Rock Mass Properties Transcript

This transcript describes the YouTube video "Rock Mass Properties - Dr. Evert Hoek Lecture"

The video begins with a black screen. The Rocscience Logo fades in with the word “presents.” Both fade to black.

Text appears “Practical Rock Engineering Lecture Series by Dr. Evert Hoek. Copyright 2016” and then fades to black.

Text appears “Rock mass properties Lecture 4.” Transitions to the first slide showing a rock face of a mountain with zigzag striations and foliage near the top. The title “Rock mass properties” remains.

Dr. Hoek: In this lecture on rock mass properties, I'm going to be discussing the issues of rock mass behavior,

Transitions to show Dr. Hoek standing in a room speaking to an audience seated in front of him. The wall behind him is wooden with an antique fireplace and picture frames hanging from the wall.

Dr. Hoek: where the intact pieces that we discussed in the last lecture, number three, are now separated by discontinuities and although tightly interlocking, have a much weaker strength performance.

The first image is projected back on the screen without the title and alternates between Evert and the picture on screen.

Dr. Hoek: In the opening slide here, you see a road cut in Greece and it's obvious from the folding and deformation that's occurred in that rock mass, that we're no longer dealing with anything like an intact material. Engineers love to have numbers to use in their calculations and there is a great pressure to categorize rock in terms of some kind of number that you can use to calculate properties. And in my opinion, that's the wrong way of approaching the problem because we're dealing with geology as given to us and so to me the starting point and the fundamental requirement in any rock engineering project

A slide appears in the top left corner with an image of a man looking down at s variety of rocks in a slope. Text on the right reads “A fundamental requirement for any design in rock engineering is a comprehensive geological model which requires the involvement of an engineering geologist or geological engineer, preferably with local experience, who understands how the origins and the characteristics of rock masses influence their engineering behaviour.” After a moment, the slide moves to fullscreen.

Dr. Hoek: is the presence of a very good well-trained engineering geologist or geological engineer or geologist with some understanding of the concepts of mechanics, who's familiar with the local geography and geology and who can understand what the process he's involved in, is aiming at.

Transitions back and Evert Hoek is shown on screen again.

Dr. Hoek: So that somebody who understands the origins and characteristics of the of the rock mass, is a really key component of that process and let's forget the numbers until later in the process.

A new slide is projected into the top left corner of a world map with yellow dots all over. After a brief moment it is expanded to fullscreen. Alternates between the slide and Dr. Hoek on screen.

Dr. Hoek: This map shows the tectonic plate boundaries and earthquake locations around the world and the yellow dots are sites that I've worked on around the world. And you'll see that they vary from benign conditions in southern Africa, which is a very stable continent; it's surrounded by a large plate. There are no major active tectonic boundaries. And so, it tends to be a very stable continent with very predictable conditions, which you'll see later in the lecture. On the other hand, a lot of the projects I've worked on are associated with major mountain changes, the Andes, the Himalayas, the Rockies, and these are totally different in character, in terms of their rock mass behavior because the rock very often has been tectonically deformed and the shear strength of the discontinuities is entirely different from those in a stable environment. And that has to be recognized. When you start a project that you might be in a totally different geological environment - one in which some author working in a different environment has written a paper, which you propose to follow. Beware that not all methodology is applicable to all rock masses. There are differences. Similarly, again as a result of plate tectonic movements, the stresses in the Earth's crust are entirely different.

The World Stress Map is shown on screen. Text reads “The World Stress Map giving orientations of the maximum horizontal compressive stresses. 2005 version form www.world-stress-map.org. The map is maintained by the Geophysical Institute of Karlsruge University in Germany.” After a moment, Dr. Hoek is again shown on screen.

Dr. Hoek: So that in a stable continent like Africa, very often we measured horizontal stresses much lower than the vertical stresses. And I started my career there and was used to horizontal stresses of about one-sixth of the vertical stress and I remember visiting Australia and everything was turned 90 degrees and I couldn't understand it until I realized that their stress field is entirely different. And particularly in the mountain chains, because of the of the mountain formation process, the compression that occurs, you can have very high horizontal stresses. And you can have spalling at relatively shallow depths because the stresses are high enough to induce it.

A slide is shown in the top left corner. It shows a schematic section of the geodynamic evolution of the Hellenides in Northern Greece. Text at the bottom reads “An example of the creation of a geological model for the Metsovo tunnel on the Egnatia Highway in Greece.” After a moment, Evert is shown on screen, and it alternates between him, and this slide being shown on screen.

Dr. Hoek: So, the first process that we have to go through in developing a rock mass model for a project is the geological model. And here's an example of one for a tunnel in Greece, the Metsovo tunnel, on the Egnatia highway which runs right across northern Greece, and it passes through the Pindus mountains, which are the tail end of the Alps in the rest of Europe. And the geologist who's made up this geological model has gone right back to the origins of that particular bit of the Earth's crust. The plate tectonic movements had taken place, the formation of a subduction zone through various geological ages, and you see the four pictures there depicting these stages of the evolution of the Earth's crust in that location.

The slide is shown, and a new diagram is added to the right with text “Sketch of the geological model of the massif of the project area of the Metsovo tunnel.” After a moment, Dr. Hoek is shown on screen.

Dr. Hoek: And finally, zooming in on the tunnel location shown on the right there, it's very important to recognize when you're setting out to design this tunnel, that it's gone through that geological history and that you have built into the rock, all of the problems that have been created by the geological process. Now, there's also a question of scale, which is tremendously important because we, in the last lecture, talked about core recovery and you're talking about core of 50 millimeter typically diameter, 100 meters long. And that's in no way representative of a slope which might be a thousand meters high or a tunnel that might be 50 meters in span. So that you have to consider the scale and it's very difficult to come up with simple numerical solutions that then incorporate an automatic adjustment for scale. And in my view, it's much better done by simply going back to basics and thinking through the problem and saying what kind of scale of problem am I dealing with? And how do I incorporate that into my into my engineering model?

A new picture is shown on screen with text on the right reading “In constructing the geological model for a project, the scale of the problem must be taken into account. Intact rock and structural discontinuity properties are important on a small scale while the overall properties of the jointed rock mass must be considered in large scale problems.”

Dr. Hoek: And this picture shows that you can go anywhere from the top intact behavior, through to very very strongly structurally controlled behavior where you have one or two discontinuities in the size of rock mass that you're dealing with.

Transitions to show Dr. Hoek speaking and alternates between him and the previously described slide.

Dr. Hoek: Which in the case of a tunnel, might be ten meters, in the case of a thousand-meter-high slope might be a hundred meters. What kind of discontinuities exist in that volume of rock and all the way then through to multiply jointed rock masses, which I'll show you examples of in a moment. So, for example, in the Chuquicamata open pit mine in Chile, the pit is currently one kilometer deep.

A slide is shown with two images. The top left shows a dump truck driving along the bottom and a large rock slope beside it. The left image shows the benches and the roadway with small vehicles at the bottom. Alternates between Dr. Hoek on screen and the slide.

Dr. Hoek: And if you look at the photograph on the upper left, you'll see that the benches which are each 18 meters high, are clearly defined by the structural features. You have wedges and blocks which fall out simply because they are defined by structural intersecting, structural features and they fall out under gravity. On the other hand, in the thousand-meter-high slope, when you stand back and look at that, it almost looks like a soil, a sandy soil, with small particles. The particles might be five meters in size but on that scale, it looks very homogeneous. We still incorporate the major features, such as faults and shear zones as discrete elements in our analyses, but we're justified there in treating the whole rock mass as a homogeneous material.

Dr. Hoek is shown on screen and a new slide is in the top left corner. An image show engineers standing in a tunnel with water pouring out from the rock above them. Text reads “Groundwater can cause construction problems in tunnels and instability in slopes.” After a moment, the slide increases to fullscreen and then transitions back to Dr. Hoek speaking.

Dr. Hoek: Water is extremely important, and it can cause huge problems in tunneling, mainly from a construction point of view is just a nuisance, you've got to get rid of it. But it also has pressure implications and in slope stability it's extremely important because it causes a reduction in the shear strength of the material due to the generation of internal pore pressures or pressures on discontinuity surfaces. So, it's really necessary to build a hydrological model in parallel to the geological model in order to end up with a complete description of rock mass behavior. Now, the starting point for most of our current thinking on rock mass behavior goes back to people like Don Deere at the University of Illinois, who proposed a very simple classification based on core recovery. And it generally assumes that you have intact blocks of hard rock,

A slide is shown on screen with an image of a rock slope with a roadway passing through it. The text on the right reads “Behaviour of the rock mass will depend upon the strength of the intact rock, the properties of the discontinuities and their frequency in relation to the size of the foundation, slope or tunnel.” Alternates between the slide and Dr. Hoek being shown on screen.

Dr. Hoek: tightly interlocked with their neighbors, if they're still in-situ and undisturbed, and that the behavior of the mass is really a function of how those blocks can move, rotate, deform, and sometimes break but primarily the behavior of the rock mass is controlled by the discontinuities, by the shear strength, and by the three-dimensional orientation of those discontinuities in the mass.

A new slide is shown on screen with two images. On the left an image of a man looking up at a rock slope. On the right, the image shows two engineers looking at blocks at risk of toppling. Text below the images reads “The stability of slopes in relatively massive rock with widely spaced outward dipping joints will be controlled by wedge or block slides. For predominantly inward dipping joints, the failure process may involve toppling of blocks or columns.”

Dr. Hoek: And at the one end of the scale, if we have massive rock with really widely spaced discontinuities, the failures are clearly structurally defined. So, you see there on the left, a wedge failure in an open pit mine bench, where two intersecting planes have created a situation where the wedge simply slides out along the line of intersection. And on the right, you see a potential toppling failure, where the major joints are dipping into the slope rather than out of it and where a tall column will tend to gradually topple over.

Transitions to show Dr. Hoek on screen again.

Dr. Hoek: The consequences of those two types of failure are quite different but they're both controlled by structural features rather than by the properties of the intact material. And similarly, in the tunnel,

A new slide is shown on screen with an image on the left of a tunnel through rock where you can see through to the other side. Text on the right reads “Tunnels in relatively massive rock with widely spaced joints can suffer instability as a result of falls of gravity driven wedges or blocks.” After a moment, the screen returns back to show Dr. Hoek.

Dr. Hoek: this is a tunnel in Wales, built perhaps a hundred and twenty years ago and through a slate quarry bench. And you can see that the structural definition of the failure there is very very strong and wedges in the roof simply fall out under gravity. There was no support applied in those days in the building of tunnels.

A new slide is shown with two images. On the left a man stands in front of a large rock slope near a road. On the right, an image of two men standing under a rock arch, one is taking a photo of the other. Text overtop of this image reads “Reconstruction of arch over the entrance to the original Olympic games stadium at Olympia, Greece (The games started 2,800 years ago).” Text at the bottom of the slide reads “Intersecting joints in hard rock can form a tightly interlocking matrix of blocks which results in a strong rock mass in which stable excavations can be excavated. These rock masses are very strong in compression but have practically zero tensile strength.”

Dr. Hoek: And again, you see on the left there, a tightly interlocking hard rock mass which although it's fairly ragged, still enables us to build a steep stable slope. Incidentally, it's very difficult to achieve perfect blasting in a rock mass like that because the near surface blocks are big enough that they'll fall out and give you a very ragged surface.

Transitions back to showing Dr. Hoek on screen. After a moment, the slide is shown on screen again.

Dr. Hoek: So, people tend to look very hard to see perfect blasting results which are not always possible to achieve because the rock doesn't allow you to do that. The slide on the right is interesting because that's a reconstruction of the arch over the original entrance to the Olympic Stadium, two thousand eight hundred years ago in Greece, and there's no cement in that arch, it simply blocks which had been cut and formed into an arch by putting them over a formwork and then taking the formwork over.

The slide cuts to show Evert speaking on screen.

Dr. Hoek: And so clearly there's a rock mass with zero tensile strength, forming a very stable and very strong structure. And you see many as you go through Europe, you see many old arch bridges built exactly like that. So, rock masses tend to be very strong in compression and extremely weak, in fact zero strength in tension. That's something we have to recognize.

A new slide is shown on screen. On the top left is an image of a man beginning to climb up a jagged rock face. Text to the right of it reads “The mechanical behaviour of a jointed hard rock mass depends almost entirely upon the orientation and shear strength of the discontinuities.” An image on the bottom right shows a different man looking up towards a rock slope which is less uniform in shape. Text to the left of it reads “Tectonically deformed weak rock masses may have extremely complex structural patterns and shear failure may already have occurred in most of the component rock pieces.” After a moment, Dr. Hoek is shown back on screen.

Dr. Hoek: And just to summarize again, in a hard rock mass, the behavior is dependent almost entirely on the orientation and the shear strength of the discontinuities. But, as shown in the lower right, if you have a tectonically deformed rock mass, typically of those that we find in the Andes or the Alps or the Himalayas, the structure has been almost completely destroyed by tectonic movement. Even the intact pieces of rock may have been broken or sheared and you have an entirely different form of behavior.

A new slide is shown. In the top right an image of grey rock with a green axe in the middle. Text to the right reads “The structural fabric in rock masses which have suffered severe tectonic deformation has been largely destroyed and interlocking block no longer exist. In addition, the intact rock pieces may have been weakened and the shear strength of the discontinuity surfaces reduced to their residual values.” An image in the lower right shows miners in a horseshoe-shaped tunnel. Text to the left reads “In such rock masses, excavations may suffer from ductile compressive failure resulting in large deformations.”

Dr. Hoek: So that, here's an example from Venezuela, where you see at the tunnel face the rock mass has been very heavily sheared. There's almost no recognizable structure to it now and on the lower right you see, the consequence of that massive deformation failure, squeezing as we call it, in a tunnel.

Transitions back to Dr. Hoek.

Dr. Hoek: Now putting all of that together, there have been many attempts to build up a series of classifications or characterizations for determining the numbers that engineers really do need for designing in rock masses. And I'm only going to deal with one of these, which is the geological strength index,

An image of the Geological Strength Index is shown. For detailed explanation of each cell in this chart please refer to https://www.sciencedirect.com/topics/engineering/geological-strength-index. A blue ellipse is highlighted with 85 written inside in the box of intact or massive rock with very good surface conditions.

Dr. Hoek: which and the prime publication for that is by myself and Paul Marinos from Athens in Greece in 2000. And what we tried to do was to develop a very simple chart, to be used visually by geologists to define the character of the rock as a number. And in intact rock as you see here that the blue ellipse with the number 85 in the center of it is four massive very sparsely jointed rock, very clean tight joints and it behaves almost as an intact material.

Transitions back to Dr. Hoek.

Dr. Hoek: Or does behave in the extreme where you have the GSI value of a hundred as intact material. And that tends to give us problems of brittle spalling in tunnels and boreholes. That's the one extreme.

The GSI chart is shown on screen again with the blue ellipse slightly lower and closer to the right than before with 65 written inside. An image of a man climbing up a rock slope is on the right with text below it “Blocky rock mass with cubical blocks defined by rough, slightly weathered joints. Small wedge or block failures create ragged excavation surfaces making it difficult to blast smooth faces. Excavations are generally stable.”

Dr. Hoek: We move on then to the blocky, strong rock, tightly jointed, tightly interlocking, strong wedges. And that the column and the row reads blocky interlocked rock masses and rough not necessarily as strong as that as the previous one but still pretty good. And that gives us a GSI value of 65.

A transition happens. The chart remains unchanged. The blue ellipse now reads 45 and is in the middle of the chart region. An image of two engineers have their hands on a rock face. Text below reads “Very blocky rock mass with multi-faceted angular blocks defined by moderately weathered joints. Very small wedge or block failures of face and also some through-going shear failures of unsupported rock masses can occur without warning.”

Dr. Hoek: And then we move on to multiple jointed rock. And this is from the Solomon Islands in northeast of Australia, a very active tectonic area, andesite here, volcanic material but very strongly jointed. So small blocks perhaps, ten or twenty centimeters in size,

Transitions back to Dr. Hoek on screen. After a moment, the previous slide appears on screen again.

Dr. Hoek: four or five joint sets, very, very tightly interlocking but they're easily disturbed. So, blasting in that would or simply call it to rattle down on the floor. And there we've got a GSI value of 45.

The slide fades into the next where the blue ellipse moves further lower and to the right with 30 inside of it. The image changes to two people looking at a rock slope. Below it, text reads “folded blocky poorly interlocked rock mass with multi-faceted irregular blocks defined by highly weathered joints. Through-going shear failures of the rock mass can occur, and installation of support is difficult and not always effective.”

Dr. Hoek: This is from Taiwan, and you've got folded sedimentary rock there, where the original layers of sandstone have been broken up into little pieces. There's a lot of weathering, these material which was originally in the joints has become quite soft and of a lower shear strength. And so there we have a GSI number of 30.

Transitions back to Dr. Hoek.

Dr. Hoek: And finally, in the material which in Greece is called Flysch which is a sedimentary material that was deposited before mountain-building and subsequently completely tectonically deformed during the mountain process, mountain building process, where there's no recognizable structure.

The GSI chart appears again, this time the blue ellipse is near the bottom right corner of the chart with an 18 in the center. The image on the right is of a man with a blue binder, looking at a rock slope. Text below the image reads “Tectonically deformed rock mass in which structure has been destroyed and discontinuity shear strength is at residual. Time-dependent ductile shear failure of overstressed rock can occur resulting in squeezing of tunnels and sloughing of slopes.”

Dr. Hoek: It's bordering on a soil but it's still fundamentally rock particles in intimate contact with each other.

Transitions back to Dr. Hoek on screen. A slide appears in the top right corner and after a moment enlarges to fullscreen. It shows a graph, x axis is minor principal stress and y axis is major principal stress. The chart on the right charts GSI on the x-axis and Deformation modulus on the Y-axis. Text in the bottom right of the slide reads “For the range of rock masses shown in the previous figures, the GSI values together with the intact rock properties can be used to estimate the triaxial compressive strength and the modulus of deformation of each rock mass.”

Dr. Hoek: That then allows us to develop a series of curves. And I'm not going to go into the detail of these in this lecture, there's much too much detail to discuss. But you see there, going from GSI of 85, the top, the very strong almost intact material, very high compression, it's just very low tensile strength if any. And for all the other materials, where you have GSI of 65, 45, 30, and 18, I'm suggesting, that there is no tensile strength at your disposal. Zero.

The screen transitions back to show Dr. Hoek.

Dr. Hoek: So that you still have compressive uniaxial compressive strength, but as soon as you move into tension, it drops to zero. And that's fundamentally important to have that kind of information available for our numerical models and analyses, that we then use later for engineering design.

Transitions back to the previous slide.

Dr. Hoek: On the right, is a different curve and this is the deformation modulus of rock masses with different GSI's. There are similar curves for other types of classification, and they all end up with roughly the same results. But this is one that that I have used for many years.

Transitions back to Dr. Hoek on screen.

Dr. Hoek: So, we have both strength and deformation characteristics, which we can then move forward into design. Now, GSI cannot be used for a whole number of things. It cannot be used for intact rock or sparsely jointed rock such as the wedge failures I showed you earlier in the photograph. It has no meaning when used in that environment. None. It cannot be used for broken or transported materials,

A new slide is shown with an image on the right showing a dam and a slope that has failed and debris has fallen below into water. Text to the left reads “GSI cannot be used for structurally controlled failures or for transported materials such as rockfill, waste rock, rock masses that have failed, uncemented conglomerates, sand, sandy soils and clays.” After a moment, Dr. Hoek is shown on screen again.

Dr. Hoek: as you see here this is a dam foundation excavation in Greece and you see the benches very nicely cut and designed at the top there, for which GSI is appropriate. That's a reasonable tool. And then the spilt material and the waste rock, the broken rock for which GSI is absolutely not applicable. Transported material, waste rock, soil, sandy soil, clay, none of those can be dealt with by GSI. There's one area which is marginally possible and that is compacted rockfill. Where you've taken a broken rock and compacted it back almost, almost to an interlocking rock, but not quite. And that's important in dam design and in fact this slide,

A new slide shows a man in a suit standing next to a large scale triaxial cell. After a moment, Dr. Hoek is shown on screen again.

Dr. Hoek: is of a very large triaxial cell for testing one meter diameter, 2-meter-long core specimens of compacted rockfill used in the snow mountains hydroelectric project on Australia between 1949 and 1974. They built a huge number of dams and underground facilities. And so, this was a key piece of equipment for testing those core samples. And the behavior of that material is not too dissimilar from a weak rock and concrete fits into the same general spectrum of behavior and there are similar cells for testing concrete.

An image appears in the top left-hand corner of the screen. After a moment, enlarges to fullscreen. The image on the left shows two men with large coal cutting machine. On the right, a man looks at a coal pillar.

This is a series of tests, this was carried out in South Africa in the 1960's, after a massive mine, coal mine collapse. And we mobilized a group of people to try and understand the strength of coal on a real scale. Not little laboratory samples which are very difficult to do in coal anyway but at a scale approximately equal to the field scale of coal mining pillars. And on the right, there you see a coal pillar which was about four feet high and two feet by two feet in size and on the left you see a coal cutting machine which is just a great big chainsaw, enormous chainsaw, which is used to actually cut the coal. And that was used to cut the coal specimens which were then loaded by the array of hydraulic jacks you see there and instrumented.

A new slide is shown with two new photos of the coal pillar. A closeup on the right showing a failed coal pillar and two miners looking at it with machinery on the left. Text below reads “Drilling for installation of in situ stress measurement equipment in the coal pillar test. The pillar was eventually tested to destruction as shown on the right.”

Dr. Hoek: And before testing the pillar we drilled in stress measuring cells, so that we could measure the stresses during loading and on the right there you see a failed coal pillar. Those types of tests are extremely expensive.

Transitions back to Dr. Hoek

Dr. Hoek: And you can't really do enough of them to build up a database of sufficient information to do probabilistic analysis, but occasionally they're useful to do. We seldom do this kind of thing anymore, but we still do a lot of plate jacking tests or jacking test to measure modulus density and I'll show you some of those in a moment.

A new slide shows an illustration on the left showing an in-situ shear test set up in an adit in a dam foundation. An image on the right shows a man holding up a block of rock looking at it with a flashlight. After a moment, the screen returns to Dr. Hoek.

Dr. Hoek: This is a dam project in China, where in-situ sheer testing was done and you'll see the equipment there on the left, the vertical blue hydraulic jack is to apply the normal load on the specimen, which is simply cut out as you mine the trial tunnel in which the test is carried out. And the inclined blue one applies the shear load and a set of instrumentation to measure the deformations and so on. On the right you see the shear surface, which had been tested and to get one point on a shear strength curve. I remember visiting France many years ago with a dam engineer, Pierre Londe, and he showed me a large test that he'd done, and he said this test cost me a lot of money and he said it gave me one point on a curve and I'll never do that again. So, we don't tend to do too much of this. We tend to do it now as I described in the previous lecture in the series by measuring the friction angle and adding the roughness or the characteristics of the shared surface should I say.

A new slide is shown on screen. The photo on the left show a man looking at a large silver piece of equipment resembling a large circular saw. On the right a man looks towards an older looking piece of cutting equipment. Text below reads “Slot cutting equipment and hydraulic flat jack for in situ determination of rock mass deformation modulus. Laboratorio Nacional de Engenharia Civil, Lisbon, Portugal used this method in the 1970’s.” Alternates back to Dr. Hoek and the slide.

Dr. Hoek: This was equipment used in Portugal in the 1970's and what they did here was to drill a hole and then lower a saw that the hole was to accommodate the drive mechanism for the saw and cut a vertical slot into which was then lowered a flat jack consisting of two welded steel plates close to each other and pump in hydraulic fluid under pressure so that you're deforming the wreck the rock mass and measuring its deformation modulus. The difficulty is, with flat jack testing of this kind, and there have been several attempts to do it, is you're very close to the surface. So, the rock tends to be disturbed before you're doing your test anyway. So much more typically and this is still –

A new slide is shown on screen with an image on the left of a man in an underground excavation with a large piece of equipment. On the right is an image with machinery performing a horizontal test. Text reads “Testing equipment for determining the modulus of deformation in a rock mass. This test on the left was carried out in an exploration adit for an underground hydroelectric power cavern in Taiwan. The horizontal test illustrated above was in a test adit in the Drakensberg Pumped Storage Project in South Africa.” Alternates between Dr. Hoek and the slide on screen.

Dr. Hoek: that's a recent photograph in a hydroelectric project. We tend to do this with large hydraulic jacks. So, you create an excavation at the trial tunnel or an adit as we call it, and you put into it a large array of hydraulic jacks. And down the center between the four jacks there, extensometers are inserted, so that you can measure the deformation of the rock mass over a considerable depth below the loading pad of the jack. And there's one a vertical jack and another one horizontal jack on a project in South Africa. And that type of testing is still done and is still very important because it calibrates the rock mass on-site, in location. And here's a case history that wraps up this discussion because we really need to have some means of confirming that all of the information we've gathered together and analyzed and come up with a failure criterion and done piecemeal testing, pulls together and gives us a workable rock mass model which actually does perform in the field. And so, I'm going to go through a little case history here of the Drakensberg pumped-storage project in South Africa,

A picture appears of the Drakensberg Mountain in South Africa, from a mountain top with foliage and dry grass overlooking brown hued mountains. They extend further into the background and become smaller and fade to a blue colour. After a moment, this image expands to fullscreen and then returns to Dr. Hoek on screen.

Dr. Hoek: which was designed in the early 70s and completed construction by the late 1970's. It's in the Drakensberg mountains in South Africa. It was a storage project, which means that during the generation of electricity if you have a surplus, you pump water into an upper reservoir and when you have a demand, you lower the water down through the turbines and generate electricity. And it was in a very stable predictable area of geology, bedded sedimentary materials: sandstone, siltstone, mudstone layers. And because it was the first major underground excavation in South Africa, it was decided with the agreement of the client, to do two major in-situ tests. One of them was this,

An image is shown on screen of 4 miners in an underground excavation work room. Text below reads “Monitoring piezometers and extensometers during a full-scale pressure test on a concrete lined pressure tunnel during construction of the Drakensberg Project.”

Dr. Hoek: where we took a three-meter diameter concrete line tunnel, bulk headed it off and loaded it under full pressure and measured. We had many extensometers and piezometers in the rock mass around it and in the concrete and measured the rock mass behavior under test.

A new slide is shown with a drawing on the left of yellow, pink, red and white horizonal stripes showing the bedding layers and spacing. An image on the right shows a bulldozer in an underground powerhouse cavern. Text below it reads “Geological cross-section and underground powerhouse cavern during construction of the Drakensberg Project, South Africa.” After a moment it transitions back to Dr. Hoek.

Dr. Hoek: We also did a test to simulate the creation of the arch of the cavern. And there you see horizontal bending in yellow and red, representing a geological model of the actual rock mass and the profile of the cavern that was to be mined. And on the right, you see a photograph of the cavern when it was mined with a trapezoidal roof and vertical walls. And as I say, this was the first major cavern constructed in South Africa. And so, it was decided to make a model of that, full-scale.

A new slide shows a Finite element model of the construction stages for a test excavation of the arch and upper sidewalls of the underground powerhouse for the Drakensberg Project. Below it a plot shows the agreement between the measured and calculated displacements in one of the extensometers installed in the rock above the arch before excavation commenced.

Dr. Hoek: And so, at the end, the one at the end of the cavern, we constructed the trapezoidal arch, left a pillar in the center, left jacks in place to keep everything in plan and put support in. And we had extensometers, the green line running up from the center of the crown there, is an extensometer with multiple points. And so, we could gradually release the load on the jacks, release the tension in the rock bolts and see how the whole structure performed.

Transitions back to Dr. Hoek.

Dr. Hoek: And at that time, our models were very very crude, we only had elastic numerical models and so we were not able to do non-earlier analysis and our analyses were fairly crude. The cavern was successfully constructed and it's still working forty years later, but almost 30 years later I went back in 2007, and re-analyzed the extensometer behavior and you see the graph on the bottom there,

The previous slide is shown on screen again. After a moment, it returns to Evert.

Dr. Hoek: Where the dotted lines are measured, and the solid lines are calculated in the colors correspond. And so, in rock engineering, that's pretty good. You couldn't do a lot better. So, we had confidence then that this model actually worked. That confidence was tested recently because about 10 years ago, they started working on a sister scheme about 30 kilometers away. They needed another pump storage scheme and so a new scheme called originally Brown-Hoek, now called Ingula, was designed and the question arose, could we use the data from Drakensberg, in a rather similar, not identical, but similar rock mass setup for the design of the Ingula underground excavations. And I'd never went to site, but I corresponded a lot with the designers because I'd worked on Drakensberg. And they then went ahead and designed a cavern using very similar, very similar rock mass model to that which I had derived from my back analysis. Basically, what it showed was that if we treated the horizontal layers of rock as intact beams with weak bedding surfaces, we got a very good agreement with the actual behavior as measured. And that's the Ingula cavern under construction.

A photo of an underground powerhouse excavation with machinery and mesh along the side of the wall. After a moment, a new slide is shown with an illustration of the Ingula project instrumentation arrays in the powerhouse and transformer hall. Alternates between this slide and Dr. Hoek.

Dr. Hoek: These are the sequences and I'll be talking in lecture number six about the design of large caverns, and about the need to sequence very carefully, design your structure. Because these days, we use rock as the engineering material. In the old days, we tended construct a hole and then build a building inside it, because the engineers didn't trust the rock. Now we know enough about rock to use the rock itself as the engineering material and to support it with cables and rockbolts. And I'll talk about that in lecture number six. But those are the layout of the extensometers and this excavation sequence, and they're two references to two papers describing this.

Four charts are shown with dotted and solid lines on each. A) Machine Hall crown center line B) Machine Hall side heading above crane beam level C) Machine Hall sidewall above operating floor level D) Transformer Hall crown center line. Text at the bottom reads “Comparison between calculated and measured displacements in typical extensometers in the Ingula caverns.”

Dr. Hoek: And here are the measured deformations, the upper left and the lower right, are deformations measured and calculated for the crown and they're almost perfect agreement. The upper right and the lower left are measurements in the side walls which are a little bit more difficult because you have a very tall, irregular side wall with different cutouts in it and so you get slight variations in its behavior in terms of modeling.

Transitions back to show Dr. Hoek on screen.

Dr. Hoek: But again, in terms of rock engineering, these are excellent results and to get this far, I think is pretty good. Thank you for your attention.

Screen fades to black.

Text appears “Practical Rock Engineering Lecture Series is sponsored by Rocscience. Software tools for rock and soil.” Fades to black.

Text appears “Filmed on location at University Club Toronto, Ontario, Canada October 2015. Technical Production by Harvey Montana Productions.” An image of a painting showing sheep in a pasture is below with text www.seanharveydp.com

This fades to black and ends the video.

Go back to Hoek's Corner Video Lecture Series page.

Back to top