Large Underground Excavated Caverns Transcript

This transcript describes the YouTube video "Large Underground Excavated Caverns - Dr. Evert Hoek Lecture"

Rocscience logo fades in overtop a black screen.

Transitions to text on screen “Presents Practical Rock Engineering Lecture Series by Dr. Evert Hoek copyright 2016” followed by “Large underground excavated caverns. Lecture 6.”

Transitions to Dr. Evert Hoek standing in a large room speaking to several seated people in the room facing him.

Evert Hoek: This lecture, number six in the series, deals with large underground excavated caverns and I emphasize excavated caverns because I'm not going to deal with natural caverns at all in this series of lectures and what you see in the opening slide

Transitions to a photo of a finished powerhouse cavern

Evert Hoek: is a large underground powerhouse after it's completed and we're going to go through the process of how you get there in this lecture.

Transitions to Dr. Hoek on screen.

Evert Hoek: I've chosen large underground hydroelectric projects because it's an easy topic to talk about. It's a nice, confined situation but the same methodology applies to any kind of underground cavern that you're excavating, and you'll see some examples later on.

An image is overlaid in the top right corner of an excavated cavern of a hydroelectric plant.

Evert Hoek: This slide shows a typical cross section of an underground powerhouse.

Image covers the whole screen.

Evert Hoek: And what you have is water entering the headrace tunnel as shown by the upper red arrow there. It splits up to the different turbines, in this case six turbines passes through the turbines and then emerges from the tailrace tunnel and the other excavations are to do with transformer galleries and access points.

Transitions back to Evert Hoek with the same image in the top right corner.

Evert Hoek: So, you end up with a fairly complicated three-dimensional structure and the caverns can be large typically 25 meters span and up to 40 meters in height and perhaps a hundred meters or more long.

Image disappears from the top right corner.

Evert Hoek: Now, I want to pause it as this moment to talk about the access tunnel that you see there.

The same image is displayed on the whole screen.

Evert Hoek: This is not a typical tunnel in that for most tunneling projects where we're simply going from A to B in a tunnel, the whole object of the tunnel design is to create a production-line scenario.

Transitions back to Dr. Hoek standing.

Evert Hoek: So you try and find what's common in terms of the rock types, support systems and so on, so that you maximize the routine elements of the operation, so that the crew coming in every day knows what they've got to do and you can achieve very high advanced rates through treating it as a production line and you do it generally with minimal information because a long tunnel you simply can't get enough geological information and technical information along the route. You do an averaging process and you do the best you can and then go for it. In this case, we're entirely different we have a large excavation in one piece of rock, one volume of rock, which is a very high investment process so they're going to be very expensive pieces of machinery in there. It's got to operate for a hundred years or so. So you can afford the luxury of putting an access tunnel in before construction so it becomes an exploration tunnel and that gives us the opportunity of getting in and really taking a good look at the rock doing some measurements and so we have a much better base of information for our design.

Transitions to slide showing two photos. On the left, a man looking at surface exposure of jointed sandstone at the site of the Mingtan Pumped Storage Project in Taiwan. On the right, a man looking at underground exposure of inter-bedded sandstone and siltstone in an exploration tunnel.

Evert Hoek: So, for example, here is a surface exposure for a project in Taiwan called Mingtan which I'm going to discuss as an example and on the left you see the surface exposure of the bedded siltstone sandstone series tilted bedding.

Transition back to Dr. Hoek

Evert Hoek: and seeing that on surface gives you one impression,

Transition back to the previous slide with the photos.

Evert Hoek: but underground in the exploration tunnel you get an entirely different impression of the rock mass that's what you've got to work with underground and that's a very important advantage.

Transitions back to Dr. Hoek speaking.

Evert Hoek: It also gives the advantage to do stress measurements for example,

A new slide is shown on screen picturing in-situ stress measurements. Two pictures show miners in underground tunnels.

Evert Hoek: as shown on the upper left or jacking test to measure the deformation modulus of the materials so that you can if the preliminary contract is planned correctly, you can do a lot of work to build up a very solid database.

Transitions back to Dr. Hoek

Evert Hoek: The particular project I was dealing with as I say is the Minturn project in Taiwan and I was asked to go there to advise them primarily in the first place on the shape of the cavern.

An image showing alternative cavern shapes considered for the Mingtan appear in the top right corner.

Evert Hoek: and three caverns had been proposed and in general the shape of the cavern is defined by the by its function.

The same image appears full screen.

Evert Hoek: So, in this case you have the turbine if you look at the left-hand picture there you'll see water going through the turbine. You have the turbine itself in red there. You have a series of rooms for switchgear from four valves and so on, on the left-hand side which you've got accommodate and very importantly you have an overhead crane because the turbine components might weigh up to 100 tons each.

Transitions back to Dr. Hoek speaking.

Evert Hoek: And so they typically the overhead crane has a capacity of 2 or 300 tons.

Transitions back to the same image.

Evert Hoek: So, the first cavern they asked me to look at was the traditional cavern that was typically done I would say in the 1950s and 1960s. In those days we didn't have any precise tools

Transitions back to Dr. Hoek speaking with the same image in the top right-hand corner.

Evert Hoek: for analysis. The design was basically on experience and rule of thumb and the tendency was to treat the cavern excavation as a hole in the rock into which you put a building the building being the powerhouse structure.

The image disappears and transitions to a different angle of Dr. Hoek speaking.

Evert Hoek: So, it's like taking a surface powerhouse and putting it into a hole underground and therefore the crane beams were constructed of concrete and put underground and the construction sequence was typical of a surface building and in order to support the failed rock above the arch the easiest concept used was to put a concrete arch in.

Transitions back to the image of the three shapes of arches considered for the Mingtan project.

Evert Hoek: So that was let's say the 1950s and 1960s and some of them are still used with justification. The second proposal was an elliptical cavern and this is to optimize the shape of the cavern in terms of the in-situ stress field and the third option was what we today call a letterbox cavern, which is just the Cavern as tight as possible around the components that go in there.

Transition back to Dr. Hoek speaking.

Evert Hoek: So, in order to study this problem I looked at some case histories and I did some analyses.

An image of a slide appears with a photo of the inside of Tamut 2 underground powerhouse.

Evert Hoek: The cases history I looked at was the Tamut 2 underground powerhouse in the Snowy Mountains Project in Australia and this was done way back in the in the 70’s and you see there a large underground cavern with concrete arches in the roof and way back in the background you can see the crane being constructed on concrete beams or the crane rails and the crane being inserted.

Transitions back to Dr. Hoek speaking.

Evert Hoek: So that is typical of construction of that era, and I then decided to do some analyses. Now, we have tools that enable us to do extremely sophisticated analysis, but I wanted to pause and put in a general point here and this was written by my colleague Pierre Londe.

Slide shown on the screen with text. All text will be explained verbally. Source Hoek, E. and Londe, P. 1974. The design of rock slopes and foundations. General Report on Theme III. Proc. Third Congress International Society for Rock Mechanics, Denver. 1(A), 613-752.

Evert Hoek: in a general report we did for the International Congress for rock mechanics in Denver in 1974 and Londe said “the responsibility in the design engineer is to judge soundly rather than compute accurately” and

Transitions to Dr. Hoek standing with the slide in the top right corner.

Evert Hoek: So, to me in general but in particular here, the computational tools we have are aides to rather than a replacement for judgment and very very powerful they are in doing that. So you don't have to go to extreme limits to calculate all sorts of sophistication, a simple model will do. You have to ask yourself before you start

Transition to the slide fullscreen.

Evert Hoek: Have you really defined the problem? Do you know what you're trying to solve? Do you have sufficient information to make the model at whatever stage simple or complicated? What programs should you use? obviously is an important question and how should you present the results?

Transitions back to Evert Hoek speaking.

Evert Hoek: and this is something that in engineering is often a disaster scene, where these analyses are done and then a presentation is made to a Board of Directors or a client who has no idea what you're talking about and so you really have to think hard about how to get the information across.

Evert changes to the next slide showing the analysis of failure and displacements for a cavern with a concrete roof arch. Showing with rainbow factor of safety contours.

Evert Hoek: So in this case, I did very simple analyses which were non-linear allowing me to look at failure and the first one I looked at was the cavern with a concrete arch. The concrete arch worked very well in some of the early projects which tended to be in hard rock, where the

Transitions back to Dr. Hoek speaking with the image in the top right corner.

deformation modulus of the rock was higher than or equivalent to that of concrete. This was a sedimentary series of softer rocks which were much more deformable and what happened here is that as the cavern went down,

The image is shown fullscreen.

Evert Hoek: the construction and pull the construction say stages were modeled the sidewalls tend to bulge in and the top of the cavern pinches the arch and breaks it and in fact there have been cases where exactly this has happened and

Transitions back to Dr. Hoek speaking.

Evert Hoek: so that immediately ruled out this as an option. Before we leave this line it's worth looking at a couple of things. What I've shown here are the deformation

Transitions back to the slide.

Evert Hoek: vectors that you can see the little blue arrows there and you see tiny little crosses and circles. Circles representing tensile failure, crosses representing shear failure in the materials. The other thing you notice is that the little cavern which is the transformer cavern tends to be sucked towards the big cavern just because the big cavern is a dominant component in the whole analysis and

Transitions back to Dr. Hoek.

Evert Hoek: that leads us to do some very simple analyses which produce the conclusion that ideally your small cavern should be at least as far away from the big cavern as the height of the big cavern. That gives us a nice short distance because the electrical engineers want a short distance for their chart for conveying the current from generators to the transformers and we want a big distance, so this is a good compromise. The next one I looked at was the elliptical cavern.

A new slide is shown of the analysis of failure and displacements for an elliptical cavern with rainbow colours showing the displacements.

Evert Hoek: and this is clearly better. There is a smaller zone of failure around it. The deformations are significantly smaller 164mm as compared with 184mm in the previous one and the problem that the Taiwanese client had was,

Transitions back to Dr. Hoek on screen.

Evert Hoek: that they said we've never built a cavern like this before we're not sure we have the skill to actually do the profile so it was their least preferred option of the two remaining ones and what we finally settled on was just a letterbox Cavern.

A new slide is shown onscreen of the analysis of failure and displacements for a letterbox shaped cavern.

Evert Hoek: plain and simple. and one thing to point out at this point is that we're going away from the concept of a cavern that is built with support like concrete arches. We're using the rock material as the material and it's supported by Rock bolts and cables there are no internal elements in here

Goes back to Dr. Hoek on screen.

and you'll see later on some examples of cavern's that have no internal support at all. So, the rock is the material that we're dealing with and that's an important topic in this whole series of lectures that we are dealing with rock as an engineering material. The other issue with the Mingtan project was that there were 21 faults that we had to pass through in the cavern.

A slide is shown on screen with the final orientation of the Mingtan cavern complex in relation to bedding faults and shears.

Evert Hoek: and this is one of them shown there and you'll see that the cavern has been rotated slightly in order to pass through it at right angles. The original cavern alignment was determined by hydraulic considerations to give the water a shape straight passed through the turbines and I said that's not acceptable and we rotated the cavern 22 degrees in order to give you a straight shot at the at the faults rather than the water a straight shot at thing at the Cavern

The video goes back to Dr. Hoek standing with the same image in the top right corner and then disappears to focus on Evert.

Evert Hoek: and this gave us the opportunity of going through the 21 or 22 faults without problem. in fact, a lot of work was done in preparation because the preliminary contract was a large one and there was sufficient funding to allow us to go in and improve the rock above the cavern arch so that when the main contractor came in he had good rock to deal with which made the contracting process much simpler and the final support design is illustrated in this slide

Slide is shown on screen with text: Excavation and support sequence for the Mingtan powerhouse cavern. Installation of untensioned grouted 50 tonne capacity cables downwards from gallery and upwards from construction adits to support rock mass above the roof arch. Excavation of the roof arch from a central heading, installation of cables with 50 mm thick shotcrete and additional rockbolts where required. Excavation of 2.5 m vertical benches and installation of 112 tonne tensioned and fully grouted cables on a 3x3 m grid spacing with intermediate rockbolts. Completion of cavern excavation with 150mm total thickness of shotcrete. Access to instrumentation and for minor repair work provided by a temporary overhead crane.

Evert Hoek: which I'm not going to go through in detail but which the viewer can pause the program and read and it involved basically suspending the roof from a gallery shown by the red arrow, ten meters above the arch and from that gallery untensioned grouted cables were drilled in downwards to suspend the roof and when they were subsequently exposed, when the cavern was excavated, the ends were cleaned often and face plates and clamps were put on to form a clamped cable. In addition, we constructed two side galleries as you see there by the red arrow and from these cables were put outwards and these are tension cables. These were 50-tonne cables and then as we went down the side walls that cable capacity was increased to 112-tonne capacity typically these are stressed

Transition back to show Dr. Hoek speaking.

Evert Hoek: to about 30 percent of their ultimate capacity when they put in and then they pick up another 30 percent or so as that as the deformation takes place, so they stress to about two-thirds of their of their capacity. And in this case because we at this stage had computing facilities available, we were able to set up programs on-site and to monitor the displacements as we went down and actually alter the tension in different parts of the sidewall as the cavern was excavating and

Screen switched to the previous slide.

Evert Hoek: So, you see the cavern being progressively exploited downwards in a series of stages. These were 2.5-meter vertical benches and it's each stage the cables were installed and finally we had crane beams anchored to the wall and the temporary construction crane placed there to give us working capability during construction.

Switch back to Dr. Hoek on screen.

Evert Hoek: The client insisted still on using concrete column cranes I couldn't persuade him to suspend the crane beams entirely and say the crane beams were put in traditionally in this case.

Photos shown on screen.

Evert Hoek: This is a photograph of a gallery on the left, 10 meters above the cavern from which the downward grouted cables were placed and, on the right, you see the upward tension grouted cables being placed in the two side galleries. This was a very effective solution we really had no problem with the arch of the cavern throughout the whole project.

Evert Hoek is shown on screen.

Evert Hoek: This is the first stage of excavation, and it looks a little bit rough there. As I've mentioned in jointed rock and remember this is a is a dipping bedded sandstone

Photo shown on screen of an Excavated cavern arch supported by 50 tonne capacity cables.

Evert Hoek: and other sedimentary series, shales and so on, and it's difficult to get good blasting so that the

Transitions to Dr. Hoek with the image in the top right corner.

Evert Hoek: cavern blasting is not quite as nice as I would like to see but it hasn't been cleaned up and shotcreted at this stage and the cables are still projecting.

The image in the top right corner is replaced with two more images and then the image moves fullscreen.

Evert Hoek: And then we go on to the sidewall cables and you see on the upper right there, drilling in the holes for the cables and the upper left rather and the lower right the tensioning of these cables which are done wire by wire to tension them up to the designed installation capacity.

A new slide is shown on screen with text cavern under construction for the Mingtan Pumped Storage Project in Taiwan.

Evert Hoek: And there's the cavern getting down towards it's lower benches and you see the construction crane rail at the top there and dimly in the background you can see the construction crane that was used.

Dr. Hoek is shown on screen.

Evert Hoek: In fact in one case it was very important because we found movements in one of the areas of the cavern or a little bit too high and we went up there and there was actually a wedge that had been defined and it happened to fall in such a geometry that what didn't have sufficient cables through it so it was very easy to remedy that by installing additional cables in that location so having access to the roof and I mean meaningful access that you could actually get a drilling machine up there was very important and finally for this particular cavern, we had many instruments are lonely the cavern lengths in different locations.

A new slide is shown on screen. Source Hoek, E and Moy, D. 1993. Design of large powerhouse caverns in weak rock. In Comprehensive rock engineering, (ed. J.A. Hudson) 5, 85-110. Oxford: Pergamon.

Evert Hoek: and what you see there are the top graph is movements in the extensometers and the lower graph is load in load cells attached to the cables and you see a lot of action in a lot of movement during the benching as things are developing and then finally once the benches have moved down the cavern the movements and the roads become very stable.

Evert is again shown on screen.

Evert Hoek: that's a period of a year or so and that cavern has performed extremely well since it was commissioned. Now, talking about cavern interiors and support,

New slide shown on screen with images of caverns.

Evert Hoek: This is a project in Argentina called the Rio Grande pumped storage project and it was the first cavern that the Argentinians have constructed, and the geologists believe this was the best piece of granite in the world and so no provision was made for support. None. Until they realized that there were occasional wedges that came out and they had to deal with them.

Cuts back to Evert speaking

Evert Hoek: But I was a consultant on the project and the company I worked for was involved and we had a team down there and we had to justify every rock bolt and they were only about 900 reports in the entire cavern nothing else, so you see an effectively unsupported cavern there

Cuts back to the previous slide

Evert Hoek: of 25-meter span and 40 meters height and the thing about this there was no cranage.

The image moves to the top right corner revealing Dr. Hoek speaking

Evert Hoek: So, there were no temporary construction cranes at all we had to access the instruments from a from a catwalk suspended from rock bolts in the roof which was a terrifying thing to do and we had no other access

The slide in the top corner disappears and Dr. Hoek is alone on screen.

Evert Hoek: until they came in and built the surface structure underground in this hole and it worked very well but that's not in my opinion the best way to do it. You're not making the best use of the rock as an engineering material to support the structure.

A new slide is introduced with another image of an underground cavern.

Evert Hoek: Here's another example which was a compromise this is in Greece where they had a crane beam bolted to the sidewall for the temporary crane, but they insisted on concrete columns which were put on after the construction had gone down to the bottom they went back and put concrete columns in so it’s sort of compromise between one and the other.

A new slide with another image of an underground cavern.

Evert Hoek: Here's another example this is the Drakensberg project in South Africa, where the crane beams you see them on the left there, were actually anchored with additional cables to the side walls and those were for both temporary and permanent cranes.

A new slide is shown with an image of an underground cavern with a temporary crane.

Evert Hoek: So, you see there that temporary crane in place and we could construct the permanent crane in a niche of the end of the of the gallery during construction so that it was available fairly early

Dr. Hoek is shown on screen with the image in the top right corner. After a few seconds the image disappears.

Evert Hoek: in the process to move heavy bits of equipment and waste rock. And the final example here is the ultimate, an elliptical Cavern.

A new slide is shown on screen with an image of an elliptical cavern. Photograph courtesy of Dr. Sigmund Babendererde.

Evert Hoek: with suspended crane beams for both temporary and permanent the Sinkarak hydroelectric project in Indonesia. I was not involved in this, but my friend Sigmund Babendererde provided me with this photograph and that is using rock as the engineering material in the full sense everything is supported by the rock.

Transitions back to Dr. Hoek on screen.

Evert Hoek: Now, we have other categories of underground excavation that are important for us to deal with in this lecture and one of them is very shallow underground excavations.

A new slide is shown on screen of a shallow underground excavation with engineers and equipment inside.

Evert Hoek: This is an underground station in the Oporto metro subway system in Portugal and the span is 18-meters and the depth below surface is about 20 meters, and this is in the bottom of it is in granite but the top is in weathered granite weathered to residual soil and so it becomes very important.

Transitions back to Dr. Hoek speaking. And then after a few seconds returns to the slide.

Evert Hoek: to very carefully consider the construction sequence and what you see there on the left is a drill that's capable of drilling 12-meter holes for jet grouting so you can only do this in clay free materials like weathered granite because otherwise the graft won't flow so the 12-meter hole is drilled and then the drill is pulled out and jet grouting is done to effectively create a concrete arch and then on the side,

Transitions back to Dr. Hoek with the image in the top right corner. After a few seconds the slide disappears. After a few more seconds, we return to the slide in fullscreen.

Evert Hoek: the tunnel, the side had it is extended eight meters so there still an overlap and then it's flipped over and grouting is done from the side at it while excavation of the central adit is carried out and it's a very effective technique but it requires very careful planning. The curved wall that you see there is sacrificial. That's removed after it's carried out its function of supporting the different excavations during construction.

Dr. Hoek is briefly shown on screen and then transitions to a new slide with graphics of a simplified shape and excavation sequence.

Evert Hoek: Here’s an example from Greece there were three stations Academiia, Omonia and Acropolis in the Athens Metro that were constructed about 15 meters underground and the caverns were 16 or so meter span and some of these were in very critical locations so it was very important to control the ground displacement and you'll see these were done in 10 excavation stages using basically grafted rock bolts, untensioned grouted rockbolts so that they attention themselves as thee rock deformed and shotcrete lining and then finally a concrete lining in place

Transitions back to Dr. Hoek on screen briefly and then a new slide is introduced with an image with coloured contours, mostly blue, green, yellow, and orange.

Evert Hoek: And in this case I did an analysis this is many one of the many analyses that were done where you see the sequential excavation showing a phreatic surface. It's difficult to see but it's a red line six meter long grouted untensioned cables they were 4 and 6-meter long cables used in different rockbolts rather than cables and 25 meter shotcrete lining and so that's stage one and that's stages one two and three.

Stages 2 and 3 are shown on consecutive slides showing the change in colour producing more yellow, orange, and red colours.

Evert Hoek: Leaving the central pillar in place until the sides are fully stabilized and then finally the central pillar is replaced.

The new image on screen shows the final stage with text Approximately 20 mm maximum surface displacement after excavation. It then transitions back to showing Dr. Hoek on screen.

Evert Hoek: and you end up with a stable profile and about 20 millimeters of displacement on the surface. This was a very effective technique used on all three of the stations.

A new slide with photo is shown with text Excavation of the central pillar in the Athens Metro Akropolis underground station.

Evert Hoek: and this is a photograph provided by Attiko Metro of you see the two curved temporary walls there shotcrete walls and the central pillar being replaced and they're applying lattice girders into the shotcrete lining there in preparation for the final lining.

A brief moment where Evert is on screen and then it changes to the new slide with a drawing of the Norwegian Olympic Ice Hockey Cavern at Gjovik.

Evert Hoek: Here's another example that comes from Norway and this is an ice hockey Cavern of 61 meters span 24 meters height and depth below surface 45 meters. Now this is in in hard granitic type rock and so we're looking here at a totally different situation from the previous examples which were in soft ductile material.

Transition back to Dr. Hoek on screen.

Evert Hoek: This is interlocking blocks and it was possible because the horizontal stresses in Norway in this region were high.

An image appears in the top right corner showing the UDEC-BB model of the Gjovik cavern. After a moment the image fills the entire screen.

Evert Hoek: So this is a UDEC Barton-Bandis model that was used by Barton and his colleagues to do the analysis and the comment they make is that it was only possible because the horizontal stresses are about twice the vertical stress so it effectively clamps the whole structure together and you can go in and in a number of carefully planned excavation stages, remove the blocks, support them with bolts and cables and you end up with a very stable successful excavation.

Transitions to Dr. Hoek back on screen.

Evert Hoek: So, this is another extreme of shallow large caverns underground. This is the biggest underground cavern that's been constructed to date for civil engineering purposes and one thing that comes out of this is a thought about groundwater and experienced people will tell you that if you walk into a very wet rocky tunnel, you know the horizontal stresses alone because the like this one,

A slide is shown on screen with an image of water ingress in a tunnel. Miners stand in the tunnel.

Evert Hoek: water pouring out through the joints and the rock the stresses are low enough that the joints are open. If you go into a tunnel and when I've worked under in many years in Vancouver the seawall Capilano tunnel,

Transitions back to Dr. Hoek.

Evert Hoek: we have a hand of water above us of 500 meters the tunnel is bone-dry because the horizontal stresses are twice the vertical stress, so everything is clamped up. So, the movement of water and the in-situ stresses are very closely allied. So that brings me to the end of a very brief summary of a complex subject of underground cavern design, which clearly justifies more videos to deal with individual cases but hopefully this has given you a good overall picture of how we go about designing underground caverns. Thank you for your attention.

The screen fades to black. Text appears- Practical Rock Engineering Lecture Series is sponsored by Rocscience. Software tools for rock and soil. This fades to black.

Text pops up Filmed on location at University Club. Toronto, Ontario, Canada October 2015. Technical Production by Harvey Montana Productions. An image of a painting of sheep and text

This ends the video.

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

Back to top