Introduction
Methane (CH4) emission from hydroelectric reservoirs at the global scale
was recently revised downward, and it would represent only 1 % of
anthropogenic emissions (Barros et al., 2011). This latter estimate is mostly
based on CH4 diffusion at the reservoir surface and to a lesser extent
on CH4 ebullition, which are the two best documented pathways to the
atmosphere. However, emissions from the drawdown area (Chen et al., 2009,
2011) and emissions downstream of dams (Galy-Lacaux et al., 1997; Abril et
al., 2005; Guérin et al., 2006; Kemenes et al., 2007; Chanudet et al.,
2011; Teodoru et al., 2012; Maeck et al., 2013) were poorly studied and are
not taken into account in the last global estimate (Barros et al., 2011).
Some authors attempted to include these two pathways in the global estimation
of greenhouse gas emissions from reservoirs (Lima et al., 2008; Li and Zhang,
2014), and it increased drastically the emission factors of reservoirs.
The downstream emissions include the so-called degassing which occurs just
below the turbines. It is attributed to the high turbulence generated by the
discharge of the reservoir water into the river below the dam and the large
pressure drop that the water undergoes while being transported from the
bottom of the reservoir to the surface of the river below the dam. It also
includes emissions by diffusion from the river below the dam. Downstream
emissions were first reported at the Petit Saut Reservoir (Galy-Lacaux et
al., 1997), and this pathway was later confirmed in some Brazilian reservoirs
(Guérin et al., 2006; Kemenes et al., 2007). When all emission pathways
from tropical or temperate hydroelectric reservoirs (disregarding the
drawdown emissions) are taken into account, downstream emissions could
contribute 50 to 90 % of total CH4 emissions (Abril et al., 2005;
Kemenes et al., 2007; Maeck et al., 2013). At two other sites located in
Canada and in the Lao People's Democratic Republic (Lao PDR) where this
pathway was studied, downstream emissions were found to contribute less than
25 % when it exists (Chanudet et al., 2011; Teodoru et al., 2012).
According to the differences from one reservoir to the other, it appears that
the factors controlling downstream emissions from reservoirs must be
identified in order to propose realistic estimations of the global emissions
from reservoirs, including downstream emissions.
In the present study, we quantified emissions downstream of the Nam Theun 2
Reservoir (NT2R) located in Lao PDR on the basis of a high temporal (weekly
to fortnightly) and spatial (23 stations) resolution monitoring of CH4
concentration. The significance of the aerobic CH4 oxidation in the
dynamics of CH4 in the reservoir and the downstream rivers was also
evaluated. We characterized the seasonal patterns of downstream emissions
and evaluated the contribution of this pathway to CH4 emissions by
ebullition (Deshmukh et al., 2014) and diffusive fluxes at
the surface of the reservoir (Guérin et al., 2015). We
finally discuss the contribution of downstream emissions according to the
reservoir hydrodynamics and the design of the water intake by comparing our
results to previously published studies.
Material and methods
Study area
The NT2 hydroelectric Reservoir was built on the Nam Theun River located in
the subtropical region of Lao PDR. The NT2 hydroelectric scheme is based on a
trans-basin diversion that receives water from the Nam Theun River and
releases it into the Xe Bangfai River through a 27 km long artificial
downstream channel (Fig. 1; see Descloux et al., 2014, for a detailed
description of the study site). Below the powerhouse, the turbinated water
first reaches the tailrace channel (TRC1 in Fig. 1) and the water is then
stored in an 8 Mm3 regulating pond (RD in Fig. 1) located around
3.5 km below the powerhouse. The regulating pond also receives water inputs
from the Nam Kathang River (3 % of its volume annually). Daily, the water
discharge of the Nam Kathang River that reaches the regulating pond is
returned to the downstream reach of the Nam Kathang River, below the
regulating pond. The remaining water from the regulating pond is released
into the artificial downstream channel. To prevent potential problems of
deoxygenation in the water that passed through the turbines, an aeration weir
was built midway between the turbines and the confluence to the Xe Bangfai
River (AW in Fig. 1). A continuous flow of 2 m3 s-1 (and
occasionally spillway release) is discharged from the Nakai Dam (ND in
Fig. 1) to the Nam Theun River. Annually, the NT2 Reservoir receives around
7527 Mm3 of water from the Nam Theun watershed, which is
twice the volume of the reservoir (3908 Mm3), leading to a residence
time of nearly 6 months.
Map of the Nam Theun 2 Hydroelectric Reservoir (Lao People's
Democratic Republic).
Typical meteorological years are characterized by three seasons: warm wet
(WW; mid-June–mid-October), cool dry (CD; mid-October–mid-February) and
warm dry (WD; mid-February–mid-June). During the CD season, the reservoir
water column overturns, and during the WW season, sporadic destratification
occurs, allowing oxygen to diffuse down to the bottom of the water column
(Chanudet et al., 2012; Guérin et al., 2015). Daily average air
temperature varies between 12 ∘C (CD season) and 30 ∘C (WD
season). The mean annual rainfall is about 2400 mm and occurs mainly
(80 %) in the WW season (NTPC, 2005).
The filling of the reservoir began in April 2008; the full water level was
first reached in October 2009 and stayed nearly constant until the power
plant was commissioned in April 2010. After the commissioning, during the
studied period the reservoir surface varied seasonally and reached its maxima
(489 km2) and minima (between 168 and 221 km2, depending on the
year) during the WW and WD seasons, respectively.
Sampling strategy
A total of 23 stations were monitored weekly to fortnightly in order to
determine physico-chemical parameters and the CH4 concentrations and
emissions in pristine rivers, the reservoir, and all rivers and channels
located downstream of the reservoir. In the reservoir, two stations were
monitored (RES1 and RES9, Fig. 1). Station RES1 is located 100 m upstream of
the Nakai Dam and RES9 is located ∼ 1 km upstream of the water intake
which transports water to the turbines.
Below the powerhouse, the water was monitored at nine stations: in the
tailrace channel (TRC1), regulating pond (REG1), artificial downstream
channel (DCH1, DCH2, DCH3, and DCH4), and the Xe Bangfai River (XBF2, XBF3,
and XBF4). Owing to the existence of the above-listed civil structures
downstream of the powerhouse, three sections were defined in order to
calculate emissions and degassing downstream of the powerhouse, the
regulating pond, and the aeration weir (Fig. 1). The influence of the water
released from the regulating pond on the Nam Kathang River was evaluated by
the monitoring of two pristine stations (NKT1 and NKT2) upstream of the
regulating pond and three stations (NKT3–NKT5) below the regulating pond
(Fig. 1).
Below the Nakai Dam, four sampling stations (NTH3-NTH5 and NTH7) were used
for the monitoring of the Nam Theun River. Section 4 refers to the Nam Theun
River section located between stations NTH3 and NTH4 (Fig. 1).
Additionally, we monitored the pristine Xe Bangfai River (XBF1) upstream of
the confluence with the artificial channel and one of its pristine
tributaries (Nam Gnom River: NGM1) and a pristine tributary of the Nam Theun
River (Nam Phao River: NPH1) downstream of the Nakai Dam.
During various field campaigns (March 2010, June 2010, March 2011, June 2011,
and June 2013), aerobic methane oxidation rates (AMO) were determined at
three stations in the reservoir (RES1, RES3, and RES7; Fig. 1). Additionally,
AMO was also determined in the reservoir at the water intake (RES9) in June
2013.
Experimental methods
In situ water quality parameters
Oxygen and temperature were measured in situ at all sampling stations with a
multi-parameter probe Quanta® (Hydrolab,
Austin, Texas) since January 2009. In the reservoir, the vertical resolution
of the vertical profiles was 0.5 m above the oxic–anoxic limit and 1–5 m
in the hypolimnion, whereas it was only measured in surface waters (0.2 m)
in the tailrace channel, downstream channel, and rivers.
Methane concentration in water
The CH4 concentrations at all stations have been monitored between May
2009 and December 2012 on a fortnightly basis. Surface and deep-water
samples for CH4 concentration were taken with a surface water sampler
(Abril et al., 2007) and a Uwitec water sampler, respectively.
Water samples were stored in serum glass vials, capped with butyl stoppers,
sealed with aluminium crimps and poisoned (Guérin and Abril,
2007). A N2 headspace was created and the vials were vigorously shaken
to ensure an equilibration between the liquid and gas phases prior to
CH4 concentration gas chromatography (GC) analysis. The concentration
in the water was calculated using the solubility coefficient of
Yamamoto et al. (1976).
Aerobic methane oxidation
In the reservoir, water samples for AMO rate measurements were collected in
the epilimnion and in the metalimnion (at the oxicline). At RES9, the samples
were taken in the middle of the water column since the water column was well
mixed. AMO was also performed at TRC1 (immediately downstream of the
powerhouse). The water was collected in 1 L HDPE bottles, homogenized,
oxygenated and redistributed to twelve serum vials (160 mL). Each vial
contained 60 mL of water and 100 mL of air. Vials were covered with
aluminium foil to avoid the effect of light on any bacterial activity and
incubated in the dark (Dumestre et al., 1999; Murase and Sugimoto, 2005) at
20 to 30 ∘C, depending on in situ temperatures. According to in situ
concentration of CH4 in the water, different amounts of CH4 were
added by syringe while withdrawing an equal volume of air from the headspace
with a second syringe in order to obtain concentrations of dissolved CH4
in the incubated water ranging from in situ to 4 times in situ. Incubations
were performed with agitation to ensure continuous equilibrium between the
gas and water phases. Total CH4 concentrations in the vials were
measured five times in a row at a 12 h interval, and oxidation rates were
calculated as the total loss of CH4 in the vial (Guérin and Abril,
2007). The oxidation rate for each concentration was the average value of the
triplicates with standard deviation (±SD).
The kinetics parameters of aerobic methane oxidation obtained from the
experiment were combined with the in situ CH4 concentration profiles in
order to calculate the integrated aerobic methane oxidation in the oxic water
column. As the aerobic methane oxidation rates we obtained were potential,
CH4-ox were corrected for two limiting factors, the oxygen
availability and the light inhibition as described in Guerin and
Abril (2007). The final equation to compute in situ oxidation rates
(CH4-ox, mmol m-2 d-1) is
CH4-ox=CCH4⋅SCH4-ox⋅CO2/(CO2+Km(O2))⋅d⋅I(z),
with CCH4, the CH4 concentration; SCH4-ox, the specific
CH4-ox; CO2, the oxygen concentration; Km(O2), the Km of
O2 for CH4 oxidation, d, depth of the water layer and I(z), the
inhibition of methanotrophic activity by light as defined by Dumestre et al. (1999) at the Petit Saut Reservoir. Finally, the CH4 oxidation rates
were integrated in the oxic water column, from the water surface to the
limit of penetration of oxygen.
Gas chromatography
Analysis of CH4 concentrations were performed by gas chromatography (SRI
8610C gas chromatograph, Torrance, CA, USA) equipped with a flame ionization
detector. A subsample of 0.5 mL from the headspace of water sample vials was
injected. Commercial gas standards (10, 100 and 1010 ppmv, air liquid
“crystal” standards and mixture of N2 with 100 % CH4) were
injected after analysis of every 10 samples for calibration. The detection
limit is 0.3 ppmv in the headspace and the accuracy is around 4 %
allowing the determination of nanomolar concentrations in water samples,
depending of the volume of the vials and headspace. Duplicate injection of
samples showed reproducibility better than 5 %.
Calculations
Estimation of diffusive fluxes from surface concentrations
The diffusive CH4 fluxes downstream of the powerhouse (sections 1–3 in
Fig. 1) and downstream of the Nakai Dam (NTH3-7, Fig. 1) were calculated with
the thin boundary layer (TBL) equation (Liss and Slater, 1974) from the
difference between the water surface CH4 concentrations and the average
CH4 concentration in air (1.9 ppmv) obtained during eddy covariance
deployments (Deshmukh et al., 2014) combined with a gas transfer
velocity (k600) as follows:
F=kT×ΔC,
where F is the diffusive flux at the water–air interface, kT the gas
transfer velocity at a given temperature (T), and ΔC =Cw–Ca the concentration gradient between the water
(Cw) and the concentration at equilibrium with the overlying
atmosphere (Ca). Afterward, the kT was computed from k600
with the following equation:
kT=k600×(600/ScT)n,
with ScT the Schmidt number of CH4 at a given temperature (T;
Wanninkhof, 1992) and n=1/2 for turbulent water (Borges et al., 2004;
Guerin et al., 2007).
The artificial channel and the Nam Theun River downstream of the dam are
closed for navigation because of the potential high water level changes due
to reservoir management and because of the presence of zone of very high
turbulence immediately downstream of the powerhouse and downstream of the
regulation pond. In the artificial channel, water current velocity never
exceeds 1 m s-1 and averaged 0.5 m s-1. Floating chamber
measurement was not possible for the accurate determination of the k600.
On a few occasions, k600 was calculated from floating chamber
measurements (Deshmukh et al., 2014) and concomitant CH4 water surface
concentrations in the turbulent waters downstream of the powerhouse
(section 1 at stations TRC1 and REG1), in the Xe Bangfai River downstream of
its confluence with the artificial channel (XBF2), and in pristine rivers
(XBF1, Nam On River, and Nam Noy River). The gas transfer velocity reached up
to 45 cm h-1 and averaged 10.5 ± 12.1 cm h-1 (data not
shown). This is very similar to the average k600 value obtained using
the formulation k600–wind speed relationships from Guerin et al. (2007)
obtained downstream of the Petit Saut Reservoir and in small estuaries of the
same size with similar water currents like the Scheldt by Borges et
al. (2004). We therefore kept 10 cm h-1 as a conservative estimate of
the k600 in the artificial channel downstream of the NT2R. The gas
transfer velocity for the artificial channel, the Xe Bangfai River, and
downstream of the Nakai Dam (NTH3-7) was kept constant over the whole period
of monitoring since the average of the results obtained by the formulations
of Borges et al. (2004) and Guerin et al (2007) was
10.06 ± 1.48 cm h-1 according to the limited variation of the
monthly average wind speed (1.8 ± 0.46 m s-1).
Degassing
Although the so-called “degassing” usually occurs only below dams
(Galy-Lacaux et al., 1997; Abril et al., 2005; Kemenes et al., 2007; Maeck et
al., 2013), degassing occurs at four sites at NT2R: (1) the Nakai Dam,
(2) the turbine release in the tailrace channel, (3) the regulating
dam, and (4) the aeration weir using
the following equation:
Degassing=(Cupstream-Cdownstream)×water discharge,
where Cupstream is the CH4 upstream of the site where
degassing might occur and Cdownstream is the CH4 concentration
in the water downstream of the degassing site. On each of these structures,
the degassing was calculated using the water discharges and the difference of
CH4 concentration between the stations: (1) NTH3 located below the Nakai
Dam and RES1, (2) TRC1 located below the turbines and RES9, (3) DCH1 below
the regulating dam and REG1, and (4) DCH3 below the aeration weir and DCH2
(Fig. 1). In addition, degassing was calculated for the occasional spillway
releases from the Nakai Dam.
The estimation of the concentration upstream of the degassing sites was
different for the four sites. For the degassing below the turbines and below
the regulating dam, the average of the vertical profile of CH4
concentrations at RES9 and REG1 were considered as concentrations before
degassing, respectively. Surface concentration at DCH2 was considered for
the degassing at the aeration weir. For the degassing below the Nakai Dam,
since the continuous flow of 2 m3 s-1 was released from the
surface water layer, we considered the average CH4 concentration in the
upper 3 m water layer at RES1 located ∼ 100 m upstream of dam.
For the spillway release of the Nakai Dam, as the spillway gate is located
at 12 m below the maximum reservoir water level, the degassing due to
spillway release was calculated using the average CH4 concentration in
the upper 15 m water layer at RES1.
Vertical profiles of temperature, oxygen, and methane concentrations
at stations RES1 and RES9 in the Nam Theun 2 Reservoir during the three
seasons in 2010, 2011, and 2012.
Results
Temperature, O2 and CH4 concentrations in the reservoir (RES1 and
RES9)
Before the commissioning of the power plant, the vertical profiles of
temperature and oxygen and CH4 concentrations at stations RES1 located
at the Nakai Dam and RES9 located at the water intake were similar (Fig. 2).
As already shown in Chanudet et al. (2015) and Guérin et al. (2015), the
reservoir was thermally stratified with higher temperature at the surface
than at the bottom during the WD (surface: 26.8 ± 2.7 ∘C and
bottom: 18.9 ± 1.6 ∘C) and WW (surface:
28.0 ± 1.6 ∘C and bottom: 21.5 ± 1.7 ∘C)
seasons and it overturns during the CD season
(average = 23.2 ± 3.9 ∘C; Fig. 2). During the WD and WW
season, the epilimnion was always oxygenated with surface O2
concentrations ranging from 14 to 354 µmol L-1 (5 to
137 % saturation) and the hypolimnion was anoxic. In the CD season, the
reservoir water column was poorly but entirely oxygenated during a few
weeks/month (127 ± 93 µmol L-1). In the WD and WW
seasons, the CH4 concentrations ranged between 0.02 and
201.7 µmol L-1 in the epilimnion and between 0.02 and
1000 µmol L-1 in the hypolimnion. In the CD season, the
CH4 concentrations are only slightly higher in the hypolimnion than in
the epilimnion. After the starting of turbines, the hydrodynamics of the
water column at RES1 followed the same seasonal pattern as described before,
whereas the CH4 vertical profiles of concentrations at RES9 located
upstream of the water intake were homogeneous from the surface to the bottom.
At RES9 during the years 2010 to 2012, the temperature was constant from the
bottom to the surface whatever the season and the water column was always
oxygenated (O2= 166 µmol L-1; Fig. 2). During this
period, CH4 concentration peaked up to 215 µmol L-1 with
averages of 39.8 ± 48.8, 29.9 ± 55.4 and
1.9 ± 4.3 µmol L-1 during the WD, WW, and CD seasons,
respectively. For the two stations, the average CH4 concentrations over
the water column were always the highest in the WD season, intermediate in
the WW season and the lowest in the CD season. At the two stations, the
average concentrations were significantly higher in 2009 and 2010 than they
were in 2011 and 2012. The average CH4 concentrations at NT2R were in
the range reported for tropical reservoirs flooded 10–20 years ago (Abril et
al., 2005; Guérin et al., 2006; Kemenes et al., 2007).
Emissions downstream of the Nakai Dam
CH4 and O2
concentrations below the Nakai Dam
Downstream of the Nakai Dam (NTH3) after the commissioning, the average
O2 concentration was 224 µmol L-1, that is 87 %
saturation, and the concentration increased further downstream. When
excluding the periods of spillway releases, the CH4 concentration at
NTH3 ranged from 0.03 to 6 µmol L-1 (average:
0.94 ± 1.2 µmol L-1), with the highest CH4
concentrations in the WW season and the lowest in the CD season (Fig. 3a).
High CH4 concentrations (up to 69 µmol L-1) were
occasionally observed when CH4-rich water was released from the
spillway, especially in 2009. Ten kilometres downstream of the Nakai Dam, CH4 concentration decreased down
to 0.41 ± 0.32 µmol L-1 at NTH4 and NTH5 without any
clear seasonal pattern (Fig. 3a).
Methane concentrations and emissions downstream of the Nakai Dam at
the Nam Theun 2 Reservoir between 2009 and 2012. (a) Time series of
CH4 concentrations at stations NTH3 and NTH4, (b) diffusive
fluxes at stations NTH3 and NTH4, (c) emissions by diffusive fluxes
in section 4 (between NTH3 and NTH4), (d) degassing due to spillway
release below the Nakai Dam, (e) degassing below the Nakai Dam due
to the continuous water discharge of 2 m3 s-1, and
(f) total emissions by degassing and diffusion downstream of the
Nakai Dam.
The concentrations observed below the Nakai Dam at stations NTH4 and NTH5
were similar to the CH4 concentrations found in the pristine Nam Phao
River (NPH1) in the watershed and 40 % lower than the CH4
concentrations at station NTH7 located 50 km downstream of the dam. They
were 2 orders of magnitude lower than the concentrations observed downstream
of 10–20-year old reservoirs in Brazil and in French Guiana (Guérin et
al., 2006; Kemenes et al., 2007).
Diffusive fluxes below the Nakai Dam
The average diffusive flux downstream of the Nakai Dam was
3.3 ± 3.9mmol m-2 d-1 for the year 2010 and fluxes
decreased down to 1.9 ± 2.5 and
1.4 ± 0.9 mmol m-2 d-1 for the years 2011 and 2012,
respectively (Fig. 3b). Ten kilometres downstream from the Nakai Dam at NTH4
and at NTH5 downstream of the confluence of the Nam Phao River, the CH4
fluxes decreased down to 1.14 ± 0.92 mmol m-2 d-1 on
average (Fig. 3b). As for the concentrations, no seasonal or interannual
trends were found. At station NTH4 located 10 km downstream of the dam, the
CH4 emission was similar to that found in pristine rivers of the
watershed, and it was 2 orders of magnitude lower than the emissions observed
downstream of 10–20-year old reservoirs (Guérin et al., 2006; Kemenes et
al., 2007).
Considering that the CH4 emissions from the Nam Theun River below the
dam can be attributed to the reservoir over a maximum length of 10 km and a
constant width of 30 m, annual emissions below the Nakai Dam decreased from
20 to 1 Mg–CH4 month-1 between 2009 and 2012, respectively
(Fig. 3c). The very high emissions in 2009 were due to spillway releases (see
below).
Degassing below the Nakai Dam
Due to the low water discharge at the Nakai Dam (2 m3 s-1),
CH4 emissions by degassing reached a maximum of
0.1 MgCH4 d-1 at NTH3 (Fig. 3e). The occasional spillway
releases occurred mostly in 2009 before the commissioning of the power plant
and in the CD after the commissioning. They led to very intense degassing (up
to 72 Mg–CH4 d-1, August 2009, Fig. 3d). In total, 99 % of
the degassing below the Nakai Dam is due to the spillway releases in 2009,
which represent 32 % of total emissions downstream of the Nakai Dam
during the study (2009–2012). Total degassing below the Nakai Dam was very
significant in 2009 due to the spillway releases, and it dropped below
3 Mg–CH4 month-1 when only 2 m3 s-1 were released
for the years 2010 to 2012.
Emissions downstream of the powerhouse
CH4 and O2 concentrations below the powerhouse
Downstream of the turbines at station TRC1 after the commissioning, the
average O2 concentration was 174 ± 58 µmol L-1,
that is, 67 ± 20 % saturation. After the commissioning of the power
plant, the O2 saturation downstream of station DCH4 located 30 km below
the turbines was always around 100 % saturation in the artificial
downstream channel. Just below the regulating dam, in the Nam Kathang River
(NKT3), the average O2 concentration was 237 µmol L-1,
that is, 93 % saturation. There was no marked
interannual change in the O2 concentration.
Methane concentrations and emissions downstream of the powerhouse of
the Nam Theun 2 Reservoir between 2009 and 2012. (a) Time series of
CH4 concentrations at stations TRC1, DCH1, DCH3, and DCH4,
(b) diffusive fluxes at stations TRC1, DCH1, DCH3, and DCH4,
(c) emissions by diffusive fluxes in sections 1, 2, and 3 (see
Fig. 1), (d) degassing downstream of the powerhouse, the regulating
dam and the aeration weir, and (e) total emissions by degassing and
diffusion downstream of the Nakai Dam.
Surface CH4 concentration at station TRC1, which is located below the
turbines and receives water from the homogenized water column in the
reservoir (RES9), varied by 4 orders of magnitude, from
0.01 µmol L-1 (August–February, WW and CD seasons) to
221 µmol L-1 (June, end of the WD and beginning of the WW
season; Fig. 4a). The seasonal pattern of the CH4 concentrations at TRC1
mimicked the concentrations at RES9. In 2010, the surface CH4
concentration decreased from 117 ± 71 µmol L-1 at TRC1
to 1.55 ± 1.15 µmol L-1 at DCH4 in the WD season and
from 88 ± 84 to 1.26 ± 1.59 µmol L-1 in the WW
season. In 2011 and 2012, the average CH4 concentrations just below the
turbines at TRC1 were 4-fold (33.4 ± 32.0 µmol L-1) and
9-fold (9.8 ± 29.6 µmol L-1) lower than in 2010 for the
WD and WW seasons, respectively. At DCH4, the surface CH4 concentration
drops to 1.1 ± 2.4 µmol L-1 (WD) and
0.3 ± 0.5 µmol L-1 (WW) in the years 2011 and 2012,
that is, similar to what was observed in 2010. Whatever the years, in the CD
season, surface CH4 concentrations were lower than
14.5 µmol L-1 along the 30 km long watercourse
(0.02–14.5 µmol L-1).
On average, at station DCH4 (30 km below the turbines) and at station XBF4
located 90 km below the confluence of the downstream channel and the Xe
Bangfai River, the CH4 concentrations were 0.54 ± 0.95 and
0.3 ± 0.4 µmol L-1, respectively. These concentrations
are the same as those found in the pristine Xe Bangfai River
(0.78 ± 0.86 µmol L-1 at station XBF1).
At station NKT3 located in the Nam Kathang River just below the regulating
dam, the average surface CH4 concentration was
0.87 ± 0.77 µmol L-1. At station NKT5 located 15 km
downstream of the regulating dam, the average CH4 concentration was
1.34 ± 2.09 µmol L-1. These concentrations are not
statistically different from the concentrations found in the pristine Nam
Kathang Noy River (0.42 ± 0.49 µmol L-1 at the NKT1
station), the pristine Nam Kathang Gnai River
(1.01 ± 1.73 µmol L-1 at the NKT2 station) and the
pristine Nam Gnom River (1.08 ± 1.45 µmol L-1 at NGM1),
all located in the same watershed.
Diffusive fluxes below the powerhouse
In 2010, in section 1, the flux was 198 ± 230 mmol m-2 d-1,
which was 2 times higher than in section 2
(94 ± 102 mmol m-2 d-1; Fig. 4b). In section 3 (below the
aeration weir), fluxes were 15 times lower than the fluxes in section 1
(12.7 ± 18.6 mmol m-2 d-1). After the confluence with the
Xe Bangfai River, CH4 fluxes dropped down to
0.95 ± 0.76 mmol m-2 d-1 for the next 30 km. For the
years 2011 and 2012, the average diffusive fluxes below the powerhouse
decreased by a factor of 4 as compared to 2010. In 2010, most of the
diffusive fluxes occurred from the middle of the WD season until the late WW
season (155 ± 127 mmol m-2 d-1), whereas diffusive fluxes
in the CD season were 100 times lower
(1.4 ± 1.1 mmol m-2 d-1). In 2011 and 2012, most of the
emissions occurred during the WD season
(61.9 ± 50 mmol m-2 d-1), whereas emissions were 20-fold
lower during both the WW and CD seasons
(3.7 ± 3.9 mmol m-2 d-1).
As observed for the concentrations, emissions downstream of DCH4 in the
downstream channel (1.5 ± 2.7 mmol m-2 d-1) and at NKT3
downstream of the regulating dam in the Nam Kathang River
(2.03 ± 2.23 mmol m-2 d-1; Fig. 4b) were not significantly
different from those calculated for the pristine Xe Bangfai River
(2.2 ± 2.6 mmol m-2 d-1 at station XBF1), Nam Kathang Noy
River (station NKT1) and Nam Kathang Gnai River (station NKT2;
1.98 ± 4.01 mmol m-2 d-1).
The average diffusive flux for sections 1 to 3 during the monitoring was
12 ± 22 mmol m-2 d-1, which is 7 times lower than the
diffusive flux along the 40 km reach below the Petit Saut Dam
(90 mmol m-2 d-1; Guérin and Abril, 2007) 10 years after
impoundment, and 12 times lower than the diffusive flux along the 30 km
reach downstream of the Balbina Dam (140 mmol m-2 d-1; Kemenes
et al., 2007) 18 years after impoundment.
The sum of the CH4 emissions by diffusion from sections 1, 2, and 3
(Fig. 1) peaked at 333, 156, and 104 Mg–CH4 month-1 at the end
of the WD beginning of the WW season in 2010, 2011, and 2012, respectively
(Fig. 4c). Diffusion was negligible for more than half of the year. The
results clearly show that emissions decrease with time within the first 4
years after flooding.
Degassing below the powerhouse
The degassing mainly occurred within 3 to 5 months around the transition
between the WD and WW seasons (Fig. 4d). Below the powerhouse (TRC1), the
degassing reached up to 385 Mg–CH4 month-1 at the end of the WD
season and beginning of the WW season in 2010, just after the turbines were
operated (Fig. 4d). Below the regulating dam, the degassing was almost 3
times higher (1240 Mg–CH4 month-1) than below the turbines, and
the degassing from the release to the Nam Kathang River was
55 Mg–CH4 month-1 in the WD season. Even if CH4
concentrations at DCH2 were 50 % lower than at TRC1, up to
756 Mg–CH4 month-1 were still emitted at the aerating weir. This
shows the very high degassing efficiency of the aeration weir (up to
99 %), especially in the WD season (Descloux et al., 2015). Therefore,
most of the degassing emissions occurred below the regulating dam and at the
aerating weir.
In 2010, most of the degassing occurred from April to August, whereas it
occurred only from March to June in 2011 and 2012. The annual degassing
emissions almost deceased by a factor of 4 in 2011 and 2012 compared to 2010
(Fig. 4e).
Aerobic CH4 oxidation in the reservoir and downstream of the powerhouse
and the Nakai Dam
In the reservoir, the potential AMO rates increased linearly with the
CH4 concentration (Fig. 5a, b, c) in both epilimnetic and metalimnic
waters at stations RES1, RES3, and RES7. The AMO rates in the middle of the
well-mixed water column at station RES9 were not statistically different from
the AMO rates in the metalimnion at the other stations of the reservoirs.
Therefore, the AMO rates from RES9 were plotted versus the initial CH4
concentration together with AMO rates from the metalimnion. The slope of the
linear correlation, or the so-called specific oxidation rate (SOR, d-1)
in the metalimnion, was similar for the CD and WD seasons
(SOR = 0.88 ± 0.03 d-1; Fig. 5a). In the epilimnion the SOR
was twice higher in the WD season (5.28 ± 0.43 d-1) than in the
CD season (2.24 ± 0.41 d-1; Fig. 5b, c). Overall, the SOR in the
epilimnion was 2- to 4-fold higher than the SOR in the metalimnion.
Downstream of the powerhouse, the SOR was 1.47 ± 0.07 d-1, that
is, intermediately between the observation in the epilimnion and the
metalimnion (data not shown). The values of SOR observed at the NT2R are in
the same range as those reported at the Petit Saut Reservoir
(2.64–4.13 d-1; Dumestre et al., 1999; Guérin and Abril, 2007) and
boreal experimental reservoirs during the summer period (0.36–2.4 d-1;
Venkiteswaran and Schiff, 2005).
Linear relationships between methane (CH4) concentrations and
aerobic methane oxidation in the (a) metalimnion, (b) the
epilimnion in the cool dry season, and (c) the epilimnion in the
warm dry season at the Nam Theun 2 Reservoir.
The depth-integrated oxidation rates ranged from 0.16 to
931 mmol m-2 d-1 at RES9 and from 0.13 to
310 mmol m-2 d-1 at RES1 upstream of the Nakai Dam. Overall, for
the years 2010, 2011, and 2012, the average integrated oxidation rate at RES9
is 122 mmol m-2 d-1, that is, more than 3 times higher than the
average integrated oxidation rate at RES1 (35 mmol m-2 d-1).
Since oxidation occurs from the surface to the bottom of the water column at
RES9 and mostly around the oxicline at RES1, the depth-integrated oxidation
rates were 5–20 times higher at RES9 than at RES1 during the WD season, and
no clear tendency can be drawn for the WW and CD seasons (Table 1). At RES9,
the total amount of oxidized CH4 decreased from 5 to
1 Gg(CH4) yr-1 between 2010 and 2012, whereas it ranged between
0.4 and 0.7 Gg(CH4) yr-1 without a clear trend at RES1 (Table 1).
Depth-integrated methane oxidation rates (mmol m-2 d-1)
and annual amount of oxidized CH4 (Gg(CH4) yr-1) at stations
RES9 and RES1 of the Nam Theun 2 Reservoir. The depth-integrated CH4
oxidation rates are given for each season: cold dry (CD), warm dry (WD), and
warm wet (WW) for each year.
RES9
RES1
Year
Season
mmol m-2 d-1
Gg(CH4) yr-1
mmol m-2 d-1
Gg(CH4) yr-1
2010
CD
11.6 ± 5.5
2.8 ± 1.0
WD
444.1 ± 106.1
5.2 ± 1.2
18.2 ± 6.5
0.7 ± 0.2
WW
442.3 ± 93.6
96.3 ± 29.8
2011
CD
1.0 ± 0.2
7.5 ± 2.7
WD
128.2 ± 46.2
1.0 ± 0.5
5.3 ± 2.4
0.4 ± 0.2
WW
46.9 ± 31.8
50.2 ± 26.3
2012
CD
33.9 ± 9.6
34.7 ± 11.3
WD
94.1 ± 19.4
1.2 ± 0.3
41.9 ± 21.8
0.6 ± 0.2
WW
80.7 ± 24.2
26.13 ± 5.3
Discussion
Spatial and temporal variations of downstream emissions
Before the power plant was commissioned in April 2010, only a few m3 of
water was discharged at the powerhouse for testing the turbines and most of
the water was discharged at the Nakai Dam. The continuous water discharge at
the Nakai Dam was about 2 m3 s-1 and occasionally, water was
spilled in order to prevent dam overflow. The continuous discharge at the
Nakai Dam mimics the lowest annual water flow in the Nam Theun River before
it was dammed. Since it expels CH4-poor water
(0.95 µmol L-1) from the surface associated with a very low
discharge, subsequent degassing and diffusive emissions below the Nakai Dam
were lower than 4 Mg–CH4 month-1 in 2010 just after the
commissioning and lower than 1 Mg–CH4 month-1 in 2012 (Fig. 3e).
Degassing was 4-fold higher in 2010 than in 2012 because of the very high
CH4 concentrations in the water column resulting from the long residence
time of water in the reservoir before the first water releases. In 2011, the
concentrations were lower than in 2012 due to the high water discharges from
the inflows that decreased the CH4 concentrations by dilution
(Guérin et al., 2015). The spillway releases reached up to
5309 m3 s-1 and water from the top 15 m of the water column
having an average concentration around 100 µmol L-1 at RES1
were released at these occasions. During these short releases, up to
3000 Mg–CH4 month-1 were released in 2009 (Fig. 3d). After the
commissioning, the spillways were used only twice in October 2010 and
September 2011. The diffusive fluxes in the Nam Theun River below the Nakai
Dam were only highly significant during the spillway releases, when it
reached up to 20 Mg month-1 in 2009. After the commissioning, the
diffusion ranged between 0.2 and 1.5 Mg–CH4 month-1 (Fig. 3c)
and contributed only a few percent of total downstream emissions below the
Nakai Dam (Fig. 3f). Emissions below the Nakai Dam are low compared to
emissions below the powerhouse because, except during spillway releases, only
a small amount of water is discharged downstream, and this water has a low
CH4 concentration since surface water is released. However, we show here
that short spillway releases with high water discharge and moderate CH4
concentrations could contribute up to 30 % of downstream emissions in 4
years.
Downstream of the powerhouse, maximum yearly emissions were dominated by
degassing (Fig. 4e). They ranged between 1 and 3 Gg month-1 and had a
clear seasonal pattern. Emissions below the powerhouse peaked during the WD
season until the beginning of the WW season, when the CH4 concentration
in the hypolimnion of the reservoir is up to 1000 µmol L-1
(Guérin et al., 2015) and concentration at RES9 higher than
100 µmol L-1. Emissions were negligible in the late WW and
during the CD seasons when hypolimnic concentration in the reservoir and
concentration at RES9 decreased down to 5 µmol L-1
(Guérin et al., 2015). Due to the accumulation of CH4 in the
reservoir in the absence of turbining until commissioning, emissions
downstream of the powerhouse in 2010 were higher than in 2011 and 2012, and
lasted from the commissioning to the beginning of the next CD season in 2010.
After the commissioning, the high emissions downstream of the powerhouse
occurred within 3–5 months in the WD season and the very beginning of the WW
season. During the wet 2011 year, emissions became negligible after the first
rainfalls. For all years, downstream emissions were negligible in the CD
season. These results show the very high seasonal variations over 3–4 orders
of magnitude for downstream emissions as already observed in tropical
reservoirs flooding primary forest (Abril et al., 2005; Kemenes et al.,
2007). However, we show in this monomictic reservoir that downstream
emissions are negligible most of the year, and this is mostly due to the
seasonal overturn in the CD and some sporadic destratification events and
dilution of the hypolimnoion in the WW season. Overall, these results
highlight the fact that the precise determination of downstream emissions
cannot be done on the basis of discrete sampling one to four times in a year.
It requires weekly to monthly monitoring in order to (1) capture the hot
moment(s) of emissions and (2) determine their duration. For instance,
downstream emissions reported for the Nam Ngum and Nam Leuk reservoirs
located in the same region were obtained at the beginning of the WD season
when downstream emissions are moderate and during the CD and WW seasons, when
no emissions occur (Chanudet et al., 2011). Therefore, emissions were
probably underestimated since the peak of downstream emissions at the end of
the WD season–beginning of the WW season was missed.
Methane emissions from the Nam Theun 2 Reservoir between 2009 and
2012.
Gg(CH4) yr-1
2009
2010
2011
2012
Emission from reservoir
Ebullition1
11.21 ± 0.16
14.39 ± 0.11
14.68 ± 0.10
12.29 ± 0.09
Diffusion at RES9 only2
0.02 ± 0.01
2.33 ± 0.21
0.86 ± 0.12
0.66 ± 0.11
Diffusion at RES1 only2
0.06 ± 0.03
0.09 ± 0.07
0.01 ± 0.00
0.01 ± 0.00
Total diffusion2
4.45 ± 1.01
9.34 ± 2.32
3.71 ± 0.81
4.95 ± 1.09
Total emissions from reservoir
15.66 ± 1.02
23.73 ± 2.32
18.39 ± 0.82
17.25 ± 1.09
Emissions from downstream
Degassing (continuous release)
0.49 ± 0.03
8.48 ± 0.74
1.83 ± 0.41
1.67 ± 0.31
Degassing (spillway release)
7.20 ± 0.90
0.92 ± 0.39
0.14 ± 0.00
0.00 ± 0.00
Diffusion
0.10 ± 0.02
1.33 ± 0.03
0.32 ± 0.02
0.33 ± 0.03
Total downstream emissions
7.79 ± 0.90
10.73 ± 0.83
2.29 ± 0.41
2.00 ± 0.32
Total emissions (reservoir + downstream)
23.45 ± 1.36
34.46 ± 2.46
20.67 ± 0.92
19.24 ± 1.14
Downstream emissions (%)
33
31
11
10
1 Deshmukh et al. (2014); 2 Guérin et al. (2015).
Contribution of downstream emissions to CH4 gross emissions
Table 2 reports CH4 emissions by ebullition and diffusion at the surface
of the reservoir from Deshmukh et al. (2014) and Guérin et al. (2015),
respectively. These estimates take into account the seasonal variations of
the reservoir water surface and the variations of depth. Between June and
December 2009, the spillway releases contributed to 30 % of total gross
emissions from the NT2R. In 2010, downstream emissions (degassing +
diffusive fluxes) contributed to more than 30 % of total gross emissions
(disregarding drawdown emissions). In 2011 and 2012, downstream emissions
contributed to about 10 % of total gross emissions. This contribution of
downstream emissions to total emissions is low compared to tropical
reservoirs located in South America (Abril et al., 2005; Kemenes et al.,
2007). Disregarding the first 2 years of monitoring (2009 and 2010) during
which the quantification highly depends on the management of the reservoir,
the contribution of downstream emissions to total emissions is even lower
than in boreal reservoirs (Teodoru et al., 2012). The low downstream
emissions arise from the fact that the reservoir is monomictic. Each time the
reservoir overturns in the CD season, 1–3 Gg of CH4 are emitted to the
atmosphere within a few days and up to a month, which purge the reservoir
water column (Guérin et al., 2015). As a consequence, bottom
concentrations decrease from 500 to less than 5 µmol L-1
during these events, and the amount of CH4 transferred from the
reservoir to the downstream reaches decreases by 2 orders of magnitude and
stays low during 8 to 9 months, before the CH4 concentration in the
reservoir increases again. Monomictic reservoirs like Nam Theun 2, Nam Leuk,
Nam Ngum in Lao PDR (Chanudet et al., 2011), the Three Gorges Dam in China
(Li et al., 2014) and the Cointzio Reservoir in Mexico (Némery et al., 2015) are common in the subtropics
and especially in Asia, where 60 % of the worldwide hydroelectric
reservoirs are. Although CH4 emissions below amictic reservoirs like
Petit Saut and Balbina are high and very significant in the total emissions
(Abril et al., 2005; Kemenes et al., 2007), low emission downstream of
monomictic/dimictic/polymictic reservoirs is likely to be a general feature.
The thermal stratification of hydroelectric reservoirs has to be taken into
account for the estimation of global downstream emissions from hydroelectric
reservoirs. Therefore, global estimates of CH4 emissions from
hydroelectric reservoirs that include downstream emissions (Lima et al.,
2008; Li and Zhang, 2014) calculated on the basis of the results from
Amazonian reservoirs (Abril et al., 2005; Guérin et al., 2006; Kemenes et
al., 2007) must be considered with caution as also pointed out by Narvenkar
et al. (2013).
Consequence of outgassing and aerobic CH4 oxidation at the water intake
for the emissions below the powerhouse
In addition to the dynamics of the thermal stratification of the NT2R, the
design of the water intake contributes to lowering the emissions downstream
of the powerhouse. After the power plant was commissioned, the water column
at station RES9 was always completely mixed from the top to the bottom, as
revealed by the vertical profiles of temperature. Consequently, O2
penetrated down to the bottom of the water column and CH4 concentration
were higher than 100 µmol L-1 from the top to the bottom of
the water column in the WD season and at the beginning of the WW season. The
overturn of the water column at RES9 results from the artificial mixing due
to the advection of water caused by the water current generated by the water
intake localized around 11–20 m under the water surface, depending on the
water level. The water intake is responsible for the mixing of the whole
water column over an area of 3 km2 according to the hydrodynamic model
of Chanudet et al. (2012). This mixing has a strong effect on both the
outgassing (Guérin et al., 2015) and the aerobic oxidation of CH4
around the water intake and on the oxidation of CH4 below the
powerhouse.
In the area of influence of the water intake where RES9 is, large amounts of
CH4 (up to 600 mmol m-2 d-1) are emitted by diffusive
fluxes at the end of the WD season–beginning of the WW (Guérin et al.,
2015). The artificial mixing at RES9 generated a hotspot of CH4
emissions where diffusive fluxes are 15 to 150 times higher than at other
stations in the reservoir for the years 2010 to 2012 (Guérin et al.,
2015). The emissions at RES9 correspond to 20 to 40 % of the total
downstream emissions (Table 2). Therefore, a very significant amount of
CH4 that could be emitted downstream is emitted at the reservoir
surface, and this contributes to lower downstream emissions.
However, the mixing at the water intake has a strong impact on aerobic
CH4 oxidation. The vertical mixing allows O2 to penetrate down to
the bottom in the vicinity of the water intake and enhances both oxidation at
the water intake and downstream of the powerhouse. On average,
depth-integrated CH4 oxidation at RES9 upstream of the water intake is 1
order of magnitude higher than at station RES1 upstream of the Nakai Dam,
where the water column is thermally stratified. Over the 3 km2 area
representative of RES9 between 2010 and 2012, aerobic CH4 oxidation
consumed an amount of CH4 that is equivalent to 50 % of total
CH4 downstream emissions (Tables 1 and 2). In the absence of artificial
mixing, aerobic CH4 oxidation would only remove an amount of CH4
that is equivalent to the amount of CH4 removed by oxidation at RES1
that is average, that is, 11 % of total downstream emissions over the 3
years of monitoring (Tables 1 and 2). Total downstream emissions were
therefore lowered by 20 % due to the enhancement of aerobic CH4
oxidation at RES9 if we compare total downstream emissions to total
downstream emissions plus the amount of CH4 that would not be oxidized
in the absence of mixing (oxidation at RES9 minus oxidation at RES1). In
addition, aerobic methane oxidation in the downstream channel might be
enhanced too since water from RES9 being transferred to the artificial
downstream channel is better oxygenated than it would be in the absence of
artificial mixing.
Overall, the design of the water intake that mixes the whole water column
decreases virtually downstream emissions since part of the CH4 is
outgassed at the reservoir surface instead of being transported and emitted
downstream. The very positive consequence of this artificial mixing at the
water intake is that the mixing allows O2 to penetrate down to the
bottom of the water column, enhancing aerobic methane oxidation both at the
water intake and in the river/channel downstream of the powerhouse. Roughly,
CH4 emissions from the NT2 Reservoir are lowered by 40 % or more due
to the artificial mixing of the water column at the water intake.
Conclusions
This first quantification of CH4 emissions downstream of a subtropical
monomictic hydroelectric reservoir shows that emissions are negligible during
most of the year due to low CH4 concentrations in the hypolimnion. They
occurred only during 2–4 months at the end of the warm season–beginning of
the wet season and globally contribute 10 % of total emissions as
observed during normal reservoir operation years (2011 and 2012). The
monitoring of downstream emissions before and just after the commissioning
(2009 and 2010) after a period with long water residence time in the
reservoir (up to 5 years) with occasional use of spillways stresses that
reservoir management can have very significant impact on emissions by
enhancing diffusive emissions and downstream emissions resulting from the use
of spillways.
Emissions downstream of the Nam Theun 2 Reservoir have a low contribution to
total emissions also because a very significant amount of CH4 that could
be emitted downstream of the reservoir is (1) emitted upstream of the water
intake and (2) is oxidized in the vicinity of the water intake because of the
artificial mixing it generates. This artificial mixing contributes to
improving the water quality downstream of the turbines since the water that
passes through is well oxygenated (70 % saturation). The other positive
consequence is that it generates a hotspot of aerobic methane oxidation that
contributes to the oxidation of 20 % of the CH4 that would
potentially be emitted at the water intake or downstream of the turbines.
This study shows that downstream emissions from future or existing reservoirs
could be significantly mitigated by the adoption of water intake design or
the installation of devices enhancing artificial water column
destratification and oxygenation upstream of the turbines.
On the basis of these results, different from those previously published, we
recommend that estimates at the global scale of emissions below dams take
into account the mixing status of reservoirs, the water residence time and
depth of the water intake, and its impact on the oxygenation of the water
column immediately upstream of the turbines.