Rock Mech Rock Eng (2016) 49:3293–3305

DOI 10.1007/s00603-016-0970-8

ORIGINAL PAPER

Stress Changes and Deformation Monitoring of Longwall Coal

Pillars Located in Weak Ground
Bin Yu1 • Zhenyu Zhang2 • Tiejun Kuang1 • Jinrong Liu1

Received: 26 July 2015 / Accepted: 1 April 2016 / Published online: 8 April 2016
Ó Springer-Verlag Wien 2016

Abstract Coal pillar stability is strongly influenced by the concentrations in the chain pillar. Convergence measure-
site-specific geological and geotechnical conditions. Many ments in the tailgate during longwall mining further indi-
geological structures such as faults, joints, or rock intrusions cated the evolution characteristics of coal pillar
can be detrimental to mining operations. In order to evaluate deformation, clearly showing that the gateroad deformation
the performance of coal pillars under weak roof degraded by is mainly induced by the longwall extraction it serves. When
igneous rock intrusion, stress and deformation monitoring predicting the future pillar loads from the monitored data,
was conducted in the affected tailgate areas of Nos. 8208 and two stress peaks appeared across the 38-m-wide tailgate coal
8210 longwalls in Tashan coal mine, Shanxi Province, pillar, which are separated by the lower stress area within the
China. The measurements in the 8208 longwall tailgate pillar center. This 10-m-wide elastic pillar core area indi-
showed that the mining-induced stresses in 38-m-wide coal cates that the coal pillar may be narrowed to 30 m to
chain pillars under the overburden depth of 300–500 m improve coal recovery. The measurements further indicate
started to increase at about 100 m ahead of the 8208 long- that, if the headgate of the next panel can be developed after
wall working face and reached its peak level at approxi- the adjacent gob becomes stable, the coal pillar width may
mately 50 m ahead of the longwall face. The peak stress of be further reduced to 20 m. This study is applicable for the
9.16 MPa occurred at the depth of 8–9 m into the pillar from chain pillar design, the gateroad secondary support design
the tailgate side wall. In comparison, disturbance of the ahead of the longwall mining face, and the gateroad prepa-
headgate block pillar area was negligible, indicating the ration of the next longwall panel under similar geological
difference of abutment pressure distribution between the and geotechnical conditions.
tailgate and headgate sites where the adjacent unmined
longwall block carried most of the overburden load. How- Keywords Stress development
Deformation
Coal
ever, when the longwall face passed the headgate monitoring pillar
Pillar design
Longwall mining
site by 360–379 m, the pillar stress increased to a peak value
of 21.4 MPa at the pillar depth of 13 m from the gob side
mainly due to stress redistribution in the chain pillar. In 1 Introduction
contrast to the headgate, at the tailgate side, the adjacent
goaf was the dominant triggering factor for high stress In mining practice, pillars of ore body are often used for the
purpose of ground control (González-Nicieza et al. 2006;
Peng 2008). For example, in underground coal mining,
& Zhenyu Zhang various coal pillar sizes and shapes are used to enable safe
[email protected] and efficient coal extraction. In pillar mining used to mine
1 under surface infrastructures, coal pillars serve to minimize
Datong Coal Mine Group Company, Datong 037003, Shanxi,
China the ground subsidence and maintain the long-term stability
2 of overlying strata (Chen et al. 2014). In room and pillar
State Key Laboratory of Coal Mine Disaster Dynamics and
Control, Chongqing University, 174 Shazhengjie St., mining of relatively horizontal and thin or medium-thick
Shapingba District, Chongqing 400044, China coal seams, coal pillars are temporarily left to support the

123
3294 B. Yu et al.

overlying strata and then partially or fully recovered, analyzed the serial and parallel rock pillar stability of dif-
depending on the actual roof conditions (Hustrulid and ferent mechanical properties, showing that a rock pillar of
Bullock 2001). Together with the top coal caving tech- higher uniaxial compressive strength in serial pattern can
nique, longwall mining has been the most common method lead to the sudden failure of adjacent pillars being of lower
used in thin or thick coal seams with relatively undisturbed uniaxial compressive strength, while in parallel pattern,
and structure-free strata. In light of the longwall method pillar failure always occurs in rock pillars of low uniaxial
advantages that include high advance and coal recovery compressive strength or lower elastic modulus. Using
rates, the longwall roadways must be well protected by FLAC3D, Wang et al. (2013) investigated the mechanical
adequately designed chain pillars (Shabanimashcool and Li state of a coal pillar due to longwall mining and the results
2012). With the advance of longwall mining, the periodic indicated that the stress peak appeared at a distance of 7.5 m
weighting of the overlying rock strata induces repeated in front of the longwall mining face. This predicted stress
cyclic high loads on coal pillars. This dynamic loading peak position is consistent with that measured in strip
significantly impacts on the integrity of coal pillars and mining (Chen et al. 2014). Li et al. (2015) studied the width
brings challenges for pillar stability. to height ratio of yield pillars in a coal mine in China and
The long- or short-term pillar stability may be required for concluded that decreasing the pillar width to height ratio
different mining purposes. A successful coal pillar can transfer from 2.67 to 1.67 can effectively eliminate the pillar bumps.
the overlying load into floor strata. However, failed coal pillars It has been known that pillar stability is strongly influ-
may redistribute the load to adjacent pillars, resulting in their enced by the site-specific geologic conditions and actual
failure as well. This may continue until a new load equilibrium mining layout. Therefore, the field measurements of pillar
is reached. Starting from the gateroad preparation to longwall performance during mining is necessary and irreplaceable
mining, coal pillars have to experience the loading due to for pillar design, even though it is costly and time-con-
in situ stress and mining-induced stress (Singh et al. 2011), suming. On one hand, the field measurements can evaluate
where the mining-induced static and dynamic pressure affects the pillar performance and help the engineers to better
the caving characteristics of overlying strata. understand the pillar behavior. On the other hand, the
At mine sites, many research efforts have been made to measured results and their interpretation can significantly
monitor the mechanical behavior of coal pillars. Mark et al. contribute to pillar design improvements of future work.
(1988) measured the changes of vertical stress in three Taking a longwall top coal caving face at Shanxi Province,
longwall yield pillars and found that the average pillar load China, as the example, the mechanical response of coal pillars
reduced significantly once pillars yield, implying the under complex geological conditions in longwall panels are
necessity of considering the time effect in narrow pillar investigated here. Specifically, the top layer of targeted Nos.
design. In the field, Haramy and Kneisley (1990) confirmed 3–5 coal seams is seriously affected by the magma intrusion
that the yield pillars can effectively mitigate pillar bumps. of lamprophyre, resulting in a weak roof structure. Moreover,
Quantitatively, Newman (1989) analyzed the changes in magmatic intrusion complicated the Nos. 3–5 coal seams by
pillar stress in terms of pillar strain, showing that the compounding coal seam structure and degrading coal quality.
Young’s modulus of coal in the prefailure state varies within For mining under such complex conditions, the key issues
the range of 208–661 MPa, while reducing to 55–208 MPa include rational pillar size determination for the prevention of
after failure. Therefore, the commonly assumed value of roof failure and pillar bumps, and pillar stability evaluation
Young’s modulus in pillar design overestimates the real during consecutive longwall panel mining, where its core
cases. For coal pillars laying on weak floor, Chugh et al. topic is to analyze stress and deformation characteristics of
(1990) developed an approach based on the ultimate bearing coal pillars under such complex geological conditions with
capacity and settlement of coal pillars. Singh et al. (2011) the advance of the mining face.
found that the cover depth and the caveability of overlying In this study, factors affecting coal pillar strength were
strata significantly influenced the characteristics of mining- firstly compiled. Then, the data from the field coal pillar
induced stress, implying that the mining-induced stress monitoring located under complex geological conditions at
during the pillar extraction may vary with the site-specific Tashan coal mine, Shanxi Province, China, were analyzed.
geological conditions. More recently, Chen et al. (2014)
evaluated the long-term bearing capacity of a strip coal pillar
and found that the stress peak of a stable coal pillar (typical 2 Factors Affecting Pillar Strength
saddle stress profile) occurs at pillar depths of 6–8 m.
In numerical modeling, Mohan et al. (2001) demon- The mechanical integrity of a solid block can be deter-
strated failed and stable pillar cases in Indian coal mines mined by its load bearing capacity, stresses it undertakes,
using the strain softening model. Using Realistic Failure and boundary conditions. The mechanical behavior of coal
Process Analysis 2-D (RFPA2D), Wang et al. (2011) on the engineering scale is known to be dominated by

123
Stress Changes and Deformation Monitoring of Longwall Coal Pillars Located in Weak Ground 3295

geological structures, such as cleats, joints, fractures, and have been developed in terms of pillar width to height
bedding planes. Therefore, the factors affecting coal pillar ratio to facilitate pillar design (Hustrulid 1977; Peng
stability can be attributed to the mechanical properties of 2008). However, as the development of each strength
coal, geology, pillar profile, and dimensions. formula is based on the site-specific geological condi-
tions and different assumptions, the empirical formulae
2.1 Mechanical Properties of Coal Pillars should be treated with caution. For unknown geologies
and pillar conditions, the selected safety factors should
As a kind of soft rock, coal has a typical Protodyakonov be higher.
coefficient f range of 2–4 and tends to be heterogeneous in Due to the site-specific geological conditions and the
the field. Its bearing capacity in loading is determined by increase of mining depth, it is still a challenge to design
the mechanical properties of coal, such as strength and coal pillars economically and safely. To ensure stability of
deformability. coal pillars and mitigate roof collapse and occurrence of
In the field, the strength of coal within a coal pillar rock bumps throughout the panel service time, the com-
varies due to the existence of local structures. The pillar is, bination of stiff (wide) pillars and yield (narrow) pillars is
therefore, weaker than the smaller samples tested in the sometimes more desirable, rather than choosing only one
laboratory. For a critically loaded coal pillar consisting of type of coal pillar.
weak and strong layers, failure would first initiate along the
weak layers (Wang et al. 2011). It is worth noting that 2.3 Other Parameters Influencing Coal Pillar
Kaiser et al. (2011) discussed the rock mass strength in Strength
depth and proposed that a distinct S-shape stress–strain
failure envelope for the entire range of confinements may The coal pillar strength is not only affected by its
suit the brittle rock at depth. If this is true, the widely used mechanical properties, geometrical profile, and dimen-
Mohr–Coulomb and Hoek–Brown failure criteria may sions, but also by other parameters such as the properties
significantly underestimate the strength of coal pillar core. of roof–coal or floor–coal interfaces and coal seam dip-
To validate this speculation, Buzzi et al. (2014) examined ping angle. After mining, the designed coal pillar has a
the triaxial strength of Australian coal under low confine- potential to expand laterally and the frictional force
ment and confirmed the non-convex failure criterion under developed along roof–coal and coal–floor interfaces can
a low level of confinement. This is of great significance for provide horizontal confinement (Gale 1998). A strong
coal pillar stability at greater mining depth. However, the interface can provide high confinement to prevent the coal
triaxial strength of coal under a high level of confinement pillar expanding laterally. With the increase in cohesion
typically experienced for coal pillar core at depth are not and frictional resistance along the roof–coal and coal–
involved and further studies are needed to describe the floor interfaces, the magnitude and range of the confining
strength of coal pillar core at depth. stress within the coal pillar increases more rapidly. The
Regarding deformability, Newman (1989) suggested influence of roof–coal and coal–floor interface properties
that the commonly assumed values of Young’s modulus in on pillar confinement distribution are schematically
pillar design may be significantly overestimated, implying shown in Fig. 1.
that the actual deformation of coal pillars would be larger It is noted that increase of the width to height ratio of
than expected when approaching the critical pillar loads. pillars can also increase the pillar confinement, as the
contact area between the roof–coal and coal–floor inter-
2.2 Coal Pillar Profile and Dimensions faces increases with the width to height ratio of coal pillars
when the pillar height is invariable, hence enlarging the
In the laboratory, the effect of specimen size and shape cover range of confinement on coal pillars.
on rock strength has been extensively studied using non- In non-horizontal coal deposit, its dipping angle also has
standard profiles and dimensions (Darlington et al. 2011). a significant influence on coal pillar stability. It is found
It is well established that the strength of rock specimens that the width of the non-elastic zone at the underside of
reduces significantly with the increase in specimen size, the pillar increases parabolically with the coal seam incli-
generally following an exponential function. On the field nation angle, while the width of the non-elastic zone at the
scale, the engineering behavior of coal pillars is also upside of the pillar decreases exponentially (Wei 2014).
significantly influenced by the coal pillar profile and In addition, there are some other factors impacting the
dimensions (Bieniawski and Van Heerden 1975). It is coal pillar strength, such as coal saturation (Poulsen et al.
generally accepted that the strength of coal pillars 2014), fracture density, and orientation (Peng 2008). For
increases with the pillar size but decreases with pillar details of their impact mechanisms on the coal pillar
height. Many empirical strength formulae of coal pillars strength, the reader can refer to the relevant literature.

123
3296 B. Yu et al.

Fig. 1 Confinement
distribution of coal pillars
developed by different roof–
coal and coal–floor interface
conditions

3 Geological Background and Underground thickness. The geological column chart of Nos. 3–5 coal
Monitoring Instrumentation seams in the 8210 mining face is shown in Fig. 3.

In order to study stress changes and deformation of weak 3.2 Instrumentation Plan
roof coal pillars under complex geological conditions, field
measurements were conducted in Tashan coal mine, China, During the longwall mining, a single-entry system was
where the coal seam was seriously damaged by magmatic employed with the 38-m-wide chain pillar. Two sets of
intrusion. monitoring instruments were installed and monitored for this
study. One set of measurement instruments was installed in
3.1 Geological Conditions each of the tailgates of Nos. 8210 and 8208, respectively
(Fig. 4). In Fig. 4, the ellipses are used to schematically
Tashan coal mine is located in the middle-east edge of the represent the collapsed rock masses after mining and the gob
east wing of Datong coalfield, Shanxi Province, China, and denotes the mined out void. The purpose of the former set of
the mineable coal deposit is combined coal seam Nos. 3–5. instruments is to evaluate pillar stability impacted by adja-
The thickness of coal seam Nos. 3–5 varies from 11.1 to cent 8208 (first) longwall face excavation, while the latter is
31.7 m, with an average of 19.4 m. The coal seam dip to investigate the pillar stability disturbed by current 8210
angle varies from 1° to 3°, with a cover depth of (second) longwall face excavation. Therefore, by integrating
300–500 m. Due to magmatic intrusion, a 6-m metamor- two sets of instruments, the evolution of stress and defor-
phic coal band exists in the upper layer of the coal seam, mation of weak roof and coal pillars can be obtained during
and the hardness and joint distribution varies from the consecutive longwall face mining.
upper to the lower portions of the coal seam. In the lower The measurements include stress at different pillar
part, a few joints are present, while in the upper part, the depths and convergence of roof to floor and rib wall of the
number of joints is extreme and the coal tends to be softer, 8208 tailgate and headgate. Here, the pillar depth denotes
containing several rock parting bands. the distance horizontally measured from the tailgate rib
The exploratory boreholes reveal that the thickness of wall to the stress meter instrument. Continuous monitoring
Nos. 3–5 coal seams at the 8208 mining face ranges of stress and deformation was carried out by setting up an
8.74–16.57 m with an average of 14.90 m. It contains ten online monitoring system to achieve the characteristics of
rock parting bands as shown in the geological column pillar stress changes and deformation during the whole
chart shown in Fig. 2. process of consecutive longwall mining.
At the 8210 longwall face, the cored boreholes show Seven borehole stress meters were installed in the tail-
that the structure of the coal seam in this panel is also gate coal pillar of the 8208 longwall face and the other two
complex. In this area, the coal seam varies from 6.41 to were installed in the coal block of the 8208 longwall face
20.28 m in thickness, with an average of 12.29 m, and prior to mining. The stress meters on the pillar side were
contains 0–29 layers of rock partings, averaging 0.94 m in installed at pillar depths of 3, 6, 8, 9, 10, 11, and 19.5 m,

123
Stress Changes and Deformation Monitoring of Longwall Coal Pillars Located in Weak Ground 3297

Fig. 3 The geological column chart of Nos. 3–5 coal seams at the
8210 mining face
Fig. 2 The geological column chart of Nos. 3–5 coal seams at the
8208 mining face
4 Results and Discussions

respectively, and the stress meters in the unmined longwall 4.1 Pillar Stress
block were installed at depths of 21 and 25 m, respectively.
In the tailgate of the 8210 longwall face, ten stress meters 4.1.1 Pillar Stress Results in the Tailgate of the 8208
were installed in the tailgate pillar at depths of 3, 6, 9, 12, Longwall Face
15, 18, 21, 25, 30, and 30 m, respectively. Boreholes of
56 mm in diameter were drilled 2 m above the floor with Figure 5 shows the pillar stress distribution at the tailgate
2-m spacing. The detailed layout of stress meters can be of 8208 when the longwall face approached. Note that the
schematically seen in Fig. 4. pillar depth was measured from the tailgate rib of the 8208
The set of points to measure the convergence of roof to side. It can be seen that, when the longwall 8208 face was
floor and rib walls within the tailgate were installed at an 216–228 m away, the largest stress increase of 0.66 MPa
approximate spacing of 5 m. Ten convergence stations was measured at the pillar depth of 8 m. With further
were set up in the 8208 tailgate and 20 in the 8210 tailgate advance of the longwall face to 108–120 m away, the
as shown in Fig. 4. largest stress increase of 5.9 MPa was measured at the

123
3298 B. Yu et al.

Fig. 4 Horizontal view of the

layout of stress, deformation of
pillars, and convergence
measurement of tailgate

pillar depth of 8 m, while the adjacent stress increase of readings reduced to 0.1 or 0 MPa, suggesting that the pillar
1.73 MPa was at the pillar depth of 9 m. On this occasion, yielded.
the stress readings at other depths were less than 0.6 MPa. The vertical stress measurements indicate that, when the
When the longwall face was 51–63 m away, the peak stress 8208 longwall face was approximately 200 m away, the
increase of 9.16 MPa was measured at 8 m into the pillar, front abutment pressure did not influence the pillar load.
while the second largest stress increase of 1.62 MPa Therefore, the vertical load increase did not occur further
emerged at the pillar depth of 3 m. However, when the than 200 m ahead of the longwall face. When the longwall
longwall face 8208 was 6–18 m away, all the stress meter face was 100 m away, the pillar stress increased at pillar
depths of 0–9 m, with the largest increment appearing at
pillar depths of 8–9 m. However, stress at pillar depths of
more than 10 m did not change. The results presented in
Fig. 5 also imply that, when the longwall face was 50 m
away, significant disturbance due to longwall mining
occurred, causing the pillar to yield, and the vertical
loading was transferred deeper into the pillar. At this stage,
the pillar stress increased at all monitoring sites. This
mining disturbance became stronger when the longwall
mining face was approximately 10 m away, but the stress
reduced to a very low value, indicating that the coal pillar
was severely damaged. Therefore, it can be concluded that
the impact range of longwall mining under such geological
conditions is approximately 100 m in front of the working
face and the severely disturbed zone develops at 0–50 m in
front of the longwall face. In the case of the 38-m-wide
Fig. 5 Vertical stress increase at different pillar depths from the 8208
coal pillar, the peak pillar stress increase occurred at depths
tailgate of 8–9 m from the tailgate rib side.

123
Stress Changes and Deformation Monitoring of Longwall Coal Pillars Located in Weak Ground 3299

4.1.2 Pillar Stress Distribution in the Longwall Block monitoring site, the largest stress remained at the pillar
Ahead of the 8208 Longwall Face depth of 25 m, but its value reduced to 6.79 MPa, followed
by 1.90 MPa at the pillar depth of 15 m, while the stress
Stress change ahead of the 8208 longwall face is presented above the base level at the rest of the monitoring positions
in Fig. 6. It shows that, when the longwall face advanced was less than 1.00 MPa. The developed peak stress further
from 177 to 17 m away, the vertical stress increase at depths decreased when the longwall face advanced closer to the
of 21 m and 25 m (from the tailgate) was very low, with the monitoring sites. For example, it reduced to 4.83 MPa
largest value of 0.8 MPa when the longwall face was when the longwall face was 47–66 m away and 4.39 MPa
56–61 m away. This indicates that the coal block distur- when the longwall face was 23–42 m away. During the
bance ahead of the longwall mining was very slight, as the process of longwall face advancing from 169 to 47 m
solid longwall coal block is much larger than the tailgate away, the highest stress was consistently located at the
coal pillar. Comparatively, the developed stress at the depth pillar depth of 25 m.
of 25 m was higher than that at the pillar depth of 21 m. When the longwall face of 8208 passed the monitoring
In general, it can be seen that the developed stress on sites, the increase in coal pillar stress appeared low. When
unmined coal block increases with the approach of the the longwall face passed the instruments by 42–61 m, the
longwall face, but the stress at 17–22 m ahead of the increase in peak stress above the pillar was only 1.90 MPa,
longwall face reduced significantly. The reason for such a located at the depth of 21 m from the pillar side. With the
stress reduction was that the coal around the borehole has continual advance of the longwall face, the increase in peak
probably yielded. stress in the coal pillar transferred deeper to 25 m from the
pillar side and its value increased gradually to 2.11 and
4.1.3 Tailgate Pillar Stress Distribution Around the 8210 2.40 MPa when the longwall face passed the measuring
Longwall Face site by 85–104 and 125–144 m, respectively. However,
when the longwall face passed the monitoring site by
Figure 7 shows the development and evolution of pillar 360–379 m, the increase in peak pillar stress soared to
stress change at the tailgate of 8210 when the 8208 long- 21.40 MPa. This sudden increase in the pillar load was
wall face approached and passed the monitoring positions. approximately 20 times greater than the stress increase
Note that the pillar depth in Fig. 7 was measured starting when the longwall face passed the site by 85–104 m, while
from the 8210 tailgate side, which is different from Figs. 5 its position remained at the pillar depth of 25 m.
and 6. It can be found that, when the longwall face was In mining production, the coal pillar of the 8208 tailgate
150–170 m away, the largest stress increase developed at had already experienced repeated loading and unloading
the pillar depth of 25 m of a magnitude of 11.02 MPa, due to the tailgate development and longwall face mining
followed by 2.64 MPa at the pillar depth of 9 m and of the 8206 and 8208 panels. However, at this stage, the
1.60 MPa at the pillar depth of 15 m. In other monitoring coal pillar of the 8210 tailgate only experienced the min-
sites, the stress increase was less than 1.00 MPa. When the ing-induced stress caused by the 8210 tailgate development
longwall face advanced to a distance of 77–96 m from the

Fig. 7 The maximum stress experienced by the tailgate of the 8210

Fig. 6 The pillar stress distribution on the side of unmined coal block pillar when the 8208 longwall face approached and passed the
when the longwall face approached monitored site

123
3300 B. Yu et al.

and 8208 longwall mining. Each loading and unloading

process lead to the stress change above the coal pillar.
Furthermore, due to mining excavation, the boundary
conditions on two sides of the coal pillar adjacent to the
tailgates of 8208 and 8210 were distinctly different. Such
differences resulted in the inconsistent distribution of
abutment pressure in front of the longwall face. Comparing
results in Figs. 5, 6, and 7, it can be seen that the peak
abutment pressure at the tailgate of the 8208 coal pillar
occurred 50 m in front of mining face, while at the tailgate
of 8210, it occurred at more than 150 m in front of the
mining face.
In addition, the monitoring results shown in Figs. 5 and
7 indicate that, with the advance of the 8208 longwall face,
stress changes above the coal pillar adjacent to the unmined
panel 8210 were lower as the longwall face approached Fig. 8 The 8210 tailgate pillar stress change versus the distance from
closer, while the stress changes increased when the long- the 8210 tailgate when the longwall face of 8208 passed the
wall face was farther away. However, for the coal pillar monitored site
adjacent to the tailgate of 8208, the stress gradually
increased until the borehole wall became severely damaged
when the longwall face of 8208 approached the monitoring
locations. This indicates that the mining impact of the
longwall face of 8208 on the coal pillar of the tailgate of
8210 was small and the dominant triggering factor for
stress increase came from the loading induced by the 8208
longwall mining. When the longwall face of 8208 passed
the monitoring instruments by more than 360 m, the pillar
stress adjacent to the headgate increased due to a major
stress redistribution.
Taking the maximum stress of each measuring point as
shown in Fig. 7, the peak stress in each hole was plotted in
terms of the distance measured from the 8210 tailgate and
is shown in Fig. 8. In the same manner, the pillar stress
variation in terms of the distance to the longwall face were
plotted by taking the maximum stress at a specific pillar
depth. Figure 9 shows the peak stress variation of the 8210 Fig. 9 The stress variation of the 8210 tailgate pillar at the depth of
25 m when the longwall face of 8208 approached and passed the
tailgate pillar at the pillar depth of 25 m when the longwall
monitored site
face of 8208 approached and passed the monitoring site.
The stress across the 8210 tailgate pillar shown in Fig. 8
indicates that the peak stress appeared to be asymmetric instruments by 120 m, followed by a gradual stress
across the pillar and there is a single maximum peak stress increase. These measurements indicate that the lagged
near the 8208 gob side. The peak stresses were located distance for stabilization of the overlying rock strata is
approximately at pillar depths of 21–30 m from the 8210 120–130 m in the rear of the longwall working face. The
tailgate side, i.e., 8–17 m from the 8208 headgate side. results are of great significance for gateroad development,
Therefore, the designed 38-m-wide coal pillar can effec- implying that the gateroad driveage for the next panel
tively withstand the vertical loads caused by the 8208 should lag the mining face by more than 120 m in order to
longwall panel mining. avoid heavy strata deformation.
Figure 9 shows that the pillar stress at the depth of 25 m Similarly to Figs. 8 and 9, Figs. 10 and 11 show the
from the 8210 tailgate rib started to increase rapidly at peak stress increase of the 8208 tailgate pillar in cross-
70 m in front of the longwall face and reached the highest section and stress changes at the depth of 8 m into the pillar
level 40 m ahead of the longwall face. The stress then with the mining advance. The results show that the stress
began to decrease gradually and reached the lowest level distribution of the 8208 tailgate pillar is not symmetric and
when the longwall face of 8208 passed the measuring there is a single peak stress near the 8208 gob side located

123
Stress Changes and Deformation Monitoring of Longwall Coal Pillars Located in Weak Ground 3301

tailgate may also experience similar mining disturbance to

the coal pillar of the tailgate 8208 once mining of the 8210
longwall panel is conducted. Assuming that the coal pillar
of the tailgate of 8210 follows a similar behavior during the
subsequent longwall 8210 mining and remains elastic in
pillar core, the probable peak pillar stress can be predicted
by superimposing the results of Figs. 8 and 10. This pre-
dicted loading is presented in Fig. 12.
It can be seen that the predicted pillar stress profile in
Fig. 12 has two stress peaks that are not symmetric about
the center of the coal pillar. The high stress peak on the
right-hand side in Fig. 12 is located at 8 m away from the
8210 tailgate pillar wall and would be caused by mining
disturbance of the 8208 panel, while the relatively low
peak on the left-hand side 13 m away from the 8208 tail-
Fig. 10 The peak stress of the 8208 tailgate pillar along the pillar gate pillar wall would be caused by mining of the 8210
width when the longwall face of 8208 approached the instrument site panel. The distance between these two stress peaks is
approximately 10 m. An elastic core of low stress appears
between these two stress peaks. It should be noted that the
above conclusions are based on the assumption that the
coal pillar of the tailgate of 8210 follows a similar behavior
to the previously monitored sites and remains elastic in
pillar core.

4.2 Tailgate Deformation

4.2.1 Tailgate Deformation of the 8208 Longwall Face

The measured roof to floor convergence and convergence

rate of the tailgate of 8208 during longwall mining are
presented in Fig. 13. The results show that the roof to floor
convergence rate of the tailgate of 8208 increased signifi-
cantly at 100 m ahead of the longwall face, with daily
convergence of more than 20 mm. The convergence
Fig. 11 The peak stress of the 8208 tailgate pillar at the depth of 8 m increased further to 100–150 mm/day at 50–60 m in front
when the longwall face of 8208 approached the monitored site of the longwall face. During the period when the longwall
face advanced from 100 m away to the measuring instru-
approximately at the pillar depth of 8 m from the 8208 ments, the cumulative roof and floor convergence ranged
tailgate. Therefore, the 38-m-wide coal pillar performed between 1000 and 1400 mm. The results of the roof to floor
well throughout its service time. cumulative convergence and the convergence rate of the
The results presented in Fig. 11 show that, during the tailgate of 8208 indicate that the mining disturbance on the
8208 longwall face mining, the pillar stress of the tailgate of 8208 tailgate coal pillar started approximately 100 m ahead
8208 started to increase 150 m ahead of the working face of the longwall face and accelerated at 50–60 m ahead of
and reached its peak value at 60 m in front of the working the longwall face.
face, then the stress rapidly reduced to 0.1 MPa at 10 m in Figure 14 shows the rib wall convergence and the con-
front of the longwall face, indicating severe damage. vergence rate of the tailgate of 8208 when the longwall
In mining production, the coal pillar of the tailgate of face of 8208 approached the instruments. The data indicate
8210 may also experience similar mining disturbance to that the convergence rate of the 8208 tailgate rib wall
that of the tailgate 8208 coal pillar once the mining of the started to increase at 70 m ahead of the longwall face and
8210 longwall face is in progress. Therefore, the devel- accelerated to 100 mm/day at the distance of 30–40 m
opment and evolution of pillar stress during consecutive ahead of the mining face, up to 150 mm/day as the long-
8210 longwall mining can be predicted from the monitored wall face neared the instrumented site. During the period
results. In mining production, the coal pillar of the 8210 when the longwall face advanced from 100 m away to the

123
3302 B. Yu et al.

measuring points, the total rib wall convergence ranged up Fig. 15. The relatively small roof to floor convergence of
to 600–1000 mm. This indicates that the mining distur- the tailgate of 8210 began to occur within the distance of
bance of the adjacent longwall panel on the 8208 tailgate 0–20 m ahead of the 8208 longwall face and continued to
rib wall started at the distance of approximately 70 m increase up to 160 m behind the longwall face, but the
ahead of the longwall face and became dramatic at impact of mining-induced stress on the pillar deformation
40–50 m ahead of the longwall face. These results are was minor. The roof to floor convergence rate only fluc-
different from the characteristics of the roof to wall con- tuated within the range of 1.0–4.2 mm/day. When the
vergence shown in Fig. 13. longwall face of 8208 passed the instruments by more than
200 m, the roof to floor convergence rate of the tailgate of
4.2.2 Tailgate Deformation of the 8210 Longwall Face 8210 became very small, less than 1 mm/day.
Figure 16 shows the rib wall convergence rate of the
The roof and floor convergence rate of the tailgate of 8210 tailgate of 8210 when the 8208 longwall face approached
with the advance of the 8208 longwall mining is shown in and passed the measuring sites. It can be seen that the rib
convergence of the tailgate of 8210 initiated at 20 m ahead
of the 8208 longwall face; however, the magnitude of the
rib wall convergence rate was small, fluctuating between
1.0 and 2.7 mm/day. Similarly to the roof to floor con-
vergence rate, the rib convergence rate further reduced to
less than 1.0 mm/day 200 m past the longwall face.
Figures 15 and 16 indicate that, during the 8208 long-
wall panel mining, the deformation of the tailgate of 8210
was relatively small. Therefore, the designed coal pillar can
satisfy the service requirement for the 8210 longwall panel.

5 Implications on Pillar Stability and Pillar Design

During the 8208 longwall mining, the deformation of the

tailgate of 8210 was very small. During the period when
the measuring sites were located 50 m ahead of the long-
Fig. 12 Prediction of the maximum stress increase in the cross- wall face to being passed by 500 m, its cumulative con-
section within the 8210 tailgate vergence of the roof to floor was in the range 8–85 mm,

Fig. 13 Roof and floor

convergence and convergence
rate of the tailgate of 8208 when
the longwall face of 8208
approached

123
Stress Changes and Deformation Monitoring of Longwall Coal Pillars Located in Weak Ground 3303

Fig. 14 Rib wall convergence

and convergence rate of the
tailgate of 8208 when the
longwall face of 8208
approached

Fig. 15 The roof to floor convergence rate of the tailgate of 8210

with the advance of the 8208 longwall mining
Fig. 16 The rib wall convergence rate at the tailgate of 8210 with the
advance of the 8208 longwall mining
while the cumulative convergence of rib walls was in the
range 25–81 mm. Therefore, it can be concluded that the
induced deformation of the tailgate of 8210 was very small In the pillar zone that was significantly impacted by
during the 8208 longwall mining. This implies that the mining-induced stress as shown in Figs. 13 and 14, bend-
38-m-wide coal pillar can easily withstand the dynamic ing and cracking of single hydraulic props, which are used
pressure caused by the adjacent longwall panel mining. At to hold the roof, were observed. Therefore, upgrading of
the tailgate of 8208, heavy deformation occurred and some the secondary strata support in the gateroads ahead of the
single hydraulic props were damaged when the longwall longwall mining face is crucial for ground control. To
face approached the measuring sites. The data showed that avoid damaging single hydraulic props, it is recommended
the main deformation and damage of the tailgate of 8208 to replace the currently used hydraulic props with flexible
was induced by the 8208 longwall panel extraction and supports which can yield substantially, in order to allow for
exacerbated by weak roof conditions and igneous rock large roof and floor convergence. To maintain stability of
intrusion. surrounding rock strata, the density of single hydraulic

123
3304 B. Yu et al.

props should be intensified and/or cable and bolting support Assuming that the pillar mechanically remains linearly
should be used. elastic during the whole consecutive mining process, two
The predicted pillar stress in the 8210 pillar during the stress peaks develop in a 38-m-wide coal pillar. Between
consecutive longwall mining of the 8208 and 8210 working the two stress peaks, a 10-m-wide low stress elastic pillar
faces as shown in Fig. 12 indicates that the distance core exists, suggesting that it is feasible to narrow the coal
between the two stress peaks is 10 m and there may be the pillar from 38 to 30 m. The coal pillar width under such
possibility to narrow the coal pillar from 38 to 30 m to conditions may be further reduced to 20 m if the gateroad
improve the coal resource recovery rate. Furthermore, if preparation for subsequent panels can be developed after
the gateroad development for subsequent longwall panel stabilization of the overlying strata of the adjacent gob.
preparation can be done after stabilization of the overlying
strata of the adjacent gob, the coal pillar can be further Acknowledgments The study was partially financially supported by
the National Natural Science Foundation of China (No. 51404168)
reduced to approximately 8–10 m if only considering the and Fundamental Research Funds for the Central Universities (No.
impact of adjacent panel mining. For this case, the coal 106112015CDJXY240005), and the authors are thankful for the
pillar would be located in the low stress zone caused by the support from the Tashan coal mine during the field measurement.
adjacent mined out panel, as the peak abutment stress Special thanks go to Dr. Jan Nemcik, University of Wollongong,
Australia, for the proofreading.
(right-hand side in Fig. 12) was located at the depth of
13 m away from the 8208 gob side. However, the second
pillar stress peak due to the panel mining that it serves References
(left-hand side in Fig. 12) was located at approximately
8 m from the gob side. Therefore the coal pillar width may Bieniawski ZT, Van Heerden WL (1975) The significance of in situ
be further reduced to 20 m, but it can still resist the dis- tests on large rock specimens. Int J Rock Mech Min Sci
turbance of longwall panel mining that it serves. Geomech Abstr 12:101–113. doi:10.1016/0148-9062(75)90004-2
Buzzi O, Sieffert Y, Mendes J, Liu X, Giacomini A, Seedsman R
(2014) Strength of an Australian coal under low confinement.
Rock Mech Rock Eng 47:2265–2270. doi:10.1007/s00603-013-
6 Conclusions 0493-5
Chen SJ, Guo WJ, Zhou H, Shen B, Liu JB (2014) Field investigation
of long-term bearing capacity of strip coal pillars. Int J Rock
The field monitoring of pillar stress changes and gateroad Mech Min Sci 70:109–114. doi:10.1016/j.ijrmms.2014.02.022
deformations provided valuable data for the pillar design Chugh YP, Pula O, Pytel WM (1990) Ultimate bearing capacity and
for future longwall panels. settlement of coal pillar sub-strata. Int J Min Geol Eng
Field monitoring shows that the influence of longwall 8:111–130. doi:10.1007/BF00920499
Darlington WJ, Ranjith PG, Choi SK (2011) The effect of specimen
panel mining on coal pillars in weak ground begins size on strength and other properties in laboratory testing of rock
approximately 100 m in front of the working face, while and rock-like cementitious brittle materials. Rock Mech Rock
the significant mining disturbance occurs at 0–50 m ahead Eng 44:513–529. doi:10.1007/s00603-011-0161-6
of the longwall mining face. The peak pillar stress on the Gale WJ (1998) Coal pillar design issues in longwall mining. In: Aziz
N (ed) Coal operator’s conference, Wollongong, Australia.
tailgate side appeared at the depth of 8–9 m from the gob University of Wollongong & the Australasian Institute of Mining
side wall. Comparatively, its disturbance on the unmined and Metallurgy, pp 133–146
longwall block is relatively small, as the unmined long- González-Nicieza C, Álvarez-Fernández MI, Menéndez-Dı́az A,
wall coal block is much larger than the tailgate coal pillar. Álvarez-Vigil AE (2006) A comparative analysis of pillar design
methods and its application to marble mines. Rock Mech Rock
It was found that the increase of stress in the headgate Eng 39:421–444. doi:10.1007/s00603-005-0078-z
coal pillar was very low when the adjacent longwall Haramy KY, Kneisley RO (1990) Yield pillars for stress control in
working face passed the monitored site by less than longwall mines—case study. Int J Min Geol Eng 8:287–304.
125 m. However, when the longwall face passed the doi:10.1007/BF00920642
Hustrulid WA (1977) A review of coal pillar strength formulas. Rock
instrumentation site by 360–379 m, the peak stress of the Mech Rock Eng 10:111. doi:10.1007/BF01261806
pillar increased to 21.4 MPa due to the overall panel Hustrulid WA, Bullock RL (2001) Underground mining methods.
stress redistribution. This indicates that the disturbance Society for Mining, Metallurgy, and Exploration, Littleton
due to the adjacent longwall working face on the headgate Kaiser PK, Kim BH, Bewick RP, Valley B (2011) Rock mass strength
at depth and implications for pillar design. Min Tech
coal pillar is small and the dominant stress increase in the 120:170–179
coal pillar occurs on the tailgate side. The deformation of Li W, Bai J, Peng S, Wang X, Xu Y (2015) Numerical modeling for
tailgates during consecutive longwall mining confirmed yield pillar design: a case study. Rock Mech Rock Eng
the evolution characteristics of the mechanical states of 48:305–318. doi:10.1007/s00603-013-0539-8
Mark C, Listak JM, Bieniawski ZT (1988) Yielding coal pillars—field
pillars and also indicates that the main deformation and measurements and analysis of design methods. In: Proceedings
damage of gateroads is induced by adjacent panel of the 29th US symposium on rock mechanics, Minneapolis,
extraction. Minnesota, June 1988, pp 261–270

123
Stress Changes and Deformation Monitoring of Longwall Coal Pillars Located in Weak Ground 3305

Mohan GM, Sheorey PR, Kushwaha A (2001) Numerical estimation Singh AK, Singh R, Maiti J, Kumar R, Mandal PK (2011) Assessment
of pillar strength in coal mines. Int J Rock Mech Min Sci of mining induced stress development over coal pillars during
38:1185–1192. doi:10.1016/S1365-1609(01)00071-5 depillaring. Int J Rock Mech Min Sci 48:805–818. doi:10.1016/j.
Newman DA (1989) In situ yield behaviour of a coal pillar. Int J Min ijrmms.2011.04.004
Geol Eng 7:163–170. doi:10.1007/BF01554344 Wang SY, Sloan SW, Huang ML, Tang CA (2011) Numerical study
Peng SS (2008) Coal mine ground control. China University of of failure mechanism of serial and parallel rock pillars. Rock
Mining and Technology Press, Xuzhou Mech Rock Eng 44:179–198. doi:10.1007/s00603-010-0116-3
Poulsen BA, Shen B, Williams DJ, Huddlestone-Holmes C, Erarslan Wang H, Jiang Y, Zhao Y, Zhu J, Liu S (2013) Numerical
N, Qin J (2014) Strength reduction on saturation of coal and coal investigation of the dynamic mechanical state of a coal pillar
measures rocks with implications for coal pillar strength. Int J during longwall mining panel extraction. Rock Mech Rock Eng
Rock Mech Min Sci 71:41–52. doi:10.1016/j.ijrmms.2014.06.012 46:1211–1221. doi:10.1007/s00603-012-0337-8
Shabanimashcool M, Li CC (2012) Numerical modelling of longwall Wei G (2014) Study on the width of the non-elastic zone in inclined
mining and stability analysis of the gates in a coal mine. Int J coal pillar for strip mining. Int J Rock Mech Min Sci
Rock Mech Min Sci 51:24–34. doi:10.1016/j.ijrmms.2012.02.002 72:304–310. doi:10.1016/j.ijrmms.2014.09.013

123