Transcript ppt - Kreidenweis Research Group

Mercury emissions from laboratory
combustion of wildland fuels:
speciation and emission factors
D. Obrist, H. Moosmüller, R. Schürmann; Desert
Research Institute, Reno, NV
C. Wold, E.N. Lincoln, P. Freeborn, and W.-M. Hao;
USDA Forest Service, Fire Sciences Laboratory,
Missoula, MT
S. Kreidenweis; Colorado State University, Fort
Collins, CO
Additional support:
NSF SGER # ATM-0632780
FLAME meeting
February 2006
Overview
• Atmospheric Hg deposition is the major source of Hg
contamination in remote systems
• Hg emissions from wildfires globally an important
source to atmospheric Hg load (>500 tons/year?)
• Speciation of emissions unclear
– if GEM (Hg[0]): emissions likely enter global atmospheric
Hg pool
– if RGM (Hg[2]) or Hg-P: emissions likely deposit locally
Project goal:
To measure Hg emissions from controlled
laboratory combustion of wildland fuels
To calculate Hg emission factors and Hg speciation
during biomass combustion
Importance of atmospheric Hg
• Terrestrial systems are
main source of Hg in
(most) aquatic systems
Lorey & Driscoll (EST 1999)
• Main input of Hg to
terrestrial ecosystems occurs
via atmospheric deposition
Role of plants and biomass
• Plants contain significant amounts of Hg (leaf ~24 g kg-1)
• Above-ground biomass Hg originates from the atmosphere
 Litterfall and plant senescence important pathways for
atmospheric Hg deposition
Millhollen et al (EST 2006)
Role of plants and biomass
Obrist (Biogeochem in press)
• Total atmospheric Hg Pool: ~ 2,500 to 5,000 Mg
(Mason et al. 1994; Banic et al. 2003)
Role of plants and biomass
“Plant Hg Pump” strong
enough to affect
atmospheric Hg levels?
Obrist (Biogeochem in press)
Data from Ebinghaus et al. (Atmos
Environ 2002) and Baker et al. (Atmos
Environ 2003)
Role of plants and biomass
Dreher & Follmer (Water Air Soil Poll 2004)
• 30,000 Mg Hg sequestered in US forest soils?
– 115-210 Mg/yr anthropogenic/industrial US Hg emissions
(EPA 1999, UNEP 2002)
• How stable are vegetation and soil Hg pools?
Fires and Hg emissions
• Basically all fuel Hg is released during biomass burning (>98%
Friedli et al. 2001,2003)
• Low volatilization temperature for mercury (100-300 deg C;
Biester and Schily 1997) facilitates the release
• Natural Hg emissions through wildfires are significant
– 250-430 Mg yr-1 (Sigler et al, EST 2003)
– 800 to 850 Mg yr-1 (Andreae & Merlet GBC 2001; Friedli et al. GBC 2003)
– Emissions can be much higher when organic soil pools are affected
during wildfires (e.g., 15 x higher in circumboreal systems; Turetsky et
al; GRL 2006)
• Few data on Hg speciation of wildfire emissions
– Hg-0 (elemental gaseous mercury): long atmospheric residence time
between 6 to 12 months, will undergo long-range transport processes
– RGM/Hg-P (reactive gaseous mercury/particulate-phase mercury):
shorter residence times, will deposit within 300 to 400 km distance from
emission sources
Fires and Hg emissions
Results from previous studies
– No detectable RGM emissions (Friedli et al., 2003)
– Laboratory: generally <1 % of emission as Hg-P, but up to 13%
during combustion of green, coniferous needles and deciduous
vegetation (Friedli et al., 2001, 2003) – unclear why
– Field: 13% of Hg-P in a boreal wildfire in Quebec but only 3% to
5% of Hg-P in a temperate forest fire (Friedli et al. 2003)
Project goal:
To measure Hg emissions from controlled
laboratory combustion of wildland fuels
To calculate Hg emission factors and Hg speciation
during biomass combustion
Methods
• Hg-0 measurements
– 2537A elemental mercury analyzer
(Tekran Inc., Toronto, Canada)
– ¼” Teflon tubing & filters (2 m and
0.2 m)
– Sampling flow rate: 1.5 l min-1
– Time resolution: 2.5 minutes
– Twice daily calibrations
• Hg-P measurements
– Cleaned, pre-heated (>800 deg
C) 47-mm diameter quartz fiber
filters
– Flow rate: 50 lpm
– One filter/3 replicate burns
– Six filters for dynamic/static
blanks
– Analysis: digestion in BrCl,
SnCl2 reduction, CVAFS analysis
(Forntier Geosciences, Seattle,
WA)
Results: Analysis of stack concentrations
• Example Burn #
Combustion of Biomass
37:
Ponderosa
Pine Needle
Stack
smoke concentrations
Litter
20.0
Hg-0
550
17.5
15.0
CO2
12.5
500
10.0
17.50
15.00
12.50
10.00
7.50
5.00
5.0
2.50
400
0.00
7.5
-2.50
450
Time before/after ignition
• Calculation of emissions:
– Baseline
– Integration of measured concentrations above baseline
Hg-0 (ng/m-3)
600
-5.00
CO2/CO (ppmv)
650
Results: Temporal patterns of Hg-0
Combustion of Biomass
• Example Burn # 37:
Ponderosa Pine Needle Litter
Stack smoke concentrations
GEM
8
150
CO2
6
100
4
CO x 10
50
2
0
Time after ignition
17:30
15:00
12:30
10:00
07:30
05:00
02:30
0
-3
 Hg-0 (ng/m )
10
00:00
 CO2/CO (ppmv)
200
Results: Analysis of stack concentrations
2.5
R = 0.4381
-3
C loss (g m )
y = 0.0071 weight loss
• Quality
check: total C loss versus
2
2
2.5
1.5
y = 0.0071
R2 = 0.4381
-3
1
C loss (g m )
2
0.5
1.5
1
0.5
0
0
0
50
0
100
50
100
150
150
Mass loss (g)
200
200
250
250
Mass loss (g)
25%25%50%
50% 75%
75%
100%
100%
– Good relationship between measured C loss and mass loss
– Need to confirm using C losses based on fuel/ash C contents
Results: Temporal patterns of Hg-0
Hg-0 loss over time
16
16000
14
14000
12
12000
C loss (ppm)
-3
Hg-0 loss (ng m )
• Hg-0 loss over time
10
8
6
4
10000
8000
6000
4000
2000
2
0
0
-2000
-2
0
0
20
40
Fire Time (minutes)
10
20
30
40
50
Fire Time (minutes)
– 57% of Hg-0 emissions occurred in the first 2.5 minutes (60 burns)
– No detectable emissions after 15 minutes
– But: most C also lost at the beginning
Results: Temporal patterns of Hg-0
• Hg/C loss ration over time
6.5E-03
C // Hg
lossratio
ratio
C loss
Hg
5.5E-03
4.5E-03
3.5E-03
2.5E-03
1.5E-03
5.0E-04
-5.0E-04
2.5
5
7.5
10
12.5
Time (minutes)
Needle Litter / Duff Fire
Leaves
/ Needles
15
Branches
– High Hg losses during initial combustion phases mainly due to
corresponding high C losses
2Results: Hg to C loss ratio
• Emission factors and fuel types
1
6
Needle Litter
Chamise Branch/Leaves
5
Hg-0 loss (ng m-3)
Hg-0 los
y = 1.4271x
2
R = 0.7623
0
-1
0
4
Needle Duff
3
y = 1.4271x
2
R = 0.7623
2
1
0.5
0
-1
0
0.5
y = 0.8252x
R2 = 0.5526
1
1.5
-3
2
2.5
C loss (gm )
1
1.5
C loss (g m-3)
Branches
Leaves / Needles
Outliers
(Leaves / Needles)
Branches
Leaves
/ Linear
Needles
Outliers
Linear (Branches)
Linear (Leaves / Ne
– Emission factors: seem very low compared to expected fuel Hg
concentration
– Need confirmation using fuel/ash Hg analyses
Results: Speciation of emissions
• Effects of fuel types
30
Particulate-phase Hg
(% of total Hg)
25
20
15
10
5
0
Needle Fresh
Leaf Fresh
Leaf Dry
Rice
Branch Fresh
Needle Dry
Branche Dry
-5
Fuel type
–  3.6% Hg-P of total Hgtot
– Very high (up to 23%) Hg-P contributions in fresh needles!
Results: Speciation of emissions
• Effects of flaming/smoldering combustion
Particulate-phase Hg
(% of total Hg)
25
20
15
smoldering - dominated
flame - dominated
10
5
0
-5
0.01
0.1
1
10
Flaming / Smoldering ratio
– Smoldering/flaming ration does have an effect on speciation
(P<0.01)
Results: Speciation of emissions
• Effects of fuel moisture content
Particulate-phase Hg
(% of total Hg)
25
20
15
10
y = 0.2416
R2 = 0.2857
y = -0.0025
R2 = 0.001
5
0
-5
0
20
40
60
80
100
Fuel Moisture (% of DM)
– Fuel moisture content certainly affects speciation
– Fuel moisture ‘threshold’ of significant Hg-P contributions: 40-50%
Conclusions
• Temporal loss of Hg during combustion
– Initial high release of Hg is due to initially high C losses
– Hg/C loss ratio surprisingly constant
• Measured emission factors seem very low
– Control with fuel and ash Hg contents and mass loss
• Speciation of emissions (Hg-P and Hg-0)
– Particulate Hg emissions (very!) important in fresh
needles and leaves
– Fuel moisture threshold when Hg-P emissions become
large
– Flaming/smoldering combustion also seems to control
speciation (related to fuel moisture)
Results: Hg to C loss ratio
• Fuel moisture effect
6
Needle Litter
Chamise Branch/Leaves
0.5
Hg loss (n gm-3)
5
4
Needle Duff
3
2
1
1
0
0
1.5
-3
C loss (g m )
0.5
1
1.5
2
2.5
C loss (g m-3)
25%
50%
25%
50%
70%
70%
100%
100%
Outliers
Outliers