This document discusses the design of crane beams in underground hydroelectric caverns. It describes the evolution from early designs that used free-standing columns to later designs that anchored crane beams to cavern walls. Anchoring the beams provides advantages like earlier construction and more space in the cavern. Recent designs have even suspended beams from inclined rock surfaces or overhanging arches using anchors. With proper engineering, suspended beams can perform as well as free-standing designs while providing construction and design benefits.
This document discusses the design of crane beams in underground hydroelectric caverns. It describes the evolution from early designs that used free-standing columns to later designs that anchored crane beams to cavern walls. Anchoring the beams provides advantages like earlier construction and more space in the cavern. Recent designs have even suspended beams from inclined rock surfaces or overhanging arches using anchors. With proper engineering, suspended beams can perform as well as free-standing designs while providing construction and design benefits.
This document discusses the design of crane beams in underground hydroelectric caverns. It describes the evolution from early designs that used free-standing columns to later designs that anchored crane beams to cavern walls. Anchoring the beams provides advantages like earlier construction and more space in the cavern. Recent designs have even suspended beams from inclined rock surfaces or overhanging arches using anchors. With proper engineering, suspended beams can perform as well as free-standing designs while providing construction and design benefits.
This document discusses the design of crane beams in underground hydroelectric caverns. It describes the evolution from early designs that used free-standing columns to later designs that anchored crane beams to cavern walls. Anchoring the beams provides advantages like earlier construction and more space in the cavern. Recent designs have even suspended beams from inclined rock surfaces or overhanging arches using anchors. With proper engineering, suspended beams can perform as well as free-standing designs while providing construction and design benefits.
This document has been prepared on the basis of several consulting projects in which the issue of crane beam design was considered. It has not been published and there is no intention to publish it. However, the document is being made available for discussion purposes and your comments and contributions would be welcomed.
Evert Hoek February 2004 Hoek Anchored crane beams in hydroelectric caverns Page 2
Anchored crane beams in hydroelectric caverns
Introduction
In designing an underground powerhouse the choice of the crane and its associated supporting structure can have a major impact on the cavern sidewalls and on the construction sequence. In many cases the crane structure is designed by structural engineers who treat the underground cavern as a building and design the crane supporting structure to be independent of the surrounding rock. This approach is based upon the concept that the crane supporting structure should be free-standing so that it is not influenced by nor can it influence the surrounding building. While this approach is appropriate for surface buildings it is misguided in the case of underground caverns since it ignores the enormous load carrying capacity of the rock mass and it results in designs that are inefficient in terms of the overall design and construction of the cavern.
An alternative approach is to attach the crane beams to the cavern walls by means of a tensioned and grouted anchor system. This has the advantage that the crane beams can be constructed during the early benching operation in the cavern and they are then available for use by a small construction crane and for early assembly of the main crane. The availability of these cranes during almost the entire construction process can be of great benefit in the cavern excavation, access to the roof structure and installation of equipment in the base of the cavern. The attachment of the crane beams to the cavern walls also frees space in the cavern for accommodating other services or for reducing the cavern span.
A common fear amongst structural designers is that the rock will move during and after construction and that this will cause misalignment of the crane rails. In fact, this movement can be calculated reasonably precisely and is in the order of millimetres. Provision of simple slotted attachments between the crane beams and crane rails allows the gauge to be adjusted as required based on measurements of displacements as the cavern is excavated.
Evolution of crane beam designs
Before the 1970s most underground powerhouses incorporated free-standing columns to support overhead cranes. Typical examples of such designs are illustrated in Figures 1 and 2 which show cross sections for the Ritsom power cavern in Sweden and the Poatina cavern in Tasmania. It is interesting that, in spite of innovative treatments of the cavern shape and arch support, the designers of these caverns did not take advantage of the load carrying capacity of the sidewalls to assist in the support of the crane beams. Hoek Anchored crane beams in hydroelectric caverns Page 3
Figure 1: Ritsom underground powerhouse in Sweden with a span of 17.5 m and a column mounted crane. (Holmstrm, 1978)
Figure 2: Poatina underground powerhouse in Tasmania (Australia) with a span of 14 m and two 75 ton cranes supported on columns. (Endersbee and Hofto, 1963)
Hoek Anchored crane beams in hydroelectric caverns Page 4
An interesting development is illustrated in Figures 3 and 4 which show the initial and final designs for the arch and crane beams for the Paulo Alfonso IV underground powerhouse in Brazil. Freire and Souza (1979) point out that the initial design was based on a reinforced concrete arch which was to be 0.9 m thick at the crown and 1.8 m thick at the haunches. The overhead travelling cranes would be supported on an L shaped reinforced concrete beam, one flange being fixed into the arch and the other resting on recesses formed during excavation of the rock.
In order to reduce construction time, an alternative design was developed utilizing 9 m long rockbolts on a 1.5 m grid, tensioned to 22.5 tons, with a 10 to 15 cm thick shotcrete lining to support the cavern arch. The reinforced concrete crane beams were 3.5 m high tapering from 1.5 m wide at the top to 0.6 m at the bottom as shown in Figure 4. These beams were anchored against the inclined rock surface of the cavern walls by tendons spaced 0.9 m apart and stressed to 132 tons.
Freire and Souza (1979) list the advantages of this design, which was adopted for the cavern construction, as follows:
1. Elimination of the haunch recesses to support the reinforced concrete roof arch, thereby avoiding zones of stress concentration during various excavation phases. 2. Elimination of temporary support since the final support was applied immediately. 3. A reduction of the volume of rock excavation equal to the volume of the proposed reinforced concrete arch. 4. A reduction in the span of the cavern roof, giving improved stability. 5. A reduced demand for reinforced concrete and avoidance of problems associated with the erection of formwork and steel reinforcement. 6. A reduction of construction time. 7. A reduction in construction costs.
Note that these advantages may not apply in other projects with different rock mass conditions and construction schedules. I have also encountered situations in which the Owner will not accept the apparent risks associated with suspended crane beams, in spite of the advantages listed above.
An even bolder suspended crane beam design is illustrated in Figure 5 that shows the details for the Kvilldal underground power cavern in Norway. In this case, passive (ie untensioned) grouted anchors and rockbolts were used to hold the crane beam against the rock surface. As in the previous case, the rock surface was inclined in order to provide a reaction against downward vertical movement of the beams. Stokkeb and Tndevold (1978) point out the advantage of this design in that it permits the construction of the crane beams without the need for high scaffolding and that it provides support for a temporary crane that can be used during construction. Hoek Anchored crane beams in hydroelectric caverns Page 5
Figure 3: Initial design of the roof arch and crane beam support for the Paulo Alfonso IV underground powerhouse in Brazil. (Freire and Souza, 1979)
Figure 4: Final design for the roof arch and crane beam support for the Paulo Alfonso IV underground powerhouse in Brazil. (Freire and Souza, 1979) Hoek Anchored crane beams in hydroelectric caverns Page 6
Figure 5: Anchored crane beams in the Kvilldal power cavern, Norway (Stokkeb and Tndevold, 1978).
Figure 6: Two 403 ton cranes supported on rock ledges in the 26.5 m span La Grande 2 underground powerhouse in Quebec, Canada. (Murphy and Levay, 1985)
Hoek Anchored crane beams in hydroelectric caverns Page 7
When the rock mass is of exceptionally high quality, as in the case of the La Grande 2 underground powerhouse described by Murphy and Levay (1985), there may still be merit in using rock ledges to support the crane beams as illustrated in Figure 6. This design demands a very high level of construction control in order to avoid damage to the rock ledges but, if these can be excavated successfully, the need for anchoring to tie the beams against the wall is avoided.
In one case in which I was involved it proved impossible to persuade the Owners designers that it was safe to suspend the crane beams as illustrated in Figures 4 and 5. They insisted that a solid column of rock was required to support the beams and decided on a design rather similar to that shown in Figure 6.
In my opinion, the design of a suspended crane beam is simply a matter of mechanics. The static and dynamic forces applied to the crane rails can be estimated with a high degree of confidence and it is then only a matter of calculating the forces that have to be applied by means of the anchors to provide the required level of security. As will be shown later, some recent designs involve suspending the crane beams on overhanging roof arches where no advantage is taken of the frictional resistance of the rock/concrete interface.
One possible source of concern in these designs is related to long term corrosion of the anchors. This can be dealt with by providing a high level of corrosion protection or, in the extreme, by installing new anchors at time intervals specified by the designers.
Drakensberg Pumped Storage Project South Africa
This cavern was constructed during the late 1970s and, at that time, it was one of the largest caverns to be constructed in weak rock. The rock mass consists of interbedded sandstones, siltstones, shales and mudstones which, overall, comprise a weak rock mass. Bieniawski published his Rock Mass Rating system in 1973 and Drakensberg was the first major project to which it was applied. He predicted that the rock mass was too weak to accommodate a large cavern.
The engineers responsible for the project design developed a design that was subsequently constructed successfully and has continued to perform without any problems for the past 25 years (Bowcock et al, 1976). A minimum cavern span was critical in this design and, by using crane beams anchored to the vertical cavern walls (Figures 7 and 8), a span of 16.5 m was achieved. The height of the cavern walls was also minimised by placing each of the four turbines in its own pit with pillars between each pit providing support for the walls in the lower part of the cavern. Hoek Anchored crane beams in hydroelectric caverns Page 8
Figure 7: Drakensberg cavern showing formwork for installation of the cast-in-place crane beams.
Figure 8: Drakensberg cavern showing the cast-in-place crane beams installed and anchored and benching operations being carried out in the cavern. Hoek Anchored crane beams in hydroelectric caverns Page 9
Figure 9: Drakensberg cavern with benching almost complete. The cast-in-place crane beams can be seen on the walls. Formwork has been stripped and finishing completed in the far distance but formwork is still in place for construction of the beam over the major gallery seen in the middle of the picture. Note the shotcrete window at the junction between the flat roof and the inclined upper walls of the arch.
Figure 10: A lightweight construction gantry was provided for access to the cavern roof. This ran on temporary rails on top of the crane beams. Hoek Anchored crane beams in hydroelectric caverns Page 10
The choice of the trapezoidal cavern arch profile followed similar reasoning to that adopted for the Poatina hydroelectric project (Figure 2). In both cases the rock mass in which the caverns were excavated consists of weak horizontally bedded sedimentary rocks which, if allowed to fail, tend to form a trapezoid cavity above the opening. Advantage was taken of this process and the arch shape was chosen to be similar to the stable configuration to which failure would propagate.
In passing I would like to add that I would not recommend following the same path today. With the numerical analysis methods and the reinforcing techniques that are now available I consider that it would be entirely feasible to construct a more conventional arch shape in these weak bedded rock masses.
Returning to the question of the trapezoidal arch; in the case of the Poatina cavern it was necessary to drill stress relief holes along the intersection between the flat roof and the inclined upper wall of the arch (Endersbee and Hofto, 1963). These closely spaced holes created a plastic hinge that minimised the stresses in the remaining portions of the arch. Numerical analysis of the Drakensberg arch behaviour showed that the overstressing would not be as severe as in the case of Poatina but that there would be a concentration of deformation sufficient to damage the shotcrete at these intersection lines during benching down of the cavern. Consequently, windows were left in the shotcrete lining as shown in Figure 9. After benching down had been completed, the shotcrete windows were closed by the application of additional shotcrete. A lightweight construction gantry running on the top of the crane beams, shown in Figure 10, provided access for this final shotcrete placement and also for other finishing tasks on the arch.
Thissavros hydroelectric project, northern Greece
The Thissavros hydropower and pumped storage project on the Nestos river in northern Greece involved construction of a 172 m high rockfill dam and an underground power house with 300 MW installed capacity, excavated in an alternation of granitic gneiss and biotite para-gneiss, traversed by pegmatite and aplite veins. Foliation, joints, faults and shears created a hazard of wedge failures in the crown and the sidewalls of the caverns and this was dealt with by rockbolting and shotcrete.
The cavern design provided for cast in place crane beams which, before the main benching work started, were rockbolted to the walls and connected to the cavern roof as shown in Figure 11. The beams supported a construction gantry used for installation of rockbolts and shotcrete for roof and wall support and for concrete works. These crane beams were later incorporated into the main column crane support as shown in Figure 12 (Anastassopoulos et al, 2004). Hoek Anchored crane beams in hydroelectric caverns Page 11
Figure 11: Rockbolted crane beams in the powerhouse cavern of the Thissavros hydro- electric project in northern Greece. These crane beams were used to support a light crane which was used during construction. The stub columns under the beams were later incorporated into the concrete columns used to support the main crane.
Figure 12: Final construction of the column structure used to support the beams for the main crane in the Thissavros underground power cavern. Hoek Anchored crane beams in hydroelectric caverns Page 12
Singkarak hydroelectric project - Indosesia
The the 175 MW Singkarak Hydropower Project in West Sumatra, Indonesia, began operations in 1998.
The photograph reproduced in Figure 13 was provided by Dr Siegmund Babendererde who was a member of the Consulting Board on this project. It can be seen that the cavern has an elliptical shape and that anchored crane beams have been used. Note that these beams are anchored on overhanging portions of the elliptical arch and hence no credit can be taken for the frictional resistance of the concrete/rock interface. The rock mass in which the cavern was excavated is gneiss of reasonable quality.
Figure 13: Underground powerhouse cavern for the Singkarak hydroelectric project in Indonesia.
Hoek Anchored crane beams in hydroelectric caverns Page 13
Cirata hydroelectric project Indonesia.
The 1008 MW Cirata hydroelectric project in West Java, Indonesia, was commissioned in 1998. The powerhouse cavern is the largest in Indonesia and one of the largest in the world. It measures 253 m long, 35 m wide, 49.5 m high and is 109 m below surface. The caverns is an elliptical shape and the crane beams are anchored to the overhanging portion of the arch as illustrated in Figure 14.
Figure 14: Underground powerhouse for the Cirata underground powerhouse in Indonesia. This cavern has a span of 33 m and is in relative poor quality rock. The cavern shape is elliptical and anchored crane beams are used as illustrated. Photograph provided by Professor S. Sakurai of Kobe University in Japan.
Hoek Anchored crane beams in hydroelectric caverns Page 14
References
Anastassopoulos, K., Hoek, E., Milligan, V. and Riemer, W. 2004. Thissavros hydropower project: Managing geotechnical problems in the construction. Proc. Fifth International Conference on Case Histories in Geotechnical Engineering. New York, NY. April, 2004. Bieniawski, Z.T. 1973. Engineering classification of jointed rock masses. Trans S. Afr. Inst. Civ. Engrs 15, 335-344. Bowcock, J.B., Boyd, J.M., Hoek, E. and Sharp, J.C. 1976. Drakensberg pumped storage scheme, rock engineering aspects. In Exploration for Rock Engineering (ed. Z.T. Bieniawski) 2 , 121-139. Rotterdam: Balkema. Endersbee, L.A., and Hofto, E.O. 1968. Civil engineering design and studies in rock mechanics for Poatina underground power station, Tasmania. J. Instn. Engineers, Australia. Vol. 35. 187-209. Freire, F.C.V., and Souza, R.J.B. 1979. Lining, support and instrumentation for the cavern of the Paulo Alfonso IV power station, Brazil. Proc. Second International Symposium of Tunnelling, London: Inst. Min. Matall. 12 p. Holmstrm. A. 1978. The 320 MVA generator for Ritsom. Water Power and Dam Construction. August, 1978. 28-30. Murphy, D.K., and Levay, J. 1985. Rock engineering on the La Grande Complex. Quebec. Canadian Tunnelling. 1985, 129-141. Stokkeb, O. and Tndebvold, E. 1978. Designing for year-round efficiency at Ulla-Frre. Water Power and Dam Construction. August 1978. 23-27.