Cost savings in dams: low dams

Posted on August 16, 2013 in Low Dams

Printable version Printable version


Committee on Cost Savings in Dams

Bulletin : “Cost Savings in Dams”


3.2  Low Dams

Worldwide there are 3 000 dams higher than 60 m and 100 more are completed annually, mainly in rockfill or concrete (or RCC). Most are gated and founded on rock; their cost is usually over USD 100 million. The chapters above have analysed opportunities of cost saving for these high dams.

The data and opportunities of cost savings are very different for most dams lower than 30 m which are generally ungated earthfill dams Worldwide there are some 100 000 dams between 10 and 30 m high (less than half are classified as “large dams”) and the present annual rate of construction is in the range of 1 000.  Some are in large rivers and/or store over 100 hm3. But the great majority is built in catchment areas under 100 km2 with an average flow under 1m3/s. Their main purpose is water storage for irrigation and with storages of between 0.5 and 20 hm3.  Many are not founded on rock. Their average cost is in the range of USD 1 million, with a large part devoted to the spillway, which is usually ungated. Specific comments will be given in another bulletin for low dams in large rivers and the comments below refer mainly to low dams in catchment areas under 100 km2.

Some specific suggestions for cost savings for high or low dams may also apply to some aspects of dams between 30 and 60m high.

3.2.1  New low earthfill dams: 

Some 90% of dams lower than 30 m built on medium or small rivers are earthfill dams. Most existing ones were built in North America before 1920 or in Asia between 1960 and 1980 without heavy equipment.  Compaction was limited and floods were often underestimated. About 3% of these dams failed, 1% by internal erosion and 2% by flood overtopping. The rate of failure was well over 1 in 1 000 per year.

Virtually all dams are now built with heavy equipment, with better compaction and with better technical knowledge.  The risk of failure has been much reduced but remains significant for the lowest dams.  Internal erosion applies particularly to long dams or to defects around embedded pipes. As the consequences of failure are lower than for high dams a risk of failure by very exceptional floods is usually accepted, but climate warming and increased downstream population may justify future increased discharge capacities of their spillways.


As most future dams will be in countries with 6 or 8 months of dry season, the dam, or at least that part in the river section, may be built with a low cost for cofferdams or river diversion. But if the associated catchment area is prone to a few months a year of intense rainfall, even in small catchments, significant flows can develop and the spillway can therefore become a major part of the investment. Not only will the spillway itself be costly but, for example, if a nappe depth of 2 m is required over a free flow spillway, this may represent a 30% storage loss for a 20 m high dam. This therefore also represents a main opportunity for cost saving.

For such catchment areas, the extreme floods discharge is essentialy linked to the catchment area S and the regional climate. The maximum world registered values for such extreme floods have been:

For many future dams, especially in Asia, the Probable Maximum Flood (PMF) may be in the range of half the above figures and the flow to be discharged before failure in the range of, or higher than, 20 m3/s/km2, or 1 000 times the average flow.

Failure by overtopping happens usually for a nappe depth of 0.50 m over the crest of a dam formed in cohesive material such as clay but will occur at smaller nappe depths for non-cohesive gravel crests.

The risk of failure by overtopping is therefore low for a dam in cohesive materials in a catchment area of few km2 but for most future dams in catchment areas between 5 and 100 km2 the flow to be discharged before overtopping will be between 200 and 2 000 m3/s.

As the gap between the spillway sill and the dam crest is usually between 2 m (for a flow per m of about 5 m3/s) and 5 m (for a flow per m of 20 m3/s), a spillway length in the range of 50 to 100 m may be required with traditional designs.

It is possible to construct such a spillway more economically by “folding” the crest into the form of a classic labyrinth.  Effective multiplication factors of three are easily achievable, for the same nappe depth, when using a labyrinth shape spillway instead of a traditional Creager profile. Over 100 labyrinth spillways exist worldwide and new shapes to reduce the cost may be built on top of traditional gravity dams. Details of these are given in Appendix 2. Such labyrinths are relatively easy to construct and typically require approximately 0.2 m3 of reinforced concrete per m3/s of spillway discharge. The cost increase is much less than the cost saving of spillway length, especially in countries with low labour costs. It is also possible to use this solution to increase storage or reduce dam height.

It is also possible to increase overall spillway capacity, while at the same time increasing effective storage in the reservoir, by the use of simple concrete fuse plugs (see Appendix 3). The cost is negligible for new dams and they would typically be designed to fail at floods with probabilities of, say, between 1 in 100 or 1 in 1.000. More expensive fuse devices (fusegates) have a labyrinth shape and may be cost effective for large discharges.

It is also possible, whether or not associated with the above solutions, to effect low cost improvements to the crest of embankment dams in order to increase effective spillway capacity. Embankment dam crests can often be raised by small amounts, say between 0.5 and 1 m, by steepening the upper part of the embankment.  This facilitates higher flow depths, and hence discharges, over the spillway while maintaining the same freeboard to the dam crest.  The use of wave parapets can also be used in a similar way. It may be possible to focus such improvements on mainly the central and highest part of the embankment because a failure where the dam is 5 or 10 m high is less dangerous.

Bottom outlets

The spillways of small dams are usually ungated single small bottom outlets are used for water intake and/or for emergency reservoir draw-down. It may be cost effective for many future dams to increase the capacity of such bottom outlets for the reasons discussed below.

For many future reservoirs storing a small part of the annual flow, the siltation of the reservoir may happen in just a few dozen years.  The reservoir life may be multiplied by 2 or 3 by keeping the reservoir empty during the first part of the rainy season and by sluicing most of sediments. This is often used in China and no permanent operator is necessary.

For low dams with significant reservoir volume and ungated spillways and storing most of the annual flow, it should be possible to use the reservoir for reducing downstream flood peaks in the annual probability range between 1 in 10 and 1 in 200, provided the bottom outlet is large enough. Such an outlet could be used to increase outflow discharges at the beginning of floods and would reduce associated flood peaks by 20 or 30%. The capacity of such bottom outlets should be 10 or 20% of the total spillage capacity. Operators would be necessary for only a few days per year according to weather forecasts. Any unintended lack of operators or wrong operation would have limited impact.

Dam Body

Embankment dam bodies generally require tens of thousands or hundreds of thousands of m3 of earthfill. Cost savings may be based upon reducing the unit costs of such materials rather than the quantities.  There may be many ways in which this can be achieved, such as:

– Simplified cross sections and best choice of materials and borrow pits.

– Improved access facilities.

– Well adapted specifications and quality control.

– Optimised schedules of works. As far as possible the dam body should be built within a dry season with an even rate of earthmoving.

– Working with contractors to ascertain least cost alternatives.

General layout

The impact of the general layout on the schedule of works, and consequently on the unit costs, may be important. For example:

– For rock foundations and large spilling capacities, structures such as morning glory spillways and associated discharge culvert can be constructed and completed prior to any major earthmoving.  This will then greatly simplify subsequent earthmoving operations and minimise the cost.

– It may be possible to improve the overall layout of the works and minimise associated spillway costs by adopting a labyrinth type spillway.

– In the absence of rock foundations there is now an increasing use of concrete spillways constructed directly over embankments.  This may be especially cost effective where a dry season is sufficiently long to permit full construction.

3.2.2.  New low concrete gravity dams

– Many existing gravity dams are masonry dams.  Their rate of failure has been close to 2%, due to poor tensile strength, leakage, low density due to poor workmanship or ageing. The construction of masonry dams is labour-intensive and so it is unlikely that any masonry gravity dams will be built in the future. The comments below are thus not intended to refer to masonry dams.

– Concrete gravity dams have been much safer than masonry dams, with failure rates in the order of 0.2% based on 3 000 such dams.  Failures have not been due to poor concrete quality nor have they occurred in the main dam body.  The main reasons of accidents at low gravity concrete dams have been due to either high seepage flows under the dam in locally soft foundation areas or by sliding failure at foundation level during floods.

– Concrete gravity dams have been used extensively for spillways and water intakes but their use for forming the main dam wall has declined, compared with earthfill dams, for reasons which have been discussed elsewhere in this bulletin. However, they could be more favoured in future for several reasons:

  • The widespread prevalence of hand made, low earthfill dams in Asia has significantly reduced due to increased labour costs.
  • Re-evaluating and increasing required spilling capacities favour concrete dams.
  • The high cost of fuel will have less impact on concrete dams than on earthfill dams.
  • The design criteria and specifications for concrete dams were often overly conservative, especially in the case of low dams. Significant savings can be made if such requirements are rationalised and updated.
  • Successful gravity dams can be formed using materials such as Hardfill and CSG, rather than more expensive conventional concrete or RCC.

There are thus many opportunities of cost savings in low concrete dams.

Spillways and bottom gates

Many of the relevant comments and solutions suggested for earthfill dams may apply also to concrete gravity dams. Reducing the spillway length may not be necessary but labyrinth weirs may be useful for reducing the nappe depth, i.e. for increasing the storage or reducing the dam height and cost. They may be associated with a downstream stepped slope, thus reducing or avoiding downstream protection. Larger bottom oulets and relevant galleries may also be easily placed in concrete dams.

Reviewing designs and materials utilization

Most concrete gravity dams, including low ones, have been designed with a traditional shape, often featuring a vertical upstream face, homogenous concrete and foundation on good quality rock. Specifications for concrete quality were often costly, especially for avoiding any leakage in the dam body. The risk of sliding in case of exceptional floods was sometimes overlooked for low dams for which a few metres of increased reservoir level (and downstream level) have much more impact on safety than for high dams.  Many other solutions may be more attractive according to local foundation conditions and available materials.

The vertical upstream face may not be the best solution (see ICOLD bulletin 109, appendix 2). It may be safer and less expensive to use larger quantities of lower grade concrete to reduce overall costs. A sloping upstream face attracts vertical water load significantly enhancing sliding stability. A symmetrical profile, with both upstream and downstream faces sloping, reduces the foundation stresses thus reducing the need for very good foundations.

It may be also efficient, instead of a homogenous dam body, to use low cost concrete, based on as-dug material available directly from riverbed excavation, and to achieve upstream impermeability using a waterproof membrane or lining.

3.2.3.  Associating Roller Compacted Concrete (RCC) with earthfill

RCC has been mainly used for dams higher than 30 m where the width of the structure favours access, traffic and operation of heavy construction equipment. As the techniques and equipment required for RCC are essentially the same as the equipment required for earthfill construction it may be reasonably possible to design hybrid, low cost dam solutions making the most effective use of both materials. Two examples are represented below.  In both cases the actual construction time involved could be just a few months.

The first cross section may be especially attractive if there is a possible foundation on rock and if it is difficult to find impervious earthfill material. In this case the earthfill materials are used only as deadweight and their cost may be very low.  They may be placed upon natural ground and include materials from excavations under the RCC.  If filter zones are required either side of the RCC it should be possible to use the same materials as used to form the RCC aggregates. In the construction shown it is assumed that the RCC would be constructed and raised at the same time as the earthfill.  The unit costs for RCC are much lower than for Conventional Vibrated Concrete (CVC) while the use of an RCC core will greatly facilitate the transition to any concrete spillway or intake sections.  The solution is also environmentally favourable as the both fill may be extracted from the reservoir area and the downstream side landscaped with grass, shrubs and trees.  Little or no wave protection will be necessary upstream.

The solution shown may not be particularly attractive in seismic areas, nevertheless the concept could easily be adapted to produce seismically acceptable designs.

Solution 2 has been used for improving safety at approximately100 existing low dams in the United States. It could be used at some new dams lower than 15 m, thus avoiding separate spillways.

3.2.4.  Low arch dams

China has built 500 masonry arch dams lower than 30 m and at low cost. The shape was generally simple with vertical faces. There have been some problems of leakage and ageing due to masonry quality. In favourable narrow valleys with sound rock, such simple arch shapes using concrete could be also cost effective.

Printable version Printable version

Back to top