Page 526 - Rock Mechanics For Underground Mining

MINING-INDUCED SURFACE SUBSIDENCE

the subsidence profiles shown in Figure 16.22 that the slope of the ground surface
and the induced strains must vary from point to point across the width of the panel.
Since differential vertical movement, horizontal compressive or tensile strain, tilt and
curvature can all adversely affect surface structures and utilities, in many cases more
severely than the subsidence itself, it is essential that means be developed of predicting
the values of these variables produced by trough subsidence.
The primary parameters of interest in this regard are:

the maximum subsidence, S max ;

the maximum ground tilt, G max ;

the maximum tensile and compressive ground strains, +E max and −E max ; and

the minimum radius of ground curvature, R min .

16.5.2 Empirical prediction methods
For many years, the most comprehensive and widely used method of subsidence
prediction was the empirical method developed by the then National Coal Board
(NCB) in the United Kingdom and described in the Subsidence Engineers’ Handbook
(National Coal Board, 1975). The NCB approach used a series of graphs relating the
major variables defined in Figure 16.22 and section 16.5.1. It was found that strain
and tilt were proportional to the maximum subsidence and inversely proportional to
the cover depth resulting in the following expressions:

+E max = 1000 × K 1 × S max /h
−E max = 1000 × K 2 × S max /h
G max = 1000 × K 3 × S max /h

where K 1 , K 2 and K 3 are constants of proportionality.
The curvature, 1/R, is directly proportional to strain and indirectly proportional to
the depth of mining so that

1/R min = K 4 × E max /h

where K 4 is another constant of proportionality.
The NCB method was widely used in the UK where the maximum subsidence was
said to have been predicted to within 10% in the great majority of cases. The fact that
the method gave such satisfactory predictions of subsidence profiles in collieries over
a wide geographical area was probably due to the fact that the nature and properties
of the Carboniferous strata involved were similar over the entire mining area. The
method did not account for the influence of major geological features such as faults
intersecting the panel or the deforming strata. Because of differences in geology and
rock mass properties, and the generally site-specific nature of empirical correlations,
attempts to apply the NCB correlations to longwall coal panels in other parts of the
world met with variable success (e.g. Hood et al., 1983, Galvin, 1988, Alejano et al.,
1999). However, many of the concepts developed by the NCB have been found to
be applicable elsewhere and have been used in other locally developed empirical
methods (e.g. Holla and Barclay, 2000).

508