Journal of Loss Prevention in the Process Industries 22 (2009) 528–532

Contents lists available at ScienceDirect

Journal of Loss Prevention in the Process Industries

journal homepage: www.elsevier.com/locate/jlp

A statistical analysis of coal mine accidents caused by coal dust

explosions in China
Yuan-Pan Zheng a, *, Chang-Gen Feng a, Guo-Xun Jing b, Xin-Ming Qian a, Xin-Juan Li c, Zhen-Yi. Liu a,
Ping Huang a
a
State Key Laboratory of Explosion Science and Technology, Beijing Institute of Technology, Beijing 100081, China
b
School of Safety Science and Engineering, Henan Polytechnic University, Jiaozuo 454000, China
c
International Office of Henan Polytechnic University, Henan Polytechnic University, Jiaozuo 454000, China

a r t i c l e i n f o a b s t r a c t

Article history: One hundred and six coal dust explosion accidents that occurred in China between the years of 1949 and
Received 1 September 2008 2007 were investigated through statistical methods so as to review the overall situation and provide
Received in revised form quantitative information on coal dust explosions. Statistical characteristics about accident-related factors
12 February 2009
such as space, time, volatile ratio of coal dust, ignition sources, and accident categories were analyzed.
Accepted 12 February 2009
Some special conclusions have been achieved. For example, most dust explosions were found to have
occurred only in certain provinces and during certain months of the year. Such results may be helpful to
Keywords:
prevent dust explosions in coal mines. Moreover, comments were made on SMS (safety management
Coal mine
Coal dust explosion system) and safety culture. In conclusion, countermeasures were proposed in accordance with the results
Accident of statistical studies, including the change of safety check time.
Statistics Crown Copyright Ó 2009 Published by Elsevier Ltd. All rights reserved.
Safety
Countermeasure

1. Introduction Zlochower, 2000), factors that influence explosions, and methods to

mitigate explosions (Amyotte, Pegg, Khan, Nifuku, & Tan, 2007).
The coal industry is of great importance to worldwide energy Unfortunately, few efforts involving macroscopically statistical study
supply, especially to that of China. However, there are always some on coal dust explosions have been made. Only a few scientists have
safety problems to be solved urgently, such as coal dust explosion recently focused on the statistical features of such accidents, e.g.
accidents. These accidents do harm to both coal production itself study on coal dust particle size survey of US mines (Sapko, Cashdollar, &
and miners’ lives and health. On November 27th, 2005, Dongfeng Green, 2007). We believe that safety issues of any kind should be of
Coal Mine, one of China’s national key mines, suffered a major coal systematic problems and so should the coal dust explosion accidents
dust explosion, which claimed the lives of 171 workers. What’s that occurred in coal mines. Technical strength of natural science
more, this accident was the 21st in which more than 100 itself can’t solve safety issues alone without the cooperation of safety
employees’ lives were lost (Wang & Li, 2001). management system (SMS) and safety culture.
Coal dust explosions are not only featured by the severe A good knowledge of statistical features of certain accidents,
consequences, but also by the high occurrence probability for therefore, is the basic requirement to implement SMS. In other
mining enterprises. In China, 87.32% of the 532 national key mines words, major hazard identification must be performed. Statistical
face the risk of coal dust explosion, and the proportion becomes study on accident cases would be a powerful tool to meet this
larger for town-owned and private coal mines. Scientists all over requirement. For this purpose, we collected all the related coal dust
the world have already paid much attention to coal dust explosions explosion accidents that occurred in China from the years of 1949–
and obtained some knowledge about it, such as explosion-related 2007. The research focused on statistical features such as space,
characteristics (Cashdollar, 1996; Cashdollar, Weiss, Montgomery, & time and influencing factors.
Going, 2007; Continillo, Crescitelli, Fumo, Napolitano, & Russo,
1991; Liu, Sun, & Chen, 2007; Sapko, Weiss, Cashdollar, & 2. Statistical features of space

Space features were analyzed from two levels: the provincial

* Corresponding author. Tel.: þ86 1068949594. level and municipal level (a lower regional jurisdiction next to
E-mail address: [email protected] (Y.-P. Zheng). province in China).

0950-4230/$ – see front matter Crown Copyright Ó 2009 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jlp.2009.02.010
 Y.-P. Zheng et al. / Journal of Loss Prevention in the Process Industries 22 (2009) 528–532 529

Table 1 explosions even in the same province. For example, statistical

Statistical features of accidents at the provincial level. features of explosion accidents in different cities of Shandong
Provinces Timesa Casualtiesa Timesb Casualtiesb Timesc Casualtiesc Province are shown in Table 2.
Sichuan 1 124 2 15 3 139 In Henan Province, almost all the coal dust explosion accidents
Jiangxi 1 16 1 6 2 22 happened in Pingdingshan City, and methane participated in nearly
Heilongjiang 4 68 8 310 12 378 all the coal dust explosions. Whereas, in Heilongjiang Province,
Jilin 4 167 1 79 5 246
most of the explosion accidents happened without methane and
Hebei 6 155 4 52 10 207
Shanxi 7 490 4 761 11 1251 half of them took place in Qitaihe City.
Liaoning 4 55 3 43 7 98
Shandong 2 81 17 493 19 574
3. Statistical features of time
Inner Mongolia 4 156 2 20 6 176
Guizhou 2 246 1 3 3 249
Shaanxi 2 149 0 0 2 149 Time statistical analyses were given at three levels: year, month
Jiangsu 2 122 6 187 8 309 and hour levels. Time statistical features are shown in Fig. 2 in
Hunan 2 79 1 35 3 114 terms of statistics of accident cases: Fig. 2(a) – year level, Fig. 2(b) –
SinKiang 1 12 3 33 4 45
Henan 7 594 1 62 8 656
month level, and Fig. 2(c) – hour level. Note that y axis in Fig. 2
Total 49 2514 54 2099 103 4613 represents occurrence times of accidents.
Note: Superscripts a, b and c refer to explosion with methane, without methane and
the sum the two types, respectively. 4. Statistical features of influencing factors for
coal dust explosion
2.1. Statistical features of space at the provincial level
Influencing factors for coal dust explosion include quite exten-
In total, there are 15 provinces in China suffering from coal dust sive contents such as concentration of oxygen and granularity of
explosion accidents. Statistical data are shown in Table 1. Coal dust coal dust. However, influencing factors in our study just referred to
explosions include two types: explosion with methane and explo- volatile ratio and ignition sources because of the lack of enough
sion without methane. Both accident occurrence times and casu- related statistical data.
alties were listed for the two kinds of explosions. Risk of a coal dust
explosion can be assessed by comparing accident occurrence times 4.1. Statistical features of volatile ratio
or casualties. In Fig. 1, the most risky provinces for coal dust
explosions in China were more clearly shown from the point view The volatile ratio is a value established by the former United
of both accident occurrence times and accidental casualties. States Bureau of Mines to evaluate the explosibility of coals based
on large-scale tests in the Experimental Coal Mine. To calculate the
2.2. Statistical features of space at the municipal level volatile ratio, a proximate analysis must be performed in the
laboratory on a sample of coal. This analysis determines the volatile
Statistical characteristics of coal dust explosions at the provin- matter and fixed carbon quantities of coal along with moisture and
cial level are provided in Section 2.1, but those are not enough to ash. The volatile ratio is defined as the volatile matter divided by
constitute space-oriented major hazard identification. More the summation of volatile matter and fixed carbon of coal.
detailed information for each province is also needed, because According to case statistics, the maximum volatile ratio in the
different cities have different conditions that may cause coal dust past accidents was 55.59% and the minimum was 17.85%. Existing

Fig. 1. Risk chart at the provincial level.

530 Y.-P. Zheng et al. / Journal of Loss Prevention in the Process Industries 22 (2009) 528–532

Table 2 into five bands according to safety management convention in

Statistical features of accidents at the municipal level for Shandong Province. China. Band I, Band II, Band III, Band IV and Band V indicate that
Cities Timesd Casualtiesd Timese Casualtiese Timesf Casualtiesf there are 1–3, 3–10, 10–30, 30–100 and more than 100 casualties in
Xinwen 1 48 2 197 3 245 an accident, respectively. Specific statistical features related to
Zaozhuang 1 33 9 168 10 201 casualty scale are shown in Table 5.
Linyi 0 0 2 44 2 44
Yanzhou 0 0 1 2 1 2
Other 0 0 0 0 0 0 5.2. Statistical features of accident categories
Total 2 81 14 411 16 492

Note: Superscripts d, e and f refer to explosion with methane, without methane and Accident categories referred to the general reasons leading to
the sum the two types, respectively. certain accidents. There were three accident categories according to
the past accident records: accidents caused by breach of duty, by
natural hazards and intended demolishment. Concrete statistical
volatile ratios were classified into five groups so as to clarify
features of accident categories are presented in Table 6.
statistical characteristics. The range of each group is less than 10%,
as shown in Table 3.
6. Analysis and discussion

4.2. Statistical features of ignition sources 1. According to the space analysis at the provincial level (Fig. 1),
Shanxi, Henan, and Shandong Provinces risk the heaviest
Ignition sources in past accidents were divided into the casualties caused by coal dust explosions. However, if accident
following five types: open flames, electrical sparks, flame of blast- occurrence times were considered, Shandong, Heilongjiang
ing, accidental explosions of initiators & pyrotechnics, and sparks and Shanxi Provinces would face the greatest risk. Zaozhuang
caused by friction or strike. Open flames referred to fire of matches, City and Pingdingshan City were the most vulnerable in
cigarette smoking and coal self-ignition, etc. Accidental explosions Shandong and Henan Provinces from the perspective of space
of initiators & pyrotechnics meant explosions of detonators and so analysis at the municipal level. Accident cases also showed that
on. Detailed statistical features of ignition sources are shown in mining face and heading face were the two sites where most
Table 4. coal dust explosions took place when more microcosmic space
levels were considered.
5. Other statistical features 2. As shown in the time analysis at the year level in Fig. 2(a), coal
dust explosion accidents peaked in 1958, from 1970 to 1976,
5.1. Statistical features of casualty scales of accidents around 1980, from 1992 to 1993, and from 1999 to 2005. This
macroscopical trend is consistent with that of the whole
Casualty scales in this study referred to the death toll of industry safety situation of China from 1949 to 2007(Jing &
employees in a certain accident. Casualty scales are usually divided Zheng, 2005). Coal dust explosion accidents ascend sharply

Fig. 2. Statistical features of time.

 Y.-P. Zheng et al. / Journal of Loss Prevention in the Process Industries 22 (2009) 528–532 531

Table 3 Table 5
Statistical features of volatile ratios. Statistical features of casualty scales.

Volatile ratios/% [17.85, 20] [28.93, 29.50] [30, 39.64] [40, 48.85] [55, 55.59] Casualty scales V IV III II I
Occurrence times 9 3 39 16 3 Occurrence times 12 36 35 12 6
Proportions/% 12.86 4.29 55.71 22.86 4.29 Casualty number 2146 1754 637 67 9
Proportion of times/% 11.88 35.64 34.65 11.88 5.94
Proportion of casualties/% 46.52 38.02 13.81 1.45 0.2

from November of the previous year to February of the

following year at the monthly level in Fig. 2(b). On the contrary,
Table 6
accidents descend rapidly from March to September. This
Statistical features of accident categories.
phenomenon may have something to do with the seasonal
change of coal output and moisture content of the air, but the Accident Occurrence Proportion of Casualty Proportion of
categories times times/% number casualties/%
authors believe output change plays a much more important
Breach of duty 43 80.63 1977 94.18
role than does moisture content change. Generally, ground
Intended 2 3.7 76 3.62
moisture content of the air in winter is relatively lower than demolishment
that in other seasons, and lower moisture content is good for Natural hazard 9 16.67 46 2.19
coal dust explosibility. However, water spray is the most
common method for dust control in underground coal mines.
Zhou, & Li, 2001). So prevention of accumulation and ignition of
They are installed to suppress dust at all major sources
methane would also be very important to prevent coal dust
including cutting, loading, transporting and support advancing.
explosions.
Water spray can greatly wet the underground air flow besides
5. Accidents of Band IV and Band V took a proportion of 47.52%
suppressing dust. Geothermal energy and groundwater
from the point of view of occurrence times of accidents. And
seepage are also factors helpful to keep the underground air
the total proportion would reach 82.17% if occurrence times of
humid even in winter. Coal is the first key energy in China and
accidents of Band III were included. Detailed information is
occupies more than 70% in energy consumption. What’s more,
presented in Table 5. Statistical data suggested that conse-
more than 60% of coal output is used to generate electricity. In
quences of most coal dust explosion accidents are disastrous:
winter, more coal output is needed because waterpower
a large number of lives would be lost once a coal dust explosion
decreases sharply in China while the country’s heating demand
happens.
reaches a peak. Many coal companies usually tend to increase
6. According to accident categories, more than 80% of accidents
output regardless of safety issues. There are four spans when
resulted from breach of duty. Thus, most coal dust explosion
most accidents occur in a single day in Fig. 2(c): 8–9 a.m., 11–
accidents could not be seen as inevitable. It would be
12 a.m., 16–17 p.m. and 23–24 p.m. Note that three of the four
a propitious time to strengthen safety management and staff
spans correspond with shift time in most coal mines in China.
training.
3. More than 55% of coal dust explosions had their volatile ratios
in between 30% and 40%; about 82.86% in between 30% and
56%. When volatile ratio was lower than 17%, there was no
7. Conclusions
explosion caused by coal dust. National regulation of China for
coal dust prescribes that coal dust is prone to explode when its
Coal dust explosion accidents are not rare in China and have
volatile ratio is higher than 10%. This regulation can be
caused great loss to both China’s international image and coal
considered a little bit conservative from the perspective of this
enterprises, especially to coal miners’ lives and health. However,
statistical analysis.
such accidents are not inevitable because most of them are induced
4. Flame of blasting was the first key ignition source, and nearly
by breach of duty. Fortunately, some statistical conclusions useful
half of the explosions were induced by it. The second important
for preventing dust explosions have been presented in our statis-
ignition source was electrical spark, and more than one third of
tical research.
explosions were caused by it. These may result from the fact
Because casualties caused by coal dust explosions are highest in
that a large number of coal mines in China usually use blasting
Shanxi, Henan and Shandong, much more importance should be
methods rather than automatic machineries to excavate coal.
attached to these three provinces regarding the prevention of
From the point view of methane, methane is easier to be
explosions. The whole safety situation related to coal dust explo-
ignited than coal dust under the same condition (Cashdollar,
sions would get much better once coal dust accidents in these
1996). Statistical data showed that methane was usually
provinces were reduced. Similarly, the whole safety condition in
involved in big coal dust explosion accidents. Obviously,
Shandong and Henan Provinces would change for the better if coal
methane itself can be part of the fuel in the whole explosion of
dust explosions in Zaozhuang City and Pingdingshan City were
coal dust. What’s worse, flames of methane may become
under effective control.
powerful ignition sources for coal dust clouds, and the
Special safety checks for preventing coal dust explosions should
minimum explosive concentration (MEC) of coal dust decreases
be practiced in the previous year’s last 2 months and the following
sharply with the increase of methane content in the air (Hu,
year’s first 2 months. Meanwhile, check procedures should be
performed especially strictly during these 4 months. Likewise, it is
recommended that daily safety patrols of coal mines be arranged
Table 4
around shift times.
Statistical features of ignition sources.
Collieries should make efforts to seek new safe explosives and
Flame of Electrical Initiators & Sparks of Open pyrotechnics or guarantee that every step of blasting accords with
blasting sparks pyrotechnics strike flames
corresponding operation regulations. The fundamental method to
Occurrence times 43 31 6 4 11 get rid of blasting flames is adopting machinery mining as much as
Proportions/% 45.26 32.63 6.32 4.21 11.58
possible.
532 Y.-P. Zheng et al. / Journal of Loss Prevention in the Process Industries 22 (2009) 528–532

Strengthening safety management and avoiding breach of duty Acknowledgement

as much as possible would be effective countermeasures to reduce
coal dust explosions. After all, over 94% of casualties in dust acci- The authors gratefully acknowledge the financial support from
dents were caused by breach of duty. the National Natural Science Foundation of China (Grant No.
To sum up, SMS and safety culture badly need to be put into 50674041).
practice to prevent coal mines from being hit by coal dust explo-
sions. SMS (including management of technical affairs) is hardware
References
of safety guarantee, and safety culture is software. The two coun-
termeasures should work together rather than be used separately. Amyotte, P. R., Pegg, M. J., Khan, F. I., Nifuku, M., & Tan, Y. X. (2007). Moderation of
Unfortunately, there is a clear fact that the average production dust explosions. Journal of Loss Prevention in the Process Industries, 20(4–6),
technique and safety consciousness of employees in coal mines of 675–687.
Cashdollar, K. L. (1996). Coal dust explosibility. Journal of Loss Prevention in the
China is not high enough to meet the requirements of work safety. Process Industries, 9(1), 65–76.
This actuality counteracts many efforts aiming for work safety. Cashdollar, K. L., Weiss, E. S., Montgomery, T. G., & Going, J. E. (2007). Post-explosion
Many qualified job seekers refuse to work in coal mines mainly observations of experimental mine and laboratory coal dust explosions. Journal
of Loss Prevention in the Process Industries, 20(4–6), 607–615.
because coal mining is a high risk occupation in China. And those Continillo, G., Crescitelli, S., Fumo, E., Napolitano, F., & Russo, G. (1991). Coal dust
mentioned above make for a vicious circle: more accidents caused explosions in a spherical bomb. Journal of Loss Prevention in the Process Indus-
by unqualified employees will lead to a higher probability of tries, 4(4), 223–229.
Hu, S. P., Zhou, J. G., & Li, Z. Z. (2001). Characteristics and control of coal dust
qualified people refusing to engage in coal mining. To break explosion. Coal Mining Technology, 2001(1), 63–65.
through such a circle, besides emphasizing safety culture, high-tech Jing, G. X., & Zheng, Y. P. (2005). Application of system methods in safety
equipment and techniques should be strengthened. For instance, management of coal mine in China. In S. C. Li, Y. J. Wang, & P. Huang (Eds.),
Progress in safety science and technology (pp. 1837). Beijing: Science Press
intrinsic safety processes and facilities should be applied as much Beijing.
as possible so as to reduce the dependence of work safety on Liu, Y., Sun, J., & Chen, D. (2007). Flame propagation in hybrid mixture of coal dust
human behaviors to a minimum. The last point worth emphasizing and methane. Journal of Loss Prevention in the Process Industries, 20(4–6),
691–697.
is that coal mining enterprises must invest time, money and
Sapko, M. J., Cashdollar, K. L., & Green, G. M. (2007). Coal dust particle size survey of
employees in safety training and implement a system of rewards US mines. Journal of Loss Prevention in the Process Industries, 20(4–6), 616–620.
and punishments for safe behavior. In this way, more qualified Sapko, M. J., Weiss, E. S., Cashdollar, K. L., & Zlochower, I. A. (2000). Experimental
employees would enter the coal mining field and coal mine safety mine and laboratory dust explosion research at NIOSH. Journal of Loss Preven-
tion in the Process Industries, 13(3–5), 229–242.
would naturally step into a virtuous cycle with the safety conditions Wang, J., & Li, W. (2001). China’s coal mine accidents and comments of safety
of the coal industry improved. specialists. Beijing: Coal Industry Press.