Documente Academic
Documente Profesional
Documente Cultură
James I. Mathis
Zostrich Geotechnical
Abstract
The proper design and evaluation of the catch bench angle, inter-ramp slope angle, and overall slope angle, individually as well as
in combination, are required for successful excavation and economic optimization of a rock slope. In many slopes at least one, if not
more, of the above controlling angles are essentially ignored, resulting in a slope properly designed for one facet of the excavation
but ignoring the other components. Bench face angles can be accurately described statistically utilizing engineering predictions
from the rock mass discontinuity network and discontinuity shear strengths. Together with the required bench width, the bench
controlled inter-ramp angle is determined. Inter-ramp angles can be accurately determined by careful construction of a structural
geologic model, noting location and orientations of discrete intermediate and large planes of weakness for the excavation in
question. The location and orientation of the overall slope is dependent upon the slope as determined by the bench controlled
inter-ramp angle and the stability controlled inter-ramp angle. Given advances in data collection and analytical techniques and
continuing moves to increase mining safety while simultaneously attempting to minimize excavation costs, the only possible way
to truly optimize slopes is through rigorous analytical methods combined with probabilistic techniques.
INTRODUCTION
A rock slope consists of up to three stability controlled slope components. These are bench face angle, inter-ramp angle, and overall
slope angle (Figure 1). Depending on the situation, these can be excavation specified or maximum attainable angles. It is, however,
critical that the interaction of each of these components is incorporated in the resulting slope design.
Lets clarify this interaction. Assume an open pit mine with a substantial overall slope height. Now, assume a bench height of 15m,
a required catch bench width of 8m with a stability determined face angle of 70. Assume the inter-ramp slope has been determined
to be acceptably stable at an angle of 55 with a ramp width of 33m and a mean height of 150m. Structural considerations determine
that the overall slope shall not exceed an angle of 44.
What to all these numbers tell us? Well, in order to maintain the required catch bench angle, the inter-ramp angle cannot exceed
the geometrical constraints imposed by the bench geometry. In this case, the bench determined inter-ramp angle is:
tan-1 (15m/(8m+(15m/tan(70)))
= 48
(1)
Now, it is obvious that the bench controlled inter-ramp angle (48) is less than the angle at which the inter-ramp slope has been
determined to be stable (55), thus the slope must be designed to accommodate the required bench geometry.
How about the inter-ramp versus overall slope? The inter-ramp angle cannot exceed 48, as noted above. If the ramp width is
added, the overall slope, as dictated by inter-ramp constraints is:
tan-1 (150m/(33m+(150m/tan(48)))
= 42 (2)
The overall slope was determined to be sufficiently stable at 44. Yet, the bench geometry dictates the inter-ramp angle. This angle,
together with the inter-ramp height and the required ramp width impose an overall slope angle of 42. Therefore, the bench geometry,
for this specific case, dictates the overall slope angle as well.
Of course, some variables can be modified. Bench height can at times be adjusted, as can excavation methods (especially blasting),
ramp widths can be adjusted based on equipment selection, artificial support may be contemplated, etc. In fact, a multitude of
possibilities exist to adjust the individual components of the slope geometry. Still, in order to reach that point one must first understand
how one engineers each of the critical slope components: bench, inter-ramp, and overall slope angles.
Slope Stability
Figure 1 - Bench controlled inter-ramp angle, inter-ramp angle, and overall slope angle.
For a standard bench design, this information is then compiled and the following conducted:
Identify potential failure modes;
Determine the population of potential structural orientations that may occasion bench failure;
Slope Stability
Calculate the stability of the structurally defined failure blocks by standard kinematic analysis;
Adjust the bench face angle until an appropriate safety factor is realized against sliding and/or;
Calculate potential reinforcement for the sliding blocks, if required;
Determine the requisite bench width to retain failed rock from the bench face.
Yet, a variety of questions arise with this standard design. Amongst these are:
Most benches are drilled and blasted vertically. To what minimum angle will the bench fail to and how much material will fall from the
crest?
What is the distribution of face angles and how will this distribution of face angles affect the bench catch width design?
How was the variability of the shear strengths incorporated in the analysis?
What is the impact of the discontinuity length and spacing on the face angle and what is the sensitivity of the design to these
parameters?
What would the impact be of utilizing the entire structural orientation distribution instead of point values?
How much backbreak can be eliminated by drilling angle holes and is it warranted?
Catch bench face angle design, as conducted by this author, utilizes a rock fabric simulation to determine bench face angle reliability
(2). Discontinuity spatial characteristics obtained from a rigorous sampling method and obtained either from physical or photogrammetric
mapping are simulated in a three dimensional, Monte Carlo generated, discontinuity model (Figure 3). The three dimensional model is then
cut by a simulated bench face and statistical failure analyses of wedge and plane shear failures are conducted on the daylighting features
that transect the bench crest. This provides not only the bench face angle distribution as a function of bench height, but also provides a
large number of simulated face profiles for analysis and allows for the effect of an excavated face angle of something less than 90.
Rock fall is analyzed using simulated face profiles to determine/verify the required bench width to accommodate rock fall. Note that rock fall
described herein is material physically falling to the bench, not volumetric failure accumulation as considered by some engineers.
Of course, the analysis incorporates known structural domains (areas of similar geologic structure including lithologic variations),
variations in face orientation (design sectors), and the complete discontinuity shear strength distribution (peak and residual) in the design.
Blasting effects are accommodated as adjustments in the discontinuity spatial characteristics. At times, rock reinforcement may be
considered to modify the bench face angle distribution.
Once the bench face angles distributions are defined, the reliability of the bench face angle is utilized for selection of the design bench
geometry. The face angle reliability can vary from 70% for areas not often frequented by man or machine to values >90% for areas where
bench failure may substantially impact operations or potentially endanger personnel. For open pit mines a reliability of 80% appears to be
somewhat standard as this appears to contain most rock that escapes a single bench.
Face angles at the chosen reliability, and segregated by external effects (excellent vs. poor excavation techniques, etc.) are compiled
into a table. This table includes the structural domain and design sector (face orientation). Note that this table can also include varying
bench height and catch bench width. A geometrically constrained bench controlled inter-ramp angle, as noted in the introduction to this
article, is then calculated.
Slope Stability
Note that the described methodology answers all of the questions posed above, including the effect of discontinuity length (persistence)
and center density (spacing). Note further these questions can only be answered using probabilistic techniques.
One of the aforementioned points, discontinuity persistence, is absolutely critical to proper face angle design. As can be seen in Figure
4, discontinuities that are assumed to be continuous through the bench will not honor the rock mass. Continuous structures would result
in a bench face angle that would not change as a function of bench height. Of course, this is not the case, as double and triple benches
are nearly always steeper than single benches due to the interaction of discontinuity persistence occasioning failures as they transect the
bench crest. Thus, only a method honoring the mapped discontinuity persistence will provide an accurate estimate for bench face design
if the discontinuity persistence is substantially less than the bench height. This variation in bench face angles is demonstrated for a 10m
and 20m bench height utilizing the program Z-Fabric (Zostrich Geotechnical) (Figure 5).
Experience has shown that it is always prudent, as well as good engineering, to validate any slope design. Verification of bench
scale performance utilizing the above described methodology has been conducted using a multitude of individual face profiles
obtained from surveying as well as photogrammetric techniques. One of the advantages of utilizing photogrammetric techniques is
that the entire imaged slope is available for face profiling to compare with the analytical bench face and slope angles. While some
blind areas may exist due to camera location, the accuracy of the slope topography is far superior to that obtained by any other easily
applied methodology. The verification process allows one to detect errors in discontinuity data collection, failure mode analysis, and
blasting practices such that the bench design may be refined.
INTER-RAMP DESIGN
As was discussed previously, bench scale design is, for most part, predicated on relatively simple failure modes with low applied
stresses on the sliding surfaces. The controlling geologic structures (rock fabric) can be dealt with statistically as has been done above.
However, inter-ramp slope design is more much complex, incorporating intermediate faults, rock fabric, and at times, rock mass
strength characteristics. Due to this varying height of the inter-ramp slope, the required slope analyzes can fall anywhere between fabric
stability analyzes for benches and the individual failure analyzes required for overall slope stability.
Slope Stability
Slope Stability
accurate. As was noted above, the direct impact of utilizing infinitely persistent bench scale (fabric) features was expressly discouraged
as it had a demonstrably detrimental effect on the accuracy of the bench face angles. In that case the persistence of the fabric features
could be obtained relatively simply. For the inter-ramp case, this may not be possible (although it has been done for some larger scale
exposures). Thus, for this specific case where there is a lack of information, an estimate may be made of the mean and variation of the
persistence of the feature or an infinite persistence determined that honors the density (spacing) of the discontinuity set.
One of the interesting facets of analyzing inter-ramp stability in the fashion described above is that the release planes for failures can be
analyzed in both a discrete and probabilistic fashion. The greater the persistence and the number of the required release planes to define
a kinematically viable failure block, the smaller the probability that block will actually exist (Figure 7).
For analytical purposes, the ramp, crest, or toe of an inter-ramp slope section provide obvious, pre-defined release planes. This is similar
to the previously discussed bench geometry where the structures defining the failure block were required to transect the crest in order for
a viable failure block to be defined.
The rapid change in wall curvature at a pit bottom (Figure 8), or a nose developed in a pit wall can also be analyzed statistically. Where
these two special slope cases have often been described as being more, or less, stable because of degrees of freedom of motion that is only
partially true. As the plan curvature of the slope wall increases, it is more difficult to create a viable geometry as the release structures must be
found in a specific locale and have a substantial, and increasing, persistence for a viable failure block to be defined. Alternatively, for a nose,
the requirements for persistence and location of the release structures are substantially reduced. These can both be addressed analytically,
using probabilistic methods, with the aforementioned interpretative structural work thus improving slope stability assessments.
Economic constraints
An inter-ramp slope analysis, if conducted using statistical methods, will be either expressed as a probability of failure, the expected
number of failures, or both. The location and extent of the failure will have some economic impact on the operation.
Slope Stability
For example, if the inter-ramp scale failure is on a slope where it will have little impact either on mining operations, traffic, or facilities,
then the economic impact of inter-ramp failure is minimal. The probability of failure can likely be quite high without substantial economic
consequences.
However, if the design inter-ramp slope is below a haul road to the bottom of a pit with no alternative methods for access if the ramp is
obliterated by failure, the failure is an end of mine life event with large economic consequences. In this situation, the probability of failure,
including the accuracy of the failure estimate, must be carefully considered. Slope angles may be flattened to account for this risk in
specific portions of the pit specifically to address this factor.
Figure 8 - Effect of structure persistence for inter-ramp failures as a function of slope geometry
All inter-ramp slope failures carry an impact on operation. Thus, inter-ramp slope designs should always consider not only the stability
of the slope, but the impact on operations if the slope is to fail. This can only be done with techniques specifically addressing the volume
and probability of slope failure. Many operations fail to consider this in their design rendering them exposed to risks which should have
been assessed during slope design.
Inter-ramp designs are conducted as are bench designs, with accommodation of structural domains, lithologic bounds, design
sectors, etc. Existing inter-ramp slopes can be back analyzed and compared to theoretical values. Of course this is more difficult than
for benches as the number of inter-ramp failures will provide a less reliable sample than multiple benches provide. However, such an
analysis is certainly worthwhile, especially as any failures provide information not only regarding potential failure geometries in future
walls but on the shear strength of the features/zones involved.
Slope Stability
CONCLUSION
One may argue that this is not the way most rock slopes are currently designed. In fact, for an open pit mine the geotechnical engineer
is often asked for the overall stable slope angle. Bench face and inter-ramp angles are then back calculated from that angle. While this
can be done at times, and approximate overall slope angles may be utilized to generally locate the pit extents and depth this is the reverse
of this authors approach.
The methodology for determining the influence of slope components on each other is not the only concept that may appear upside down.
As was demonstrated throughout the article, many analyses simply cannot be conducted without the use of probabilistic techniques. This
includes not only bench design, where all the components of the discontinuity spatial characteristics, including persistence, are analyzed in
conjunction with the discontinuity shear strength distribution, but also inter-ramp and overall design, where the probability of discontinuity
density (spacing) and persistence comes into play not only for the main sliding plane but associated failure release planes.
Economics are a critical, and generally understated, portion of slope design. Again, these can generally only be addressed with
probabilistic failure analyses and the associated distribution of the probability of failure occurrence and volume. In order to truly optimize
the slope, economics must be included.
REFERENCES
1. Mathis, J.I. (1988) Development and verification of a three-dimensional rock joint model. Doctoral Thesis 1988:63 D, University of
Lule, Sweden, May.
2. Mathis, J.I. (2002) Bench face design in rock. http://www.edumine.com
3. Mathis, J.I. (2007): Pit slope design and structural analysis at the Jericho diamond mine utilizing digital photogrammetry, Slope
Stability 2007, Perth, W. Australia, Sept, 2007, pp93-104
Slope Stability