Worldwide there are now an estimated 45,000 large dams - structures rising 15m or more above their foundations. Through the 1990s, investment in developing new dams totalled close to $43bn/year.
But the future for grand scale dam construction looks bleak. A recent report by the World Commission on Dams shows that as a rule dams make poor economic sense, are socially inequitable and environmentally destructive.
Dams and development - a new framework for decision making is the outcome of two years' work by the WCD, an organisation which claims political neutrality and which has called for evidence on the pros and cons of dams from every conceivable corner. WCD says its report is the summation of findings put forward by dam designers, contractors and operators, human rights organisations, environmental scientists, geological, hydrological and climatological experts, banks, donor and charity organisations, governments and green activists.
Despite this balance and objectivity, chapter by chapter Dams and development builds a catalogue of dams' shortcomings. Arguing the validity of dams will, in future, be tough.
Dam development comes under attack for sloppy accounting. The normal margins of error in cost planning, performance and investment payback calculations are not tolerated in any other branch of infrastructure procurement, the WCD states. It also means dams frequently fail to deliver real economic benefit.
'As large dams and their associated infrastructure may cost billions of dollars, financial overruns have important consequences for public and private budgeting. Projects also often derive tariff charges based on cost estimates, so under-estimates will undermine financial viability or efforts to recover costs, ' the report concludes.
Construction cost overruns for large dams average 56%. A quarter of all dams break the budget by up to 20%, 15% by up to 50%, a tenth by up to 100% and an astonishing 17% of dams exceed planned capital costs by more than 100%.
The WCD notes that project costs frequently soar as a result of unforeseen ground conditions, demanding redesign of foundations, subterranean galleries and tunnels. Dams are frequently built in remote locations adding logistical complexity.
Schedules slip - a survey of dam projects backed by the World Bank shows an average 28% delay in delivery.
But these factors are as old as dam building itself. And amazingly, 'little to no provision has been made to improve the estimates', says the report.
As a rule, dams underperform physically, the report concludes.
This is particularly true for single purpose structures - dams designed for irrigation, water supply or flood alleviation, for example.
A sample selection of large scale irrigation projects shows that the irrigated area is on average 30% less than planned when a dam is commissioned. A quarter of projects deliver water to no more than 65% of the intended distribution area. Over the first three decades of operation, the irrigated area grows towards an average peak of around 95% of that planned.
Small dams with correspondingly smaller reservoirs perform better than large dams.
Problems with irrigation projects, the WCD points out, do not always lie with dams. Distribution channels are frequently inadequate or the drainage rate of soils miscalculated. However, the reach of irrigation projects seldom exceeds expectation.
And shortfalls in distribution are compounded by inefficiencies in surface irrigation systems, ranging between only 20% to 40% in India, Morocco and Pakistan, and 25% to 60% in Taiwan, Japan and Israel. In a fifth of all cases worldwide irrigation has increased soil salinity. But in some locations, for example Turkmenistan, up to 80% of irrigated land has become saline.
Dams for bulk water supply fare no better. A quarter of dams deliver less than half of projected water supply and 70% of dams fail to meet targets over time. After 30 years of operation average performance is only 60% of that planned. Data gathered by the WCD suggests that all water provision in excess of targets 'can be ascribed to multipurpose dams. Even where it is not planned, demand for water supply from dams built for other purposes emerges over time'.
As with dams built for irrigation, the shortcomings of dams built for water supply can be ascribed to inadequate distribution - poorly developed supply networks and leakage en route to the consumer. Social inequalities also play a role, with wealthy sections of society consuming disproportionately large quantities of water.
Meanwhile, flood control dams offer only a partial solution to the problem of flooding.
Worse, 'some dams have increased the vulnerability of riverine communities to floods'.
The WCD observes that property developers move ever closer to rivers as perceived flood risk diminishes. New buildings are erected on land that is still marginal. They are liable to annual flooding, and are at extremely high risk from exceptional flooding, caused when river flows exceed the dam's storage capacity. The WCD also draws attention to the hazards posed to downstream communities by breaches of flood defence dams. 'The failure rate of large dams has been falling over the last four decades, ' the report notes. But 0.5% of structures built since 1951 have collapsed.
At the same time, loss of annual flood cycles leads to degradation of agricultural land in the affected areas, damaging the local economy.
Hydroelectric dams are the only single purpose structures that come out of the WCD's report at all well. On average they come close to achieving their productivity targets - many exceed target thanks to added generating capacity. Hydropower schemes offer economic rates of return and generally offer investors pay-back.
The downside of hydropower, though, is that it can only be considered a 'green' way of generating electricity in temperate climates. Decomposition of plant debris lying at the bottom of reservoirs in tropical and subtropical regions results in the emission of greenhouse gases such as methane. It is now apparent that reservoirs supplying hydropower plants can produce more greenhouse gas per kW/h generated than burning the equivalent of fossil fuels.
All dams suffer from the perennial problem of sedimentation - it is estimated up to 1% of the world's reservoir capacity is lost this way each year. A tenth of all reservoirs have lost a half their active capacity due to sedimentation. Evaporation is also a growing concern. In the face of annual global warming of up to 0.4degreesC, around 5% or more water contained in some reservoirs is simply disappearing as vapour.
If dams are to be built in future, they should be multifunctional, suggests the WCD.
A hydropower scheme has potential to mitigate flooding and to provide water for irrigation or municipal supply, for example.
However, there are a host of attendant issues which proponents of new dams must take into account alongside expectations of economic and physical performance. Dams have devastated downstream ecosystems, depleting fisheries, drying wetlands, and starving agricultural land of water-borne nutrients.
Dams and reservoirs have also raised water tables, resulting in waterlogging.
Creation of reservoirs has resulted in displacement of an estimated 40M to 80M people, the great majority of whom have never been properly compensated for loss of property, livelihood or heritage, nor offered a share in the economic dividends.
Unless new dam projects are conceived and developed addressing humanitarian, social and environmental impact alongside the economic up and downsides, they will not be allowed to go ahead, predicts the WCD.
The operation of existing dams must be urgently addressed to redress current failings.
Dam building in developed countries is almost at a standstill. In Europe and the US most of the best sites for dams were developed in the course of the last century. But even in regions with huge and as yet untapped hydroelectric potential - Norway, for example - has imposed a moratorium on the construction of new dams. Public opinion, led by environmental scientists and NGOs, is powerfully anti-dam and pro-ecological conservation.
And opposition to dams is not limited to the developed West.
WCD secretary general Achim Steiner predicts that over the next decade projects in China, Brazil and India - now the world's leading dam building nations, will be thwarted by domestic and international antidam pressure. The Indian government's recent decision to go ahead with construction of the Sardar Sarovar Dam on the Narmada river provoked howls of protest.
And Turkey's plans for construction of the Ilisu hydroelectric dam on the Tigris could falter on the charge that the reservoir will displace thousands of Kurds. Blocking the Tigris will also put political relations with neighbouring Syria and Iraq onto a knife edge, say critics, and it spells economic catastrophe.
In future, the WCD will not just require international funding institutions to look far harder at whether the figures stack up. Lenders and financial backers - government export credit guarantee departments, for example - will increasingly be held to account for the impact and long term performance of schemes they support.