Sei sulla pagina 1di 10

Rock Mechanics, Fuenkajorn & Phien-wej (eds) 2007.

ISBN 978 974 533 613 1

A prediction method for time-dependent land subsidence area over room-and-pillar mining
N. Phien-wej & T. Suwanishwong Geotechnical and Geoenvironmental Engineering, Asian Institute of Technology, Thailand

Keywords: Surface subsidence, prediction methods, subsidence troughs, time-dependent property, room-and-pillar mining ABSTRACT: Surface subsidence is recognized as a problem in most countries, particularly those with significant mining and other underground resource extraction industries. Subsidence troughs induced by room and pillar mining can occur over abandoned or active mines. The knowledge about the surface activity caused by mining, and the subsidence prediction enable efficient prevention and repairs of the mining damage. Accurate and reliable prediction can influence significantly the strategy of a mine operation. As some rocks and solid minerals have strongly time-dependent properties such as potash, shale and rock salt, subsidence of the ground surface over the mine area with time is one of the key factors in planning the mining. Subsidence prediction methods are basically developed to calculate the final subsidence profile. General means of the predicting subsidence due to underground mining is 2-D modeling using numerical method such as finite element and finite different. These methods can well simulate the room and pillar subjected to various loading and timedependent conditions. However, the calculation process is quite complicated and also has some constraints in shape and dimension of room and pillar. Note that no one method can solve all the problems of subsidence due to underground mining. Each method has its own value and range of application. For a case study, the method should be selected according to its suitability given the analysis objective together with the special mining and geological conditions. In order to simplify the predicted subsidence profiling and contouring, this paper introduces a prediction method, critically reviews the assumptions of subsidence due to underground mining, the influence parameters as well as procedure of predicted subsidence profiling and contouring. 1 INTRODUCTION

Surface subsidence can be regarded as the movement of the ground surface which takes place due to the extraction of valuable mineral or fuel. Wherever mineral resources have been mined, underground voids are left behind. Overlying layers of rock and earth naturally tend to sink to fill these voids. When that sinking reaches the ground surface, it is called mine subsidence. Subsidence may occur regardless of the age of the mine, the depth, or how much mineral resources were left behind. This happens at very slow rates. Salt that might be placed

119

A prediction method for time-dependent land subsidence area over room-and-pillar mining

as backfill in the mined-out rooms will prevent sudden subsidence, and will gradually reduce the rate of subsidence until it eventually stops. In abandoned mines, troughs usually occur when the overburden sags downward due to the failure of remnant mine pillars, or by punching of the pillars into a soft mine floor or roof. Abandoned mines may collapse many decades after the mining is completed, if the mine workings were not designed to provide long-term support. The principal method of underground extraction is room and pillar mining which often gives rise to extraction panels of irregular shapes making subsidence predictions difficult because all earlier empirical approaches have been two-dimensional. In fact, such terms as width-todepth ratio of extraction, subsidence profile and even though angle of draw assume a plane strain condition. Of all the different tools available for subsidence modeling, the simplification of predicted subsidence profiling with suitable assumptions and modifications appears to be a powerful alternative for complete subsidence prediction for all shapes of extraction panels. Numerical techniques offer some possibilities for subsidence modeling, but these are invariably fraught with some difficulties. Continuum methods, (finite or boundary element, fast Lagrangian analysis of continua) completely fail, even though with elasto-plastic analysis performed using realistic rock mass strength inputs, assuming that such inputs are possible for each bed. One aspect, which is generally ignored in numerical modeling, is the possibility of each bed having different horizontal in situ stresses. Because of these difficulties, the simplification of predicted subsidence profiling method has been resorted to as a more convenient and simpler alternative. In the present work, the authors attempt to extend the procedure of the predicting subsidence profile due to mining a single panel to the situation where there are a number of adjacent panels separated by pillars as in the panel and pillar mining system. With this system, the method uses the superposition principle in an unjustified way by simply adding the predicted subsidence of each panel to obtain a total predicted profile. The input data include the data on excavations and location coordinates of grid on the surface, and the output data are their height coordinates. Consider edge effect of the room on the subsidence and the timedependent response of the materials following staged mining. 2 2.1 BASIC PRINCIPLE Mechanics of mine subsidence

There are two types of subsidence: 1) pit, also called sinkhole, and 2) sag or trough. (The term sinkhole more properly refers to solution-collapse features in limestone.) Pit subsidence is characterized by an abrupt sinking of the surface. Sag subsidence is a gentle, gradual settling of the surface. It is associated with pillar punching or pillar crushing of deep mines. The area of mine subsidence proportionally increases with increasing width of unsupported roof rock. The potential area of subsidence is equal to the extraction are plus an area surrounding the extraction area measured by an angle of draw from the vertical at the edge of the extraction area. The deeper the mine, the larger the area potentially affected by mine subsidence at the surface. The length of time for occurring of mine subsidence increases with increasing depth of mining and increasing competency of overburden. The type and amount of roof support in addition to pillars left in the mine also affect subsidence. For our 120

Rock Mechanics, Fuenkajorn & Phien-wej (eds) 2007. ISBN 978 974 533 613 1

case we consider subsidence due to time-dependent ground deformation over mined out area due to changes of state of stress without yielding of pillars or collapse of roof. It is different from the case cited above. The magnitude of subsidence is less and generally not significant. 2.2 Angle of draw

The zone of subsidence influence zone extending from the boundary of mined out area can be measured by the angle of draw. The surface position of the boundary between areas of subsidence and no subsidence is defined by the angle of draw as shown in Figure 1a. This is the angle between a vertical line drawn upward to the surface from the edge of the underground opening and a line drawn from the edge of the opening to the point of zero surface subsidence. The larger the angle of draw the wider will be the zone on the surface in which subsidence should occur. 2.3 Factors influencing subsidence

Many parameters affect the surface expression of subsidence. It seems that the most predominant factor influencing the extent and magnitude of subsidence is overburden depth and ground strength. Other influencing factors include, but are not limited to, mining method, mining height, geologic features of overlying and underlying strata, extraction ratio, strength of intact rock, state of stress, topography and time. These factors should be considered in evaluation of the mine subsidence. 2.4 Time dependent response Soils and rocks possess time dependent properties which usually result in gradual increase in deformation and reduction in strength under the constant application of shear stress. This phenomenon is usually referred to as Creep. Some rocks exhibit significant creep response such as rock salt, potash, mudstone and clayshale. Creep can be attributed to two different mechanisms; mass flow and propagation of microfractures. Mass flow behavior is commonly associated with certain evaporite rock types such as halite and potash. Creep associated with microfracture propagation has been observed in most rock types. Creep will generally occur if the applied stress is within the range associated with unstable fracture propagation. For this reason, estimates of creep response for mining founded on rock masses require specialized studies and research. 3 3.1 METHODOLOGY OF PREDICTING SUBSIDENCE Influence parameters

The influence distance from edge of mining can be simply expressed as follows: d = Z tan (1)

where d is influence distance, Z is depth of mining and is angle of draw. The illustration of these influence parameters and influence zones are shown in Figure 1a. In order to predict the maximum possible subsidence, max, finite element models are constructed to represent the geological formations in the panels under variety of room-pillar dimensions as shown in Figure 1b. A series of finite element analyses have been performed 121

A prediction method for time-dependent land subsidence area over room-and-pillar mining

on the proposed mining. Conservative approach and assumptions have been used in assigning the material properties, the model geometry and the initial and boundary conditions. A timedependent finite element program has been used in the analyses.
CL CL

Ground surface

Subsidence trough Z Room P Room P Room P Room


A B

Figure 1a. Room-and-pillar mining.

CL

CL

Ground surface

roof R P floor

Figure 1b. Model for 2-D numerical analysis (section A-B). To determine the subsidence in the area in the vicinity of mine boundary, the numerical elastic solution on characteristics of ground movement in the face advance influence zone for tunneling (Brown et al., 1983) is adopted. The solution yields the distance in the vicinity of the mine boundary within which the ground movement is deviated from the plane strain movement due to the boundary effect. The characteristics of the roof settlement at various distances from the boundary within the zone are defined. The subsidence is expressed in terms of full subsidence and the location of the profile points. There are four influence zones which will be identified denoted as zone A, B, C and D as shown in Figure 2a-2b.

122

Rock Mechanics, Fuenkajorn & Phien-wej (eds) 2007. ISBN 978 974 533 613 1

x Ground surface max max/3

B Mining

Figure 2a. Section view to illustrate the influence zone identification.

d d

d A Mining B

d C D

Figure 2b. Plan view to illustrate the influence zone identification. The equation for estimate the displacement at various distance from edge of mining are linearly expressed as shown below A = max B = (2max*x/(3d))+ ( max/3) C = (-max*x/(3d))+( max/3) D = 0 3.2 Normalization of time dependent subsidence over a room area (2.1) (2.2) (2.3) (2.4)

2-D single room numerical analysis is made to evaluate ground settlement above the room. Width and height of room, overburden depth and rock layers in roof and floor are considered in the analysis. With consideration of creep properties of the rock, the visco-elastic analysis yield relationship between settlement with elapsed time and the magnitude of settlement at any given time in long term as shown in Figure 3a. Any creep laws may be introduced in the analysis. Utilizing the principle of normalization, it is assumed that the relationship between the settlement and elapsed time at a given depth of mine can be normalized and a unique normalized curve of the relationship exist for any room dimensions provided that the states of stress in the pillar and roof remain in elastic range. 123

A prediction method for time-dependent land subsidence area over room-and-pillar mining

Mining Depth 250 m 70 60 Subsidence (cm) 50 40 30 20 10 0 0 5 10 Tim e (years) 15 20


Room H. 2 m Room H. 4 m Room H. 5 m

Figure 3a. Subsidence-times plot as a function of time.

Normalized Subsidence

1.00 0.90 0.80 0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0 5

Mining Depth 250 m

y = -0.0016x2 + 0.0707x + 0.1168 10 Time (years) 15 20

Figure 3b. Normalized subsidence over mined out area as a function of time. Once the subsidence of each room height at the same depth were normalized, the relationship between normalized subsidence and time after the start of mining can be expressed as the 2nd order polynomial equation as shown in Figure 3b. This equation will be simply used as the time-dependent property representative in the vicinity of considered room-and-pillar at particular depth in predicting subsidence. 3.3 Input parameters

Input parameters for simplification of predicting subsidence include: Coordinate of panels/extracted areas Coordinate of grid on ground surface Mine timing (time after the start of mining) Relationship/equation between displacement and mine timing Angle of draw Depth of mining

124

Rock Mechanics, Fuenkajorn & Phien-wej (eds) 2007. ISBN 978 974 533 613 1

By these input parameters mentioned above, the subsidence profiles in any considered year can be calculated and its procedure are presented in section 3.4-3.5. 3.4 Influence zones identification and calculation of subsidence profiles

As a number of coordinates, the spreadsheet was simply developed in Microsoft Excel (Microsoft Office 2003) to assist the calculation including influence zones identification and subsidence profiles calculation. The study area is divided into a grid with proper scale therefore each point will have its own coordinate. Square elements of surface grid were laid overlapping the panel up to the draw limits of considered area. By using assisting spreadsheet, firstly, influence zones of each extraction areas were automatically identified by input the coordinates of responsible extraction areas. Next, the predicted surface subsidence-time relationship was input in spreadsheet. The sequence of extraction was properly classified based on the excavated year as illustrated in Figure 4. The responsible time after the start of mining were used, called mine timing, to input in the spreadsheet. Consequently, subsidence of each coordinates was calculated corresponding to their zone of influence and mine timing. The superposition principle also was used to determine the total subsidence by summation of all corresponding subsidence within the extracted area. Subsidence usually occurs gradually when it is concurrent with mining. After cessation of mining, subsidence may continue to occur in a steady, gradual manner, or it may cease for a period, to be followed by failure at some later date. 3.5 Contouring and subsidence profile

In order to estimate subsidence profiles it is necessary to define contours of estimated subsidence. The subsidence contours over the central portions of the mine at various times after the start of mining could be based on the estimates of maximum subsidence, which is the subsidence away from any edge effects. Subsidence contour map was constructed using Surfer8.0 software by input all the coordinates and corresponding subsidence which were determined previously as shown in Figure 5a-5d. Finally, the subsidence contour can be used to generate cross-sections easily as well as the ratio between differential settlement and distance as shown in Figure 6 and Figure 7, respectively.

1st year 3rd year

1st year 3rd year 5th year

1st year 3rd year 5th year

Figure 4. Sequence of mining based on extracted year (Plan view).

125

A prediction method for time-dependent land subsidence area over room-and-pillar mining

(cm)
1000

14
800

12 10

600

8 6 4
200

400

2 0
800 1000

0 0 200 400

A
600

Figure 5a. Subsidence contour at 2nd year after the start of mining.
(cm)
1000

22 20
800

18 16
600

14 12

400

10 8 6

200

4 2

0 0 200 400

A
600 800 1000

Figure 5b. Subsidence contour at 4th year after the start of mining.
(cm)
1000

28
800

24 20 16
400

600

12 8

200

4
A
0 200 400 600 800 1000

Figure 5c. Subsidence contour at 6th year after the start of mining. 126

Rock Mechanics, Fuenkajorn & Phien-wej (eds) 2007. ISBN 978 974 533 613 1

(cm)
1000

42
800

34

600

26

400

18

200

10

0 0 200 400

A
600 800 1000

Figure 5d. Subsidence contour at 10th year after the start of mining.
Ground Surface

20 Settlement (cm)

40

2nd year 4th year 6th year 10th year

60 66 70

max (20th year)

0 (600, 0)

200

400 600 Distance (m)

800

1000 (600, 1000)

Figure 6. Settlement-distance relationship along section A-A at various years after the start of mining.
0.01

0.001 Diff. Settlement/distance

0.0001

1E-005

1E-006

2nd year 4th year 6th year 10th year

1E-007 0 (600, 0) 200 400 600 Distance (m) 800 1000 (600, 1000)

Figure 7. Differential settlement/distance ratio and distance along section A-A. 127

A prediction method for time-dependent land subsidence area over room-and-pillar mining

CONCLUSIONS

In order to predict the maximum possible subsidence, finite element models are constructed and analyzed to represent the geological formations in the panels under variety of room-pillar dimensions. To determine the subsidence in the area in the vicinity of mine boundary, the numerical elastic solution on characteristics of ground movement in the face advance influence zone for tunneling was adopted. The principle of normalization was utilized as the assumption that a unique normalized curve of the relationship exist for any room dimensions provided that the states of stress in the pillar and roof remain in elastic range. The procedures presented in this paper clearly demonstrate the ability of the simplification of prediction method for time-dependent land subsidence area over room-and-pillar mining including determination of subsidence-time relationship, influence zone identification and calculation of subsidence profile. Spreadsheet can be easily made and developed using Microsoft Excel. The subsidence contour can be well plotted using Surfer8.0 software and the cross-section can be simply depicted as shown in the illustration results. REFERENCE Brown, E.T., Bray, J.W., Ladanyi, B. & Hoek, E., (1983), "Ground response curves for rock tunnels," ASCE, Journal of Soil Mechanics and Geotechnical Engineering, Vol.109, No.1, pp. 15-39.

128

Potrebbero piacerti anche