Fatalities in Small Underground Coal Mines (1983)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

0) C^ r^ CO >•» 10 to to m 01 jQ 4* 16 • 1 01 C 3 C iH % 1 « O -H O S X JS H Ld 4J a. 0) U-l C -H 1 rH rH O*l iH in VO P" CN PO rH 1 ft O -H > § in co m v0 m o co • i CO CO CN Is* \O CN CO u in rH CT\ VO rH CN VO VO 1 CN rH rH rH f*^ '"I « O Q in _— _ «•« CN rH CO ^""1 rH "a. CN r> r» CN CN rH rH CN 1 1 in rH ^ f** Cft rH CO f) 3 ^^ rH CN CN <T •D a C f 3 O u O •H in CN CN rH PO CN Mh 1 /*, ** w1 ul u rH cu 1 in VO P* ^* PO rH ^* T3 rH VO PO ^" CN _., C in rH 5 3 OJ rH rH in rH ^, ^- (N 0 V V ^H V l*> rH rH r» -^ a ID c« w E I rH CN 01 rH *T rH rH ^ a> 0) paf oo CN oo m » 0) (0 <u ^| rH CN w jC O N JJ "I C u U 01 i rH rH O rH —, VO _., •a <o E V V H V rH rH ^* e o *- U4 «J C e "w^1 1 rH rH in fl O rH <r 0) 'H rH in CN vo 9 4J ^ rH f> CN CN <0 01 1^ JJ u O CO >0 01 0) -2 M nlJ rn" rH* ^. rH* -^ ^. -~ 10 Mid o V V T V >-l VO rH 0) 0) fjl 0) 0 C JJ N rH Is- rH ^ ^T CN in rH •^ (Q •H 1 r*"> rH CO rH x u W rH ^- rH CN CN >4-l 0) 0 N VI CO 1 2 10 3 -4 •H C Z TJ C -H C 10 O^ CN a 0) >i > 1T3 Lj 10 -H Jf rH -rl T< (0 E o o > c > a o> 01 US C 3 CO -H -: u 4J J3 -H jj o C en JJ CO Vi 10 10 rH C -H C U M < 0 U rH rH 0) JS 1U -H 0! .r- 0 W < IH u: o 04 > 2

Note that three states, Kentucky, Virginia, and West Virginia, have sizable numbers of their underground coal miners in small mines. In all other states, less than 5 percent of the underground coal mining workforce is employed in small mines. Thus the problem of small mines is concentrated in Kentucky, Virginia, and West Virginia, which in terms of both number of mines and employee-hours account for over 85 percent of the small (1-50 employee) coal mines in the United States. Size and Seam Thickness Table 3 gives a breakdown of mines by seam thickness and size. It is clear that, in general, smaller mines have thinner coal seams than do larger mines. However, within each size category there is no clear association between seam thickness and fatality rates. In particular, there appears to be no obvious relationship between seam thickness and the fatality rate in small mines. In contrast, within each thickness category the fatality rate tends to decline with increasing mine size. Thus the association between mine size and fatality rate is not due to differences between small and large mines with respect to seam thickness. Note also from Table 3 that a sizable proportion (accounting for 41 percent of the employee-hours) of the small mines have unspecified seam thickness, which means that their seam thicknesses were not reported on the MSHA quarterly reports. In contrast, mines with unspecified seam thickness account for only 9 percent of the employee-hours in mines with over 50 employees. Because of this large amount of missing information, it is not possible to give a more detailed analysis of seam thickness in small mines. Mining Method The MSHA computer files used in the Committee's earlier report did not provide information on mining method. However, MSHA inspection reports contain information on both the mining method and number of active sections (which can be regarded as a surrogate for the number of employees). Table 4 gives the joint distribution of mines by size and mining method. The data are for the period 1973-79. For example, between 1973 and 1979 there were 5,075 quarters worked in mines having two active sections. Of these mines, the relative proportion of conventional mining was 31 percent. The table shows that conventional mining is more common in mines with fewer active sections. In order for mining method to explain any sizable portion of the association between mine size and fatality rate, two conditions must hold. First, mining method must be correlated with fatality rate. Since over 90 percent of all mining in the United States is conventional or continuous mining, this means that the fatality rates corresponding to these methods must be substantially different. Second, the distribution of mining method must vary considerably across mine size. The data in Table 4 indicate that conventional mining is more common in small mines than in larger mines. However, based on other analyses (see

10 TABLE 3 Fatality Rates by Seam Thickness and Mine Size, 1975-80 (standard errors are in parentheses) Mine Size Seam Thickness 1-50 51-150 151-250 251+ Less than 48 in. No. of Mines 916 119 34 17 Hours (millions) 69.7 80.6 53.1 57.4 Rl .18 (.02) .09 (.02) .05 (.01) .05 ( .01) 49 in. -60 in. No. of Mines 197 57 37 37 Hours (millions) 21.6 48.4 66.0 140.6 *1 .10 (.03) .07 (.02) .09 (.02) .05 ( .01) 61 in. -72 in. No. of Mines 104 28 14 30 Hours (millions) 10.2 26.0 25.8 140.7 Rl .18 (.06) .09 (.03) .04 (.02) .05 ( .01) 73 in.-84in. No. of Mines 31 13 20 27 Hours (millions) 2.4 10.8 37.6 122.3 Rl .08 (.08) .13 (.05) .11 (.02) .05 ( .01) More than 84 in. No. of Mines 55 26 11 26 Hours (millions) 7.9 24.6 20.6 111.8 *1 .10 (.05) .20 (.04) .08 (.03) .05 ( .01) Unspecified seam thickness No. of Mines 2,228 81 17 12 Hours (millions) 77.8 42.5 19.9 38.0 *1 .14 (.02) .20 (.03) .07 (.03) .03 ( .01) p. 104 of our earlier report), there do not appear to be marked differences in the fatality rates of these two mining methods. Thus it does not appear as if mining method per se could explain any material amount of the elevated fatality rate in smaller mines. Mine Size and Age In Toward Safer Underground Coal Mines (pp. 100-104), we indicated that the age of a miner was strongly and negatively correlated with disabling injury rates, but that age did not appear to be correlated with fatality rates.

11 TABLE 4 Conventional Versus Continuous Mining Methods by Mine Size, 1973-79 Mine Size Number of Relative Proportion (number of sections) Mine Quarters of Conventional Mining* 1 23,560 .60 2 5,075 .31 3-4 4,188 .27 5-7 2,817 .20 8-10 1,472 .17 11 or more 846 .10 *The proportion is the number of sections using conventional mining divided by the number of sections using conventional or continuous mining. In our previous study we were able to obtain the then-current age distribution of miners in the major companies that operate underground coal mines by requesting those data directly from the coal companies. The 15 coal companies that responded represent approximately one half of the underground coal mining population of the United States. On the basis of this age data we were able to demonstrate that younger miners (age 18-24) have a disabling injury rate that is three times that of miners over 45. However, we found no correlation between age and fatality rates. This implies that even if the age distribution in small mines is different from that in large mines, it would not explain the association between mine size and fatality rates. Nevertheless, in trying to learn more about small mines we felt it important to have some indication of how age distribution varies with mine size. Data on age distribution were obtained from MSHA coal dust files for the years 1972-76 and are presented in Appendix A. We see from Tables A.I through A.5 that age distribution varies with time as shown in Figure A.I. In particular, the age of the mining population decreased during the period 1972-76, as indicated by the significant decrease in the percentage of miners 46 and older and the increase in the percentage of miners in the 18-35 age group. However, for each time period the age distribution in smaller mines is quite similar to that in larger mines. In conclusion, we find that the age distribution of miners in the small mines is not substantially different from the age distribution of all underground coal miners. Hence our previous finding with respect to the correlation of age and injury rate still applies, as does our recommendation for increased emphasis on the training of younger miners. Underreporting Whereas it is generally accepted that nearly all fatalities are reported to MSHA, hours of employment are not so completely reported. Hence, if

12 underreporting of employee-hours is more common in smaller mines, the computed fatality rate for smaller mines would tend to be inflated. If a mine reports an injury to MSHA in a quarter when it has not submitted an employment/production report, MSHA creates a dummy employment/production report for the mine in that quarter so that the injury will be accepted by the computer. Between 1978 and 1980 there were 10 fatal accidents for which dummy employment/production forms had to be created. If the fatality rate for mines that do not report hours is the same as that for other mines (i.e., R]_ = .07), then we would expect there to be 28.6 million employee-hours of unreported employment during this three-year period (because 28.6 million employee-hours and a fatality rate of .07 correspond to 10 fatalities). Since there were some 634 million employee-hours reported during this period, any underreporting would be too small to have any effect on the overall association between mine size and fatality rate. Of the 10 fatalities for which dummy employment/production reports were created, in the 6 that occurred in 1979-80 the mines had 11 or fewer employees. This suggests that nonreporting of employee-hours worked occurs mainly in very small mines. If so, the fatality rates in Table 1 for mines with 1-10 employees may be unreliable. For this reason it is preferable to consider the broader category of 1-20 employees as an entity, because there are sufficient employee-hours in this category that the fatality rate would not be severely biased by underreporting. Duration of Active Operation The Committee noted in its report (p. 84) that small mines appeared to remain in active operation for less time than do large mines. Table 5 examines this point by indicating the number of quarters of active operation by mine size for 1975-80. Note that over half (57 percent) of the small (1-50 employee) mines were operational for 6 or fewer of the 24 quarters during this six-year period. In contrast, only 4 percent of the larger mines were operational for so little time. This striking difference between small and large mines gives rise to several possible explanations for the association between mine size and fatality rates. Number of Quarters Opened Table 6 compares fatal accident rates in small (1-50 employee) mines according to the number of quarters open (i.e., with some production) during 1975-80.* Note that small mines open for one to six quarters have about twice the fatality rate of small mines open for more than six quarters. A closer inspection of the 1-6 category reveals that the fatality rates for mines open 1-2, 3-4, and 5-6 quarters are .51, .23, and .23, respectively. Thus the small mines *In this and subsequent tables we use fatal accident rates, in which accidents with multiple fatalities count as a single fatality in the calculation.

13 TABLE 5 Distribution of Mines by Size and Number of Quarters Operational, 1975-80 (column percentages are in parentheses) No. of Quarters Operational Mine Size Total No. in 1975-80 1-50 51-150 151-250 251+ of Mines 1-2 1,048 (30) 8 (2) 1 (1) 0 1,057 (26) 3-4 564 (16) 5 (2) 2 (2) 0 571 (14) 5-6 413 (12) 8 (2) 3 (2) 0 424 (10) 7-8 294 (8) 18 (6) 7 (5) 1 (1) 320 (8) 9-10 251 (7) 16 (5) 1 (1) 1 (1) 269 (7) 11-12 222 (6) 18 (6) 4 (3) 1 (1) 245 (6) 13-14 168 (5) 16 (5) 5 (4) 0 189 (5) 15-16 130 (4) 20 (6) 5 (4) 5 (3) 160 (4) 17-18 118 (3) 29 0) 5 (4) 1 (1) 153 (4) 19-20 87 (2) 32 (10) 9 (7) 5 (3) 133 (3) 21-22 89 (3) 26 (8) 10 (8) 8 (5) 133 (3) 23-24 148 (4) 126 (39) 81 (61) 126 (85) 481 (12) TOTAL 3,532 322 133 148 4,135 TABLE 6 Fatal Accident Rate in Small Mines (1-50 Employees), 1978-80, by Number of Quarters Open (standard errors are in parentheses) Fatal No. of Quar- Fatal Hours Accident ters Open Accidents (millions) Rate (Rl) 1-6 37 26.58 .28 ( .05) 7-12 23 43.77 .11 ( .02) 13-18 33 49.97 .13 ( .02) 19-24 44 70.25 .13 ( .02) that are open for very short periods have a greater fatality rate than do those open for longer periods. The same comparison was made separately for mines with 1-20 employees and 21-50 employees. For each size category the fatality rate was highest in mines open only one to six quarters. Note that the fatality rates in small mines that are open for more than six quarters are still higher than those in large mines. Thus the

14 elevated fatality rate in small mines open one to six quarters does not explain the higher overall rate of fatalities in small mines. However, it does indicate a subcategory of these mines where the fatality problem appears to be more acute. For small mines that operate intermittently, the length of a period of shutdown might have some influence on the safe operation of the mine, but shutdown time is not reported to MSHA. MSHA knows only whether the mine operated during a given quarter. Hence only intermittency of operation could be investigated, and, as we have seen, mines that open and close repeatedly tend to have a higher fatality rate. Number of Runs A related hypothesis is that small mines that open and close regularly are at higher risk for a fatal accident than are those that are open continuously. Unfortunately, our analyses are limited because we can measure continuity of operation only on a quarterly basis. Let a "run" be defined as a period of time during which a mine is open. Thus if a mine is open all of 1975, closed in 1976-79, open for the first quarter of 1980, and closed for the rest of 1980, it will have two runs. Table 7 gives the fatality rate by the number of quarters open and the number of runs. TABLE 7 Fatal Accident Rate by Quarters Open and Number of Runs in Small (1-50 Employee) Mines, 1975-80 (standard errors are in parentheses) No. of Quar- NO. of Runs ters Open 1 2 3 or More 1-6 .28 (.05) .27 (.10) .27 (.19) 7-12 .09 (.03) .12 (.05) .14 (.07) 13-18 .12 (.03) .10 (.04) .23 (.07) 19-24 .13 (.02) .09 (.03) .21 (.09) TOTAL .14 (.02) .12 (.02) .20 (.07) After controlling for the number of quarters open, there is an indication that small mines that open intermittently may be at greater risk of a fatality than small mines that do not open and close repeatedly. However, the overwhelming number of employee-hours (119.7 million) occur in mines with only one run. The intermittently operated small mines are the source of a relatively small percentage of the total employee-hours worked in small mines. This means that even if the fatal accident rate is higher in these intermittently open mines, they account for relatively few additional fatalities.

15 Declining Fatality Rates Another hypothesis that might explain the higher fatality rate in small mines is that the risk of a fatality is greatest shortly after a mine first opens and thereafter declines—i.e., that there is a learning curve that reduces the risk of fatalities with time. Such a curve, if it does exist, might take the form shown in Figure 1. < oc TIME SINCE OPENED FIGURE 1 A hypothetical learning curve relating fatality rate to time of operation. If this hypothesis were true, it would go a long way toward explaining the higher fatality rate in small mines, since these tend to stay open for less time than do larger mines. That is, it could be that a small mine and large mine that have been active over the same period have equal fatality rates, but that, overall, smaller mines have higher fatality rates simply because they remain open for less time than do larger mines. To explore this possibility, we computed changes in mine fatality rates over time, focusing only on small mines open continuously for seven or more quarters. Table 8 gives the fatality rates for different times since opening. There is no clear indication of a time trend in the fatality rate. Thus it does not appear that the higher fatality rate in small mines is due to a learning-curve phenomenon. TABLE 8 Time Trends in Small (1-50 Employee) Mines Open Continuously for More Than Seven Quarters in 1975-80 (standard errors are in parentheses) Time Since First Opened 0-6 7-12 13-18 19-24 25-48 49-96 Months Months Months Months Months Months Fatal Accident Rate .14 (.05) .08 (.03) .09 (.04) .09 (.04) .10 (.01) .19 (.01)

16 Summary The fatality rate in small mines (those with 1-50 employees) is higher among smaller mines within the 1-50 category. Specifically, the fatality rate in mines with 1-20 employees is considerably higher than that in mines with 21-50 employees. The possibility of underreporting of employee-hours in very small mines (1-10 employees) prevents any separate conclusions for this category. Small coal mines are concentrated in the states of Kentucky, Virginia, and West Virginia. In every other state, mines with 1-50 employees account for no more than 5 percent of the total employee-hours. The data on seam thickness, age of miner, and mining method indicate that none of these factors has any substantial impact on the relationship between mine size and fatality rates. Mines with fewer miners tend to have smaller coal seam thicknesses. However, there is no apparent association between seam thickness and fatalities. The age distribution of workers in small mines, estimated by indirect methods, does not appear to be substantially different from the age distribution in larger mines. There is a striking difference between small and large mines with respect to the length of time they remain open. The majority (57 percent) of small mines were active for six or fewer quarters in 1975-80, compared with only 4 percent of larger mines. Mines open for one to six quarters had a substantially higher fatality rate than did mines open for seven or more quarters. However, this phenomenon can explain little of the overall difference in fatality rates between large and small mines.