- Email: [email protected]
Evaluating the risk of coal bursts in underground coal mines
International Journal of Mining Science and Technology 26 (2016) 47–52
Contents lists available at ScienceDirect
International Journal of Mining Science and Technology journal homepage: www.elsevier.com/locate/ijmst
Evaluating the risk of coal bursts in underground coal mines Mark Christopher ⇑, Gauna Michael Pittsburgh Safety and Health Technology Center, Mine Safety and Health Administration (MSHA), Pittsburgh, PA 15236, USA
a r t i c l e
i n f o
Article history: Received 27 July 2015 Received in revised form 15 October 2015 Accepted 30 October 2015 Available online 15 December 2015 Keywords: Underground mining Coal Coal burst Mine safety Ground control
a b s t r a c t Coal bursts involve the sudden, violent ejection of coal or rock into the mine workings. They are almost always accompanied by a loud noise, like an explosion, and ground vibration. Bursts are a particular hazard for miners because they typically occur without warning. Despite decades of research, the sources and mechanics of these events are not well understood, and therefore they are difficult to predict and control. Experience has shown, however, that certain geologic and mining factors are associated with an increased likelihood of a coal burst. A coal burst risk assessment consists of evaluating the degree to which these risk factors are present, and then identifying appropriate control measures to mitigate the hazard. This paper summarizes the U.S. and international experience with coal bursts, and describes the known risk factors in detail. It includes a framework that can be used to guide the risk assessment process. Ó 2015 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
1. Background Coal bursts involve the sudden, violent ejection of coal or rock into the mine workings. They are almost always accompanied by a loud noise, like an explosion, and ground vibration. Bursts are a particular hazard for miners because they typically occur without warning. During the years 2012–2014, serious coal bursts occurred at three different U.S. room and pillar mines. These events resulted in three fatalities and two permanently disabling injuries. In all three instances, the events occurred during pillar recovery at depths exceeding 300 m. None of these three mines had previously reported a burst to MSHA. Coal bursts also occurred at three longwall mines during this same time period. Despite decades of research, the sources and mechanics of bursts are not well understood, and therefore these events are difficult to predict and control. Experience has shown, however, that certain risk factors are associated with an increased likelihood of a coal burst. A coal burst risk assessment consists of evaluating the degree to which these risk factors are present. In addition, some control techniques are effective in reducing the likelihood of an event or protecting miners from their effects. 2. Factors contributing to the risk of coal bursts The one universal characteristic of burst-prone environments is the presence of highly stressed coal. The overburden depth is ⇑ Corresponding author. Tel.: +1 412 3866522. E-mail address: [email protected] (C. Mark).
responsible for the overall level of stress, but pillar design or multiple seam interactions can concentrate stresses in distinct locations. Geology also plays a big role where strong roof and floor are characteristic of most, but not all, burst prone environments. Geologic features, including sandstone channels, faults, and seam dips, have been associated with the events. Certain mining layouts and practices also increase the burst risk, as does a past history of bursts. Each of these factors is discussed in more detail below. 2.1. Depth of cover Very few bursts have occurred at depths less than 300 m, although there were two incidents that occurred during pillar recovery under 230 m of cover during the early 1980s. Experience shows that the burst risk increases with depth. An analysis of the NIOSH Analysis of Retreat Mining Pillar Stability (ARMPS) database showed that for case histories where the depth of cover was less than 450 m, only 2% encountered bursts. For the handful of cases where the depth of cover exceeded 600 m, however, almost half encountered bursts (Fig. 1). Another study found that of 34 burst events that occurred in mines located in the North Fork Valley of Colorado, only three occurred where the overburden depth was less than 450 m, and 13 occurred at depths exceeding 600 m [1]. Consequently, the MSHA Handbook on Roof Control Plan Approval and Review Procedures includes the following statement: ‘‘pillar recovery at depths exceeding 600 m may not be appropriate due to the heightened risk of bursts at such unusual and extremely deep cover.”
http://dx.doi.org/10.1016/j.ijmst.2015.11.009 2095-2686/Ó 2015 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
48
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
Number of cases
100 Burst Squeeze collapse Success
75 50 25
60
120 180 240 300 360 420 480 540 Depth of cover (m)
600
660
Fig. 1. Distribution of pillar failures and pillar bursts with depth in the ARMPS database.
event killed a miner in 1996. At greater depths, interpanel barrier pillars have been used at several Utah longwalls [7]. In some cases, rather than leave a full barrier, mines have elected to make midpanel move around the area of deepest cover. The unmined panel provides a local interpanel barrier for the next panel [8]. The interpanel barrier effectively protects the tailgate corner from the influence of previous panels, but at greater depths the single-panel stresses on the longwall face reach the same levels as were present with abutment loads from adjacent extracted panels separated by yield pillars. After a fatal bump occurred on a longwall face near the headgate at 840 m depth of cover, one major Utah operator announced that it would consider reserves at depths exceeding 900 m to be unmineable [9].
2.2. Pillar design 2.3. Multiple seam interactions The U.S. underground coal mining industry has extensive regions where multiple seams have been mined. The interaction of the active mining with overlying and/or underlying old workings can generate stress concentrations. The severity of a multiple seam stress concentration typically depends on two factors: (1) The thickness of the interburden between the active seam and the previously-mined seam (or seams). In general, the thicker the interburden, the less likely that the interaction will result in a severe stress concentration. (2) The type of remnant structure present in the previous seam. Isolated remnants, with worked out areas on two or more sides, are the most hazardous. Remnant structures are typically created when coal is left in place adjacent to areas of full extraction. However, bursts have occurred above and beneath large remnants adjacent to smaller developed pillars [10,11]. In these cases, the smaller developed pillars apparently behaved as a pseudo gob area, transferring much of their load onto the larger pillar (Fig. 3). A burst risk assessment should take such situations into account particularly when inmine evidence suggests a stress concentration exists. As noted in Fig. 3, the large pillar was surrounded by a ‘‘pseudo gob area” consisting of smaller developed pillars that had apparently yielded and transferred load to the large pillar. Interactions between all previously mined seams should be considered in the assessment. The workings in several seams may overlap, creating very high stress zones, particularly if the interburdens separating the older workings from the active seam are thin (Fig. 4). Empirical or numerical computer models should be a part of a thorough burst risk assessment. Models such as the Analysis of Multiple Seam Stability (AMSS) or LaModel can identify potentially high stress zones due to multiple seam mining [12]. However,
(a) Conventional abutment-yield design
(b) Two-entry yield pillar design
Fig. 2. Pillar design approaches used for burst control.
Tailgate
Headgate
Tailgate
Headgate
Tailgate
Headgate
Longwall face
(c) Panel barrier design
Barrier pillar
A pillar that is properly designed and large enough to distribute the load that it carries is unlikely to be burst prone. On the other hand, a pillar that is sufficiently small and yielding, is also not burst-prone. The burst hazard is greatest for poorly designed pillars that are too small to properly distribute the loads they carry, but too large to yield. NIOSH studied the 17 largest burst events in room and pillar mines that occurred between 1980 and 2010. Each of these events resulted in extensive damage to at least several pillars. The analysis showed that 12 of the 17 multi-pillar bursts could be attributed to inadequate pillar design. These 12 events all occurred during pillar recovery mining. In nine instances, the barrier pillars were too small, were being extracted on retreat, or were not used at all. In five of the 12 cases, pillar splitting operations without a barrier pillar apparently triggered the multi-pillar burst. Barrier pillars are particularly important in room and pillar mining because they protect each new panel from the abutment loads arising from previously mined areas. The NIOSH ARMPS program was revised in 2010 specifically for evaluating production and barrier pillars in deep cover applications [2]. In longwall mining, several different pillar design strategies have been employed in burst prone conditions (Fig. 2). Conventional approach employs at least one large abutment pillar in a multi-entry system, sometimes flanked by small yield pillars. Experience has shown that appropriately sized abutment pillars can reduce the incidence of bursts [3,4]. In Utah, two-entry yield pillar layouts have been used since the 1960s, and they can virtually eliminate gate pillar bursts [5]. Mining engineers also learned to avoid critical pillars which are too large to yield non-violently yet too small to support large abutment loads. The width-toheight ratios of such burst-prone, critical pillars normally exceeded 4 or 5 [6]. While the yield pillar system typically performs well at depths up to approximately 600 m, it can concentrate the load on the tailgate corner of the longwall face, and this can result in severe face bursts near the tailgate corner of the longwall. A tailgate corner
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
49
While there is no established generic signature of burst-prone geology, the following are two examples developed from experience in different regions: Pirst Bump area 1600
Fig. 3. Location of coal burst that occurred in upper seam development workings (solid lines) located above a large pillar in a lower seam (dashed lines).
Fig. 4. Overlapping gob areas in two overlying seams (workings shown in blue and orange) probably concentrated unusually high loads on the two pillars that burst (outlined in red).
these programs have typically been used to identify potential pillar squeezes or roof instability, and may require special adjustments for burst risk evaluation. Specialized analysis techniques, such as Energy Release Rate (ERR) models, may be more useful in some applications [13]. Accurate identification of remnant structures requires reliable maps of the older workings. The burst risk assessment should also include an evaluation of the adequacy of the available maps. If mapping detail does not exist or is incomplete, then the likelihood of encountering unexpected remnants should be considered. Not all multiple seam mining increases the risk of coal bursts. The risk can actually be reduced when mining is conducted in de-stressed ground above or below an area that has been fully extracted [14]. 2.4. Geology Strong sandstone or siltstone roof and floor have been associated with coal bursts, particularly in the eastern U.S. and in Utah. In evaluating whether the roof or floor geology may contribute to the burst risk, it is important to consider: (1) The thickness of the strong sandstone or siltstone unit. Thicker units are more likely to be associated with bursts. (2) The distance between the strong unit and the coal seam. Strong units close to the seam pose the greatest risk of coal bursts. (3) The strength of the rock. Surrounding rock units associated with coal bursts typically have uniaxial compressive strengths of at least 70 MPa but can exceed 100 MPa. (4) The characteristics of the rock. Massive units with minimal bedding, jointing, or other discontinuities are more likely to be associated with coal bursts.
(1) A 4.5 m thick package of strong sandstone in the first 15 m above the mining horizon, or a 6 m thick package of strong sandstone within the first 9 m of the floor [15]. (2) A massive sandstone unit at least1.5 m thick is found within 1.3 m above the coal seam. Core logs, combined with rock mechanics testing, may be used to identify when such conditions might be present. However, since surface core holes are generally spaced too far apart to identify all zones of burst prone ground, supplemental underground test holes may be necessary [4]. Sandstone channels are suspected of creating high stress concentrations [16]. At one longwall mine in eastern Kentucky, numerous large bursts occurred beneath sandstone channels, but none were encountered when a channel was not present [4]. Because sandstone channels may be limited in extent, they may be particularly difficult to identify in widely spaced surface boreholes. Faults or fracture zones have been associated with increased burst risk in both coal and hard rock mines [17]. When mining approaches a highly-stressed fault or joint system, the ground may suddenly shift, releasing seismic energy that results in a burst [1]. Faults or joints may also partition the overburden, resulting in an unexpected concentration of overburden load [16]. The presence of steep seam dips has been observed at a number of burst sites, and rapid changes in the depth of cover due to steep topography have also been associated with bursts [18]. Coal strength is one factor that does not seem to play a significant role in the burst risk. Iannacchione and Zelanko noted that bursts have occurred in at least 25 different U.S. coalbeds, varying from strong, blocky seams to the very friable Pocahontas No. 3 and No. 4 seams [17]. Laboratory studies conducted by Babcock and Bickel showed that most coals can fail violently if they are highly stressed and the confinement is suddenly released [19]. Extensive German laboratory studies using large-scale specimens have also concluded that nearly all bituminous coals can burst. In these experiments, coal seams ranging in unconfined compressive strength from 5 to 50 MPa have all been shown to be burst-prone [15].
2.5. Mining layouts and practices Historically, more than 80% of bursts have been reported during retreat mining, with less than 20% occurring on development [17,20]. Retreat mining increases the likelihood of bursts because it concentrates abutment loads on the pillar line, gate pillars, or longwall face, and because caving overburden releases seismic energy as it breaks. Of the two widely used retreat mining methods, pillar recovery is significantly more burst prone than longwall mining. Wider panels are a factor that increases the burst risk during pillar recovery. A wider panel results in a greater front abutment load, and subjects the mining area to more overburden load. Modeling indicates that the ERR may increase approximately in proportion to the panel width (Figs. 5 and 6). As noted in Fig. 6, Curve ‘‘A” shows the ERR cut-by-cut in the critical entry for a panel with six pillars extracted, while curve ‘‘B” shows the ERR for a panel with four pillars extracted. Certain pillar recovery practices can further increase the burst risk. In particular, many bursts have occurred during barrier pillar extraction, a practice that was once wide-spread but is now seldom employed [21]. Pillar extraction by split-and-fender, or by any
50
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
Fig. 5. LaModel grids used in the Energy Release Rate (ERR) analysis.
Energy release rate
1.0 A
0.8
In another example, the Brody Mine in southern West Virginia had never experienced a coal burst prior to May of 2014. Then a strong event occurred during pillar recovery, which knocked the continuous miner operator to the ground and covered him with debris from his mid-thigh down, entrapping him for less than five minutes. Three days later, while mining the next row of pillars, an even larger event cost the lives of two coal miners. Similar precursors occurred prior to the fatal burst at the Huff Creek Mine and the major event in the North Barrier Pillar at the Crandall Canyon Mine. When a mine has experienced bursts, future mining operations with similar geology and mining methods should be considered high risk. Underground observations and monitoring are critical components of a burst risk management program. Mining crews should be trained to observe coal burst warning signs, particularly the occurrence of small bursts, which are often the best indication that an area is becoming more burst prone. A record-keeping system should be maintained and management processes developed to ensure that warning signs receive appropriate responses.
0.6
3. Conducting the risk assessment
0.4 B
0.2
Operational techniques used by longwall mines include reducing the depth of the web, reducing the speed of the shearer, unidirectional cutting, and avoiding double cuts at the gate ends [22].
Based on the evaluation of the burst risk factors listed above, an overall risk level can be assigned to future mining areas. For example, those areas with negligible burst risk might be considered green zones, areas with slightly higher risk could be yellow zones, and the areas of greatest risk might be orange zones. The initial risk assessment should be conducted before an area is developed, using available borehole logs and maps of previous mining in overlying and underlying seams. During development, underground test hole drilling may be employed to provide more detailed information on the geologic conditions. Underground mapping should be conducted prior to any retreat mining, particularly where the mapping can use rib conditions to identify the locations of significant multiple seam interactions. At each step, as new data becomes available, burst potential zones should be reevaluated and updated. The matrices shown in Tables 1 and 2 may be used to assist with the risk assessment. Each of the significant known risk factors can be rated as low, moderate, or high. Note that it is the level of the factor that is being rated, not the burst risk associated with it. The ratings of all the factors should be considered when assessing overall burst risk. The matrices are intended as generic guides that can be tailored for each site-specific burst assessment. A universal quantitative rating scale has not been developed for the factors listed in the matrix, and there is no standard method to combine the individual factor ratings into an overall burst risk rating. The assessment should clearly state the assumptions made in the process, including any ratings and weightings of individual factors, and the procedure used to estimate the overall burst risk. An experienced ground control professional can use the matrix to evaluate the overall burst risk.
2.6. A past history of bursts
4. Mitigating the risk of coal bursts
Major bursts have often been preceded by a pattern of increasing coal burst activity. Whyatt identified seven ‘‘clusters” consisting of three or more reported bursts in the 10-year period 1999– 2008 [22]. He found that: ‘‘. . . the largest (cluster) recently ended with closure of the Aberdeen Mine. Among the remaining six clusters of three or more events since 1999, two ended without apparent incident, two ended with a design change or move to a new area, one ended with a fatal accident, and one with a fire and explosion. Clusters also preceded two earlier instances, the 1998 Willow Creek fire and the 1996 fatal bump accident at Aberdeen.”
Once zones at elevated risk of bursts are identified, appropriate control techniques should be used within each zone. The most effective way to reduce a risk is to eliminate the hazard [23]. In the context of burst control, this would be achieved by not mining in the areas of greatest risk. Where avoidance is not possible, mining may be limited to development only. For example, within a pillar recovery panel, a few pillars or rows of pillars might be left in place beneath a remnant structure that was considered to generate a high level of risk during pillar recovery. Or for a longwall area, a longwall face move could be made to avoid a portion of a panel
0
2
4 6 8 10 Retreat mining cut number
12
14
Fig. 6. Energy release rate (ERR) comparison for the models shown in Fig. 5.
similar technique which involves taking the initial cuts from the most highly stressed part of the pillar core, has been another high risk procedure. As a result, the MSHA Roof Control Handbook states that ‘‘at depths exceeding 300 m, pillar splitting should not be conducted on the pillar line.” Mining in pillar points, where pillars are surrounded on two or more sides by extracted pillars, also adds to the risk of bursts. Pillar points can be created when the center pillars in a row are mined last, as occurs with some cut sequences used with continuous haulage during pillar recovery. Operational techniques that can be used to reduce the likelihood of a burst during the process of pillar extraction include: (1) Narrow lift mining: the risk in extracting highly stressed coal is reduced by taking lifts that are just one-half the width of the continuous mining machine cutting head. Mining in this fashion allows the remaining pillar more time to yield and redistribute the load. (2) Avoid mining directly into the core of a highly stressed pillar: start pillar recovery at the most inby portion of the pillar and then systematically and progressively work in the outby direction.
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
51
Table 1 Coal burst risk analysis matrix for pillar recovery. Risk factor
Depth of cover Pillar design Multiple seam interaction Roof condition Floor condition
Level of factor Low
Moderate
High
<360 m Meets NIOSH or other criteria, including barrier pillars Stress shadow or AMSS condition = ‘‘Green” Weak shale or similar, no massive strata within 15 m Claystone or similar, no massive strata within 15 m
360–450 m
>450 m Does not meet NIOSH or other criteria
AMSS condition = ‘‘Yellow”
Inadequate maps or remnant surrounded by gob (AMSS condition = ‘‘red”) Strong, thick, and massive strata near the seam
Typical Western US or Central Appalachian stratigraphy Typical Western US or Central Appalachian stratigraphy
Other geologic factors Pillar recovery method Panel width Past history of bursts
Development only, or partial pillar recovery <100 m No burst history in the seam
Typical Christmas tree or outside lift pillar recovery 100–150 m Burst history in the seam
Strong, thick, and massive strata near the seam Sandstone channels, faults or fracture zones, seam dips, rapid topographic changes Closing in center (continuous haulage), barrier pillar extraction, split-and-fender pillar recovery >150 m Burst history in the mine
Table 2 Coal burst risk analysis matrix for longwall mining. Risk factor
Depth of cover Pillar design Multiple seam interaction Roof condition Floor condition
Level of factor Low
Moderate
<360 m Development only, meets NIOSH or other criteria AMSS condition = ‘‘Green”
360–600 m >600 m Longwall mines should use yield, abutment-yield, or interpanel barrier pillars as appropriate for depth and geology AMSS condition = ‘‘Yellow” Inadequate maps or remnant surrounded by gob (AMSS condition = ‘‘Red”) Typical Western US or Central Strong, thick, and massive strata near the seam Appalachian stratigraphy Typical Western US or Central Strong, thick, and massive strata near the seam Appalachian stratigraphy Sandstone channels, faults or fracture zones, seam dips, rapid topographic changes Burst history in the seam Burst history in the mine
Weak shale or similar, no massive strata within 15 m Claystone or similar, no massive strata within 15 m
Other geologic factors Past History of Bursts
No burst history in the seam
with a high burst risk. Pillar design is the primary engineering control for minimizing the risk of pillar failures and coal bursts during retreat mining under deep cover. The different strategies and methodologies for pillar design were described above. Special cut sequences have been used to reduce the burst risk during the pillar recovery process [24]. Such techniques are difficult to implement, however, and are not entirely reliable. A deep cover retreat mine that attempted to employ a cut plan to mitigate bursts in the early 2000s quickly concluded that it was ‘‘not productive or economic” and instead opted not to retreat mine in the areas of deepest cover [25]. A variety of administrative controls, physical barriers, and personal protective equipment have been employed to mitigate hazards from small burst events [22,18,23]. These are of little value in protecting miners from the effects of large bursts, however, so where an identified zone of significantly elevated burst likelihood exists, they cannot be relied upon to reduce the risk posed by coal bursts to an acceptable level. Similarly, seismic monitoring, destressing, water infusion, and caving control have been used in some coal and hard rock mines around the world. They have been rarely trialed in the U.S., and the results have been mixed. None can currently be considered available for routine use. References [1] Mark C, Phillipson SE, Tyrna P, Gauna M. Characteristics of coal bursts in the north fork valley of the Gunnison River, Colorado. In: Proceedings of the 30th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2012. p. 1–12. [2] Mark C. Pillar design for deep cover retreat mining: ARMPS version 6. In: Proceedings of the 3th international workshop on coal pillar mechanics and design. Morgantown (WV); 2010. p. 106–21.
High
[3] Gauna M. Coal pillar design for deep conditions: an operations approach. In: Proceedings of the workshop on coal pillar mechanics and design, USBM IC 9315; 1992. p. 214–24. [4] Hoelle J. Coal bumps in an Eastern Kentucky coal mine 1989 to 1997. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 14–19. [5] DeMarco MJ, Koehler JR, Maleki H. Gate road design considerations for mitigation of coal bumps in western U.S. longwall operations. In: Proceedings of Mechanics and mitigation of violent failure in coal and hard-rock mines. Bureau of Mines (WA): Spokane, Department of the Interior; 1995. SP 01-95, p. 141–65. [6] DeMarco MJ. Yielding pillar gate road design considerations for longwall mining. In: New technology for longwall ground control, USBM SP 01-94; 1994. p. 19–36. [7] Gilbride LJ, Hardy MP. Interpanel barriers for deep western U.S. longwall mining. In: Proceedings of the 23rd international conference on ground control in mining. Morgantown (WV): West Virginia University; 2004. p. 35–41. [8] Maleki H. Caving, seismicity, and mine design in four Utah coal mines. In: Proceedings of the 25th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2006. p. 268–76. [9] Foy P. Utah’s deep coal operators face heavy regulation. Associated Press news item; 2008. [10] Gauna M, Phillipson S. Evaluation of a multiple seam interaction coal pillar bump. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 51–9. [11] Newman D. A case history investigation of two coal bumps in the Southern Appalachian Coalfield. In: Proceedings of the 21st international conference on ground control in mining. Morgantown (WV): West Virginia University; 2002. p. 90–7. [12] Heasley KA. A new laminated overburden model for coal mine design. In: Proceedings of the new technology for ground control in retreat mining NIOSH IC 9446. Pittsburgh (PA); 1997. p. 60–73. [13] Sears MM, Heasley KA. An application of the energy release rate. In: Proceedings of the 28th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2009. p. 10–6. [14] Baltz R, Hucke A. Rockburst prevention in the German coal industry. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 46–50.
52
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
[15] Bräuner G. Rockbursts in coal mines and their prevention. New York (NY): Taylor & Francis; 1994. p. 152. [16] Agapito JFT. Goodrich RR. Five stress factors conducive to bursts in Utah, USA, coal mines. In: Proceedings of the 19th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2000. p. 93– 100. [17] Iannacchione AT, Zelanko JC. Occurrence and remediation of coal mine bursts: a historical review. In: U.S. Department of the Interior, U.S. Bureau of Mines. Special Publication 01-95; 1995. p. 27–68. [18] Maleki H, Rigby S, McKenzie J, Faddies T. Historic mine designs and operational practices used in deep mines for controlling coal bumps. In: Proceedings of the 45th US rock mechanics/geomechanics symposium. Alexandria (VA): American Rock Mechanics Association; 1984. [19] Babcock CO, Bickel DL. Constraint: the missing variable in the coal burst problem. In: Proceedings of the 25th U.S. Symposium on rock mechanics. Evanston (IL); 2011. p. 639–47. [20] Mark C. Coal bursts in the deep longwall mines of the United States. In: Proceedings of AusRock 3rd Australasian ground control in mining
[21] [22]
[23]
[24]
[25]
conference. Sydney (Australia): Aus IMM and University of New South Wales; 2014. Holland CT. Cause and occurrence of coal mine bumps. Min Eng 1958:994–1004. B. Whyatt J. Dynamic failure in deep coal: recent trends and a path forward. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 37–45. Iannacchione AT, Tadolini SC. Coal mine burst prevention controls. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 20B–28. Iannacchione AT, Varley FD, Brady TM. The application of major hazard risk assessment (MHRA) to eliminate multiple fatality occurrences in the U.S. minerals industry. Pittsburgh (PA): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication, IC 9508; 2008. Newman D. Coal mine bumps: case histories of analysis and avoidance. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 1–7.
Contents lists available at ScienceDirect
International Journal of Mining Science and Technology journal homepage: www.elsevier.com/locate/ijmst
Evaluating the risk of coal bursts in underground coal mines Mark Christopher ⇑, Gauna Michael Pittsburgh Safety and Health Technology Center, Mine Safety and Health Administration (MSHA), Pittsburgh, PA 15236, USA
a r t i c l e
i n f o
Article history: Received 27 July 2015 Received in revised form 15 October 2015 Accepted 30 October 2015 Available online 15 December 2015 Keywords: Underground mining Coal Coal burst Mine safety Ground control
a b s t r a c t Coal bursts involve the sudden, violent ejection of coal or rock into the mine workings. They are almost always accompanied by a loud noise, like an explosion, and ground vibration. Bursts are a particular hazard for miners because they typically occur without warning. Despite decades of research, the sources and mechanics of these events are not well understood, and therefore they are difficult to predict and control. Experience has shown, however, that certain geologic and mining factors are associated with an increased likelihood of a coal burst. A coal burst risk assessment consists of evaluating the degree to which these risk factors are present, and then identifying appropriate control measures to mitigate the hazard. This paper summarizes the U.S. and international experience with coal bursts, and describes the known risk factors in detail. It includes a framework that can be used to guide the risk assessment process. Ó 2015 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
1. Background Coal bursts involve the sudden, violent ejection of coal or rock into the mine workings. They are almost always accompanied by a loud noise, like an explosion, and ground vibration. Bursts are a particular hazard for miners because they typically occur without warning. During the years 2012–2014, serious coal bursts occurred at three different U.S. room and pillar mines. These events resulted in three fatalities and two permanently disabling injuries. In all three instances, the events occurred during pillar recovery at depths exceeding 300 m. None of these three mines had previously reported a burst to MSHA. Coal bursts also occurred at three longwall mines during this same time period. Despite decades of research, the sources and mechanics of bursts are not well understood, and therefore these events are difficult to predict and control. Experience has shown, however, that certain risk factors are associated with an increased likelihood of a coal burst. A coal burst risk assessment consists of evaluating the degree to which these risk factors are present. In addition, some control techniques are effective in reducing the likelihood of an event or protecting miners from their effects. 2. Factors contributing to the risk of coal bursts The one universal characteristic of burst-prone environments is the presence of highly stressed coal. The overburden depth is ⇑ Corresponding author. Tel.: +1 412 3866522. E-mail address: [email protected] (C. Mark).
responsible for the overall level of stress, but pillar design or multiple seam interactions can concentrate stresses in distinct locations. Geology also plays a big role where strong roof and floor are characteristic of most, but not all, burst prone environments. Geologic features, including sandstone channels, faults, and seam dips, have been associated with the events. Certain mining layouts and practices also increase the burst risk, as does a past history of bursts. Each of these factors is discussed in more detail below. 2.1. Depth of cover Very few bursts have occurred at depths less than 300 m, although there were two incidents that occurred during pillar recovery under 230 m of cover during the early 1980s. Experience shows that the burst risk increases with depth. An analysis of the NIOSH Analysis of Retreat Mining Pillar Stability (ARMPS) database showed that for case histories where the depth of cover was less than 450 m, only 2% encountered bursts. For the handful of cases where the depth of cover exceeded 600 m, however, almost half encountered bursts (Fig. 1). Another study found that of 34 burst events that occurred in mines located in the North Fork Valley of Colorado, only three occurred where the overburden depth was less than 450 m, and 13 occurred at depths exceeding 600 m [1]. Consequently, the MSHA Handbook on Roof Control Plan Approval and Review Procedures includes the following statement: ‘‘pillar recovery at depths exceeding 600 m may not be appropriate due to the heightened risk of bursts at such unusual and extremely deep cover.”
http://dx.doi.org/10.1016/j.ijmst.2015.11.009 2095-2686/Ó 2015 Published by Elsevier B.V. on behalf of China University of Mining & Technology.
48
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
Number of cases
100 Burst Squeeze collapse Success
75 50 25
60
120 180 240 300 360 420 480 540 Depth of cover (m)
600
660
Fig. 1. Distribution of pillar failures and pillar bursts with depth in the ARMPS database.
event killed a miner in 1996. At greater depths, interpanel barrier pillars have been used at several Utah longwalls [7]. In some cases, rather than leave a full barrier, mines have elected to make midpanel move around the area of deepest cover. The unmined panel provides a local interpanel barrier for the next panel [8]. The interpanel barrier effectively protects the tailgate corner from the influence of previous panels, but at greater depths the single-panel stresses on the longwall face reach the same levels as were present with abutment loads from adjacent extracted panels separated by yield pillars. After a fatal bump occurred on a longwall face near the headgate at 840 m depth of cover, one major Utah operator announced that it would consider reserves at depths exceeding 900 m to be unmineable [9].
2.2. Pillar design 2.3. Multiple seam interactions The U.S. underground coal mining industry has extensive regions where multiple seams have been mined. The interaction of the active mining with overlying and/or underlying old workings can generate stress concentrations. The severity of a multiple seam stress concentration typically depends on two factors: (1) The thickness of the interburden between the active seam and the previously-mined seam (or seams). In general, the thicker the interburden, the less likely that the interaction will result in a severe stress concentration. (2) The type of remnant structure present in the previous seam. Isolated remnants, with worked out areas on two or more sides, are the most hazardous. Remnant structures are typically created when coal is left in place adjacent to areas of full extraction. However, bursts have occurred above and beneath large remnants adjacent to smaller developed pillars [10,11]. In these cases, the smaller developed pillars apparently behaved as a pseudo gob area, transferring much of their load onto the larger pillar (Fig. 3). A burst risk assessment should take such situations into account particularly when inmine evidence suggests a stress concentration exists. As noted in Fig. 3, the large pillar was surrounded by a ‘‘pseudo gob area” consisting of smaller developed pillars that had apparently yielded and transferred load to the large pillar. Interactions between all previously mined seams should be considered in the assessment. The workings in several seams may overlap, creating very high stress zones, particularly if the interburdens separating the older workings from the active seam are thin (Fig. 4). Empirical or numerical computer models should be a part of a thorough burst risk assessment. Models such as the Analysis of Multiple Seam Stability (AMSS) or LaModel can identify potentially high stress zones due to multiple seam mining [12]. However,
(a) Conventional abutment-yield design
(b) Two-entry yield pillar design
Fig. 2. Pillar design approaches used for burst control.
Tailgate
Headgate
Tailgate
Headgate
Tailgate
Headgate
Longwall face
(c) Panel barrier design
Barrier pillar
A pillar that is properly designed and large enough to distribute the load that it carries is unlikely to be burst prone. On the other hand, a pillar that is sufficiently small and yielding, is also not burst-prone. The burst hazard is greatest for poorly designed pillars that are too small to properly distribute the loads they carry, but too large to yield. NIOSH studied the 17 largest burst events in room and pillar mines that occurred between 1980 and 2010. Each of these events resulted in extensive damage to at least several pillars. The analysis showed that 12 of the 17 multi-pillar bursts could be attributed to inadequate pillar design. These 12 events all occurred during pillar recovery mining. In nine instances, the barrier pillars were too small, were being extracted on retreat, or were not used at all. In five of the 12 cases, pillar splitting operations without a barrier pillar apparently triggered the multi-pillar burst. Barrier pillars are particularly important in room and pillar mining because they protect each new panel from the abutment loads arising from previously mined areas. The NIOSH ARMPS program was revised in 2010 specifically for evaluating production and barrier pillars in deep cover applications [2]. In longwall mining, several different pillar design strategies have been employed in burst prone conditions (Fig. 2). Conventional approach employs at least one large abutment pillar in a multi-entry system, sometimes flanked by small yield pillars. Experience has shown that appropriately sized abutment pillars can reduce the incidence of bursts [3,4]. In Utah, two-entry yield pillar layouts have been used since the 1960s, and they can virtually eliminate gate pillar bursts [5]. Mining engineers also learned to avoid critical pillars which are too large to yield non-violently yet too small to support large abutment loads. The width-toheight ratios of such burst-prone, critical pillars normally exceeded 4 or 5 [6]. While the yield pillar system typically performs well at depths up to approximately 600 m, it can concentrate the load on the tailgate corner of the longwall face, and this can result in severe face bursts near the tailgate corner of the longwall. A tailgate corner
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
49
While there is no established generic signature of burst-prone geology, the following are two examples developed from experience in different regions: Pirst Bump area 1600
Fig. 3. Location of coal burst that occurred in upper seam development workings (solid lines) located above a large pillar in a lower seam (dashed lines).
Fig. 4. Overlapping gob areas in two overlying seams (workings shown in blue and orange) probably concentrated unusually high loads on the two pillars that burst (outlined in red).
these programs have typically been used to identify potential pillar squeezes or roof instability, and may require special adjustments for burst risk evaluation. Specialized analysis techniques, such as Energy Release Rate (ERR) models, may be more useful in some applications [13]. Accurate identification of remnant structures requires reliable maps of the older workings. The burst risk assessment should also include an evaluation of the adequacy of the available maps. If mapping detail does not exist or is incomplete, then the likelihood of encountering unexpected remnants should be considered. Not all multiple seam mining increases the risk of coal bursts. The risk can actually be reduced when mining is conducted in de-stressed ground above or below an area that has been fully extracted [14]. 2.4. Geology Strong sandstone or siltstone roof and floor have been associated with coal bursts, particularly in the eastern U.S. and in Utah. In evaluating whether the roof or floor geology may contribute to the burst risk, it is important to consider: (1) The thickness of the strong sandstone or siltstone unit. Thicker units are more likely to be associated with bursts. (2) The distance between the strong unit and the coal seam. Strong units close to the seam pose the greatest risk of coal bursts. (3) The strength of the rock. Surrounding rock units associated with coal bursts typically have uniaxial compressive strengths of at least 70 MPa but can exceed 100 MPa. (4) The characteristics of the rock. Massive units with minimal bedding, jointing, or other discontinuities are more likely to be associated with coal bursts.
(1) A 4.5 m thick package of strong sandstone in the first 15 m above the mining horizon, or a 6 m thick package of strong sandstone within the first 9 m of the floor [15]. (2) A massive sandstone unit at least1.5 m thick is found within 1.3 m above the coal seam. Core logs, combined with rock mechanics testing, may be used to identify when such conditions might be present. However, since surface core holes are generally spaced too far apart to identify all zones of burst prone ground, supplemental underground test holes may be necessary [4]. Sandstone channels are suspected of creating high stress concentrations [16]. At one longwall mine in eastern Kentucky, numerous large bursts occurred beneath sandstone channels, but none were encountered when a channel was not present [4]. Because sandstone channels may be limited in extent, they may be particularly difficult to identify in widely spaced surface boreholes. Faults or fracture zones have been associated with increased burst risk in both coal and hard rock mines [17]. When mining approaches a highly-stressed fault or joint system, the ground may suddenly shift, releasing seismic energy that results in a burst [1]. Faults or joints may also partition the overburden, resulting in an unexpected concentration of overburden load [16]. The presence of steep seam dips has been observed at a number of burst sites, and rapid changes in the depth of cover due to steep topography have also been associated with bursts [18]. Coal strength is one factor that does not seem to play a significant role in the burst risk. Iannacchione and Zelanko noted that bursts have occurred in at least 25 different U.S. coalbeds, varying from strong, blocky seams to the very friable Pocahontas No. 3 and No. 4 seams [17]. Laboratory studies conducted by Babcock and Bickel showed that most coals can fail violently if they are highly stressed and the confinement is suddenly released [19]. Extensive German laboratory studies using large-scale specimens have also concluded that nearly all bituminous coals can burst. In these experiments, coal seams ranging in unconfined compressive strength from 5 to 50 MPa have all been shown to be burst-prone [15].
2.5. Mining layouts and practices Historically, more than 80% of bursts have been reported during retreat mining, with less than 20% occurring on development [17,20]. Retreat mining increases the likelihood of bursts because it concentrates abutment loads on the pillar line, gate pillars, or longwall face, and because caving overburden releases seismic energy as it breaks. Of the two widely used retreat mining methods, pillar recovery is significantly more burst prone than longwall mining. Wider panels are a factor that increases the burst risk during pillar recovery. A wider panel results in a greater front abutment load, and subjects the mining area to more overburden load. Modeling indicates that the ERR may increase approximately in proportion to the panel width (Figs. 5 and 6). As noted in Fig. 6, Curve ‘‘A” shows the ERR cut-by-cut in the critical entry for a panel with six pillars extracted, while curve ‘‘B” shows the ERR for a panel with four pillars extracted. Certain pillar recovery practices can further increase the burst risk. In particular, many bursts have occurred during barrier pillar extraction, a practice that was once wide-spread but is now seldom employed [21]. Pillar extraction by split-and-fender, or by any
50
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
Fig. 5. LaModel grids used in the Energy Release Rate (ERR) analysis.
Energy release rate
1.0 A
0.8
In another example, the Brody Mine in southern West Virginia had never experienced a coal burst prior to May of 2014. Then a strong event occurred during pillar recovery, which knocked the continuous miner operator to the ground and covered him with debris from his mid-thigh down, entrapping him for less than five minutes. Three days later, while mining the next row of pillars, an even larger event cost the lives of two coal miners. Similar precursors occurred prior to the fatal burst at the Huff Creek Mine and the major event in the North Barrier Pillar at the Crandall Canyon Mine. When a mine has experienced bursts, future mining operations with similar geology and mining methods should be considered high risk. Underground observations and monitoring are critical components of a burst risk management program. Mining crews should be trained to observe coal burst warning signs, particularly the occurrence of small bursts, which are often the best indication that an area is becoming more burst prone. A record-keeping system should be maintained and management processes developed to ensure that warning signs receive appropriate responses.
0.6
3. Conducting the risk assessment
0.4 B
0.2
Operational techniques used by longwall mines include reducing the depth of the web, reducing the speed of the shearer, unidirectional cutting, and avoiding double cuts at the gate ends [22].
Based on the evaluation of the burst risk factors listed above, an overall risk level can be assigned to future mining areas. For example, those areas with negligible burst risk might be considered green zones, areas with slightly higher risk could be yellow zones, and the areas of greatest risk might be orange zones. The initial risk assessment should be conducted before an area is developed, using available borehole logs and maps of previous mining in overlying and underlying seams. During development, underground test hole drilling may be employed to provide more detailed information on the geologic conditions. Underground mapping should be conducted prior to any retreat mining, particularly where the mapping can use rib conditions to identify the locations of significant multiple seam interactions. At each step, as new data becomes available, burst potential zones should be reevaluated and updated. The matrices shown in Tables 1 and 2 may be used to assist with the risk assessment. Each of the significant known risk factors can be rated as low, moderate, or high. Note that it is the level of the factor that is being rated, not the burst risk associated with it. The ratings of all the factors should be considered when assessing overall burst risk. The matrices are intended as generic guides that can be tailored for each site-specific burst assessment. A universal quantitative rating scale has not been developed for the factors listed in the matrix, and there is no standard method to combine the individual factor ratings into an overall burst risk rating. The assessment should clearly state the assumptions made in the process, including any ratings and weightings of individual factors, and the procedure used to estimate the overall burst risk. An experienced ground control professional can use the matrix to evaluate the overall burst risk.
2.6. A past history of bursts
4. Mitigating the risk of coal bursts
Major bursts have often been preceded by a pattern of increasing coal burst activity. Whyatt identified seven ‘‘clusters” consisting of three or more reported bursts in the 10-year period 1999– 2008 [22]. He found that: ‘‘. . . the largest (cluster) recently ended with closure of the Aberdeen Mine. Among the remaining six clusters of three or more events since 1999, two ended without apparent incident, two ended with a design change or move to a new area, one ended with a fatal accident, and one with a fire and explosion. Clusters also preceded two earlier instances, the 1998 Willow Creek fire and the 1996 fatal bump accident at Aberdeen.”
Once zones at elevated risk of bursts are identified, appropriate control techniques should be used within each zone. The most effective way to reduce a risk is to eliminate the hazard [23]. In the context of burst control, this would be achieved by not mining in the areas of greatest risk. Where avoidance is not possible, mining may be limited to development only. For example, within a pillar recovery panel, a few pillars or rows of pillars might be left in place beneath a remnant structure that was considered to generate a high level of risk during pillar recovery. Or for a longwall area, a longwall face move could be made to avoid a portion of a panel
0
2
4 6 8 10 Retreat mining cut number
12
14
Fig. 6. Energy release rate (ERR) comparison for the models shown in Fig. 5.
similar technique which involves taking the initial cuts from the most highly stressed part of the pillar core, has been another high risk procedure. As a result, the MSHA Roof Control Handbook states that ‘‘at depths exceeding 300 m, pillar splitting should not be conducted on the pillar line.” Mining in pillar points, where pillars are surrounded on two or more sides by extracted pillars, also adds to the risk of bursts. Pillar points can be created when the center pillars in a row are mined last, as occurs with some cut sequences used with continuous haulage during pillar recovery. Operational techniques that can be used to reduce the likelihood of a burst during the process of pillar extraction include: (1) Narrow lift mining: the risk in extracting highly stressed coal is reduced by taking lifts that are just one-half the width of the continuous mining machine cutting head. Mining in this fashion allows the remaining pillar more time to yield and redistribute the load. (2) Avoid mining directly into the core of a highly stressed pillar: start pillar recovery at the most inby portion of the pillar and then systematically and progressively work in the outby direction.
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
51
Table 1 Coal burst risk analysis matrix for pillar recovery. Risk factor
Depth of cover Pillar design Multiple seam interaction Roof condition Floor condition
Level of factor Low
Moderate
High
<360 m Meets NIOSH or other criteria, including barrier pillars Stress shadow or AMSS condition = ‘‘Green” Weak shale or similar, no massive strata within 15 m Claystone or similar, no massive strata within 15 m
360–450 m
>450 m Does not meet NIOSH or other criteria
AMSS condition = ‘‘Yellow”
Inadequate maps or remnant surrounded by gob (AMSS condition = ‘‘red”) Strong, thick, and massive strata near the seam
Typical Western US or Central Appalachian stratigraphy Typical Western US or Central Appalachian stratigraphy
Other geologic factors Pillar recovery method Panel width Past history of bursts
Development only, or partial pillar recovery <100 m No burst history in the seam
Typical Christmas tree or outside lift pillar recovery 100–150 m Burst history in the seam
Strong, thick, and massive strata near the seam Sandstone channels, faults or fracture zones, seam dips, rapid topographic changes Closing in center (continuous haulage), barrier pillar extraction, split-and-fender pillar recovery >150 m Burst history in the mine
Table 2 Coal burst risk analysis matrix for longwall mining. Risk factor
Depth of cover Pillar design Multiple seam interaction Roof condition Floor condition
Level of factor Low
Moderate
<360 m Development only, meets NIOSH or other criteria AMSS condition = ‘‘Green”
360–600 m >600 m Longwall mines should use yield, abutment-yield, or interpanel barrier pillars as appropriate for depth and geology AMSS condition = ‘‘Yellow” Inadequate maps or remnant surrounded by gob (AMSS condition = ‘‘Red”) Typical Western US or Central Strong, thick, and massive strata near the seam Appalachian stratigraphy Typical Western US or Central Strong, thick, and massive strata near the seam Appalachian stratigraphy Sandstone channels, faults or fracture zones, seam dips, rapid topographic changes Burst history in the seam Burst history in the mine
Weak shale or similar, no massive strata within 15 m Claystone or similar, no massive strata within 15 m
Other geologic factors Past History of Bursts
No burst history in the seam
with a high burst risk. Pillar design is the primary engineering control for minimizing the risk of pillar failures and coal bursts during retreat mining under deep cover. The different strategies and methodologies for pillar design were described above. Special cut sequences have been used to reduce the burst risk during the pillar recovery process [24]. Such techniques are difficult to implement, however, and are not entirely reliable. A deep cover retreat mine that attempted to employ a cut plan to mitigate bursts in the early 2000s quickly concluded that it was ‘‘not productive or economic” and instead opted not to retreat mine in the areas of deepest cover [25]. A variety of administrative controls, physical barriers, and personal protective equipment have been employed to mitigate hazards from small burst events [22,18,23]. These are of little value in protecting miners from the effects of large bursts, however, so where an identified zone of significantly elevated burst likelihood exists, they cannot be relied upon to reduce the risk posed by coal bursts to an acceptable level. Similarly, seismic monitoring, destressing, water infusion, and caving control have been used in some coal and hard rock mines around the world. They have been rarely trialed in the U.S., and the results have been mixed. None can currently be considered available for routine use. References [1] Mark C, Phillipson SE, Tyrna P, Gauna M. Characteristics of coal bursts in the north fork valley of the Gunnison River, Colorado. In: Proceedings of the 30th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2012. p. 1–12. [2] Mark C. Pillar design for deep cover retreat mining: ARMPS version 6. In: Proceedings of the 3th international workshop on coal pillar mechanics and design. Morgantown (WV); 2010. p. 106–21.
High
[3] Gauna M. Coal pillar design for deep conditions: an operations approach. In: Proceedings of the workshop on coal pillar mechanics and design, USBM IC 9315; 1992. p. 214–24. [4] Hoelle J. Coal bumps in an Eastern Kentucky coal mine 1989 to 1997. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 14–19. [5] DeMarco MJ, Koehler JR, Maleki H. Gate road design considerations for mitigation of coal bumps in western U.S. longwall operations. In: Proceedings of Mechanics and mitigation of violent failure in coal and hard-rock mines. Bureau of Mines (WA): Spokane, Department of the Interior; 1995. SP 01-95, p. 141–65. [6] DeMarco MJ. Yielding pillar gate road design considerations for longwall mining. In: New technology for longwall ground control, USBM SP 01-94; 1994. p. 19–36. [7] Gilbride LJ, Hardy MP. Interpanel barriers for deep western U.S. longwall mining. In: Proceedings of the 23rd international conference on ground control in mining. Morgantown (WV): West Virginia University; 2004. p. 35–41. [8] Maleki H. Caving, seismicity, and mine design in four Utah coal mines. In: Proceedings of the 25th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2006. p. 268–76. [9] Foy P. Utah’s deep coal operators face heavy regulation. Associated Press news item; 2008. [10] Gauna M, Phillipson S. Evaluation of a multiple seam interaction coal pillar bump. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 51–9. [11] Newman D. A case history investigation of two coal bumps in the Southern Appalachian Coalfield. In: Proceedings of the 21st international conference on ground control in mining. Morgantown (WV): West Virginia University; 2002. p. 90–7. [12] Heasley KA. A new laminated overburden model for coal mine design. In: Proceedings of the new technology for ground control in retreat mining NIOSH IC 9446. Pittsburgh (PA); 1997. p. 60–73. [13] Sears MM, Heasley KA. An application of the energy release rate. In: Proceedings of the 28th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2009. p. 10–6. [14] Baltz R, Hucke A. Rockburst prevention in the German coal industry. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 46–50.
52
C. Mark, M. Gauna / International Journal of Mining Science and Technology 26 (2016) 47–52
[15] Bräuner G. Rockbursts in coal mines and their prevention. New York (NY): Taylor & Francis; 1994. p. 152. [16] Agapito JFT. Goodrich RR. Five stress factors conducive to bursts in Utah, USA, coal mines. In: Proceedings of the 19th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2000. p. 93– 100. [17] Iannacchione AT, Zelanko JC. Occurrence and remediation of coal mine bursts: a historical review. In: U.S. Department of the Interior, U.S. Bureau of Mines. Special Publication 01-95; 1995. p. 27–68. [18] Maleki H, Rigby S, McKenzie J, Faddies T. Historic mine designs and operational practices used in deep mines for controlling coal bumps. In: Proceedings of the 45th US rock mechanics/geomechanics symposium. Alexandria (VA): American Rock Mechanics Association; 1984. [19] Babcock CO, Bickel DL. Constraint: the missing variable in the coal burst problem. In: Proceedings of the 25th U.S. Symposium on rock mechanics. Evanston (IL); 2011. p. 639–47. [20] Mark C. Coal bursts in the deep longwall mines of the United States. In: Proceedings of AusRock 3rd Australasian ground control in mining
[21] [22]
[23]
[24]
[25]
conference. Sydney (Australia): Aus IMM and University of New South Wales; 2014. Holland CT. Cause and occurrence of coal mine bumps. Min Eng 1958:994–1004. B. Whyatt J. Dynamic failure in deep coal: recent trends and a path forward. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 37–45. Iannacchione AT, Tadolini SC. Coal mine burst prevention controls. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 20B–28. Iannacchione AT, Varley FD, Brady TM. The application of major hazard risk assessment (MHRA) to eliminate multiple fatality occurrences in the U.S. minerals industry. Pittsburgh (PA): U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Institute for Occupational Safety and Health, DHHS (NIOSH) Publication, IC 9508; 2008. Newman D. Coal mine bumps: case histories of analysis and avoidance. In: Proceedings of the 27th international conference on ground control in mining. Morgantown (WV): West Virginia University; 2008. p. 1–7.