2
mine areas had a 6-panel district design with 700-ft-wide
panels but had challenges with respect to ventilation and
induced seismicity (Khademian et al., 2024). Thus, the new
layout was intended to improve upon the previous designs
as each panel can be sealed right after longwall recovery,
benefiting ventilation efficiency and potential seismicity in
the mined-out panels. Extensive instrumentation was con-
ducted to monitor pillar pressure and roof and floor defor-
mations during the first panel mining.
Mining the first panel proved this design efficient how-
ever, some challenges arose when approaching areas with a
weak fireclay floor. During mining of the first panel, an area
with weak fireclay floor (low load-bearing capacity floor)
showed extensive floor heave slowing the operation. Also,
the pillar pressure monitoring showed that the abutment
load from the first panel bridged across the barrier pillar
(220ft wide) increasing the pillar pressure in the tailgate
of the second panel by about 2,000 psi with signs of floor
heave and rib damage.
In this part of the research, the pillar and floor instru-
mentation results were used to initially analyze design per-
formance. Geomechanical models were then developed to
evaluate alternative designs that can remedy the current
challenges. The evaluation is based on pillar pressure and
floor heave in the alternative designs, but it is planned to
evaluate the designs with respect to the induced seismicity
potential.
DESCRIPTION OF MINE SITE
The longwall mine discussed in this paper operates within
the Pocahontas #3 coal seam in western Virginia. This seam
is part of the Pocahontas formation, which consists of a
thick lens of sediments such as sandstone, siltstone, shale,
coal, and claystone (Englund &Briggs, 1974). Located
in the Virginia overthrust belt, the mine is near the Keen
Mountain fault, a strike-slip fault with compressional over-
thrusting (Molinda, 2003). Despite some small synthetic
thrust faults within the mining area (with a minor offset of
up to 5 ft (1.5 m)), the Keen Mountain fault does not seem
to affect the mine. The overall mine roof lithology features a
sequence of shales varying from 0 to 25 ft (7.6 m) in thick-
ness directly above the Pocahontas #3 coal seam (Figure 1).
Above this shale lies the first sandstone unit, known as
Sandstone #1 (SS1), which ranges from 0 to 35 ft (10.7)
meters in thickness and is characterized by thin to medium
bedding with shale or mica streaks. Above the SS1 unit lies
a shale parting, typically ranging from 0 to 5 ft (1.5 m) in
thickness, followed by another sandstone unit, SS2. While
the shale parting is present in most of the mining area, it
is absent in some regions, resulting in one large sandstone
unit. The SS2 unit is generally thicker, cleaner, and more
massive compared to SS1, with thick to massive bedding.
The thickness of the SS2 unit usually varies between 30 ft
(9.1 m) and 75 ft (22.8 m).
The geomechanical properties used in the model are
listed in Table 1 where E is Young’s modulus or deformabil-
ity modulus, υ is Poisson ratio, C is cohesion, F is friction
angle, and T is tensile strength of rock. A Mohr-Coulomb
strain-softening constitutive law was used for defining
inelastic response of rock to stress. To this end, cohesion of
the rock was reduced to 6% of its initial values after 0.003
strain.
The friction angles were kept constant. The rock tensile
strength was reduced to zero after 0.003 strain. The choice
Figure 1. Generalized stratigraphic column of the mine area
Table 1. Geomechanical properties used in the 3DEC model
Rock Type E (GPa) υ
C
(MPa) F (degree)
T
(MPa)
Coal 1.7 0.3 2.0 31 0.7
Dark shale 5.8 0.2 15.2 32.8 7.0
Hard
Sandstone
28.4 0.1 23.5 43.9 9.2
Shaly Sand 21 0.1 18.8 38.3 7.9
Sandy Shale 16 0.2 19.1 32.8 7.0
Shale 10.3 0.1 18.8 38.3 7.9
Sandstone 28 0.1 19.6 43.9 9.2
Fireclay 4.3 0.3 8.7 27.7 0.7
mine areas had a 6-panel district design with 700-ft-wide
panels but had challenges with respect to ventilation and
induced seismicity (Khademian et al., 2024). Thus, the new
layout was intended to improve upon the previous designs
as each panel can be sealed right after longwall recovery,
benefiting ventilation efficiency and potential seismicity in
the mined-out panels. Extensive instrumentation was con-
ducted to monitor pillar pressure and roof and floor defor-
mations during the first panel mining.
Mining the first panel proved this design efficient how-
ever, some challenges arose when approaching areas with a
weak fireclay floor. During mining of the first panel, an area
with weak fireclay floor (low load-bearing capacity floor)
showed extensive floor heave slowing the operation. Also,
the pillar pressure monitoring showed that the abutment
load from the first panel bridged across the barrier pillar
(220ft wide) increasing the pillar pressure in the tailgate
of the second panel by about 2,000 psi with signs of floor
heave and rib damage.
In this part of the research, the pillar and floor instru-
mentation results were used to initially analyze design per-
formance. Geomechanical models were then developed to
evaluate alternative designs that can remedy the current
challenges. The evaluation is based on pillar pressure and
floor heave in the alternative designs, but it is planned to
evaluate the designs with respect to the induced seismicity
potential.
DESCRIPTION OF MINE SITE
The longwall mine discussed in this paper operates within
the Pocahontas #3 coal seam in western Virginia. This seam
is part of the Pocahontas formation, which consists of a
thick lens of sediments such as sandstone, siltstone, shale,
coal, and claystone (Englund &Briggs, 1974). Located
in the Virginia overthrust belt, the mine is near the Keen
Mountain fault, a strike-slip fault with compressional over-
thrusting (Molinda, 2003). Despite some small synthetic
thrust faults within the mining area (with a minor offset of
up to 5 ft (1.5 m)), the Keen Mountain fault does not seem
to affect the mine. The overall mine roof lithology features a
sequence of shales varying from 0 to 25 ft (7.6 m) in thick-
ness directly above the Pocahontas #3 coal seam (Figure 1).
Above this shale lies the first sandstone unit, known as
Sandstone #1 (SS1), which ranges from 0 to 35 ft (10.7)
meters in thickness and is characterized by thin to medium
bedding with shale or mica streaks. Above the SS1 unit lies
a shale parting, typically ranging from 0 to 5 ft (1.5 m) in
thickness, followed by another sandstone unit, SS2. While
the shale parting is present in most of the mining area, it
is absent in some regions, resulting in one large sandstone
unit. The SS2 unit is generally thicker, cleaner, and more
massive compared to SS1, with thick to massive bedding.
The thickness of the SS2 unit usually varies between 30 ft
(9.1 m) and 75 ft (22.8 m).
The geomechanical properties used in the model are
listed in Table 1 where E is Young’s modulus or deformabil-
ity modulus, υ is Poisson ratio, C is cohesion, F is friction
angle, and T is tensile strength of rock. A Mohr-Coulomb
strain-softening constitutive law was used for defining
inelastic response of rock to stress. To this end, cohesion of
the rock was reduced to 6% of its initial values after 0.003
strain.
The friction angles were kept constant. The rock tensile
strength was reduced to zero after 0.003 strain. The choice
Figure 1. Generalized stratigraphic column of the mine area
Table 1. Geomechanical properties used in the 3DEC model
Rock Type E (GPa) υ
C
(MPa) F (degree)
T
(MPa)
Coal 1.7 0.3 2.0 31 0.7
Dark shale 5.8 0.2 15.2 32.8 7.0
Hard
Sandstone
28.4 0.1 23.5 43.9 9.2
Shaly Sand 21 0.1 18.8 38.3 7.9
Sandy Shale 16 0.2 19.1 32.8 7.0
Shale 10.3 0.1 18.8 38.3 7.9
Sandstone 28 0.1 19.6 43.9 9.2
Fireclay 4.3 0.3 8.7 27.7 0.7