REP38979

• Permit No: C1996083
• IBM Index Class Name: Report
• Doc Date: 5/8/2006
• Doc Name: 2005 Bi-Annual Subsidence Report
• From: J.E. Stover & Associates Inc
• To: DMG
• Permit Index Doc Type: Subsidence Report
• Media Type: D
• Archive: No

u
u
4. SUBSIDENCE MODELING
4.1. Methodology
Surface subsidence is the readily observable
manifestation on the ground surface of the
displacement field surrounding the underground
portion of the mine. Predicting subsidence
magnitude, therefore, constitutes a particular
solution of the overall problem of finding the
induced displacement field. To study subsidence
phenomena and estimate the magnitude of
subsidence, a number of empirical, physical, and
numerical methods have been used.
Empirical methods, including profile functions,
influence functions, and graphical methods were
proposed by the British National Coal Board. These
methods involve the analysis of existing subsidence
from an area to predict future subsidence effects.
These methods are based on the mathematical fit of
a considerable number of measured subsidence
profiles. They apply to geologic conditions in the
area where they were developed and require
adjustments if they are applied to different strata
conditions.
In this study, weused athree-dimensional influence
function method while accounting for site-specific
conditions using the subsidence monitoring data
from the study area. These methods have become
very popular for the prediction of subsidence and
surface strains within the last two decades [12, 13,
14]. They are superior to graphical methods because
they can be used to model an entire longwall block
while allowing an examination of the sensitivity of
results to variations in seam thickness, pillar
designs, panel dimensions, and overburden thick-
ness.
4.2. Model Calibration
BRL monitored the long-term surface response to
longwall mining over the D1 to D9 panels. The
results of this monitoring were used to establish site-
specific modeling parameters while considering
regional data within the Somerset basin.
Panels DS through D9 were selected for model
calibration considering closer spacing of monuments
A -8
and frequency of monitoring. Initially, we focused
on the western portion ofthe longwall block because
of closer spacing of monuments and lack of past
mining activities in the US Steel and King mines. As
illustrated in figure 1, U.S. Steel and the operator of
the King Mine practiced irregular mining in the B
and C seams to the east side of the longwall block,
creating some uncertainty due to reactivation of
movement in the old workings. Because ofrather the
uniform overburden thickness over the eastern
portion of the panels, we used a fixed cover for
initial modeling while comparing measured and
calculated deformation for the D7, D8, and D9
panels.
Figure 7 compares measured and calculated results
for select monuments located toward the center of
the longwall area about an approximate north-south
cross section. Good agreement is shown. The model
appears to underestimate subsidence at station 22D
and overestimate it at 14C. The difference is
probably caused by the changes in topographic
conditions at these monument locations.
Additional modeling included digitizing the
overburden contours and thus incorporating the
variability in topographic conditions to improve
goodness of fit. Slight improvements between
observed and calculated values were observed
initially using constant subsidence factor and angle
of draw.
-~~ 2IA
o • fvleaaiaetl Data
zaro zsao aooo aaoo aooo uoo soon m saoo
- - - - - - -
~~ D~ PareA 18G Da Parel C9 Pam
_y - - - - - ~ - - -
a-= - --- -- ~ -----
0
~ _ _ ~-
a-0 - - - _ _
'z - ~ - -_- - -__
•
s - -- - ---
~ 22D
-a
6ststt (%)
Fig. 7. Compared calculated and measured subsidence
along anorth-south cross section, constant cover,