Hydropower — water, dams and ecosystems

Civil works, sediments, rivers, flexible electricity and decommissioning.

Before electricity, the river

Two ways to harness water, two scales of transformation

A hydroelectric plant converts into electricity either the potential energy of a water-level difference, or the kinetic energy of water already in motion. Large reservoir dams hold back an enormous volume of water behind a wall, creating a height difference that can be controlled at will, allowing flexible generation by adjusting turbined flow to match grid demand at any moment. Run-of-river plants, by contrast, harness the river's natural flow with almost no storage, altering downstream flow and level much less, but also with far less ability to adjust generation at will — they depend on whatever the river carries at each moment.

Building a large dam: one of the biggest civil-engineering transformations there is

Building a large dam requires temporarily diverting the river's course during construction, excavating foundations able to withstand the pressure of millions of tonnes of water, and pouring massive amounts of reinforced concrete — among the largest concrete structures ever built by humans in many cases —, plus manufacturing flow-control gates, penstocks, turbines, generators and the high-voltage lines connecting the plant to the grid. All that infrastructure is built to last generations: many hydroelectric dams keep operating well beyond their original century of design life, something exceptional compared with almost any other generation technology.

No fuel, but a deep territorial cost

During operation, a hydroelectric plant burns nothing and can respond to demand changes within minutes — an operational advantage that makes it very valuable for stabilizing a grid with a lot of variable solar and wind. But the territorial cost of creating the reservoir can be enormous and, in practice, irreversible in the short term: permanent flooding of entire valleys, forced displacement of people who lived on that land, interruption of fish migration routes, and disruption of the natural sediment transport that used to fertilize floodplains downstream.

The surprising climate problem of some tropical dams

In warm climates, the vegetation and organic matter flooded under a new reservoir decompose, under low-oxygen conditions, releasing methane — a greenhouse gas far more potent than CO₂. The magnitude of this effect depends heavily on the amount of flooded biomass and the local climate: it is practically negligible in many mountain reservoirs in temperate climates, but in certain tropical lowland reservoirs with dense vegetation it can end up producing, per unit of energy generated, more greenhouse gas emissions than an equivalent coal plant — a specific and very well documented case is developed in this leaf's data tab.

What the river stops carrying downstream

A dam also retains the sediment the river would naturally carry toward its mouth. That retention slowly reduces the reservoir's own useful capacity — it fills with sediment over the years — and can impoverish the soils and deltas located downstream, which depended on that periodic supply of nutrients and material. Drought and climate change add further uncertainty to any hydroelectric plant's expected output, because its generation depends directly on how much water the watershed receives each year.

The end of useful life: renew, decommission or rethink

Turbines, generators and gates can operate for decades with proper maintenance, but at end of life a decision must be made between modernizing an existing dam's equipment, decommissioning it entirely — a complex and rare engineering operation that also releases at once all the sediment accumulated over its whole operating life — or keeping it running beyond its original design with whatever limitations that implies. A hydroelectric dam is not a facility that simply "switches off": it has already permanently reorganized an entire watershed, and any decision about its future inherits that transformation.

The prudent criterion for judging any hydroelectric project

Comparing hydroelectric plants using only megawatts per hectare of reservoir is misleading. A more complete criterion includes the actual annual energy generated, shared uses of the impounded water — irrigation, urban supply, flood control —, impact on river ecosystems and migratory species, the dam's long-term structural safety, sediment management, and available alternatives. Modernizing already-existing plants' equipment usually adds generation with far less territorial transformation than building a new dam over a still-intact valley.

The oldest plant is also the most territorial

In favor: burns no fuel, can start up and adjust output within minutes — very useful for stabilizing a grid with a lot of variable solar and wind —, and some reservoirs offer one of the highest energy-returned-on-energy-invested ratios of any technology.

To watch, in a way that is hard to reverse: a large dam transforms an entire river — floods valleys, can displace population, interrupts fish and sediment passage, and changes downstream temperature and flow. Unlike a solar or wind plant, almost none of that territorial impact is reversible in the short term if the facility is decommissioned.

The risk the global statistic hides: dam safety is, overall, very high, but a catastrophic failure has extreme consequences — see the data tab for the case that dominates the entire global hydroelectric mortality statistic.

In warm climates, a climate surprise: organic matter left flooded under a tropical reservoir can decompose releasing methane, a greenhouse gas far more potent than CO₂; in some documented cases, this makes certain tropical dams emit more greenhouse gas per unit of energy than an equivalent coal plant.

Before building a new dam: always compare with modernizing existing plants (repowering), which can add generation with far less territorial transformation than flooding a still-intact valley.

The numbers, with their source

A single dam failure dominates the entire global statistic

Hydropower's historical mortality per unit of energy is ~1.3 deaths per TWh counting the whole history since 1965 — a figure dominated almost entirely by a single event: the failure of the Banqiao dam (China, 1975), which caused approximately 171,000 deaths. If that exceptional event is excluded, hydropower's mortality falls to ~0.04 deaths per TWh, comparable to nuclear, solar and wind. The lesson is not that hydropower is dangerous in general, but that a catastrophic dam failure is one of the very few energy accidents capable of killing hundreds of thousands of people at once — something no other electricity source has caused in modern history.

Source: Our World in Data — What are the safest and cleanest sources of energy?.

When a tropical reservoir emits like a coal plant: Balbina, Brazil

The Balbina reservoir, formed in 1987 in the heart of the central Amazon, has been studied in depth as a reference case for methane emissions from tropical dams. Just the methane emissions measured downstream of the dam — methane released after passing through the turbines — are equivalent to close to 3% of all the methane released by the entire central Amazon floodplain. Counting the reservoir's whole set of greenhouse gases, the annual total is roughly equivalent to half the fossil CO₂ emissions of São Paulo, a metropolis of millions. Research on lowland Amazonian dams concludes that some can emit more greenhouse gas per unit of energy than an equivalent coal plant, precisely because of the decomposition of flooded vegetation in a warm climate.

Source: Kemenes et al. — Methane release below a tropical hydroelectric dam, Geophysical Research Letters.

The comparative table

What this table is missing

Methane emission figures vary enormously between a temperate mountain reservoir (low emissions) and a tropical lowland reservoir with dense flooded vegetation (potentially very high emissions): there is no single "hydropower emissions" figure that applies equally to every reservoir in the world.

Sources

• The International Hydropower Association maintains the Hydropower Sustainability Standard.
• The IPCC assesses hydropower alongside other renewables in Renewable Energy Sources and Climate Change Mitigation.
• On historical mortality dominated by a single dam failure: Our World in Data — Safest and cleanest sources of energy.
• On methane in tropical reservoirs, the reference study on Balbina (Brazil): Geophysical Research Letters — Methane release below a tropical hydroelectric dam.