectable in coal dust samples with as little mass as 0.2 milligrams. This factor is ofconsiderable importance since many respirable coal mine dust samples obtained in die sur face coal mining industry are of low mass, yet have greater than five percent ofquartz. This system should allow for the analysis of such samples. The computerization of the data handling will, likewise, fliifnfnatg output and eliminate tedious and redundant taglre which are currently performed manually ABSORBANCE FREQUENCY, cm-1 Figure 3. Infrared scan on dispersive IR ofa coal mine dust sample showing manually drawn baselines and peak locations. SUMMARY The United States Congress realized the hazard associated with coal miners' exposure to quartz and, when issuing the Coal Mine Health and Safety Act of 1969 and die subsequent Coal Mine Safety and Health Act of 1977, stipulated that ex posure to quartz be controlled. Exposure to quartz is con trolled by reducing the applicable dust standard when the dust is found to contain quartz levels in excess of five percent. To determine the quartz content ofrespirable coal mine dust, MSHA utilizes an infrared spectrophotometer to measure the absorbance ofinfrared energy by quartz in a dust sample. This analysis is conducted following die destruction of die com bined sample and filter matrix by a low temperature ashing process and subsequent filter redeposition ofthe ash contain ing the quartz. Since the mineral kaolinite interferes with the quartz determination, a correction is made to die result obtained. In order to automate the processing of samples and obtain a lower level of detection, the use of a FTIR to analyze coal 1146 MIRROR DISPLACEMENT Figure 4. Interferogram of a pure quartz sample. Strategiesfor Mine Dust Measurement Figure 5. Absorbance spectrum from FT1R of a coal mine dust sample with kaolinite and quartz-kaolinite frequencies indicated. mine dust for quartz content is being investigated. It is an ticipated that a quartz mass of 10 micrograms will be detec table in respirable dust samples with masses as small as 0.2 milligrams. REFERENCES 1. Bloomfield, JJ., etal: Anthraco-Silicosis Among Hard Coal Mines. US Public Health Service Bulletin 221, pp 118 (1935). 2. Flinn, R.H., Siefeil, J.E., Brinton, HE, Jones,J.L., Franke, R.W.: Soft Coal Miners Health and Working Environment. USPublic Health Service Bulletin 270 (1941). 3. Anderson, C.C.: Collaborative Tests of Two Methods for Determin ing Free Silica in Airborne Dust. DHHS (NIOSH) Publication No. 83-124, Contract No. 210-79-0059 (1983). 4. Talvitie,N.A.:DetenninafionofQuammPresenceofSilicatesUsing Phosphoric Acid. Analytical Chemistry 23:623 (1951). 5. Talvitie,N.A.,Hyslop,F.:CokjrimetricDetennjnatioaofSiliceousAtmospberic Contaminants. Am. Ind. Hyg. Assoc. J. 19:54-58 (1958). 6. Goldberg, S.A., Raymond, L.D., Taylor, C.D.: Bureau ofMines Pro cedures for Analysis ofRespirable Dust from Coal Mines. Am. Ind. Hyg. Assoc. J. 34:200-205 (1973). 7. Kacsmar, P.M., Tomb, T.F.: Comparison ofAlpha Quartz Materials Used as Calibration Standards. Coal Mine Dust Conference Proceedings, Morgantown (1984). 1147 Strategies for Mine Dust Measurement EFFECT OF THE MEASURING STRATEGY ON THE DETERMINATION OF THE RESPIRABLE DUST CONCENTRATION IN THE BREATHABLE AIR AT UNDERGROUND WORKPLACES BERNARD PRINZ, Dipl.-Ing. TH R. Stolz Ruhrkohle AG, Essen, FRG INTRODUCTION The dust conditions at die underground workplaces are per manently discussed worldwide. These discussions are always based on die absolute values ofthe respirable dust concentra tions in die breathable air indicated in mg/m3 (Figure 1). They provide information on die effective limits for the per missible respirable dust concentrations and on the results of the statutory (hist measurements. These values are also used to reflect the state ofdie pneumoconiosis prevention and dust suppression in die various hardcoal mining countries. In the following it is intended to show that the respective in dicated absolute values are not suitable for a comparison of die dust load of miners in different countries, die reason for this being that the rules for die determination of the values, i.e. die measuring strategies, are not included in the discus sions. The existing different measuring strategies have, however, significant effects on die magnitude ofthe measured absolute values. Thus, there is no uniform basis for an objec tive comparison. The measuring results are significandy in fluenced by die following parameters of die measuring strategy: die position of the measuring point; die time required for an individual measurement; and the frequency of die measurements. Measuring Strategy in Different Countries According to the measuring strategy effective in the Federal Republic of Germany (FRG) (Figure 2) since 1954 the respirable dust concentration has to be measured at the loca tion of a working area at which die maximum dust concen tration has to be expected. In this context it is generally as sumed that in working faces this location is situated, seen in ventilation direction, at the face end respectively the end of the working area. Measurements are taken once a month under normal operating conditions. The measurement period cor responds to the time the miners stay at their workplaces. Over a period of five years each the preset limits for the dust ex position ofthe miners must not be exceeded. Higher individual shift values which have to be compensated over the 5 year period are, however, permissible. The measuring strategy in the FRG is based on the following considerations: 1. For die people employed in the environment of the measuring point, the measuring result is sufficiently accurate. 1148 2. The impact ofdust on people employed on the intake side upstream die measuring point is overrated by the "high risk method.'' 3. By overrating higher urgency is attributed to die measures for a prevention of dust impact on the employees. 4. One monthly measurement over a 5 year period is suf ficient as the dust load ofeach miner is determined with sufficient accuracy by 60 measurements in five years. In Great Britain (GB) measuring values ofa fixed measuring point located in the return airway approx. 70 m behind die free are used to assess die dust conditions at the workplaces in die face. The fact that it is only there that die measuring results are no longer influenced by the coarse dust or the unequal distribution ofthe respirable dust in die air is given as a reason for the choice of this location. The partial sedimentation of the dust between the face and die measuring point is con sidered by correction factors. Measurements are taken in monthly intervals. As in the FRG the time of a measurement corresponds to the time the miners stay at the workplace. The number of measurements in one month depends on die size ofthe measured individual fine dust concentration. At values < 15 mg/m3 one measurement per month is sufficient. At values > 8 mg/m3 die average has to be calculated from up to five subsequent measurements in one week. The main point in the measuring strategy ofthe Soviet Union (USSR) is die monitoring of dust suppression in the free. When cutting coal with shearers die air-borne dust concen tration with grain sizes of up to 74 /un without preseparator is determined directly behind die shearer,when ploughing the coal it is determined at the face end. The strategy for these dust measurements has die main objective to improve die ef ficiency of dust suppression measures. The measuring time per measurement amounts to a few minutes during the coalget ting process. The measurement is repeated in monthly inter vals if the measuring result shows a value ^ 10 mg/m3. At higher values the measurement is repeated directly after im proving dust suppression. The measuring strategy in die United States ofAmerica pro vides for a measurement ofthe acting respirable dust concen tration by means of "personal dust samplers" directly at the employee. For the measurement the person exposed to die highest dust load in die face may be chosen as representative for all employees ofone face. This measuring method called "designated occupation" is also based on the "high risk" Strategies for Mine Dust Measurement Talarabla Dost Caacaatratioa U/a3 BUHfl 8.0 Oast Fraction raspirabla Qaartz Kalaatiaa jas SAAB 4.0 raspirabls aa USA 2.0 raspirabla Tas GB 5.0 raspirabla a a USSB 2.0 iabalabla rat Li.il ....... .. D.:t i. Cl Mi..: Figure 1. Figure 2. process as the measuring strategy in the FRG. In faces the workplace of the shearer operator is mainly chosen as "designated occupation." This is justified if nobody is employed behind the shearer for more than two hours during the shift. The respirable dust concentration at the workplace or the measuring point is assessed in two-monthly intervals by the averege of5 measurements taken in 5 subsequent production shifts. Ifthe limit is exceeded additional measurements have to be carried out in the following production shifts and a new average has to be calculated from 5 subsequent measuring values. The measuring series is interrupted if one average reaches or remains below the limit. The measuring time of each individual measurement corresponds to the shift length, i.e. working time plus travelling time. The effect ofthe measuring strategy on the size ofthe measur ing values can be illustrated by means of an example for respirable dust measurements in 10 faces of Ruhrkohle AG (Figure 3). In each ofthese faces several measuring points were installed in regular intervals. The respirable dust concentration was measured over a longer time period in the first production shift ofeach day with the miners at their workplaces. In die diagram the monthly averages ofthe respirable dust concentrations in mg/m3 are listed on the ordinate, the face length in % on the abscissa. At the face entry, i.e. at face meter "0", the respirable dust concentration in die intake air of the face was listed. In nine of the ten faces the respirable dust concentrations in crease in different magnitudes towards die face end where they reached their highest values. The different increase is gov erned by work sequence, machine type, support, ventilation volume, ventilation velocity, etc. In one case--i.e. face 8--in contrast, the initial concentration is already so high that the sedimentation over the face length is higher than the concen tration increase caused by the coalgetting operations. According to the measuring strategy ofthe USA all employees in the face would be exposed to the dust concentrations ofthe intake air flow, i.e., the initial values ofthe graphs on the ex treme left of the diagram, under the prerequisite that: 1. The shearer operator stands on die intake air side in front of the machine; 2. The dust produced by coalgetting is blown away from the site of the shearer operator; and 3. Bypasses the chock fitters. A possible slight increase in the dust concentrations in die in take air flow towards the workplace of the operator by tur bulences is negligible in this approach. Applying the German measuring strategy these velues are con trasted by the concentrations of the fixed measuring points which in contrast to the USA are, however, located at the face end. This means (Figure 4) that in a comparison ofthe values up to 9 times higher values have to be assigned to the employees due to the German measuring strategy with the measuring point at the face end compared to die American strategy. Even in case of a subdivision of the face into two monitoring sections with measuring points in the center and at the end of the face up to 6 times higher values are still calculated for the employees in die lower face section accord ing to the German measuring strategy. On average the con centrations at the face entry and the face end differ by the 1149 Strategies for Mine Dust Measurement factor 3.9 and the concentrations at the face entry and the face center still by the factor 2.4. The determination of the measuring values in Great Britain is again significantly different from the determination ofthe measuring values in the USA and the FRG. On the one hand additional dust sources between the face and the measuring point are registered, cm die other die measuring result is cor rected by a factor for sedimentation which was developed specifically for British mines. It may, however, hardly be ap plied worldwide. CONCLUSION From die mentioned comparison it may be derived that both the dust limits and the absolute respirable dust concentration figures cannot be referred to in a comparative representation of the dust conditions in different countries. Also the dust suppression measures applied in the different countries have to be seen under this aspect. The measuring strategies of the FRG, GB and the USSR call for measures reducing the dust concentration in the entire return air sec tion. In the United States dust suppression may center on the intake air section up to die coalgetting machine (Figure 5). This becomes particularly clear in the "shearer clearer" pro cess. The dust produced by the coalgetting operations is kept away from die measuring point. Without doubt this process has the advantage of reducing die dust load for machine operators and chock fitters. Faca la. 0SA/FB6 Hud - Tail 0SA/FB6 Hial * Cntri 1 1 : 2.1 1 : K5 2 1 : 5,4 1 : 2.8 3 1 : 2.7 1 : 1.7 4 I : 4.3 1 : 2.0 5 1 : 6.0 1 : 3.1 6 I : 3.1 1 : 1.8 7 1 : 8.8 1 : G.O 8 1 : 0.7 1 : 0.8 9 1 : 2.3 1 : 1.9 10 1 : 3.4 1 : 2.3 1 3.9 S 1 : 2.4 Fastar ! Baspirafcla Caal Mias Oast UMNIIC ii Facts: USA/ FB6 Figure 4. Arbeits scbut* 1150 Figures. These comments were intended to show that: 1. The different measuring strategies will inevitably have to result indifferent limits; 2. The measuring values and limits determined by one measuring strategy can only be compared in its scope of validity; 3. Identical absolute values ofthe different countries do not describe also identical dust conditions or dust impact; 4. Limits provide for a statement on the pneumoconiosis risk of die employees only in their scope of validity. The comments ofDr. Bauer (FRG) on the impact ofdifferent measuring devices, tyndallometer, cycloneseparator, horizon tal elutriator, on the result of respirable dust measurements underline die mentioned reservations against a comparison of measuring values and limits. These comments are not intended to be an assessing state Strategies for Mine Dust Measurement ment on the measuring strategies but are only meant to ex plain die fact that measuring values and limits cannot be com pared as long as they are based on different measuring strategies. The uncritical comparison ofmeasuring values and limits from different measuring strategy scopes involves two dangers: 1. That a race towards actually desirable but technically not feasible limits is started; and 2. That die statement on die pneumoconiosis risk ofa min ing region in relation to the dust impact is wrong ifrisk determinations are taken over from other measuring strategy scopes. For an international comparison ofthe dust load to which die miners in die hardcoal mines are exposed, it is thus required to use identical reference measuring equipment and to apply an identical reference measurement strategy. 1151 Strategies for Mine Dust Measurement RESPIRABLE DUST AND FREE SILICA VARIATION IN MINE ENVIRONMENTS THOMAS A. HALL,* M.H.S. Morton Com,* Ph.D. S. Zeger.t Ph.D. C.C. Law.t M.S. Division of Environmental Health Engineering, Department of Environmental Health Sciences, School of Hygiene and Public Health, Johns Hopkins University, 615 N. Wolfe Street, Baltimore, MD 21205, USA fDepartment of Biostatistics, School of Hygiene and Public Health, Johns Hopkins University, 615 N. Wolfe Street, Baltimore, MD 21205, USA Regulations promulgated and enforced by die Mine Safety and Health Administration (MSHA) require that coal mine operators control respirable mine dust to prescribed concen trations.5 Specifically, coal mine operators must regularly sample (bimonthly) respirable mine dust (RMD) in working mine sections. MSHA on a less frequent basis (once a year) also samples and evaluates RMD and its free silica (FS) con tent in working coal mines. Sampling is performed by MSHA and mine operators in order to: 1) establish permissible RMD levels in working mine sections when free silica is present; and 2) demonstrate compliance with permissible exposure limits prescribed in regulations. In non-coal mines the sam pling frequency is less well defined. Since passage ofthe Coal Mine Safety and Health Act of 1969 and die Mine Safety and Health Act of 1977, enormous resources have been focused on controlling RMD in mines with highly satisfactory results. The vast majority ofU.S. coal and non-coal mines consistently meet die appropriate RMD Permissible Exposure Limits (see Formula 1) promulgated by MSHA.12 hi the past five years, more inspector RMD samples have been analyzed for free silica. This has occurred because the analytical technique MSHA uses for the detection offree silica in coal mine dust has been refined and improved resulting in lower detection limits. The use ofthe "improved" analytical technique has suggested to many that MSHA has placed in creased emphasis on enforcement ofdie coal mine respirable dust (containing free silica) standard. The standard for respirable dusts containing free silica used in coal mines in vokes a "sliding scale" to determine the allowable RMD con centration. For % FS concentrations >5, Formula 1 is used to calculate the permissible RMD concentrations in coal mines. RMD, mg/M3 = 10 (for % FS >5) % FS (1) The purpose ofthis investigation was to gain insight into the extent ofFS variation in RMD samples collected from a sam ple of U.S. coal and non-coal mines. A second goal of this study was to determine the factors (mining operation variables, Ac.) associated with this variation. Specifically, the 1152 following questions were addressed: How large is the sampling and laboratory error in measurements of respirable mine dust concentration (RMD), free silica (FS) and percent free silica (% FS)? Is the sampling and laboratory variability different in per sonal and machine samples, across occupations or mines? Is exposure to RMD and FS systematically different across occupations or mines? How large is the temporal variability in RMD, FS and % FS? Thirteen mines, seven coal and six non-coal mines initially offered opportunities for dust sampling in this study. Of die thirteen mines originally volunteering for the investigation, ten (six coal and four non-coal) provided samples for analysis. Each participating mine was required to collect six air samples per day for five consecutive days. The six daily air samples were divided among five occupations with one miner wear ing two samplers (paired sample). A total of374 personal and area samples were collected in die participating mines dur ing 55 sampling days. Mine dust technicians from the participating companies were used to collect the air samples. Before these individuals were allowed to take part in die investigation they had to participate in a workshop presented by the study authors. Additionally, each participating mine was subjected to a site visit during foe sample collection period to insure that foe prescribed tech niques for sample collection were being used. After collec tion all dust samples were forwarded to, and analyzed by an independent, accredited laboratory. Results of laboratory analyses of samples were transmitted to JHU for statistical analyses and interpretation of results. All samples were collected using Mine Safety and Health Ad ministration (MSHA) prescribed procedures with some minor modifications. The samples were analyzed at two commer cial laboratories for respirable dust and free silica using foe P7 analysis routine. The onsite dust technician or industrial hygienist responsible for sampling completed a standardized questionnaire. Data from questionnaires were analyzed by JHU investigators, as were foe analytical results of dust samples. LITERATURE REVIEW The Mine Safety and Health Administration requires that coal mine operators conduct extensive sampling for respirable mine dust and airborne free silica. The goal of this sampling is to measure progress toward achieving promulgated dust standards and thus reduce die occurrence ofpulmonary disease among the mining population. MSHA's strategy for control ling exposure to pneumoconiosis-producing dusts employs a sampling scheme which utilizes a worst-case scenario. Although there is extensive scientific and technical literature which addresses the variability of measured mine dust con centrations resulting from the dust sampling process, few studies have sought to define die variability associated with sampling for respirable dust and its free silica content in mine environments. Factors affecting variability of airborne free silica dust, such as occupation, production rates, equipment operating time, and other mine and production variables have not been examined. The most widely publicized investigation of measured dust concentration variability is a GAO report to Congress.6 In this report, the GAO indicated that under certain conditions the error associated with respirable mine dust samples could be as great as 50%. An investigation by the Bureau ofStandards studied respirable mine dust sampling and analysis.8 While focusing specifical ly on sampling and analysis (gravimetric) for respirable mine dust, each step in the sampling process was examined, e.g. dust weighing, pump flow variation, etc. It was concluded that under tightly controlled conditions with a "well-trained" technician, die average standard deviation associated with die process was 0.39 mg/M3, or 19% (@ the 2 mg/M3 RMD concentration). In 1976, NIOSH found that in high risk mine sections (those which had been repeatedly found to be in violation of the 2 mg/M3 standard) the coefficient of variation for RMD measurements was 91.6%.9 In 1980, the National Research Council concluded that uncer tainties associated with spatial and temporal variation in RMD estimates from machine mounted samplers precluded this method for estimating personal exposures.10 In 1983, a literature review by investigators at the Johns Hopkins School ofHygiene and Public Health concluded that the factors responsible for the variation in RMD had not been quantitated for free silica and estimates of free silica were at least as unreliable as those of RMD.3 More specifically stated, "Because of the unavailability of data on free silica variation in coal mine respirable dust, the representativeness of a single sample analyzed for free silica can not be assessed. " The authors went on to state that die use ofa single air sample to determine free silica content of mine en vironments is meaningless. Page and Jankowski compared RMD measurements made us ing a real-time aerosol monitor (RAM-1) and a standard gravimetric sampler at a longwall mining operation.11 The authors reported ratios ofpaired RAM-Gravimetric sampler results, expressed as concentrations of RMD ratios of 0.41 Strategies for Mine Dust Measurement to 1.63. The authors attributed this variation to differences in the aerosol cloud being sampled, air flow velocity at the filter face and cyclone orientation. Burkhart, et al, in a presentation at die American Industrial Hygiene Conference in Dallas, Texas reported data from a limited number ofair samples collected from bituminous coal mines in West Virginia.2 The authors reported %FS concen trations ranging from 2 to 30% in five samples collected on five consecutive days. The samples reported were personal air samples collected on the operator of the continuous min ing machine. Die source of this variability was not discussed. Breslin, et al, in a Bureau ofMines Circular reported that for both personal and fixed-point (area) samplers the coefficient of variation for RMD was typically less than 20%.1 Kissel], et al, reviewed several factors thought to contribute to RMD and FS variability.7 The authors, while not specifically evaluating potential contributions to variation from mine sources, concluded that sampler position, geological variation in composition of coal, production fac tors such as deep or continuous cutting and failure to control known sources such as shuttle car loading, play an important role in RMD and FS sample results. PAIRED SAMPLE ANALYSIS AND RESULTS Sampling and laboratory variability for respirable mine dust, free silica and percent free silica were studied using 23 and 20 pairs ofdust samples from coal and non-coal mines, respec tively. Paired samples were defined for this study as two samples collected on the same occupation for the same time period and located not more than 14 inches apart. For this analysis, the ratios of the RMD and FS parameters were analyzed to determine variability. % FS was analyzed using the differences between the paired values. Figures 1-3 display the cumulative frequency distributions of RMD, FS and % FS, respectively. All three dust parameters exhibit large variability. Results ofthis analysis are presented in Table I and are brief ly summarized as follows: Coal Mines The respirable mine dust ratios (larger to smaller values) exceeded 1.5 in halfofdie paired samples and 2.5 in 10% of the pairs. For free silica, 50% of the pair ratios exceeded 1.52; 10% exceeded 5.7. For % free silica, the differences (larger minus smaller) exceeded 1.3% in half of the pairs and 5.6% in one out of ten. Non-Coal Mines The variability of respirable mine dust was somewhat less in non-coal mines with 50% ofthe samples having ratios greater than 1.13. 10% of samples demonstrated ratios of 6.19. (This was due to a few extreme outlier sample pairs.) 1153 Strategiesfor Mine Dust Measurement 5 10 RATIO OF FAIRED SAMPLES (large/small) Figure 1. Cumulative distribution of RMD sample pair ratios. 1.0 >- u z o.e ' L3U o UJ 0.6 . L>lI g 0 41- 3 2 3 0.2 O 0.0 0 5 10 15 20 DIFFERENCE OF FAIRED SAMPLES (large-small) Figure 3. Cumulative distribution of % FS sample pair differences. 0.0 I 2 3 4 5 6 78 RATIO OF FAIRED SAMPLES (large/small) Figure 2. Cumulative distribution ofFS sample pair ratios. For free silica, 50% ofthe respirable mine dust sample pair ratios exceeded 1.25 and 10% exceeded 2.0. Tile variability of % FS was slightly greater in non-coal mines. Hie differences in 50% of the samples were at least 1.7% free silica; 10% had differences equal to or greater than 7.7% free silica. EFFECT OF INCREASED NUMBER OF SAMPLES ON VARIABILITY OF DOST PARAMETERS The use ofpaired samples to measure variability in RMD, FS and % FS permits die prediction of variability reductions achievable by averaging increased numbers of samples. Figures 4 and 5 demonstrate die improvement in sample variability for the mean value of RMD and % FS. These figures were calculated from all paired samples and reflect the average variability improvement. The achievement ofa standard deviation of0.2 (mg/m3) for respirable mine dust in coal and non-coal mines would require right sample pairs. (Figure 4) Inboth coal and non-coal mines, a standard deviation of 1.5% free silica can be achieved with six sample pairs. (Figure 5) 1154 Strategies for Mine Dust Measurement Table I Selected Cumulative Percentages of Coal and Hardrock Mine Dust Parameters Coal Mine Hardrock Mine RMD FS %FS RMD FS %FS (ratio) (diff.) (ratio) (diff.) 50 percentile 80 percentile 90 percentile 95 percentile 100 percentile 1.50 1.96 2.50 3.33 3.50 1.52 2.20 5.72 6.56 8.00 1.31 3.47 5.58 10.98 19.23 1.13 1.25 1.50 6.19 13.00 1.25 1.60 2.00 2.67 3.00 1.67 5.05 7.69 9.21 18.06 N (PAIRS) Figure 4. The effect of increased sample pairs on the variability of RMD estimates. Figure 5. The effect of increased sample pairs on the variability of % FS estimates. CONTRIBUTION OF STUDY PARAMETERS TO VARIATION l inear regression analysis was used to determine the contribu tion to sample variability associated with study variables, i.e. production rate, sampler location, etc. The results demon strate that for: Coal Mines Sampler location was an important contribution to die demonstrated variability. Machine-mounted samples showed an improvement in variability for all measured parameters. The improvement in variability for machinemounted samples when compared with personal samples was 40%, 20% and 5% for RMD, FS and % FS, respec tively. The improvement in % free silica variability associated with machine mounted samples was not statistically significant. Sample variability for respirable mine dust, free silica, and % free silica did not appear to be related to occupa tional category. Respirable mine dust exposure variability across mines was greater than within mine variability for occupation categories. Respirable mine dust, free silica and % free 1155 Strategies for Mine Dust Measurement silica are more dependent on production and/or dust con trol within die mine than on occupational category. Ex posure to free silica demonstrated a consistent pattern, regardless of die respirable mine dust concentration in die mine. Roofbolters were exposed to respirable mine dust levels containing 2-3% more free silica than con tinuous miner or standard shuttle car operators, and ap proximately 5% more free silica than center or offside shuttle car operators. Non-Coal Mines Regression analysis ofnon-coal mine results could not be per formed because of differences in mining methods employed by participants. These differences did not permit comparison of data between mines. COMPARISONS OF EXPOSURES BY OCCUPATION The analytical results ofall dust samples woe used to address the question ofwhether dust (RMD, FS and % FS) exposure differs across occupations within a mine, and whether there are differences in dust exposure within occupations across mines. Because only coal mines have uniform job descriptions we have focused our analysis on coal mines. The geometric mean exposure and geometric standard deviation by occupation for coal mines are presented in Table II. Figures 6-8 display the mean exposures for the three variables RMD, FS, and % FS by job classification: mine operator, bolter (double boom), shuttle car operator-standard and shuttle car operator-center Table ff Results of Air Sampling Analyses by Occupation and Mine for Coal Mines Mine Occupation RMD FS ID N Code1 (GM2, mg/M3) G.S.D.3 (GM2, mg/M3) %FS G.S.D.3 (Mean) S4 28 15 6 8 1 3 4 6 1.67 1.61 2.52 1.97 1.58 1.31 1.86 3.98 0.045 0.117 0.054 0.044 1.41 3.13 2.97 1.45 2.95 7.66 1.93 2.92 1.45 2.08 0.89 0.88 37 12 11 1 3 4 1.42 1.12 0.80 2.03 1.74 1.76 0.039 0.048 0.017 5.90 1.69 2.45 3.73 6.41 2.78 1.88 3.15 2.34 58 12 3 9 1 3 4 6 0.767 0.620 0.252 0.268 1.60 1.09 1.92 2.36 0.023 0.033 0.006 0.010 2.03 2.03 1.32 1.77 4.05 5.56 2.83 4.42 2.84 1.34 1.59 3.82 77 15 5 5 1 3 4 6 0.580 0.612 0.261 0.267 2.46 2.13 1.44 2.39 0.121 0.090 0.050 0.030 3.11 3.39 1.48 4.56 27.7 23.1 21.5 14.9 25.7 18.7 12.1 8.81 85 12 6 1 3 4 0.474 0.975 0.824 1.38 1.98 2.30 0.010 0.047 0.011 1.83 1.85 3.71 2.24 5.04 2.00 0.917 1.59 2.06 10 10 10 4 31 32 34 1.07 1.21 1.09 1.70 1.49 4.22 0.014 0.36 0.016 2.30 2.55 1.52 1.96 4.34 7.92 2.19 3.68 12.8 1. 1 -- Continuous Mine Operator, 3 - Roof Bolter, 4 - Shuttle Car Operator (Standard), and Shuttle Car Operator (center and off-side). 2. GM -- Geometric Mean 3. GSD -- Geometric Standard Deviation. 4. S -- Standard 1156 2.0 IO 1.0 0.5 MINE Z MINE 3 MINE 5 MINE 7 MINE 8 Strategies for Mine Dust Measurement RMD, m g/M MINE OPERATOR BOLTER SHUTTLE CAR DOUBLE BOOM STANDARD JOB SHUTTLE CAR CTR. OPERATOR OR OFFSIDE RMD Associated with Selected Jobs in Coal Mines, Expressed as Geometric Means (ng/M-*). Figure 6. Comparison of measured geometric mean exposures by occupation for RMD. 0.5 K> 0.05 CT> E co u. 0.005 MINE 7 MINE Z MINE 3 MINE 5 MINE 8 0.001 MINE OPERATOR BOLTER SHUTTLE CAR DOUBLE BOOM STANDARD JOB SHUTTLE CAR CTR. OPERATOR OR OFFSIDE Free Silica Associated with Selected Jobs in Coal Mines, Expressed as Geoaetric Means (mg/M^). Figure 7. Comparison of measured geometric mean exposures by occupation for FS. 1157 Strategies for Mine Dust Measurement JOB OR OFFSIDE ZFS Associated with Selected Jobs in Coal Mines, Expressed As Geometric Means. Figure 8. Comparison of measured geometric mean exposures by occupation for % FS. or off side for coal mines (Mines 2, 3, 5, 7 and 8). Figures 6-8 demonstrate that dust exposure variability across mines is greater than foe variability associated with occupations within a mine. RMD, FS and % FS levels are more depen dent on the production and/or control ofdust within the mine than on occupation. % FS is more consistent than RMD or FS across all mines ex cept mine 7 where the occupation-specific % FS averages range from IS to 20% FS over die four occupations. This is three to four times as high as in the other mines. The occupa tional exposure to free silica does have a consistent pattern regardless ofRMD concentrations in a mine. Bolters are ex posed on average to RMD containing 2 to 3 % more free silica than continuous miner and standard shuttle car operators, and about 5 % more than center oroffside shuttle caroperators who are exposed to the lowest % FS. TEMPORAL COMPONENT OF VARIATION The results of the previous sections have been used in com bination with the published precision ofour laboratory pro cedures to characterize the contributions to variability in dust parameters of: laboratory analysis; sampling; time; occupa tion and mines. The analytical lab component is the variance among repeated lab analyses for the same sample. The sum of die laboratory and sampling variances for this investiga tion was estimated from foe paired samples study. The sum 1158 of all components as well as foe contributions of occupation and mine have been estimated by ANOVA. By combining die results for foe ANOVA and foe paired sample analysis foe variability over time for a given occupation and mine can be estimated. Table HI summarizes the contributions to variability from each source for RMD, FS and % FS for coal mines in absolute units and as a percent of total. For RMD die total variance across foe 157 samples was 0.76. The analysis contributes 1%; sampling contributes 20%; variability over time for foe same occupation and mine con tributes 33 %; while variability across mines/occupations add ed foe largest fraction, 46%. The relative contributions for free silica are similar to those for RMD. For % FS, analysis again contributed little to variability although foe specific amount could not be determined from foe literature. The sampling and analysis together contributed 35 % to foe total variation; temporal variability was approximately 30% offoe total; while variation from mines/occupations was 36%. GENERAL CONCLUSIONS Occupation and mine, sampler position, laboratory analysis nH repeated sampling time contributions to sam ple variance were estimated based on the paired sample results and published values for variance associated with Strategies for Mine Dust Measurement Table m Decomposition of Variance for RMD, FS and % FS by Components: (1) Occupation and Mine; (2) Time; (3) Sampling; and (4) Laboratory for Coal Mine Data RMD Source Occupation and Mine Time Sampling Laboratory Total Variance .35 FS % of Total 46 Variance .76 %FS % of Total 49 Variance 2.8 % of Total 36 .25 33 .42 27 .15 20 .38 24 .01 1 .01 0 .76 100 1.57 100 2.3 T~ 2.7 J- 7.8 29 1 35 _1_ 100 laboratory analysis and air sampling techniques. The largest contributions to variance arose from sampling across mines and occupations, which accounted for 46% of die variability associated with respirable mine dust samples. The second important contributor to variance was die temporal variability of dust levels in mines, accounting for approximately 33% of total variability. In summary, this investigation demonstrates that die largest contribution to variability results from sampling across mines. REFERENCES 1. Breslin, J.A.,Page, SJ., Jankowski,R.A.: PredsionofPersonalSampUng ofRespirable Dust in CoalMines. Bureau ofMines RI8740, (1983). 2. Burkhart, J. Presented at tbe Annual Amer. bid. Hyg. Conference in Dallas, TX. (1986). 3. Com, M.,Breysse, P.N., Hall, T.,Cheo, G.,Risby,T., Swift, D.L.: A Crftique ofMSHA PmceAirwt for I^trrrrrirwtinn nf tVrrnigqhfo final Dust Containing Free Silica. Am. bid. Hyg. Assoc. J. 46(l):4-8 (1985). 4. Federal Register, 30 CFR, Part 57. 5. Federal Register, 30 CFR, Parts 70 and 71. 6. General Accounting Office: Improvements Still Needed in Coal Mine Dust Sampling Program andPenallyAssessmentand Collection. Report RED-76-56. December 31, (1975). 7. Kissel,F.N.,Rugg>er,S.K.,Jankowski,R.A.:HowtolinprovetheAc- curacy of Coal Mine Dust Sampling. Am. buL Hyg. Assoc. J. 47(10):602-606 (1986). 8. National Bureau of Standards. "An Evaluation ofthe Accuracy ofdie Coal Mine Dust Sampling Program Administeredby the Department of biterior, " Final Report to tbe Senate Committee on Labor and Public Welfare. Washington D.C., December, (1975). 9. National Institute for Occupational Safety and Health. CDC. Statistical Analysis of(Mining Enforcement SafetyAdministration) RespirableDust Data. (1976). 10. Ns^aasdRestsasotCcnDfal. Measurementtmd ControlofRespinMe Dust in Mines, (1980). 11. Page, S., Jankowski, R.: Correlations Between Measurements with RAM-1 and Gravimetric Samplers on LongwaD Shearer Faces. Am. bid. Hyg. Assoc. J. 45(9):610-616 (1984). 12. Watts, W. F., Parker, D. R.: Respirable Dust Levels in Coal, Metal, and Nonmetal Mines. Bureau cfMines biformation Circular 49125, U.S. Department of Interior, Washington D.C. (1987). 1159