Stability of Crown Pillar during Transition from Open Pit to Underground Mining

Abstract

Mineral deposits at shallow depth are on the verge of exhaustion and transiting to underground mining operations, for extraction of deposits at depth has now become inevitable. The depth of open pit mine is generally governed by the stripping ratio and lateral extent of lease. When open pit reaches its ultimate designed depth and the ore is still existing at greater depth transition is must and barrier pillar needs to be left in between both the workings. If underground mining is below an opencast mine then stability of barrier/crown pillar between open pit and underground mining becomes critical especially when non-caving method is to be followed. It has its own implications and maintaining safety of pit and slope along with underground workings becomes a real challenge. Ground monitoring in terms of induced stresses and displacement holds paramount importance. The most popular method in metal mining is Long Hole/Blast Hole Stoping Method in top down sequence. Earlier hydraulic fill was common and is now being slowly replaced by Paste fill. In this chapter a few challenges existing in Indian mining scenario is discussed along with review of methodologies for optimum design of crown pillar. Parameter influence is also explored for its stability.

1. Introduction

The ore that exists at shallow depth is usually extracted by surface mining. However, deposits that lies at greater depth is exploited by underground mining. Most of the opencast mine has to undergo transition from open pit to underground at certain depth when open pit mining is no longer feasible either due to economy or due to limitation in lateral extent. This transition requires a barrier pillar in between open pit and underground mine to isolate one working from another. This barrier is commonly termed as Crown pillar. It is nothing but the ore bearing rock that is left at the bottom of the open pit for support of underground workings and to extract the ore deposit existing at greater depth safely.

This barrier pillar plays a vital role during the entire mining operation and its stability analysis is of utmost important for the protection of both surface mine slopes and underground mine infrastructure. The main purpose of transition crown pillar is to protect the ground surface and underground mine from subsidence and any material inflows like water, soil and rock into the mine while underground mining activity is going on. In mine closure planning too the evaluation of the stability of crown pillar is one of the most important criteria [1]. Many metal mines in India and across the world are at greater depth more than 600 m [2]. Hence, optimization of crown thickness and stability analysis of crown pillar is quite critical for safe underground operations. Preferably, an optimum size of pillar is required to prevent undue loss of ore.

2. Existing methodologies

Apart from a few structural mechanics approach that does not consider rock mass characteristics only two Empirical approaches and Numerical Modeling approach exists for design of Crown Pillar. Initially, the thumb rule was to compute ratio of crown thickness (T) to span (S) for design. A ratio of 1:1 is considered as good rock while thickness to span ratio of 3:1 or more is considered as poor. It is also quoted as Thick cross ratio method by Sun et al. [3]. Later it was modified and linked to rock mass as T/S = 1.55 Q^−0.62 where Q is NGI’s rock mass quality index. Few structural mechanics based analytical approaches like Slab Beam Theory, Ruppenpit formula, Compressive Arch Theory, Engineering calculation method, etc. have been developed [4] but not that useful for mining scenario as it does not take into account characteristics of rock mass. Software named CPillar is developed by Rocscience that computes FoS of crown pillar using deterministic analysis, however, widely used method of pillar design is scaled span concept developed by Carter et al. [5, 6].

Carter proposed Scaled Span approach long back in late 1980’s for surface crown of near surface mined openings. Originally, it included 200 case record of near-surface openings with 30 failure cases. Later in 2002 and 2008 further modifications were made and 114 case studies were added by extending the original data base. New case studies were primarily from shallow dipping situations, including twenty-one additional failure cases for pillar reliability in long term closure planning. These cases included both hard rock situations and soft rock examples including coal [7]. Failure probability concept is also introduced in 2008. In 2014 guidelines for use of scaled span concept was published and experience from 500 cases with over 70 failures were utilized to guide design of crown pillar. Scaled crown span method have been used in a number of cases but inadvertent failure still continues to occur [7]. It probably does not involve crown pillar design during transition phase of opencast mining to underground mining, stress reorientation in rock mass on account of open pit mining and its subsequent effect on crown pillar. It is limited to deposits being mined at very shallow depths by underground mining. Influence of higher working depth, slope stability scenario and simultaneous open pit and underground mining stage where blasting in open pit is likely to disturb integrity of crown pillars is perhaps also not included.

Bakhtavar et al. [8] has developed an empirical relation to estimate crown pillar thickness between open pit and underground mining. The equation has been derived from data sets of a few case studies through multiple regression. But its application in real field has not been found reported. Use of numerical modeling like 3DEC and FLAC3D are also cited in literature. Numerical modeling is sensitive to input parameters and needs to be calibrated. Numerical code RFPA2D is used to analyze the stability of boundary/crown pillar [9]. Failure of rock mass that occurred at Yanqianshan Iron Mine is studied [10] using numerical modeling where the mine was worked by underground mining under an open pit. Kalenchuk [11] analyzed crown pillar behavior through FLAC3D to ascertain crown thickness in a coal mine [12]. Crown thickness for the transition zone of a gold mine was designed using FLAC3D and same model was used for replacement of the crown with artificial pillar [4]. FLAC3D is used to perform parametric study to investigate the effect of fault properties and horizontal-to-vertical in-situ stress ratio on surface crown pillar stability [13].

Based on the analysis of failure cases in most of the surface crown pillars, five pattern of failures mechanisms were also identified by Golder [5, 14, 15] as plug failure, chimneying, caving, unraveling and delamination. Each of this failure represents unique feature and are limited to near-surface mine openings. Failure of the surface crown is a large-scale loss that may result in subsidence or sinkhole formation.

3. Lessons from abroad

A few cases where some kind of crown pillar failure occurred is discussed to visualize the likely issues a mine can face when crown becomes unstable and also the reasons associated with it. Failure that occurred on the north pit wall of Palabora Mine, South Africa includes the collapse of the crown pillar that resulted when block caving was initiated 400 m below the open pit having a depth of 800 m from surface. From modeling study it was predicted that failure is most likely to occur on the north wall, which is mainly controlled by the underground working caving operations and the rest east, south and west walls are appeared to be in stable condition for long time [16]. It is also found that pervasive joint sets in the north wall form wedges that daylight into the caved region below the pit.

Two rock mass collapse was observed in Shirengou iron ore Mine of China, one on account of fault movement and another due to water seepage. The method of working was Short-hole shrinkage stoping method. Although design thickness of crown pillar was around 25–30 m but the actual thickness was quite less mainly due to over excavation and illegal extraction. As a result, a few goaves were just 4–8 m below pit bottom. From the detailed analysis, it was found that fault lying in the crown pillar had undergone movement due to the underground mining activity and it propagated that finally leads to collapse of the crown pillar. Prior to the failure, several abnormalities were observed in the micro-seismic parameters such as abrupt change in energy release and stress levels [17].

In Chuquicamata copper ore mine of Chile method of working used was Panel caving. The main geological aspects was the presence of regional fault and shear zone. The rock mass consists of several discontinuities and the failure was initiated due to the earthquake in December 1967. Due to this earthquake the discontinuities began to open at the south end of east wall. Additionally, work of excavation and the blasting in the mine leads to collapse of the crown pillar in the upper part i.e., north block of the slope. It was found that crown pillar failure occurs in a sequence of events and finally lead to collapse of pillar [18].

Slope wall collapse, water percolation in underground, formation of pothole and movement along fault plane were a few observations. Failure initiated gradually with small movement on slope and then extended along weaker portions. Mainly method of mining was either block caving or sub-level caving except Shirengou iron ore mine, presence of fault zones and mining near the toe of overall pit slope where barrier thickness was insufficient were identified as the causes of instability. Impact of underground blasting is also visualized. In case of caving method, it is obvious to have instabilities if other factors are not given proper attention.

The crown pillar faces a new challenge in dealing with the change in geo-mining conditions from open pit to underground. The vertical stresses acting on the pillar usually increases with increasing depth of mining, which is not commonly encountered in the case of surface crown. Hence the behavior of the pillar under these stresses must also be thoroughly examined for optimum design of crown, and if necessary, an additional support system must be provided to ensure the pillar’s stability. Critical factors must also be identified to have a proper understanding of the mining scenario and finally to have an optimum design methodology.

4. Indian geo-mining condition

Compared to coal mines, limited number of metal mines exists in India. Orebody width varies from 1 m to 75 m in the range of 5, 12, 20 and 35 m to 75 m). At few places, it is in range of 80–120 m. Orebody dip is in the range of 55 to 85°. Ultimate Pit depth varies from as shallow as 20 m to as deep as 380 m. Number of small mines in and around Central India has undergone transition at a very shallow pit bottom depth of 20–80 m mainly due to small lease area and limited lateral extent. Strike length varies from merely 100 m to as long as 2200 m. RMR varies from Fair to Good in case of orebody but for hangwall and footwall poor RMR is also observed. In Sukinda chromite mine however ore is very soft up to a certain depth and as of today no suitable technique is evolved to extract the soft ore deposit. In Rampura Agucha mine shear zone exist in orebody and a crown thickness of 70 m is kept to ensure its stability. List of the metal mines with pit bottom depth is given in Table 1.

Name of the Mine	Mineral extracted	Open pit depth (m)
Rampura Agucha	Lead-Zinc	380
Malanjkhand Copper Project	Copper	240
Sukinda Chromite Mine	Chromite	125–175
Palaspani Manganese Mine	Manganese	92–125
Dongri Bujurg Mine	Manganese	160 (Proposed)
Kandri Mine	Manganese	83–100 (approx.)
South Kaliapani	Chromite	115
Beldongri Mine	Manganese	50–55
Kathpal Chromite mine (closed)	Chromite	38
Pandharwani Manganese Mine	Manganese	50
Ramrama Manganese Mine	Manganese	22–25
Sarubil Chromite mine	Chromite
Jagantola Manganese Mine	Manganese	36
Katangjhari Mine	Manganese	36
Jam mine	Manganese	15–42
Netra Mine	Manganese	16
Nuashahi Chromite mine (closed)	Chromite	30–35 (approx.) (pit backfilled)

Table 1.

List of the metal mines with pit bottom depth.

From the analysis of literature and experience of modeling stability in different case studies various parameters that influence crown pillar stability are identified as controllable or uncontrollable factors as given in Table 2.

Controllable Factors	Un-Controllable Factors
Ultimate pit depth	Orebody geometry i.e., orebody width and dip,
Exposed width of crown at pit bottom	Rockmass quality i.e. rock mass strength, RMR
Extent of opencast working	Geological disturbances – fault or such hidden structure
overall slope angle	Presence of water
Effect of open pit blasting if simultaneous working exist	In situ stress state
Sequence of mining	stress state on account of open pit
Method of mining with Backfill	Rainwater influence/infiltration

Table 2.

Various parameters that affects crown pillar stability.

Variant of sub-level stoping like Long hole stoping or Blast hole stoping and Cut and fill method of mining is more commonly in use at Indian Mines. Top down sequence of mining has gain popularity worldwide. Similar trend has started in India too. In this method crown is subjected to load at early stage of underground mining due to drill drivage being made below it whereas in case of bottom up it is put under load at later stage of mining. Hence good strength backfill is desirable to bear the responsibility to isolate different working areas. Nowadays use of paste fill has gained importance as it helps in reducing tailing waste and providing sufficient strength to wall rocks, stope back and stope brow after filling. However, for very small mines its use cannot be economical rather a group of mines in close proximity can thought of paste use from a single plant. Vertical extensiveness of deposit does matter. Small mines having limited depth of working like 60 m – 100 m below open pit are usually adopting bottom-up sequence of mining and using bottom ash from thermal plants as fill material as sand is becoming scarce with time.

In most of the mines being small in extent measured in situ stress values are not available. Big mines does opt for such measurement in their mine. In rare cases, a few pit had been backfilled. Such fill offers confinement to the crown, but in rainy season water infiltration may cause backfill to offer dead load on crown and can lead to failure if high factor of safety is not considered. It will worsen further if rock mass is highly fractured. Hence, it is usually advisable to keep the open pit unfilled till the completion of underground workings and divert rain water to a deeper level sump or maintain a crown of at least 60 m as per Reg. 150 (3). Accessible benches and toe of the slope can be shotcreted to slow down rock surface deterioration. Detailed geotechnical mapping on regular basis is also highly essential at different stages of mining to be abreast of likely existence of discontinuity at sensitive places.

For design of this crown pillar no identified methodology for Indian mines is available. Statutory requirement of Directorate General of Mines Safety (DGMS) states that there should be a barrier of 60 m against water body/waterlogged workings or for two different workings approaching towards each other. Considering this regulation, 60 m crown had been usually left during transition phase of mining in some mines. Whereas in a few small shallow depth mines 15–20 m crown is also left to isolate open pit from underground workings. Therefore, leaving 60 m against waterlogged workings or site-specific design by numerical modeling is followed by practicing engineers.

5. Case study

A mine with steeply dipping orebody exploiting manganese ore through opencast workings is considered for small scale analysis. Depth of working has reached 103 m from surface. Being a small lease deeper level orebody exploitation is planned by prevalent cut and fill underground workings. Width of orebody ranges from 2 to 7 m. The orebody is steeply dipping at 70–85° due south. The strike length of the deposit is about 1300 m in East-West direction. From field investigations, Immediate roof Rock Mass Rating (RMR) in orebody has been estimated as 54 which is “fair category” rock mass whereas RMR of hangwall and footwall rock mass has been estimated as 61 which is “good category” rock mass. Stopes were at 45 m level interval. Overall pit slope of 40°-50° exist at the site.

Two-dimensional plain strain models were solved for surface crown design. Simulation was performed by using FLAC3D software (developed by ITASCA Consulting Group, USA) based on explicit Finite Difference method. Plain strain is a special case of three-dimensional situation. A model of size 1650 m × 1 m × 1100 m is prepared with an element size of 1.0 m in the area of interest. Existence of open pit excavation after reaching ultimate pit limit is simulated with no backfill and no accumulation of water. Initial model is run in elastic state to generate virgin stresses. The model state is then changed to Mohr model and open pit excavation is made. Since crown should be stable till underground operations are over, full height stopes up to the crown is excavated in the model to visualize stability scenario.

Since measured horizontal in situ stresses are not available for this mine, equation for mean horizontal stresses as stated in Eq. (1) based on the theory of in situ stresses [19] is being used to estimate horizontal stresses.

$$\sigma hm=\frac{\beta EG}{\left(1-\vartheta \right)}\left(H+1000\right)+\left[\frac{\vartheta }{1-\vartheta }\right]\gamma H$$E1

where, σhm = mean virgin in situ horizontal stress in MPa.

H = depth in m.

E = modulus of elasticity of rock.

ϑ = Poisson’s ratio of rock.

γ = Rock density/unit depth.

G = Thermal gradient.

β = Coefficient of thermal expansion.

Vertical stress was taken as the gravitational load due to self-weight of the overlying strata,

$$\sigma v=\gamma H$$E2

Intact Rock properties are shown in Table 3 as follows.

S. N.	Physico-mechanical Property	Orebody	Wall rock
1.	Elastic Modulus (E) in GPa	13,000	13,000
2.	Poisson’s ratio (ν)	0.25	0.25
3.	Density (Kg/m^3)	3000	2800
4.	Cohesion (KPa)	5500	5800
5.	Angle of Internal friction	35	38

Table 3.

Intact rock properties of various materials.

Simulation was performed by using FLAC3D software (developed by ITASCA Consulting Group, USA) based on explicit Finite Difference method. For simulation stope length of 60 m is considered in analysis along with a 6 m wide rib pillar and 5 m height crown in between two levels. Level interval of 45 m is considered. Due to limited information about existence of ore beyond 200 m depth, only two-level extractions is considered. To simulate the existence of opencast workings or existing pit, overall pit slope of 40° and ultimate pit depth of 103 m from the surface is considered.

From the detailed analysis, 15 m crown is found to be stable for 7 m wide orebody. Crown pillar stability status for one particular section is shown in Figure 1. All the colored blocks that are in shear or tension are unstable zone. Failure in the backfilled stope back does not extend up to the open pit level failure zone in crown.

Figure 1.

View of stability of crown pillar for 7 m wide orebody.

6. Parameter sensitivity analysis

Same model is extended for studying the effect of depth, different crown thickness, horizontal to vertical stress ratio and overall slope angle. For this comparison no backfill after stope excavation is considered. Different locations for monitoring the displacement are considered as marked on the section (Figure 2).

Figure 2.

Monitoring points location marked on cross-section.

Stress state in the crown is also analyzed. Stress ratio that is, ratio of horizontal to vertical stress of 0.63 and 0.97 is used to compare status of crown pillar. In Indian mines, open pit depth varies from 50 to 400 m. Hence, two extremes of 103 and 400 m is studied. When depth of pit is increased from 103 to 400 m and stress ratio is kept constant at 0.97, 15 m crown fails. Stress state for 15 m crown at 103 and 400 m depth is shown in Figure 3. Stress concentration in the core of crown pillar has almost doubled from the initial values to about 30 MPa at 103 m depth. High stress concentration which was experienced in corners of crown pillar at 103 m depth is now visible at pillar core in crown at 400 m depth.

Figure 3.

Stress state of 15 m crown at UPD of 103 m and 400 m respectively for overall slope angle of 40°.

When crown thickness for this case is increased to 40 m stress state becomes moderate for 103 m depth but is still high for 400 m depth although not failed as shown in Figure 4. For 40 m crown at 103 m depth stress concentration at core is 15 MPa and it increases to about 30 MPa at 400 m depth.

Figure 4.

Stress state of 40 m crown at UPD of 103 m and 400 m respectively for overall slope angle of 40°.

High stress concentration is observed near the corners at pit bottom and zone in contact with underground excavation. Hence displacement at monitoring points 3, 4, 5, 7, 8 is compared for 40 m crown at 103 m as well as 400 m depth and is presented in Figure 5. Depth appears to be a major controlling factor as high displacement is observed at all the monitoring points when depth of open pit is 400 m and as compared to 103 m depth.

Figure 5.

Displacement at selected monitoring points for 40 m crown at different depths and stress ratio of 0.97.

Similarly, displacement at monitoring points 3, 4, 5, 7, 8, 14 & 15 for 40 m crown at 103 m depth and different stress ratio of 0.63 and 0.97 is compared and is presented in Figure 6. At low stress ratio tensile movement is visible due to reduction in lateral confinement whereas uplift is possible when stress ratio is more and on footwall side slope monitoring point it increases sharply mainly due to effect of dip of the orebody.

Figure 6.

Displacement at selected monitoring points for 40 m crown at 103 m depth and different stress ratio.

Displacement at slope monitoring points (14 & 15) for 40 m crown at 103 m depth and different overall slope angle of 40° and 50° is presented in Figure 7 displacement in hangwall side remains almost same with change in overall slope angle while footwall side slope has underwent more movement when slope angle is 50°.

Figure 7.

Displacement at slope monitoring points (14 & 15) for 40 m crown at 103 m depth and different overall slope angle of 40° and 50°.

More detailed parametric study and analysis in terms of magnitude of displacement and stress state is under progress for wider range of crown thickness and other parameters. However, from obtained results depth of open pit, overall pit slope angle and stress ratio are found to have significant impact on crown pillar behavior and its design. Therefore, in addition to orebody geometry, RMR, height of stope, open pit geometry viz., overall slope angle, pit depth, influence of geological disturbances along with hydrogeological condition and in situ stress condition as well as reoriented stress on account of open pit mining should also be considered in design of a suitable thickness of crown pillar.

7. Conclusion

Determination of optimum thickness of crown pillar has been researched for many years but limited design guidelines are available. Initially, major concern was restricted to design of surface crown where underground working exists below surface infrastructure. Failure pattern for surface crown have also been analyzed. Although study was later extended to shallow dip stopes but influence of open pit working is probably not analyzed in detail. Although in literature, a few case studies are discussed where instability of crown pillar and open pit benches have occurred. Most of them were having caving method of mining below existing open pit. Hence method of mining and mining sequence also has a role in crown pillar design.

More detailed parametric study is going on but small-scale analysis is done to analyze the effect of depth of open pit, overall pit slope angle and horizontal to vertical stress ratio and are found to have significant impact on crown pillar behavior and its design. Therefore, in addition to orebody geometry, RMR, height of stope, open pit geometry viz., overall slope angle, pit depth, influence of geological disturbances along with hydrogeological condition and in situ stress condition as well as reoriented stress on account of open pit mining, mining method, sequence of mining should also be considered in design of a suitable thickness of crown pillar.

Various computer codes such as FLAC3D, UDEC, PFC, ANSYS, GEOSLOPE etc. are available to address specific mining issues. Data generated by empirical and observation techniques helps to validate the model and make the predictions more realistic. But obtaining data for each and every parameter is not always possible. Therefore, Carter (2014) has emphasized a coupled approach of numerical modeling and empirical approach for solution instead of using just one type of approach. A coupled approach of empirical, observational and analytical methods is generally beneficial.

Acknowledgments

The author is grateful to The Director, CSIR-CIMFR, Dhanbad for permission to publish this article. Views expressed are of authors and not necessarily of the institute to which they belong. This work is financially supported by Ministry of Mines. The authors gratefully acknowledge the support. We are thankful to mine management for providing the data of the mine.

Conflict of interest

The authors declare no conflict of interest.