Alison McQuillan,Ismet Canbulat,Joung Oh

School of Minerals and Energy Resources Engineering,UNSW Sydney,Australia

Keywords:Slope stability 2D 3D Open cut coal mine Limit equilibrium

ABSTRACT Strata failure is a principal hazard in open cut coal mining as it has the potential to cause multiple fatalities.Prior to the excavation of any slope,a geotechnical assessment should review the likely slope performance,including the risk of slope failure.Controls to manage this risk to an acceptable level should accompany the geotechnical analysis.A survey of 43 practising geotechnical engineers indicated that kinematic and 2D limit equilibrium analyses were the methods most commonly applied to analyse excavated slope stability.While these methods are well established and widely applied in the broad rock engineering disciplines (e.g.civil,hard rock),a recent review of over 60 slope failures suggests these methods have limited suitability for modelling the complex failure mechanisms observed in excavated coal mine slopes.Kinematic techniques do not adequately capture the rock mass component of excavated slope failure and do not provide a geospatial location of potential failure,while 2D limit equilibrium techniques do not adequately capture the 3D mechanisms of excavated slope failures.Methods which do consider the 3D mechanisms of slope failure are summarized for industry consideration and application.

1.State of the art

As part of Australian Coal Association Research Program(ACARP)funded research project C25078,to develop a new methodology for assessing coal mine slope stability,a survey was compiled and distributed to geotechnical engineers currently working as either employees or consultants in Queensland (QLD)and New South Wales (NSW)open cut coal mines.The survey included 22 questions designed to summarize state of the art practices for assessing excavated coal mine slope stability.

The survey was conducted anonymously and 43 responses were received.The number of responses and distribution in experience of survey respondents(Fig.1)was assumed to accurately represent the opinion and assessment methodologies applied by geotechnical professionals working in the Australian open cut coal industry.

The survey,generated using online survey software Survey Monkey,requested both open and closed feedback from respondents about:(1)the number of years experience as a geotechnical engineer;(2)analysis methods and failure criterion routinely applied to assess slope stability;(3)frequency and inputs of slope stability models;(4)use of empirical classification systems;and(5)methods to identify,monitor and manage geotechnical hazards[1].

Summaries of closed (i.e.yes,no or explicit multiple choice)responses,as well as notable open (i.e.please specify or other)responses are outlined and discussed in the following section.

1.1.Experience

Fig.1 shows just under half (49%)of the respondents had more than 20 years’ experience as a geotechnical engineer.

1.2.Modelling slope behavior and slope stability analysis methods

Kinematic and 2D limit equilibrium modelling techniques are the most routinely applied(Fig.2).Numerical 2D and 3D modelling techniques are less routinely applied,and at least one out of three respondents (40%)reference empirical design charts when assessing slope stability.

The majority of respondents indicated that the methods in Fig.2 are generally adequate to account for the failure types and failure mechanisms commonly observed in open cut coal mines.However,several respondents agreed that(1)kinematic analysis using stereonets is primitive in its ability to account for the spatial location and/or variability of joints;and (2)complex failure mechanisms(particularly those involving structure)are better modelled using 3D numerical techniques,yet this type of modelling is timeconsuming compared to 2D limit equilibrium techniques.

Fig.1.Survey responses to question:how many years experience do you have in geotechnical engineering?

Fig.2.Survey responses to question:what methods do you routinely use to assess slope stability at your open cut coal mine? (More than one response possible).

Modelling is generally completed on an as required basis when there is a significant change in pit geometry and/or rock mass character.Two-thirds (64%)of survey participants said they complete slope stability modelling on at least a fortnightly basis.

Geomechanical property inputs into slope stability models are generally derived from one or more of the following sources(from most common to least):(1)published properties (e.g.BMA spoil categories);(2)site-specific properties back-analysed from failed slope cases;(3)material properties derived from site-sourced laboratory-tested samples;and (4)empirical values (e.g.GSI,JRC)[2-5].

Point estimates of the average,upper and lower bound material strengths are most routinely applied to slope stability models(62%of responses).Approximate and stochastic methods are applied to a lesser extent (22% of responses).

The geological strength index is widely applied to estimate insitu rock mass strength (78% of respondents)[3,4].

Based on open responses from respondents,a reasonable estimate of defect strength and inclusion of the anisotropic nature and variability of coal measure rock is critical to modelling the structurally-controlled failures commonly observed in open cut coal slopes.Often one,or both,these factors are lacking due to difficulty in measurement and/or limited availability of funding for further sampling and testing.

The Hoek-Brown and Mohr-Coulomb failure envelopes are the most widely applied failure criterion [6](Fig.3).User-defined shear-normal relationships,the Barton-Bandis failure envelope and ubiquitous joint strain softening models are also commonly applied by respondents [6].

Fig.3.Survey responses to question:what failure criteria do you routinely apply to slope stability calculations? (More than one response possible).

Survey respondents generally agreed that the available failure criteria adequately represent the rock mass and defect behavior of excavated coal mine slopes.However,respondents commented that the anisotropy observed in coal measure rock is inadequately accounted for in the most commonly applied criterion.Further,engineering judgement must always be applied to recognize when to apply which failure criterion,to what material,in what method of analysis.

1.3.Reporting slope stability modelling results

Almost all(95%)of respondents report stability results in terms of factor of safety (FOS).Probability of failure(POF)is also used to report stability results by just over 40%of respondents(Fig.4).Stability calculations are also reported and/or communicated by the following means:(1)strength reduction factor (SRF);(2)illustrations;(3)description of stressed induced damage,deformation and stress magnitude;and/or (4)risk (i.e.likelihood and consequence)as indicated by Other (please specify)responses.

1.4.Rock mass classification systems

Open responses summarized empirical rock mass classification systems as:(1)built on years of observation;(2)easy to use;and(3)provide an appropriate starting reference when faced with limited data.However,they can be subjective and have often been developed for non-coal measure rock masses or for underground environments.The limitations of the applied empirical rock mass classification systems must be understood by those using them otherwise it can lead to a false sense of security for the inexperienced.Empirical rock mass classifications should only be applied to the same conditions under which they were developed and should be calibrated by back analysis against individual site conditions.

Fig.4.Survey responses to question:what methods do you routinely use to report the results of slope stability modelling? (More than one response possible).

1.5.Data acquisition

Defect data is overwhelmingly acquired by laser scanners (90%of respondents).Measurements are also acquired using(in order of descending number of responses):hand compasses,photogrammetry,geophysical logs and UAVs.

1.6.Geotechnical hazard identification,reporting and management

Visual inspections are the most common method to identify geotechnical hazards in excavated slopes (more than 95% of respondents),followed by comparisons of design versus as-built geometry and/or slope movement/displacement recorded by monitoring devices(e.g.tell tales,laser scanners,prisms,slope stability radars,etc.).A quarter of respondents complete inspections using airborne drones to identify geotechnical hazards.

Once geotechnical hazards are identified,a suite of methods,in addition to continued visual inspections,are used to monitor the hazard/s including (1)slope stability radar (93% of respondents);(2)laser scanners (78% of respondents);(3)peg tell tales;(4)piezometers;(5)prisms;(6)wireline extensometers;(7)inclinometers;and (8)drones (Fig.5).

Geotechnical hazards are generally managed by Trigger Action Response Plans (TARPs)using qualitative and/or quantitative triggers primarily based on historical site conditions (89% of respondents),personal experience (66% of respondents),back analysis of failed slope cases(94%of respondents)and/or based on numerical modelling predictions (36% of respondents).

1.7.Summary of survey responses

A survey of 43 Australian open cut coal geotechnical engineers found that kinematic and 2D limit equilibrium modelling techniques are the most routinely applied when assessing the stability of an open cut coal mine excavated slope design.However,several respondents commented that kinematic analysis using stereonets is primitive in its ability to account for the spatial location and/or variability of joints.Further,complex failure mechanisms (i.e.those involving a structural and rock mass component of failure)are generally better modelled using 3D numerical techniques,yet this method of modelling is time-consuming compared to 2D limit equilibrium techniques and,as such,are less routinely applied.The same study found that almost all respondents(95%)report stability results in terms of FOS.POF is used to report stability results by just over 40%of respondents.At least one out of three respondents(40%)reference empirical design charts when assessing slope stability.Respondents however cautioned that empirical systems should be calibrated by back analysis and only applied to the same conditions under which they were developed.Visual inspections are the most common method used to identify (and monitor)geotechnical hazards in excavated coal mine slopes.

Fig.5.Survey responses to question:what methods do you routinely use to monitor identified geotechnical hazards? (More than one response possible).

2.Typical excavated coal mine slope failures

To determine the most appropriate methods to assess coal mine slope stability,over 60 failed slope cases were reviewed to identify the common features in excavated coal mine slope failures [7].Case studies were collected from 25 open cut coal mines across coalfields in Queensland and New South Wales (from the Bowen,Hunter,Tarong and Callide basins)[8].

All slope cases were excavated between 2010 and 2017.All but one failed slope case were single-bench failures.The exception transcended multiple benches(Fig.6).No failed cases had dumped loads (e.g.spoil loading)behind the crest.

Failed cases were considered to be any slope that had exhibited substantial movement (i.e.,＞20 m3or approximately 50 tonnes of rock mass displacement)from its as-built geometry.

Back analyses were completed using commercially available survey,statistical and analytical software.

Reviewed case studies focused on those in competent coal measure rock only.Slope failures involving circular failures in soil-like material horizons and isolated rock falls were excluded from this review.

An overview of the distribution of case studies by slope geometry is presented in Figs.7-9.

Case studies were subjected to a rigorous review procedure which measured pre-and post-failure geological,hydrogeological,structural,geomechanical and geometrical slope conditions.At a minimum all case studies included pre-failure slope height,prefailure slope angle and photographs of geological,structural and hydrological conditions pre-and/or post-failure.Not all cases have complete measurements of failure runout distance,failure height,failure volume or orientations of defects contributing to failure where either:(1)the survey data available was cropped,or failed material had been cleaned up by operations by the time the post-failure survey was acquired;or (2)post-failure survey data was not available.

The review indicated that excavated slope failures typically have a structural component of failure in that they are generally bound vertically or sub-vertically by persistent structure;a rock mass component of failure in that they are generally bound horizontally,either at the crest or toe of failure,by a coal or carbonaceous band;and are 3D nature.

Typical excavated slope failures are highlighted in Fig.10.

The review of typical failures showed that failure occurred across a range of slope heights and slope batter angles (Fig.11).There was no cluster of failed cases at higher slopes and/or steeper batter angles as may intuitively be expected,or is alluded to in kinematic analysis where shallower slope angles have kinematically lower percentages of slope failure compared to steeper slope angles.

Also of interest is the relationship between slope angle and runout distance (Fig.12).In general,there is a negative correlation between slope angle and runout distance.The exception is a slope failure in the 70°slope angle category.This slope case failed in toppling mode.Toppling failures typically have a higher energy associated with material movement,and subsequent longer runout distance.This trend was observed in the four toppling failures included in the case study dataset (Fig.13).

Fig.6.Cross-section view.

Fig.7.Failed cases:distribution of pre-failure slope height.

Fig.8.Failed cases:distribution of pre-failure slope angle.

Fig.9.Distribution of failed cases strip geometry (i.e.,linear,concave,convex).

Fig.10.Examples of typical excavated coal mine slope failures in QLD and NSW coal mines.

As noted in Fig.12,50° category=slope angle ＜52°;55°category=53°＜slope angle ＜ 57°;60° category=58°＜lope angle ＜ 62°;65° category=63°＜slope angle ＜ 67°;70°category=68°＜slope angle ＜72°;75° category=slope angle ＞73°.

Fig.11.Relationship between slope performance(intact or failed),slope angle and slope height.

Fig.12.Slope batter angle vs failed material runout distance by slope angle category.

Fig.13.Slope height vs failed material runout distance by failure type (bullnose=failure occurred at 90° elbow in slope geometry).

As is noted in Fig.13,toppling failures exhibited longer runout distances.

This back analysis clearly showed that kinematic and 2D limit equilibrium slope stability analysis methods,the two most frequently applied techniques by practising geotechnical engineers(Fig.2),do not adequately account for the failure mechanisms observed in the field.

Specifically,McQuillan et al.state that kinematic analysis allows for a rapid assessment of the potential for structuraldriven slope instability only [9].This technique does not account for the rock mass component of slope failure,and does not provide a geospatial location of potential failure unless defect measurements plotted on the stereonet are separated into geospatially accurate geotechnical domains.This is a particularly important step to complete to confirm the actual intersection of structures modelled to form wedge or block sliding failures.Kinematic analysis further provides a conservative qualitative indication of slope instability where discontinuities are considered 100% persistent and 100% frictional (cohesive strength is generally ignored).And,the POF calculated is relative to discontinuity frequency and orientation bias that results from scanline acquisition orientation [10].

3.Discussion

Slope stability analyses can be undertaken using empirical,kinematic,limit equilibrium (2D or 3D)or numerical (2D or 3D)modelling techniques.

A survey of 43 geotechnical engineers in 2016 indicated that kinematic and 2D limit equilibrium analyses are the most commonly applied methods to assess excavated coal mine slope stability.This is due to the relative ease of model construction and rapid computation time compared to 3D models.

However,the most frequent type of failures that occur in excavated coal mine slopes are 3D in nature and have both structural and rock mass failure components.They are generally vertically bound by structure striking near parallel to the excavated slope and horizontally bound by a weak,generally carbonaceous,layer.As such,conventional kinematic and 2D limit equilibrium analysis methods cannot adequately model,or predict,the failure mechanism observed to frequently occur in open cut coal mine excavated slopes.

Methods which can adequately account for the failure mechanisms typically observed in excavated coal mine slope failure include empirical,3D limit equilibrium and 3D numerical modelling.

Examples of empirical methods suited to coal mines include those published by Jhanwar,Canbulat et al.,Sullivan and McQuillan et al.[7,11-15].Empirical methods typically provide a fast,defendable solution if they are transparent,practical and firmly tethered to reality,but may not always provide the rigour associated with more complex modelling techniques such as limit equilibrium and numerical methods [16,17].

Examples of 3D numerical modelling of coal mine slopes are published by McQuillan et al.[9,17-20].The progression from limit equilibrium to numerical models typically increases processing time and further requires additional material strength data (e.g.elastic properties)to calculate representative estimates of stress and strain [21].As such,many geotechnical engineers do not currently progress from 2D to 3D modelling.These increases in modelling requirements typically influence the uptake and routine application of 3D numerical modelling techniques by practising geotechnical engineers,where less than 15%of survey respondents indicated they use 3D numerical techniques (Fig.2).

4.Conclusions

This paper summarizes the methods currently used by practising geotechnical engineers to assess excavated coal mine slope stability in Queensland and New South Wales in Australia.A survey of 43 engineers in 2016 indicated that kinematic and 2D limit equilibrium analyses were the methods most commonly applied to analyse excavated slope stability.

Although kinematic and 2D limit equilibrium methods have been successfully applied to the civil and mining industries for decades,a recent review of over 60 excavated coal mine slope failures suggests these methods do not adequately model the failure mechanisms observed in excavated coal mine slope failures.Stereographic and 2D analysis cannot account for the 3D failure mechanisms (e.g.geospatially accurate intersecting structures)that are observed to drive failure in excavated slopes.

To reliably predict the performance (e.g.propensity for failure)and critical failure mechanism(including location)of an excavated coal mine slope,geotechnical engineers must select appropriate tools to complete slope stability assessments.Methods which can adequately account for the failure mechanisms typically observed in excavated coal mine slope failure include empirical,3D limit equilibrium and 3D numerical modelling.

Acknowledgements

The research presented in this paper is funded by the Austrlian Coal Association Research Program (ACARP),grant no.C25078.