In this paper, Chongjiang Du looks at the feasibility of incorporating a powerhouse into the hollow space of extra high hollow RCC gravity dams


When developing a hydropower project, the overall economics, including costs of the project, interest on investments and gestation period, is generally the decisive factor for comparison of alternatives. Because each dam project will differ in site-specific circumstances and has its own set of problems, the practical engineer continuously commits to provide the best engineering with a safer, cheaper, more durable and more lasting dam for the project. Therefore, various types of dams have been developed and constructed in the history of dam engineering, followed by the purposes and construction materials of dams [1]. Among these dam types, the hollow concrete gravity dams plays an important role.

Hollow concrete gravity dams were developed and constructed worldwide in the 20th century. However, this type of dam became less common as roller-compacted concrete (RCC) emerged as a new technology and material for construction of dams during the 1980s due to its low costs, fast construction and reliable performance. More attention is paid to RCC dams in planning, design and construction of a new dam project, while the option of a hollow concrete gravity dam is often put aside. Recently, major progress has been made in RCC technology and a better understanding of RCC behaviour has permitted the planning, design and construction of extra high RCC dams (dams higher than 150m). It is recognised that, for an extra high RCC dam, construction of a hollow space in the dam and incorporation of the powerhouse into this space may lead to large cost savings with limited interference with RCC construction, as opposed to building an underground powerhouse. For this reason, the feasibility of incorporating a powerhouse into extra high hollow RCC gravity dams was intensively studied for an RCC gravity dam with a height of 270m and a base width of 260m. The width of the incorporated powerhouse is 37m. The discussions in this paper have been developed partly from the feasibility study. The results of the study demonstrated that this is a feasible way of reducing the cost of new RCC gravity dams, without impairing their safety.

Powerhouse options

In planning a hydropower project with an RCC dam, it is obviously important to select a location to seat the powerhouse. The following three options are generally considered [5]:

• A surface powerhouse: usually located at the dam toe, where water is conveyed through the dam by means of penstocks;

• An underground powerhouse located in the abutments of the dam to minimize the length of the water passage; or

• An underground or surface powerhouse situated at the far downstream area to take advantage of a steep gradient of the river between the dam and the tailrace.

The choice of location is largely dependent on the project setting and site conditions, and is finally dictated by economic consideration. The layout of the powerhouse in relation to the other main features of the project is important in the overall economy of the project. A surface powerhouse at the dam toe requires enough room to place the powerhouse along the river without excessive excavation. In the case of a narrow river valley, there is often insufficient space to incorporate the powerhouse, spillways and eventual ship lifting facility in the river channel area. For this reason, designers often make use of an underground powerhouse design [5]. Although the location in this case is more flexible and the work area can be more easily separated from dam construction, underground facilities including a powerhouse cavern, a switch cavern and corresponding tunnel system need to be constructed. This can often be very expensive.

For this reason, the author proposes another option for extra high RCC dams –seat the powerhouse within the dam body. A hollow space will be constructed in the lower part of the dam interior, where the powerhouse is placed. Water will be conveyed from reservoir to the power cavern with penstocks and discharged out of the machines with the tailrace directly connected to the downstream river. With this design, the arrangement of the spillways will not be influenced by the powerhouse design.

Obviously, arrangement of the powerhouse within an RCC dam may cause interference with RCC construction. For this reason, such a version of the powerhouse would only make sense for extra large RCC dams, for which the power cavern in the dam in proportion to the large extent of the dam base becomes relatively insignificant.

It is true that the design and construction of an extra high dam is not an everyday job. Each extra high dam project will be a particular case, and specific conditions will influence the final decision on the type of dam most suitable to the site. Only a limited number of dam sites are suitable for construction of extra high RCC dams in the world. However, the large saving of project costs by this alternative may prove attractive to designers.

Hollow concrete gravity dams

A hollow concrete dam is a water barrage structure constructed of concrete with a hollow interior. The hollow dam is mostly a gravity structure, but may also be an arch-gravity or even an arch dam. It should be noted that other types of hollow dams, such as the buttress dams, the multiple arch dams and slotted gravity dams are not included in the present discussion.

Hollow gravity dams of conventional concrete were developed at the early 20th century, originally to improve stress conditions and save the concrete mass of gravity dams [4]. In comparison to solid gravity dams, hollow gravity dams rely more on their structure than their weight to resist the force of the water.

A further development of hollow dams would be to integrate the powerhouse in the hollow space. Benefits of this would include:

• Providing an option to seat the powerhouse for a narrow canyon;

• Reducing project costs by removing additional underground cavern in abutments;

• Saving concrete volume;

• Improving the stress condition in the dam concrete on the upstream side of the hollow space and the stress distribution in foundation;

• Improving heat dissipation of mass concrete; and

• Environmental protection by reducing attack to the abutments.

Until the 1980s, several hollow concrete dams were constructed, both with and without powerhouses [4]. For example, in 1957 Japan constructed the 103.6m high Ikawa hollow concrete gravity dam, with a further 13 hollow concrete dams following. In 1964, Portugal built the 87m high Bemposta arch dam with a 20m high and 14m wide cavity in the dam. In 1979, the Fengtan arch-gravity dam was completed with an integrated powerhouse in the dam –the largest concrete dam with an integrated powerhouse in dam body. Several hollow concrete dams with integrated powerhouse in the dams are listed in Table 1.

However, development and construction of hollow concrete dams lessened as RCC technology for dam construction became popular.

Large Hollow RCC Dams with Integrated Powerhouse

The use of RCC has allowed many new dams to become economically feasible. RCC is simple to produce, transport and place, and the construction process is economical and fast, which directly results in rapid rate of concrete placement, and a shorter period of construction which in turn means lower costs. Over the last three decades, RCC technology has received general recognition and acceptance in the field of dam engineering, and RCC is now successfully applied for the construction of dams throughout the world [2].

Recent advances in RCC technology and a better understanding of RCC behaviour have provided dam engineers with an opportunity to economically plan, design and construct extra high RCC dams. New techniques, materials and construction procedures for the construction of extra high RCC dams have been and are being developed. The 188m high Miel I and 216.5m high Longtan RCC dams have been completed in 2002 and 2007. In Table 2, several extra high RCC dams (over 150m high) are listed. The successful construction of the extra high RCC dams demonstrates the applicability of RCC technology and the reliability of extra high RCC dams. Based on the construction experience and available performance data from completed RCC dams, several RCC dams in a height over 200m are currently under design or are proposed for construction in the near future.

An important factor for construction of an RCC dam is that it requires a relatively large placing area with little interference so as to make full use of the advantage of rapid rates of equipment and high efficiency of labour utilization. The larger placing areas means the formworks will be reduced as well. On the other hand, the placing area should be properly restricted to ensure the RCC quality – if the placing areas are too large, the time interval between two successive layers may exceed the initial setting time of RCC mix. Experience from construction practice of high RCC dams showed that, as a rule of thumb, the appropriate placing area should be limited in the order of about 4000 to 7000m² in summer and 10000 to 15000m² in winter. Excessively large placing area may lead to high investment in equipment and difficulty in maintaining the RCC quality during placement. For this reason, the total construction area of an extra high RCC dam should be divided into several sub-placing areas (or named as Split-Level Method), as the case in construction of 216.5m high Longtan RCC dam [3].

The principal concern when it comes to integrating the powerhouse into an RCC dam is that the power cavern will take up space, while the RCC construction requires a sufficient working room to economically place and compact the RCC material. In the case of a hollow RCC dam, the base area of the dam is separated into two parts on the upstream and downstream side. However, the construction interference of a power cavern in RCC dam is a matter of the base width. When the dam is not very large, each placing area of the two parts may not be large enough to fully utilise the equipment , and this could interfere with the RCC construction. As the base width of an RCC dam increases, the relative proportion of the power cavern to the base width decreases. If the base width is large enough, the interference of construction of the power cavern with the dam RCC construction could be kept to a minimum with careful planning and appropriate management. The maximum width of a power cavern with Francis turbines is usually less than 40m, while the base width of an extra high RCC dam may exceed 140m. Therefore, there could still a placing area on each side of over 50m if arranging the power cavern in the dam body. This space is large enough to allow the RCC construction to continue with little interference. In addition, as indicated above, placement of RCC for an extra large RCC dam should be performed in several sub-placing areas, and the power cavern could provide a partition.

The main benefit to incorporating the powerhouse in the dam body is that it offers significant savings in project costs when compared with an underground powerhouse. For comparison of project costs, the quantities and costs of the following items should be separately calculated and compared:

• Civil works: including underground caverns, tunnels, surge tanks, waterways, power intakes, flood release works, power cavern in dam, penstocks, tailrace, dam foundation excavation, switch yard, access, RCC/conventional concrete for dam and powerhouses & reinforcement;

• Hydraulic steel structures: including upstream intakes, waterways, surge chamber, tailrace etc.;

• Turbines and mechanical equipment;

• Generators and electrical equipment; and

• Substations and corresponding electrical equipment

Experience from the feasibility study of an extra high RCC dam indicated that the cost savings of the incorporated powerhouse ranged from 20% to 25%.

In terms of the mass concrete used in gravity dams, the temperature rise of concrete due to the hydration heat of cementitious materials generates thermal stresses which may induce detrimental cracking. Temperature control for large RCC dams is always an important issue in construction. In principle, RCC dam should be placed with the full-scale length procedure (placing a dam block from the downstream to upstream in one section) with no longitudinal joint in the middle area of the dam blocks. Because the base width is very large, temperature control for the extra high RCC dams is very difficult. For example, in placing the RCC for Longtan Dam, pre-cooling of RCC was generally used, and in hot weather post-cooling by means of embedded cooling coins was also applied. At Miel I RCC dam, a skew longitudinal joint was arranged in the middle area of the dam and a grouting system was embedded during construction for eventual joint grouting. By placing the power cavern in the middle area of the dam, the width to be placed in the dam base is significantly reduced and can be done in two parts, which will facilitate the temperature control for RCC construction. In addition, the power cavern provides surfaces on either side for heat release of RCC mass. In this sense, inclusion of a power cavern in large RCC dams is beneficial to the temperature control.

As mentioned above, concrete in the middle area of a gravity dam is little stressed and functions more or less as filling material. Replacing this concrete with a power cavern will not considerably influence the dam stress conditions. In contrast, the concrete volume will be reduced.

To sum up, the author suggests that the incorporation of a powerhouse into extra high hollow RCC gravity dams, when the physical conditions render it possible, usually has four advantages over underground powerhouses: it is more economical; it provides another possibility to locate the powerhouse for a narrow canyon; it improves heat dissipation of mass concrete and facilitates temperature control; and it has minor impact on the landscape and environment.

It is worth noting that the powerhouse layout in hollow RCC dams is not without negative traits as far as the cost savings are concerned. Besides the concern of construction interference with RCC placement indicated above, the installation of penstocks and tailrace in the dam will cause also interference with RCC construction. However, the proportion of this interference for an extra high RCC dam is relatively small in comparison with that for a smaller RCC dam. This interference can be minimised through proper construction management. Also, the costs of the penstocks installation and higher RCC unit price are fractional in comparison with the costs of an underground powerhouse.

Construction of the sidewalls of the power cavern can be performed using formworks and grout-enriched vibrated RCC or conventional concrete. The ceiling arch may be constructed using pre-cast concrete elements. As another alternative, the mining and excavation method can be considered to construct the power cavern space. Namely, uncemented/non-cohesive materials (for example sand and gravels) are placed in the power cavern zone as RCC placement proceeds. It is only after RCC placement has progressed sufficiently above the power cavern that the fill materials can be excavated. This method provides a means to construct a power cavern with minimal interruption to RCC placement.

Stress conditions

As an example of the stress conditions around the power cavern in an extra high RCC dam, a 270m high RCC gravity dam was investigated in a feasibility study. The dam’s maximum base width is about 260m. In the river channel dam blocks, a powerhouse cavern of LxWxH=305x37x70m is arranged in the dam body to incorporate the generation units. The transformer/switchgear yard is arranged on the downstream slope of the dam, which is not considered in the stress analysis. Incorporation of such a cavity within the dam body means the stress distributions in the dam will be affected. In the concrete around the cavern, stress concentrations and tensile stresses may appear, which locally require reinforcement and higher graded concrete. For the purpose of identifying such stress zones, a two-dimensional finite element analysis is performed for the dam block. A two-dimensional view of the computational model with the element lines as used is shown in Figure 1. The maximum side length of the elements for the dam concrete is limited to 10m. In the top areas of the caverns, the mesh is finer.

The computed stresses are shown in Figure 2 under the usual loading condition, including the dead load, the usual hydrostatic pressure at upstream and downstream, the usual uplift and the silt load. From Figure 2 it can be seen that tensile stresses occur at the upstream upper corner and downstream lower corner of the powerhouse cavern within limited areas. Their values are generally less than 1.0MPa and thereby in a range that can be handled with common steel reinforcement. The principal stresses in compression are generally in vertical direction, which corresponds to the usual practice. The maximum principal stresses around the power cavern are of the order of 4.0 to 5.0MPa which lie in the range of the allowable stress of concrete. The compressive stress concentration appears at edges of the surrounding concrete of the powerhouse cavern in a value of 9.4 MPa which remains well within the allowable compressive strength of concrete. Under usual loads no tension occurs at the dam heel, and stress concentrations at the heel and at the toe of the dam are moderate. From the above calculation results, it has become apparent that the stress conditions in dam concrete are not aggravated due to the incorporated powerhouse. Moreover, because the power cavern is located in the middle area of the dam body, the stresses should not be significantly increased by seismic loading.

Chongjiang Du, lahmeyer International GmbH, Friedberger Str. 173, D-61118 Bad Vilbel, Germany


Table 1
Table 2