Methods based on numerical modelling

Integrated Subsidence Prediction Technique (ISP-Tech)

As mentioned in Chapter 2, lack of accurate input information is one of the

disadvantages of the numerical modelling techniques. Unlu et al. (2013) made

an effort to rectify this disadvantage by suggesting a methodology to import

and improve the reliability of the key input parameters required for Finite

Element (FE) modelling of subsidence from different sources of data. They

called this approach Integrated Subsidence Prediction Technique (ISP-Tech).

In this approach, they propose using geographic information system (GIS)

and mining information system (MIS) data, which combine information from

geological cross sections, geological maps, drill hole data, and etc. to derive

the necessary information for mesh building in the numerical model on var-

ious sections of the mine. Phase2, as an FE software, is then utilised to

perform elaso-plastic analysis on the generated mesh. The result from the

FE analysis on each 2D section of the mine are compiled to generate an

initial 3D subsidence map after the mining activity. This initial prediction

is compared with the GPS and/or DIn-SAR subsidence measurements. If

the results are incompatible, the FE mesh, rock mass properties and other

numerical model variables are refined and the analysis is performed again.

Once compatible results are acquired, Unlu et al. (2013) described that the

predict the surface subsidence before longwall extractions takes place.

They utilised the ISP-Tech method for prediction of mining-induced sub-

sidence in two neighbouring multi-seam cases in Turkey. They used the sub-

sidence data from a case study, which has been mined, to achieve accurate

predictions of subsidence by means of ISP-Tech and then employed the same

method for prediction of the subsidence for a to-be-extracted neighbouring

multi-seam mine.

Other numerical methods

Although various examples of numerical modelling of stress analysis, caving

process and stability of workings around multi-seam panels are available (e.g.

Zipf 2005; Suchowerska et al. 2013, 2014b), there are only a few examples of

numerical analysis of multi-seam mining-induced subsidence available in the

literature (Suchowerska et al., 2014a, 2015; Khanal et al., 2015; Adhikary

et al., 2015). In these examples, various researchers employed the Finite

Element Modelling (FEM) for multi-seam subsidence analysis, however, dif-

ferent approaches were considered to simulate strata stratification, rock mass

behaviour and caving process around the multi-seam workings. These works

will be thoroughly discussed in Chapter 5.

3.5

Summary

Multi-seam subsidence observations (Section 3.2) illustrate that the single

ference can be summarised as follows:

• Increased incremental maximum multi-seam subsidence factor (e.g. Kapp 1982; Holla and Thompson 1992; Li et al. 2007a, 2010; Holla and Bar-

clay 2000; MSEC 2012a,c; DgS 2012; Suchowerska 2014);

• Increased or decreased angle of draw from case to case (e.g. MSEC 2012a; Ghabraie et al. 2015b,a; Mills 2013);

• Change in location of maximum multi-seam incremental subsidence from case to case (e.g. MSEC 2012d; Li et al. 2010; Ashton Coal 2014;

Ghabraie et al. 2015b);

• Occurrence of localised maxima points in the subsidence profile (e.g. MSEC 2014, 2015) and

• Change in the shape of the subsidence profile, tilt and curvature (e.g. MSEC 2012a; Mills 2013; MSEC 2012c).

These differences in multi-seam subsidence profiles make it significantly dif-

ferent from that of single-seam profiles. In addition to the differences of

multi-seam and single-seam subsidence profiles, shape of the multi-seam

subsidence profiles differ notably based on the multi-seam mining configura-

tion. In other words, multi-seam subsidence is a case dependent phenomenon

(Galvin, 2016) and, unlike single-seam subsidence, a single consistent profile

cannot be utilised for various multi-seam cases.

various researchers have suggested different approaches for prediction of

except numerical based methods, are based on generalised concept of the sub-

sidence characteristics and prediction of smooth trough shaped subsidence

profiles. The generalisation of the subsidence profile does not allow for predic-

tion of irregular subsidence profiles, changes in the angle of draw and location

of the maximum subsidence, which commonly occur above the multi-seam

panel extractions.

Among these methods, only the method developed by Sheorey et al.

(2000) allow for consideration of asymmetric subsidence profiles (by altering

the location of predicted maximum subsidence) and separating the effect of

multi-seam extraction areas from single-seam areas. However, their method

is based on subsidence observations in Indian Coalfields, which cannot be

directly applied to other countries or coalfields. Also, this method does not

allow for prediction of local irregularities of the multi-seam subsidence profiles

and change in the location of maximum incremental multi-seam subsidence,

which are often observed in the multi-seam cases.

Similarly, the adapted subsidence profiles in the IPM for the two worst

case scenarios of stacked and staggered cases are generalised and are not able

to consider irregular subsidence profiles and different angles of draw from case

to case. GIFM and CISPM-MS methodologies suffer from similar deficien-

cies for prediction of multi-seam subsidence and are not able to account for

changing location of maximum subsidence, different angles of draw, effect of

changing multi-seam mining configurations and irregular subsidence profiles.

Numerical methods, on the other hand, depend significantly on the ac-

curacy of input data for prediction of multi-seam subsidence. ISP-Tech, for

suitable FE mesh and refine it based on trial and error to achieve reliable

subsidence predictions. Acquiring this information is costly and commonly

not available for lots of mine sites. In addition, these methods are mostly

based on simplified assumptions of strata behaviour and loading/unloading

of the rock masses. These deficiencies impose restrictions for application

of these methods with high level of confidence in order to achieve reliable

multi-seam subsidence predictions.