I took a deep breath after a brisk climb up the soft hill near the border to Sweden. The sun peaked out under the clouds and the light was beautifully reflected by the water of Nesjøen, while reindeer was munching away on delicious blueberries a little further down the ridge. The reservoir below contrasted to the barren land on the other side, making it more tranquil and interesting at the same time. No wonder that hydropower has a good reputation as an environmentally benign source of electricity, renewable and carbon-free. But is it really? The answer turns out to be: it depends.

One of the more remarkable environmental impacts of hydropower is the generation of the powerful greenhouse gas methane (CH4) from the rotting of organic material; plant matter inundated when the dam was filled as in the horrible case of the Balbina power station in the Brazilian Amazon (Nature, Wikipedia), organic matter transported by tributaries, and algae and phytoplankton growing in the reservoir itself. The methane is generated when the water is depleted of oxygen, as it often happens in lower layers of still water. The amounts of methane and CO2 generated from the degradation of plant matter can be substantial; in Balbina, the amount of methane released corresponds to 2 kg of CO2 equivalents per kWh of electricity generated – twice the global warming impact as that of a coal fired power station. In addition, 8 kg of CO2 is released. The huge area of inundated dead trees – 1770 km2 for a mere 250 MW in capacity – indicates that Balbina may be an extreme, an outlier.

How serious is the issue on a global scale? The Intergovernmental Panel on Climate Change (IPCC), in its special report on renewable energy published in 2011, had investigated the issue, but not offered a conclusion. Methane emissions were presented as they were measured, in grams of methane per square meter of reservoir surface. This provides an insufficient basis for comparison with other technologies. I have hence matched published measurements to the power generation and a number of potential explanatory variables such as reservoir area, latitude, and natural biological productivity of an area and analyzed the data. The findings of my work were recently published in Environmental Science & Technology. Here are some highlights:

  1. For the global average hydro power station, methane emissions correspond to 70 gCO2 equivalent per kWh, which is more than the 2-40 gCO2/kWh from the construction of the power station. This is a lot less than the impact of fossil fuel power stations, but a lot more than the impact of wind power.
  2. The emissions rate ranges by many orders of magnitude, in a distribution we call log-normal. Such a distribution has a peak near zero and a very long tail. Most hydropower stations are really benign, while the majority of emissions is created by a few plants.
  3. The best predictor for the emissions rate is the reservoir area per unit of electricity output. If the reservoir measures below 0.1 square meter per kWh/year of electricity generation, you are on the save side.
  4. Apart from the energy density, the age of the power station and the biological productivity of the area influence the emissions. Emissions go down with time at least for the first 10 years, as the biomass inundated by the reservoir decays. Naturally, one would expect that the amount of biomass and soil organic carbon originally present at the site are also important, but I had no data on that matter. However, the natural primary productivity of the area, the amount of biomass growth per land area, also plays a role, with dams on more productive land giving rise to higher emissions.

Today, a lot of hydropower plants in developing countries receive support through the so called Clean Development Mechanism (CDM), a mechanism created through the Kyoto protocol to allow rich countries to off-set their emissions. Wisely, the CDMs are issued only to hydropower plants with a high power density. My research indicates that the threshold set for the funding is able to avoid some but not all highly emitting projects.

The Vessingfoss power station at Nesjøen generates 44 GWh per year, the energy density is 0.75 m2 per kWh/y. I am not sure I could recommend the creation of this beautiful lake today. It is borderline.