Hydroelectricity

• 1. Introduction

  ›

  Hydroelectricity supplies 15% of the world's electricity. In 2023, that amounted to almost 4,210 terawatt-hours, more than every other renewable source combined, and more than nuclear power. A single technology, harnessing the weight and motion of water, quietly anchors the modern grid. But how did humanity get here, and what stands between today's dams and the future of clean power?

  The story stretches from ancient flour mills to a cavern under the Alps to a catastrophic flood in southern China. It raises a question that planners and engineers still argue over: when does a clean source of power become too costly to the land and the people around it?

• 2. From Mill to Generator

  ›

  Hydropower has ground flour and driven machinery since ancient times, long before anyone imagined electricity. In the mid-1700s, French engineer Bernard Forest de Bélidor published Architecture Hydraulique, laying out the principles behind both vertical- and horizontal-axis hydraulic machines. A few decades later, in 1771, Richard Arkwright coupled water power with continuous production through his water frame, a pairing that helped create the factory system as we know it.

  The leap to electricity came in 1878, when William Armstrong developed the world's first hydroelectric power scheme at Cragside in Northumberland, England. It powered a single arc lamp in his art gallery. Within three years, the old Schoelkopf Power Station near Niagara Falls was generating electricity. Then, on the 30th of September 1882, the Vulcan Street Plant in Appleton, Wisconsin began operating with an output of about 12.5 kilowatts. By 1886, there were 45 hydroelectric stations across the United States and Canada. By 1889, there were 200 in the United States alone.

  The pace of construction reflected a genuine hunger for power. By 1920, hydroelectricity supplied 40% of the electricity produced in the United States, prompting Congress to enact the Federal Power Act and create the Federal Power Commission. Federal agencies followed: the Tennessee Valley Authority in 1933, the Bonneville Power Administration in 1937. The Bureau of Reclamation completed the Hoover Dam in 1928, whose initial power station rated 1,345 megawatts became the world's largest hydroelectric facility in 1936.

• 3. Scale and Giants

  ›

  Hoover Dam held the record for only six years. In 1942, the Grand Coulee Dam in Washington State eclipsed it at 6,809 megawatts. Then in 1984, the Itaipu Dam opened on the border of South America and rated 14 gigawatts. In 2008, the Three Gorges Dam in China surpassed Itaipu at 22.5 gigawatts, a figure that remains the top of the global list today.

  Currently, only seven facilities worldwide exceed 10 gigawatts. Three Gorges leads at 22,500 megawatts, followed by China's Baihetan Dam at 16,000, Itaipu at 14,000, and Xiluodu at 13,860. The Belo Monte Dam in Brazil reaches 11,233 megawatts, Venezuela's Guri Dam 10,235, and China's Wudongde Dam 10,200. The largest of these facilities can generate more than double the installed capacity of the largest nuclear power stations.

  In 2022, China added 24 gigawatts of new hydropower capacity, accounting for nearly three-quarters of the global total for that year. Europe added 2 gigawatts, its largest increase since 1990. China now accounts for 30% of global hydropower generation, followed by Brazil at 10%, Canada at 9.2%, the United States at 5.8%, and Russia at 4.6%. Paraguay is a striking outlier: it produces nearly all of its electricity from hydro and exports far more than it consumes domestically.

• 4. How the Water Works

  ›

  Most hydroelectric output comes from a simple principle: water held high has potential energy. A large pipe called a penstock delivers that water from the reservoir down to the turbine. The power extracted depends on two things: how much water flows and how far it falls, a vertical measurement engineers call the head.

  Not every hydroelectric installation uses a reservoir. Run-of-the-river stations hold little or no stored water; they generate only from the flow arriving at that moment. Any excess water simply passes through unused. Pumped-storage plants work in the opposite way to a conventional dam: at times of low electricity demand, surplus power from elsewhere on the grid pumps water up into a higher reservoir, storing it as potential energy. When demand spikes, that water flows back down through the turbine. In 2021, pumped-storage schemes provided almost 85% of the world's 190 gigawatts of grid energy storage.

  At the smallest end of the scale, pico hydro systems generate under 5 kilowatts. The Intermediate Technology Development Group Pico Hydro Project in Kenya, rated at 1.1 kilowatts, supplies 57 homes with very small electric loads. Turbines as small as 200 to 300 watts can power a few homes in a developing country with a drop of only 1 meter. These systems are typically run-of-the-river, diverting a portion of flow through a pipe and returning it to the stream.

• 5. The Hidden Costs

  ›

  Large reservoirs submerge extensive areas upstream, destroying lowland forests, marshland, and grasslands. The World Commission on Dams estimated in 2000 that dams had physically displaced 40 to 80 million people worldwide. Water exiting a turbine carries very little suspended sediment, which scours riverbeds and erodes banks downstream. Turbine passages are lethal to fish: 70% of eels passing through a turbine die immediately.

  In tropical regions, the environmental toll goes further. Where part of a rainforest is inundated, plant material decays in an anaerobic environment and releases methane, a potent greenhouse gas. According to the World Commission on Dams, where a reservoir is large relative to generating capacity and the forest was not cleared before flooding, greenhouse gas emissions can exceed those of a conventional oil-fired thermal plant. That finding complicates hydropower's clean-energy credentials in equatorial zones, even as emissions in boreal reservoirs in Canada and Northern Europe typically run only 2% to 8% of any comparable fossil-fuel plant.

  Drought is another constraint the source material underlines. One study from the Colorado River found that a modest 2-degree Celsius temperature increase producing a 10% decline in precipitation might reduce river runoff by up to 40%. Brazil, which relies heavily on hydroelectricity, could see total energy production fall by 7% annually by the end of the century under those conditions. Climate change may tighten the limits of a resource that once seemed inexhaustible.

• 6. When Dams Fail

  ›

  During Typhoon Nina in 1975, the Banqiao Dam in southern China failed after more than a year's worth of rain fell within 24 hours. The resulting flood killed 26,000 people directly and another 145,000 from subsequent epidemics. Millions were left homeless.

  Geological siting errors carry their own catastrophic potential. The 1963 disaster at the Vajont Dam in Italy killed almost 2,000 people. On the 2nd of December 1959, the Malpasset Dam at Fréjus on the French Riviera collapsed, killing 423 people in the flood that followed. Smaller facilities are not immune: the earthen embankment of the Kelly Barnes Dam failed in 1977, twenty years after its power station was decommissioned, causing 39 deaths.

  The risk scales with the volume of water held back. Poor construction, natural disasters, and sabotage are all listed as potential causes of catastrophic failure. That vulnerability is one reason the IEA, in 2021, called for robust sustainability standards for all hydropower development, with streamlined rules and regulations accompanying any future expansion.

• 7. Pairing with Wind and Nuclear

  ›

  Wind power is intermittent on a daily basis even when its seasonal patterns are predictable. It can peak at night when demand is low and fall still during the day when demand is highest. A hydroelectric reservoir capable of storing weeks of output can compensate directly: peak wind offsets minimum hydro output, and minimum wind is covered by maximum hydro release. Norway, which runs on 98% hydropower, trades electricity with flatland neighbors Sweden, Denmark, the Netherlands, Germany, and the United Kingdom, whose wind capacity complements Norway's stored water.

  Nuclear plants face a different problem. Because their costs are dominated by infrastructure rather than fuel, cutting their output raises the cost per unit of energy steeply. They are suited to steady baseload generation, not to following peaks and troughs of demand. Hydroelectricity, whose turbines can reach full load from a cold start in under 10 minutes, fills that gap at lower cost. Countries that pair the two in roughly equal shares include Switzerland, Sweden, Ukraine, and Finland.

  The IEA estimated in 2021 that the combined reservoirs of all existing conventional hydropower plants can store 1,500 terawatt-hours of electrical energy through one full cycle, about 170 times more than the global fleet of pumped-storage plants. Battery storage capacity is not expected to surpass pumped storage during the 2020s, which leaves hydropower's storage role unchallenged for the foreseeable future.

• 8. Limits of Growth

  ›

  In 2022, the IEA forecast an increase of 141 gigawatts of hydropower capacity over 2022-2027, slightly below what was achieved in the previous five-year period. In the accelerated scenario, only an additional 40 gigawatts beyond that base case was deemed possible, because environmental permitting and construction timelines are long. Some nations have already reached near-maximum development: Switzerland produces 88% of its technically exploitable potential, Mexico 80%.

  In 2021, global installed hydropower capacity reached almost 1,400 gigawatts, the highest figure among all renewable technologies. Yet the IEA that same year said major modernisation refurbishments are required across the existing fleet, and called on governments to price in the multiple public benefits that hydropower plants provide, benefits that go beyond electricity to include flood control, irrigation, water storage, and tourism. The sale of electricity from the Three Gorges Dam is estimated to cover its construction costs within 5 to 8 years of full generation, a signal of the economic case for well-sited projects, even as the IEA cautioned that large hydropower dams in most countries will cost too much and take too long to build to deliver a positive risk-adjusted return without appropriate risk management measures in place.