ANNOUNCEMENTS
Rising concern over the contribution of hydroelectric reservoirs to increasing atmospheric concentration of greenhouse gases (GHGs) led to quantify net emissions from a hydroelectric reservoir, Nam Theun 2 (NT2) in subtropical region of the Lao People's Democratic Republic, Asia. We present here the first comprehensive assessment of net GHGs footprint (post-impoundment emissions minus pre-impoundment emissions) associated with the creation of a new hydroelectric reservoir. This is the result of a large scale study that has been conducted over five years (2008-2012). We have first quantified the major GHG sources and sinks of the terrestrial and aquatic components of the pre-impoundment landscape (May 2008). Then, starting from April 2009, a similar study was conducted on the reservoir and its area of influence (downstream and drawdown area). NT2 hydroelectric reservoir first reached its maximum level in October 2009 and the turbines were first operated 8 month later, in March 2010. Based on a fortnightly monitoring and five field campaigns in all seasons, emissions of major GHG (i.e. nitrous oxide (N2O), methane (CH4), and carbon dioxide (CO2)) were quantified from April 2009 to December 2011. Emissions were determined at the reservoir water surface (diffusive fluxes and ebullition), downstream of the powerhouse and the dam (degassing and diffusive fluxes), and on the soils of the drawdown area which can reach 370 km² for a 450 km² reservoir.
Methane (CH4) - Being on subtropical warm area, CH4 stands to be the most important product of anaerobic carbon mineralization at anoxic bottom of the NT2 Reservoir. CH4 emissions and their variations on time scales from daily to seasonal were assessed. For this purpose, various CH4 flux measurement techniques were investigated, including a micrometeorological station allowing for flux determination based on the eddy covariance technique (EC), floating chambers (FC), submerged funnels (SF) and thin boundary layer technique (TBL). It was shown that EC system picked-up both diffusive and bubbling fluxes at the same time. The high resolution flux sampling provided by EC allowed us to evidence the uncertainty in the magnitude of CH4 fluxes on daily and seasonal basis. A semidiurnal variation of EC fluxes was observed in four campaigns with two peaks per day - one in early morning and one in the afternoon, linked to the semi-diurnal atmospheric pressure variation. Our results suggest that the significant seasonal variation in CH4 emissions was strongly correlated with associated changes in water depth. From the bubbling measurement, we developed an artificial neuron network model which can explain up to 50% of variability of bubbling fluxes using total static pressure, variations in the water level and atmospheric pressure, and bottom temperature as inputs.
Assessing the NT2 CH4 budget, it was found that ebullition was the dominant pathway for CH4 released to the atmosphere with a contribution from 50 to 67% (2010 and 2011) of total CH4 emissions from this flat bottom young reservoir. Diffusive fluxes contributed from 15 to 22%, and degassing from 12 to 21%. An area of 2 km2 located upstream the water intake contributed to 32% of the diffusive CH4 fluxes due to the very efficient water column mixing in that zone. Because of this high mixing and subsequent outgassing before the water is actually turbined, emissions below the dam and the powerhouse contributed only a few percents of total CH4 emissions in 2011 (first full year of normal operation of the reservoir). Contrary to what was expected for an area with potentially flooded soils, the contribution of CH4 emissions from the drawdown was very low (<4%). The total CH4 emissions from the NT2 reservoirs are four times lower than emissions from same age tropical reservoirs located in South America.
The kinetics of potential CH4 production and CH4 oxidation rates were determined in laboratory controlled conditions. The area-weighted average production rates and specific CH4 oxidation rates were extrapolated at the reservoir scale. While comparing estimates of CH4 production, CH4 consumption, build-up (storage) of CH4 in the water column, we confirmed that (1) CH4 concentration in the water column (which is controlled by hydrological conditions through residence time, and dissolved oxygen level concentration) determines the quantity of CH4 being emitted by means of diffusion from the reservoir water surface and downstream, and degassing from downstream, (2) CH4 emissions from the ebullition is controlled by physical factors such as water depth, atmospheric pressure, change in the water level, change in the atmospheric pressure, and bottom reservoir water temperature.
Carbon dioxide (CO2) - We performed direct measurements of CO2 fluxes using micrometeorological eddy covariance (EC) technique and floating chambers (FC). The comparison of fluxes obtained by FC and EC techniques showed a good agreement on the reservoir. Application of the EC method revealed the importance to consider the water-air heat exchange along with thermal and CO2 concentration gradient in the water column in the process of CO2 exchange. Our results confirmed the importance of buoyancy fluxes on gas exchange. When buoyancy was negative, during low wind condition, CO2 diffusive fluxes were mainly controlled by physical processes occurring in the water column rather than by wind speed. They increased exponentially for higher speed. When buoyancy was positive, CO2 fluxes increased linearly with the wind speed. Our results show that diffusive CO2 fluxes were significantly lower during non-stratified period than during the stratified period (p < 0.0001). When the reservoir was thermally and chemically stratified, higher value of CO2 fluxes occurred at low to moderate wind speeds with surface cooling (Twater > Tair). Our results suggest that CO2 emissions are not only site-specific but also time-specific as they are governed by physical processes occurring within the water column and above the water surface at the time of measurement.
Following the work conducted for CH4, gross CO2 emissions included all major emission pathways were assessed. Diffusive CO2 fluxes at NT2 were in the higher range of diffusive CO2 fluxes reported for older tropical hydroelectric reservoirs. Our results for years 2010 and 2011 show that diffusive emissions from water surface were the main contributors (68-77%) to total CO2 emissions. Notably, diffusive CO2 fluxes from the drawdown area were in the same range as observed at the reservoir water surface, and contributed up to 25% of the annual CO2 emissions from NT2 Reservoir. Owing to physical dynamics of the reservoir and structural design, downstream (degassing and diffusion) emissions were in the lower range as reported for tropical reservoirs. There was not significant difference in total annual gross CO2 emissions between the years 2010 and 2011.
Inorganic carbon (IC) and total organic carbon (TOC) were sampled in the reservoir, the pristine rivers, and downstream of the powerhouse and the Nakai Dam. Benthic production of CO2 at the bottom of the reservoir was studied in laboratory conditions. In the pelagic water column, CO2 consumption during photosynthesis activities and CO2 production from the CH4 oxidation were also quantified. These allowed us to assess the carbon budget for the years 2010 and 2011. The annual carbon mass balance (including both CH4 and CO2) for the year 2011 indicates that the reservoir was a carbon source with an annual carbon export (atmosphere + downstream) of 401±120 and 437±108 GgC.year−1 for the years 2010 and 2011 respectively. Magnitude of carbon inputs reveals that around 85-90% of the total carbon released was fueled by flooded carbon stock at the reservoir bottom. Around 15% of the stock of carbon originally flooded has been released in the first two years after impoundment of NT2 Reservoir.
Nitrous Oxide (N2O) - Dynamics of N2O, along with inorganic nitrogen compounds i.e. ammonium (NH4 + ), nitrate (NO3 - ) and nitrite (NO2 - ) have been studied. We found that seasonal variation in N2O emissions was much stronger than the spatial one with higher N2O emissions observed during the wet season. This suggests that hydrodynamical mixing of NH4 + -rich hypolimnetic water and oxygenated epilimnetic water during the wet season leads to enhanced nitrification and subsequent high N2O emissions. Our study of the different pathways reveals that N2O was mostly emitted from the drawdown area with 53-69% of the total gross N2O emissions. The remainder (26-44%) was coming from the reservoir water surface via diffusive fluxes, with 60% of these emissions occurring during the wet season (June-September). Our results suggest that considering the drawdown area while making N2O emissions inventory is essential, whereas downstream and ebullitive emissions are nonsignificant.
Net GHG footprint - The pre-impoundment landscape was a sink of CO2, roughly neutral in terms of CH4, and a source of N2O. Post-impoundment GHG budget reveals that the same landscape had become a significant source of CO2 and CH4, but a much smaller source of N2O. For year 2010, CO2 and CH4 have contributed 60% and 35% of the total GHG budget respectively, whereas N2O accounted for less than 5%. CH4 emissions declined about 38% in the subsequent year, when CO2 emissions increased a bit in the same time, and N2O emissions remained constant. Our results indicate that upstream GHG emissions (emissions from the reservoir water surface and the drawdown area) contributed around 87% and 92% of the total GHG emissions for the years 2010 and 2011, respectively. Remainders of total GHG emissions were contributed from downstream emissions (degassing and diffusive emissions from the downstream), a percentage lower than reported for tropical hydroelectric reservoirs. The comparison of (1) total GHG emissions and (2) contribution of each pathway to the total GHG emissions from NT2 with other reservoirs evidences that the estimation of worldwide emission from hydroelectric reservoirs is still challenging.
With 2168 ± 358 and 2133 ± 276 GgCO2eq.year-1 for the years 2010 and 2011, gross NT2 emissions were about an order of magnitude higher than pre-impoundment emissions (276 ± 343 GgCO2eq.year-1 ). With net GHG emissions of 1889 ± 496 and 1854 ± 440 GgCO2eq.year-1 for the years 2010 and 2011, and given the annual power generation of 6 TWh, this leads to an GHG emission factor of 0.31 and 0.30 Mg-CO2eq.MWh-1 for the years 2010 and 2011. Thus, GHG emission factor at NT2 reservoir is lower than a typical thermal coal based power plant emission factor of 0.96 Mg-CO2eq.MWh-1 .
Shodhganga Link
