See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/266732965

CISPM-MS: a Tool to predict surface subsidence

and to study interactions associated with multi
seam mining Operations
CONFERENCE PAPER AUGUST 2012

CITATIONS

READS

61

2 AUTHORS, INCLUDING:
Yi Luo
West Virginia University
69 PUBLICATIONS 90 CITATIONS
SEE PROFILE

Available from: Yi Luo

Retrieved on: 29 January 2016

31st International Conference on Ground Control in Mining

CispmMs: a Tool to Predict Surface Subsidence and to Study
Interactions Associated with MultiSeam MiningOperations
Yi Luo, AssociateProfessor
Luo,Y.
Biao Qiu, Graduate
Qiu,B.
ResearchAssistant
MiningEngineering
West VirginiaUniversity
Morgantown,WV

ABSTRACT
Multiple-seam mining operations induce surface subsidence
basins different from those caused by mining in a single coal seam,
and also disturb the mining operations in the other coal seams. A
computer program, CISPM-MS (Comprehensive and Integrated
Subsidence Prediction Model for Multiple Seam Mining), has been
developed to predict the final surface movements and deformations
as well as the interactions associated with multi-seam coal mining
operations. The program applies a newly developed subsurface
subsidence prediction model that is able to consider the geological
stratification variations. In predicting surface subsidence, the
surface movements and deformations caused by the individual
mining operations as well as the interactive effects are all
considered. A case study is conducted in this paper for validating
theprogram.
INTRODUCTION
It is estimated that the multiple-seam coal reserves account
for about 70% of the total reserves in the United States (Mark,
Chase and Pappas, 2007). Particularly in the central Appalachian
and Western coal fields, the majority of todays coal mines are
operating above and/or beneath previously mined seams. The
effects of mining one coal seam on the other coal seams, referred
as the interactions of multi-seam mining, present one of the
major ground control challenges in the coal mining industry. The
severity of the multi-seam mining effects depends on the interburden thickness and properties, mining sequence, seam heights
and mining method applied, time interval between the mining
activities in neighboring seams, and sometimes local topographic
and hydrographicfeatures.
The mining activity in one coal seam can cause strata
movements and stress redistribution in the coal seams both above
and below. The mining-induced stress and deformations can
affect the stability of the mine structures in neighboring seams.
For example, the subsurface strain induced by an underlying
active mine may reduce the pillar confinement and the frictional
resistance along the pillar-roof and pillar-floor interfaces in the
overlying seam. Such effects reduce the pillar strength and may
potentially cause pillar failure. When the pillar failure occurs in a
sufficiently large contiguous area, it could induce additional surface

and subsurface subsidence other than that caused by the active

mine alone. This unexpected additional subsidence may influence
surface and subsurface structures and waterbodies.
In order to make accurate subsidence predictions for multiseam mining operations, the multiple seam mining interactions
should be taken into consideration. A previously developed
subsurface subsidence prediction model (Luo and Qiu, 2012) will
be used to assess the subsidence effects of an active mine on the
mine structures in the neighboring seams. In that model, a new
deformation term - void intensity is defined to express the severity
of expansion or shrinkage of a volume of rock strata under the
influence of subsidence process. The void intensity around the mine
pillars can be determined by the subsurface subsidence prediction
model (Luo and Qiu,2012).
The subsidence-induced void intensity in the subsurface
strata changes the confinement to the pillar-roof and pillar-floor
interfaces, and subsequently affects the strength of the pillars.
Particular attention will be paid to areas with positive void intensity
(expansion) that could reduce pillar strength. The possibility
for pillar failure can be judged using the reduced pillar strength
and safety factor. The failure of the pillars can cause additional
subsurface and possibly surfacesubsidence.
This void intensity-pillar strength model is then incorporated
into the newly developed multi-seam mining subsidence prediction
program, CISPM-MS. This program has a user-friendly working
environment. It allows the user to draw the mine panels and
prediction points in AutoCAD, and it provides convenient tools for
data output and postanalysis.
ENHANCED SUBSURFACE SUBSIDENCE
PREDICTIONMODEL
The subsurface deformation caused by mining activities in an
underlying coal seam could affect the pillar stability in the upper
seam(s) as both the load and the strength of the pillar are changed.
Change in the pillar load can be resulted from the differential
subsidence between the roof and floor of the pillars. Pillar strength
decreases with the confinement pressure as a result of the increased
void intensity caused by the strata subsidence process. Therefore,

31st International Conference on Ground Control in Mining

the stability of the upper seam pillars can be determined based on
the predicted subsurface movements anddeformations.

layer thickness, the overburden load, the percent of hard rock in

each layer, etc. The methods to determine these final subsidence
parameters have been discussed in a previous publication (Luo and
Qiu,2012).

The concept of the influence function method, proven to be

reliable and versatile for surface subsidence prediction, has been
employed to develop an enhanced subsurface subsidence prediction
model (Luo and Qiu, 2012). The model is capable of considering
the varying overburden stratification. This model mimics the
subsidence propagation process layer-by-layer from the strata
immediately above the coal seam to the ground surface to predict
the subsurface subsidence as shown in Fig. 1. The subsidence of a
given layer in the overburden strata is caused by the subsidence of
the layer immediately below. The final subsidence at a subsurface
point of interest is determined by integrating the defined influence
function within the proper horizontal interval. To apply the
influence function method in developing a subsidence prediction
model, the influence functions for subsidence and horizontal
displacement along a major cross-section should be defined first
and they are shown in Eqs. 1 and 2, respectively.

x,

The differential strata movements in both horizontal and

vertical directions will cause deformations in the subsurface strata.
In surface subsidence studies, the surface deformations long a
major cross-section are traditionally described by slope, strain,
and curvature. However, for applications dealing with subsurface
subsidence, the distributions of horizontal, vertical, and total
strains in the overburden strata could be much more valuable for
assessing the subsidence influences to subsurface mines structures,
hydrology, and degasification wells. The horizontal strain (ex)
is defined as the first derivative of horizontal displacement with
respect to x (Eq. 5). Sufficient horizontal strain could cause vertical
fractures or even cracks in the strata. The vertical strain (ez) is
defined as the first derivative of subsurface subsidence with respect
to z (Eq. 6). Sufficient vertical strain could cause bed separations
along the strata bedding planes or even step cracks. The total strain
(et), defined in Eq. 7, is an indicator of the severity of expansion
or shrinkage of a volume of rock strata under the influence of
subsidence process. It should be noted that the expansive type of
total strain (in positive value), reflecting the average intensity of
voids, is an indicator of the increase in porosity and permeability
for seepage flows of gases or water in the subsurface strata. It
can be an indirect indicator for reduction in confining pressure
to underground mine structures. For simplicity, the expansive
volumetric strain is also called voidintensity.

S ( x x , , zi 1 ) ai Ri
f s ( x' , z i )
e
Ri
i = 1, 2, , n(1)

x,

S ( x x , , zi1 ) ai n , Ri
f u ( x , , zi ) 2
xe
Ri h

x ( x, z )

dU ( x, z )
dx

z ( x, z )

dS ( x, z )
dz

i = 1, 2, , n(2)
The final subsidence and horizontal displacement at a prediction
point in the overburden are determined by integrating the
respective influence functions within the computing zone, as shown
in the following twoequations.

a
S ( x, zi ) i
Ri

Wi di 2 x

di1 x

S ( x x , , zi 1 ) e

x,

Ri

ai n
h Ri

Wi di 2 x

di1 x

(6)

t ( x, z ) x z x z

dx,

i = 1, 2, , n

U i ( x, zi ) 2

(5)

S ( x x , , zi 1 ) x , e

x,

Ri

(3)

dx ,

i = 1, 2, , n(4)
In these equations, ai, Ri, di1, and di2 are the final subsidence
parameters for the ith layer in the overburden strata and are very
important for subsidence prediction. Among them, ai is the
subsidence factor, Ri is the radius of major influence, di1 and di2 are
the offset distances of inflection points on the left and right sides
of the panel, respectively. These parameters are dependent on the

(7)

MULTIPLE-SEAM MININGINTERACTIONS
The subsurface deformations due to mining operations in
underlying seams have the potential to cause stability problems
to mine structures (i.e., coal pillars, entry floor, and roof) in the
upper seams (Fig. 2). Unstable pillars, roof, and floor can not only
cause significant problems to mining operations but also can create
hazardous working conditions as well as addition subsurface and
surfacesubsidence.
Multiple-Seam Interactions on Roof and FloorStability
Roof falls are the number one safety threat to underground
miners. Common roof falls in coal mines are tensile failures and
roof cutters. When an underground coal mine is affected by mining

31st International Conference on Ground Control in Mining

activities conducted in the underlying coal seam, the originally
stable mine roof could become unstable. Roof tension cracks could
be induced in zones with high void intensity, while roof cutters are
more likely to occur in high shear areas. The interactions could
also result in unstable mine floors problems such as floor cracks
in high tension zones and floor heaves in high compression zones.
The subsurface subsidence could also cause mine safety problems.
For example, the connected fracture zone between the mined coal
seams induced by the subsurface subsidence process could allow
the accumulated methane in the sealed mine areas to rush into the
active working areas to create a dangerous condition. Because of
its capability to predict the magnitude and distribution of various
subsurface deformations, the subsurface subsidence prediction
model can be used to guide mine design to avoid excessively
disturbed zones and to plan ahead on any mitigation measures to
minimize suchinfluences.
For assessing the stability of the mine roof and floor, the
predicted horizontal strain along the mine roof and floor should
be compared to their respective critical strain for initiating cracks.
For example, a tensile strain higher than 210-3 ft/ft is capable of
causing cracks in the immediate roof and floor that are made up by
shale type ofrocks.
Multiple-Seam Interactions on PillarStability
The stability of a coal pillar depends on its strength and the load
applied on it. The pillar strength is a function of its confinement
that is indirectly reflected by the pillar width to height ratio in a
single seam mining setting. However, when a pillar is disturbed
by mine subsidence, both the pillar load and its strength could
beaffected.
Change in pillar load can be related to the differential
subsidence between the roof and floor line of the pillar. The
pillar load under the influence of multi-seam influence, p,
can be calculated by Eq. 8. In the equation, sp is the pillar load
prior to the multi-seam mining effect (normally determined using
the tributary load method), Ec is the Youngs modulus of the coal
pillar, and ev is the subsidence-induced vertical strain at the pillar
location (sign convention: positive for tensile strain and negative
forcompression).

'p p Ec v

(8)

The pillar strength would decrease with the confinement

pressure that is related to the increased void intensity in the
subsurface subsidence process. The pillar strength under the
influence of subsurface subsidence, Sp, is calculated by the
followingequation.

S p' S p

(9)

The subsidence-influenced pillar strength is related to the pillar

strength prior to the multi-seam mining influence Sp which can be
determined by the Bieniawskis formula (Eq. 10), and is the pillar

strength reduction factor. The determination of this factor will be

discussed in a latersection.

S p 1 (0.64 0.36

W
)
H

(10)

In the equation, s1 is the in-situ coal strength (900 psi), W and

H are the pillar width and height, respectively. Subsequently,
the safety factor of a pillar under the disturbance of subsurface
subsidence can be determined byEquation11.

'

SF

S 'p

'p

(11)

If the pillar safety factor under the disturbance of mine

subsurface subsidence is less than a critical value, the pillar could
fail. Based on the published investigation cases (Morsey, Yassien
and Peng, 2006), it is proposed that a critical safety factor for
pillars to fail in a large area is1.0.
Void Intensity and PillarStrength
The Bieniawskis formula (Eq. 10) is widely used to estimate
the pillar strength in the process of pillar design and pillar stability
analysis. The pillar width to height ratio (W/H) contained in the
formula is an indirect measure of the confinement level of the
pillar. However, the pillar strength is also affected by other factors,
such as the structure and surface conditions of the discontinuities
inside the pillar and the roof and floor conditions. In order to
account for the effects of subsurface subsidence on the pillar
strength, the Hoek-Brown failure criterion is adopted here to
evaluate the pillar strength. The generalized Hoek-Brown (1997)
failure criterion for jointed rock masses is definedas

1' 3' ci (mb

3'
s) a
ci

(12)

In Eq. 12, 1 and 3 are the major and minor effective principal
stresses at failure, sci is the uni-axial compressive strength of the
intact rock material, and mb, s, and a are the material constants.
When considering the strength of a pillar, it is useful to have an
estimate of the overall strength of the pillar rather than a detailed
knowledge of the extent of fracture propagation in the pillar. This
leads to the concept of a global rock mass strength that could
be estimated by the following Mohr-Coulomb relationship, as
proposed by Hoek et al(2002).

Sp

2c ' cos '

1 sin '

(13)

31st International Conference on Ground Control in Mining

In Eq. 13, c and are the cohesion and angle of internal
friction. It can be further derived into Eq. 14 in the stress range of st
< 3 <sci/4.

S p' k ci

(14)

(15)

S p'

k ' ci k '

S p k ci k

(16)

The terms k and k in the previous equation are the pillar

strength reduction factors when the pillar is undisturbed (k) and
disturbed (k) by multi-seam mining. The reduced material constant
mb is a function of the unaffected material constant mi, the rocks
geological strength index (GSI), and the degree of disturbance (D)
as shown in Eq. 17. The coefficients s and a for the rock mass are
determined by Eqs. 18 and19.

mb mi e

se

GSI 100
)
9 3 D

GSI 100
)
2814 D

(17)

(18)

1 1 GSI /15 20 / 3
(e
e
)
2 6

GSI 75 0.95 ( t 10 3 )

(20)

Numerical Study of the Effect of Subsurface Deformation on

PillarStrength

m 4s a(mb 8s)(mb / 4 s) a1
k b
2(1 a )( 2 a )

considered as a mining-induced geological condition of the rock

mass. An empirical formula is established here to estimate the
GSI for the subsurface structures based on the subsurface total
straindistribution.

(19)

Coefficient D is a factor reflecting the degree of disturbance to

which the rock mass has been subjected by blasting damage and
stress relaxation. It varies from zero for undisturbed in-situ rock
masses to 1 for very disturbed rock masses. The significance of
the parameters and their values can be found in Hoek, 2004. The
Geological Strength Index (GSI), introduced by Hoek (1994) and
Hoek et al. (1995), provides a system for estimating the reduction
in rock mass strength for different geological conditions. The GSI
takes into account the geometrical shape of intact rock fragments
as well as the condition of joint faces. For underground structures,
such as tunnels, slopes, and mine openings that are easy to access
and observe the geological conditions, the GSI is determined using
the method proposed by Hoek and Brown (1997). However, in this
research, the geological conditions of the subsurface structures
under the disturbance of the subsurface subsidence are very hard
to observe. The subsurface total strain in the rock mass can be

In order to calibrate the empirical formula, numerical

simulations are also performed. The Fast Lagrangian Analysis
of Continua (FLAC) program package, capable of elasto-plastic
analysis of rock excavations with strain softening using the linear
Mohr-Coulomb failure criterion, was used in the simulation. A
FLAC3D model was developed to study the coal pillar strength
under the influence of subsurface subsidence. The model consists
of 8 ft coal seam, 30 ft roof strata, and 30 ft floor strata (Fig. 3).
The entry and crosscut are 20 ft wide. The elastic-perfectly-plastic
Mohr-Coulomb model is assigned for the rock strata, and the strain
softening Mohr-Coulomb model is assigned for the coal seam.
Roller boundary conditions are used along the sides and bottom of
the model. In order to establish the peak load for the pillar to carry,
the velocity of vertical displacement on top of the model is fixed
at a constant value of -110-5 ft/sec. The sum of reaction forces
at the base of the model is obtained via a FISH function (Itasca,
2006) to estimate the average vertical stress developed in the pillar.
The input geo-mechanical properties are the same as those used
by Lu et al. (2008). Four pillar widths (24, 40, 56, and 80 ft) for
pillar width to height ratios (W/H) of 3, 5, 7, and 10 are simulated.
The resulting stress-strain relationships are plotted in Fig. 4. The
pillar strengths based on the simulation are comparable to the
Bieniawskiequation.
To simulate the subsurface subsidence effect on the pillar
strength, the subsidence-induced horizontal strain and vertical
strain are simulated by applying the displacements on the side and
the top of the model, respectively, at a constant value of -110-5 ft/
sec. Different subsurface deformation values are simulated with
this model. The numerical simulation results are compared to the
results of the previous analytical model for validation purposes
in Fig. 5. It shows that the proposed pillar strength formula to
consider the subsurface subsidence effects agrees well with the
numerical simulationresults.
CASESTUDY
To demonstrate the capability of CISMP-MS program in the
studies of multi-seam interactions, a multi-seam mining case (Fig.
6) that caused damages to both subsurface mine structures and
to surface structures is modeled using the program. In this case,
an active room-and-pillar mine was developing its mains in the
Sewickley coal seam over a portion of the closed coal mine in the
Pittsburgh seam. The inter-burden between these two coal seams
was about 90 ft. The mine in the Pittsburgh seam had been mined
in the 1960s using the room-and-pillar method and then closed
afterwards. Pillaring operation was performed in a large portion of
the mine. The most noticeable mine structure left in the Pittsburgh
seam under the problems areas was an irregularly shaped support
area (about 250 ft long and 150 ft wide) left around a gas well.

31st International Conference on Ground Control in Mining

The spatial relationship among the two residential structures and
the layout of the active mine in the Sewickley seam are shown in
Fig. 6. The boundary between the pillared area and the area with
development mining only in the Pittsburgh seam mine are also
shown in Fig.6.
The surface subsidence damages including a ground crack and
a depression zone were reported near the surface structures. The
damages occurred on the mine structures in the active mine in the
Sewickley seam included rib spallings, cutter roof, floor cracks
and minor floor heave. The occurrence of all these surface and
subsurface structural damages coincided with an event in which
the raising mine water in the closed mine in the Pittsburgh seam
reached under the problem area. Underground observations made
in the active mine indicate that rib spalling and cutters are more
prevalent in the area above and around the boundary line in the
Pittsburgh seam. The large roof fall in the entries in the active mine
appeared to coincide with the edge of the support pillararea.

Figure 2. Subsurface subsidence profile and multi-seam

mininginteractions.

In order to study the subsurface subsidence effects of the

Pittsburgh seam on the Sewickley seam mine pillars and entries,
predictions were performed along the cross-section A-A and
in a rectangular area of BCDE as shown in Fig. 6. The predicted
subsurface subsidence and subsurface void intensity distribution
over the rectangular area BCDE in the Sewickley seam are plotted
in Figs. 7 and 8, respectively. The subsurface subsidence along
the prediction line A-A at the Sewickley seam around the support
pillar area of the Pittsburgh seam is predicted using CISMP-MS
and the results are shown in Fig.9.
The subsurface subsidence prediction indicates that the
subsurface deformation can reduce the average safety factor of the
upper Sewickley seam pillars by 11%. The original pillar safety
factor for the Sewickley seam mine is about 8.6, and the pillar
safety factor under the subsurface subsidence influence is 7.6,
which agrees well with the underground observation that the pillars
in the active mine were still intact. The resulting high pillar safety
factors and the fact that no pillar failure in the active mine strongly
suggested that the mining operation in the Sewickley seam is not
the cause for the reported surface subsidenceevents.

Figure 3. The three-dimensional discretized view of quarter

pillar (w/h=5).

Figure 1. Subsurface subsidence prediction model (h =

overburden depth; m = mining height; W = panel width; n =
number of overburdenlayers).

The maximum strain on the mine roof in the Sewickley seam at

the location near the edge of the support pillar was1.9610-2 ft/ft.
This is significantly higher than the proposed critical tensile strain
for roof cracking, 210-3 ft/ft, and more than sufficient to cause
the roof failure at in the active mine. Figure 8 shows most of the
observed massive roof falls (in cross hatch pattern), and roof cracks
in the active mine in the Sewickley coal seam are located in the
zones of high voidintensity.
CONCLUSIONS
The enhanced subsurface subsidence prediction model can be
used to analyze the interactions in multi-seam mining operations. A
number of important parameters to quantify the subsurface

31st International Conference on Ground Control in Mining

2500
W/H=3

W/H=5

strength, psi

2000

W/H=7
1500

W/H=10

1000
500
0
0

0.0005

0.001

0.0015

0.002

strain, in/in

Figure 4. Stress-strain curves under different W/Hratios.

3000
2500

pillar strength (psi)

Figure 7. Predicted subsurface subsidence in the Sewickley seam

in rectangular areaBCDE.

Analytical Model
FLAC3D

2000
1500
1000
500
0
0

0.01

0.02

0.03

Total strain (ft/ft)

0.04

0.05

Figure 5. Comparison between proposed pillar strength formula

and FLAC modelingresults.

Figure 8. Predicted subsurface void intensity in the Sewickley

seam in rectangular areaBCDE.
mining interactions has been incorporated in the subsidence
prediction model CISPM-MS for multi-seam mining operations. A
case study to illustrate its capability to assess the multi-seam
mining actions showed agreement with the fieldobservations.
REFERENCES
Hoek, E., 1994, Strength of rock and rock masses, ISRM New
Journal, 2(2), pp.416.

Figure 6. Spatial relationship among surface structures and

mains in the mines in the Sewickley and PittsburghSeams.
subsidence influences to subsurface mine structures have been
proposed. They include (1) the critical strain for causing tensile
cracks in the mine roof and floor, (2) the method to determine
subsidence-induced pillar load, and (3) the failure criteria based
method to determine pillar strength under the influence of
subsurface subsidence. The methodology to assess the multi-seam

Hoek, E., 2004, Practical rock engineeringan ongoing set

of notes, Rocscience,www.rocscience.com/hoek/corner/
Practical_Rock_Engineering.pdf.
Hoek, E. and E.T. Brown, 1997, Practical estimates of rock mass
strength, International Journal of Rock Mechanics & Mining
Science, 34(8), pp.11651186.
Hoek, E., C. Carranza-Torres, B. Corkum, 2002, Hoek-Brown
failure criterion2002 edition, Proceedings of the 5th North
American Rock Mechanics Symposium, Toronto, Ontario,
Canada, University of TorontoPress.

31st International Conference on Ground Control in Mining

Lu, J., K. Morsy, A. Ray, S.S. Peng, 2008.Effects of rock/coal
interface properties on coal pillar strength, Proceedings of the
27th International Conference on Ground Control in Mining,
pp. 262267. [http://icgcm.conferenceacademy.com/papers/
detail.aspx?subdomain=icgcm&iid=369]
Luo, Y., 1989, Integrated computer model for predicting surface
subsidence due to underground coal miningCISPM,
Dissertation, West Virginia University, UMI order No.
9020385, Morgantown, WV, 168p.
Luo, Y. and S.S. Peng, 2000, Prediction of subsurface subsidence
for longwall mining operations, Proceedings of the 19th
International Conference on Ground Control in Mining, pp.
163170.
Luo, Y. and S.S. Peng, 2010, Subsurface subsidence prediction
model and its potential applications in the study of longwall
subsidence effects on hydrologic system, SME Transactions,
Vol. 326, pp.458-465.

Figure 9. Predicted subsurface subsidence and horizontal

strain along A-A cross-section in the Sewickley seam above the
Pittsburgh seam.

Hoek, E., P.K. Kaiser, W.F. Bawden, 1995, Support of

underground excavations in hard rock, Rocscience,
rocscience.com/hoek/corner/Support_of_Underground_
Excavations_in_Hard_Rock.pdf.
Itasca Consulting Group, Inc., 2006, Flac 3D Version 3.1 Example
Applications, Minneapolis, MN, Itasca Consulting Group, Inc.,
pp. 2-12-17.
Jaiswal, A. and B.K. Shrivastva, 2009, Numerical simulations of
coal pillar strength, International Journal of Rock Mechanics
& Mining Science, 46,pp.779788.

Luo, Y. and B. Qiu, 2012, Enhanced subsurface subsidence

prediction model considering overburden stratifications, Paper
submitted to SME Publication (underreview).
Mark, C., F.E., Chase and D.M. Pappas, 2007. Multiple-seam
mining in the United States: design based on case histories,
Proceedings: New Technology for Ground Control in Multiple
Seam Mining. Ed. C. Mark and R.J. Tuchman, National
Institute for Occupational Safety and Health, (NIOSH) 2007110, pp.1528.
Morsy, K., A. Yassien and S. S. Peng, 2006, Multiple Seam
Mining Interactions A case Study, Proceedings of the 19th
International Conference on Ground Control in Mining,
pp.308314.
Qiu, B. and Y. Luo, 2011, Subsurface subsidence prediction model
and its potential applications for longwall mining operations,
Journal of Xian University of Science and Technology, Vol.
11(6), pp.823829.