case 2: Tunnel bore stable, tunnel heading not stable
10.11 Rock mechanics for underground hydroelectric power
10.11.4 Design and construction methods
Mining and railway engineers have done most of the pioneering work on the design and construction of underground openings and have provided a basis on which hydro-electric power designers have worked.
(a) Most of the problems the designers have to face have been mentioned in the previous paragraphs; summarizing all of this information it can be said that the following must all be closely investigated: geological and geo- technical information including rock strength, the rock intrinsic curve, rock jointing and faults, the residual stress pattern, the absolute value of these stresses and the ratio k; the shape of the excavations, the stress-strain pattern for different design alternatives and the different excavation stages; the hydro- dynamic conditions, surges and water-hammer waves in the system.
The problem for the designer is to adjust correctly all these data and conditions. A great variety of solutions has been developed. The turbines and generators can be vertical axis type, which is the usual solution. Santa Massenza (Italy) (see Jaeger, 1955) and Veytaux (Switzerland) have horizontal axis arrangements. A classical solution is to locate the transformers in a cavern upstream and parallel to the machine hall, the surge tanks being on the downstream side (Churchill Falls, Ruacana, etc.). Sometimes the trans- formers are located in an extension of the machine hall, such avoiding two parallel excavations (Waldeck II). Kaech put the transformers inside the machine hall at Inerkirchen, the valves being in a small cavern on the up- stream side of the machine hall: for safety reasons in case of a valve burst the flow would discharge direct into the tail-water tunnel (Jaeger, 1955).
Vertical walls have long been adopted as the normal solution in the design of machine halls but in some cases, spalling of the rock has occurred with such a design. An oval machine hall was adopted for Santa Giustina which was excavated in a plastic marl. Recent research using finite element methods of analysis has shown the advantage of oval excavations where there is a possibility of rock spalling (Waldeck II). In any case the rock vault above the cavern should be well arched. Flat vaults are likely to cause trouble.
(b) There are great difficulties in analysing the stresses in a concrete arch, if the arch is considered as supporting a rock load. How should the load be estimated? It may depend on the time gap between excavation and concreting.
The main difficulty concerns the possible displacements of the arch springs.
A flat arch has a tendency to push towards the rock mass, but experience shows that in most cases, during excavation of the high walls, the arch springs move inwardly. The finite element method usually assumes that the concrete vault is just a lining of the rock vault. The fissures which occurred on the Kariba North Bank concrete vault confirm such an assumption to be correct, as will be shown in section 16.2.
The success of the NATM in tunnelling and in supporting difficult galleries has induced designers to use similar methods for supporting rock vaults above the machine halls. Veytaux, Ruacana, Waldeck II are recent examples of this modern technique.
In table 10.10, a represents the spacing of the big anchor cables in the length direction of the cavern, b the spacing along the circumference of the rock vault, P the permanent load (smaller than the test load) on the cables
320 Underground excavations
Table 10.10 a
m b m
ab m2
P tonnes
P*
kg/cm2
P*tot kg/cm2 Veytaux (vault) 4-30 2-90 12-5 140 1-123
Lago Delio (walls) 2-97 3-00 8-9 80 0-897 p* average = 0-38 to 0-45 kg/cm2 Waldeck II (vault) 3-0 4-0 12-0 132 1-1 p*tot = 1-6 kg/cm2 and/?* the average loading of the rock due to the big anchors. In two cases the total loading p*ot on the rock caused by anchors and bolts is also given.
At Veytaux 1700 bolts of length 3-5 to 4-5 m, load 15 tonnes per bolt were used in addition to the cables. For Lago Delio, the figures are as follows:
444 cables, P = 80 to 100 tonnes, L = 17 to 30 m 405 cables, P = 35 to 60 tonnes, L = 16 to 20 m 309 cables, P = 19 to 22 tonnes, L = 15 m 6549 bolts, P = 5 to 11 tonnes, L = 3 to 5 m For Waldeck, II, for the vault:
890 cables, P = 132 tonnes, L = 28 to 30 m 64 cables, P = 100 tonnes, L = 15 to 20 m 62 cables, P = 40 tonnes
4000 bolts, P = 12 tonnes, L = 4-5 to 6-5 m Waldeck II surge tank:
45 cables, P = 132 tonnes, L = 18 to 21 m 151 cables, P = 100 tonnes, L = 17 to 20 m 22 cables, P = 35 tonnes, L = 12 to 17 m 1018 bolts, P = 10 tonnes, L = 6 m
Sometimes cables are used to consolidate the concrete arch springs or the concrete supports for the crane rails. A very good example is the Saussaz power station (Bozetto, 1974). Inward movements of the vertical walls have to be controlled to within reasonable limits. When cables are not sufficient, strong horizontal beams, either of reinforced concrete or steel, sometimes combined with concrete, are used. The case of Santa Giustina has been mentioned. The Saussaz was excavated in creeping fissured sandstone and schists (Bozetto, 1974; Jaeger, 1976). As can be seen from fig. 10.65e, deep cables were used for anchoring the arch springings and horizontal reinforced concrete beams immediately below the springings. Major reinforcement was at the generator floor level where a rigid reinforced concrete frame is stiffened by horizontal steel beams. There is a second stiffening frame with steel beams a few metres lower.
Flat jacks are used to transmit the frame deformations to the steel beams.
The pressure in these jacks adapts to the forces transmitted to the steel beams.
The pressures had to be relaxed twice to avoid buckling the beams. Inter- mediate rows of deep anchors are used to reduce the horizontal deformation of the rock walls. The 15 -m and 18-m-long prestressed cables were designed to rupture at 1601 and 1901 respectively.
There are other power stations where the rock walls had to be supported with cables, arches, horizontal beams or struts. At Lake Delio, Italy, very high walls had to be supported by permanent prestressed cables and by provisional arches or beams, which were removed after settling of the rock deformations.
At Kisenyama, Japan, dangerous horizontal displacements of the walls were stopped when heavy struts were built in to stabilize them. These struts considerably obstructed the concreting of the turbine foundations (Jaeger, 1976; fig. 10.65/).
Excavation methods for large machine halls were initially based on the conventional tunnelling procedures used for railway tunnels. Decisions on the method to be adopted were sometimes left to the contractor. Consulting engineers later adapted the 'Belgian' and 'Austrian' tunnelling procedures to large cavities. A very good example of such a technique is the Kariba South Bank machine hall excavation. In very good rock, the 'quarrying' method was adopted, ignoring the stress and strain distribution and the stored residual stresses around the cavity. The very large Storrnorfors tailrace tunnel (Sweden), excavated, on the whole tunnel width, in exceptionally good Swedish granite, was entirely successful. On the other hand, the large excava- tion for Kariba II machine hall in gneissic rock excavated nearly full-face encountered exceptional difficulties. The large Kemano machine hall (Canada) was excavated using a complicated system of vertical shafts and a series of horizontal galleries in order to minimize excavation and mucking costs.
Comments on Waldeck II (section 16.3) show modern thinking on how to proceed with large excavations in difficult rock.
Decisions on excavation methods should be discussed in detail between geologists, experts on rock mechanics, consultant and contractor.
Drainage of the large caverns is a subject which is often neglected. In most cases the machine hall is below the water-table. In addition, infiltration from the upstream side, sometimes equally from the downstream side, are to be expected. Cleft water pressure on the lining of steel-lined shafts is always a major problem to designers of shafts. All these problems should be analysed as a whole and adequate drainage of a large rock mass area may be the safest - and cheapest - solution.
In some cases the whole surface, rock vault and rock walls, are concrete lined, the lining being essential as rock support. In other cases only the con- creting of the vault is essential to rock stability, the walls being just covered with a thin lining. Thin brick walls have also been built, the space between
322 Underground excavations
rock and walls being used for rock drainage. In few cases the rock vault too remains bare, sometimes protected by a false thin concrete arch. In some Swedish underground power-stations rock vault and rock walls are entirely bare.
The design and construction of machine halls and transformer halls, which are similar in shape, the latter being slightly smaller, are not the only problems to be solved. Downstream surge tanks are often required, the volume of which may be considerable; their shape and height often cause stress concentrations or negative stresses. Hydraulic problems have to be considered when balancing the advantages and disadvantages of long or short pressure conduits versus shorter or longer tailrace tunnels, free flow or under pressure, and the location of the machine hall along the hydraulic line of conduits. The size of some tailrace tunnels, several kilometres long, may be very considerable (Storrnorfors).
For moderate inside water pressures concrete lining of pressure shafts is often acceptable. Steel lining is the most usual solution. Steel linings should be designed to withstand buckling by outside cleft water pressure (Amstutz, 1950, 1953; Jaeger, 1955a). 'Prepakt' concrete has been successfully used between rock and steel lining for the Kemano pressure shafts. Kaech, at the Maggia power-station used wet, sandy concrete of reduced crushing strength but greater imperviousness to consolidate the steel linings of the pressure shafts and avoid cleft water pressures. Drainage of the rock about the shafts is important too.
Special techniques and equipment had to be developed for the excavation and lining of vertical and inclined shafts. The choice of the angle most favourable for inclined shafts is sometimes chosen to help move the spoil by gravity.
Detailed discussions of the design and construction of three very large underground power stations are to be found in chapter 16.