New study sheds light on reservoir emissions over a long time period

Sara Mercier-Blais, Research Associate at the Université du Québec à Montréal (UQAM) writes about sources of GHG emissions from reservoirs and the timeline of those emissions from 1900 to 2060, as identified by a new study. Although new GHG emissions are expected to be released with future dam construction, the study mentions that there are ways to reduce methane emissions from existing and new hydropower plants.

The construction and operation of hydropower dams requires the creation of reservoirs to control the flow of water used to operate the turbines. These reservoirs are created by flooding the land upstream of the facility. The organic matter contained in the flooded soil decomposes quickly after flooding, releasing large amounts of greenhouse gases (GHG) in the first few years after a hydropower plant is built. These emissions then decrease and slowly stabilise as the reservoir ages.

A new study that I contributed to ‘Reservoir CO2 and CH4 emissions and their climate impact over the period 1900–2060’, published recently in Nature Geosciences, traces for the first time the long-term historical and future evolution of these emissions for the period 1900-2060. To obtain a prediction of GHG emissions for individual site-specific reservoirs, a web-interface called the G-res Tool was created and launched by a multi-stakeholder group led by the Université du Québec à Montréal (UQAM) and the International Hydropower Association (IHA) in 2017. Based on the latest scientific knowledge and available free online, it has proven to be a powerful tool to better understand and evaluate emissions from reservoirs, and the models behind the Tool are the same as those used in the study.

Several studies have already established that reservoirs (hydroelectric and others) emit GHGs, notably CO2 and methane. The intensity of emissions varies greatly from one system to another. The new study is based on estimated emissions from 9,000 reservoirs on five continents, and shows that total reservoir-induced GHG emissions peaked around 1987 – in connection with a period of large dam construction – and at that time the dominant gas was CO2.

Today, methane emissions dominate, and the contribution of this gas compared with CO2 will continue to increase in the future. The reasons for this are complex but partly because in existing reservoirs CO2 emissions decrease more quickly over time. Also, some new reservoirs expected to be built in warmer climates in the future could potentially involve higher methane emissions, depending on the site.

Methane is produced by bacteria decomposing the newly flooded organic matter in oxygen-free zones, i.e. in the deep layers of the reservoirs and sediments. These emissions tend to have a lower peak just after impoundment (flooding) as compared to CO2 but remain constant for a longer period. Methane has a higher global warming potential than CO2, but it stays in the atmosphere for less time. Therefore, its effect on the climate is more intense, but over a shorter period.

Despite total observed GHG emissions decreasing over the past 30 years, the effect of these past emissions accumulating in the atmosphere continues to have an increasing impact on the climate. This is firstly due to the large amount of CO2 released from the many reservoirs built between 1950 and 1990 that has stayed in the atmosphere and is just beginning to stabilise. Secondly, the growing methane emissions from reservoirs built in the last 10-20 years and new projects will have a future impact on the climate.

In colder regions, reservoirs tend to emit less GHG compared to tropical regions, because low temperatures lead to lower bacterial activity and therefore less gas is produced. The study also shows that the current increase in reservoir construction suggests a possible increase in methane emissions, particularly from new tropical reservoirs, due mostly to degassing or bubbling emissions. The study indicates that one-third of methane emissions from reservoirs are the result of degassing. This phenomenon occurs when the water that feeds the turbines is drawn from the deep layers of the reservoir, where the methane concentration is very high. Once the water passes through the turbines and is released downstream, the methane is released into the atmosphere.

The good news is that the design and operation of dams can be changed to mitigate a part of methane emissions. For example, by drawing water closer to the surface, we can significantly or totally remove emissions from methane degassing. In one example, in Malaysia, we observed that if the intake was located a few meters higher in the reservoir, we estimate that we could decrease the amount of methane released by about 90 per cent.

Another measure to mitigate emissions could be made through better management of the water level to limit the presence of shallower areas, which contribute significantly to methane bubbling emissions.

So, in the future, in addition to selecting new projects with a low carbon footprint, actions to mitigate methane emissions at existing and new projects are crucial to ensure sustainable hydropower modernisation and development. According to the Hydropower Sustainability Standard, a hydropower plant’s emissions over a 100-year lifecycle must be <100g/kWh for it to be considered sustainable.

The new study concludes that, on a global scale, reservoirs are currently small sources of CO2 compared to other sectors such as transportation or energy production. Methane emissions from reservoirs (hydroelectric and others) accounted for 5.2 per cent of anthropogenic methane emissions in 2020.  However, a disproportionate amount of this methane (emitted mostly from degassing and bubbling) occurs from a small fraction of dams globally, providing an opportunity for effective, targeted mitigation efforts in these systems.

For further information about G-res please visit the Tool website:

Register for training on how to use the G-res Tool in an upcoming interactive training course (virtual or in-person).

*Text partly based on an article written in French by Pierre-Étienne Caza from Actualité UQAM.

Privacy Policy