UNCERTAINTIES IN BLACK CARBON EMISSIONS AND MODEL PREDICTIONS:

A SOUTH ASIAN PERSPECTIVE

Chandra VenkataramanDepartment of Chemical Engineering

Indian Institute of Technology, Bombay, India

Collaborators: M. Shekar Reddy, Gazala Habib, Manish Shrivastava, Shubha Verma-IIT Bombay

Olivier Boucher, Jean-François Léon, Bertrand Crouzille - LOA, FranceToni Miguel, Arantza Fernandez, Sheldon Friedlander - UCLA

Presented at “Black Carbon Emissions and Climate Change: A Technical Workshop,” San Diego, USA, October 13-15, 2004

OUTLINEOUTLINE

Emissions estimation: industrial/transport, residential, open burning• Level of sectoral detail • Fuel composition and technology information• Developing / measuring emission factors• Assumptions in spatial distributions• Seasonal and interannual variability• Estimated uncertainties and strategies for their reduction

BC transport and radiative forcing using a general circulation model• Evaluation of model predictions• Sensitivity study• Optical depth and radiative forcing

EMISSIONS ESTIMATION APPROACHEMISSIONS ESTIMATION APPROACH

-emission rates (kg y-1) or densities (kg km-2 y-1) from fuel consumptions and specific emissions (or emission factors).

-at the global, national, regional, urban level (0.5-1 km for urban; 25-100 km for regional and global).

-Includes, in principle, all sources that emit into that atmosphere.

Emission Rate = Activity Data x Pollutant Emission Factor

Activity Data: Emission Factor:

- fuel used (Gg y-1) kg pollutant (Gg fuel)-1

- industrial production (Gg y-1) kg pollutant (Gg product)-1

- km y-1 travelled by vehicle g pollutant (km)-1

0 1000 2000 3000 4000 5000

Coal

Lignite

Petrol

HSDO

LDO

FO

Natural gas

LPG

Kerosene

Naphtha

Energy consumption (PJ)

Domestic TransportUtilities Iron and SteelCement FertilisersOther Inds.

Domestic3%Other Inds.

22%

Iron and Steel10%

Fertilisers8%

Cement3%

Transport17%

Utilities37%

FUEL CONSUMPTION - INDIAFUEL CONSUMPTION - INDIA

Fossil: 9,411 PJ (1 PJ = 1015 J) Biofuel/biomass: 8,213 PJ

0 50 100 150 200 250 300

Biofuels / biomass (Gg)

Fuelwood Dung-cake Crop wasteCrop waste open burning Forest fire

0 50 100 150 200 250 300

278

63

37

168

39

6%11%

47%

29%

7%

IND/TRA: SECTORAL DETAILIND/TRA: SECTORAL DETAIL

OthersLarge point sourcesUtilities (power)Iron and steelCementFertilisers & Refineries

Plant fuel use Transport

RoadRail

ResidentialRuralUrban

Industry

BC emissions @ 0.250 x 0.250 resolution

Plant Tech./ APCD

EF

Appropriate proxy

EF EF EF

BC emissions

Petroleum fuels Gaseous fuelsCoal/Lignite

Fuel Use Data/ State Level

IND/TRA: FUEL & TECHNOLOGY ANALYSISIND/TRA: FUEL & TECHNOLOGY ANALYSIS

PC-Wet bottom

15%

Not reported

15%Others*1%

PC-Dry/wet bottom

15%

PC-Dry bottom

54%ESP73%

MC3%

ESP+MC+Scrubb

er3%Scrubber

0%

ESP+MC18%

Not reported

2%

ESP+Scrubber

1%

Boiler Type Installed APCD

e.g. ELECTRIC UTILITIESe.g. ELECTRIC UTILITIES

0.00E+00 5.00E+05 1.00E+06 1.50E+06 2.00E+06 2.50E+06

HyderabadVishakapat

PatnaAhmedaba

SuratVadodaraBangalore

CochinMumbaiNagpur

PuneBhopalIndore

LudhianaJaipur

ChennaiCoimbatore

MadhuraiKanpurLucnow

VaranasiCalcutta

Delhi

Petrol Vehicles

12%

83%

4%1%

2Stroke-2 Wheeler-Passenger

2Stroke-3 Wheeler- Passenger

2Stroke-3 Wheeler- Goods

4Stroke-4 Wheeler-Passenger

Diesel Vehicles

3%6%9% 11%

34%7%

30%

LDV(Taxis)

LDV(Jeeps)

LDV(Others)

HDV(Tractors)

HDV(Buses)

HDV(Trucks)

HDV(Trailers)

IND/TRA: FUEL & TECHNOLOGY ANALYSISIND/TRA: FUEL & TECHNOLOGY ANALYSIS

e.g. TRANSPORT Vehicle count in 22 major cities

IND/TRA: EMISSION FACTORSIND/TRA: EMISSION FACTORSSize Specific PM Emission Factors for Coal Combustion in Dry Bottom Boilers

* ESP: Electrostatic PrecipitatorsA= coal ash weight percent, as fired. For example, if coal ash weight is 40%, then A=40. Source: U.S. EPA AP-42 Compilation of emissions factors for stationary sources

Cumulative Mass % < Stated Size Cumulative Emission Factor g kg-1

ParticleSize (µm)

Un-controlled

MultipleCyclone

ESP Un-controlled MultipleCyclone

ESP

1510 6 2.5 1.25 1.00 0.625

3223176221

5429143111

79675029171412

1.60A1.15A0.85A0.30A0.10A0.10A0.05A

0.54A0.29A0.14A0.03A0.01A0.01A0.01A

0.032A0.027A0.012A0.012A0.005A0.005A0.005A

BC/PM ratios for Coal Boilers: 2.2-6.4% (Henry and Knapp, 1980; Shibaoka, 1986; Veranth, 2000)

Road Transport: PM, BC/PM, OC/PM Literature reported measurements for diesel (LDV,

HDV) and petroleum (leaded, unleaded) vehicles

Rail Transport: Values reported by US EPA [1998].

Fuel/vehicle type PM avg. (g kg -1)

BC avg. (% of PM)

OC avg. (% of PM)

Diesel/ HDV a 4.62 42 36

Diesel/ LDV b 3.13 67 24

Leaded petrol 0.84 6 47

Un-leaded petrol/WCC c 0.36 23 73

aLDV: heavy duty vehicles; b LDV: light duty vehicles; c WCC: without catalytic converters

IND/TRA: EMISSION FACTORSIND/TRA: EMISSION FACTORS

IND/TRA:SPATIAL &IND/TRA:SPATIAL &SEASONAL DISTRIBUTIONSSEASONAL DISTRIBUTIONS

Assumed no seasonal variability in industrial and transport emissions.

Proxies from state to district and down to 25 km x 25 km resolution• Point sources: - district power generation, cement /

steel / industrial production.• Transport: road - vehicle population in 22 cities.

- balance district urban population. rail - district geographical area.

IND/TRA: BLACK CARBON EMISSIONSIND/TRA: BLACK CARBON EMISSIONS

BC: 90 Gg y-1

5 25 60 120 250 500 kg km-2 y-1

Bricks24%

Utilities15%

Transport59%

Other1%

•Activity data statistics : 20-40%•Emission factors :

- applicability of non-region-specific emfacs? - validation needed through measurements.- factor of 4-5 (300-400%).

•Needs : Transport > modelled emission factors.> on-road emissions measurement (mixed fleet, urban, interstate).> possible fuel-adulteration effects on emissions.Brick kilns> emissions from representative kiln types.

IND/TRA: UNCERTAINTIESIND/TRA: UNCERTAINTIES

RESIDENTIALRESIDENTIAL

• Energy surveys : high uncertainty and low representative-ness for biofuels (kg capita-

1 day-1).• User population : not estimated. • Unquantifiable uncertainties.• Highly uncertain emission factors for

biofuels.

Statewise cooking energy consumption from 6 fuel types (PJy-1)

Specific cooking energy (MJ kg-1 of food cooked)

Statewise cooking fuel use for 6 fuel types (Tgy-1)Statewise cooking fuel use for 6 fuel types (Tgy-1)

Statewise average food items consumed (kgc-1m-1) (NSS, 2001)

Cooking device

efficiencies (%)

Calorific value (MJ kg-1)

Statewise rural and urban fraction of fuel user population for 6 fuel types (Fuelwood,

dung-cake, crop waste, coal, LPG and kerosene) (NFHS, 2000)

Statewise population using 6 fuel types

Statewise rural, urban population (Census, 2001)

Statewise end use energy (PJy-1) in four cooking process and 6 fuel types

Statewise food items consumption in 4 cooking process (boiling, skillet-baking,

leavened-baking, meat cooking)

RES: SECTORAL/TECHNOLOGY DETAILRES: SECTORAL/TECHNOLOGY DETAIL

RES: BIOFUEL ACTIVITY DATA

RES: EMISSION FACTOR MEASUREMENTRES: EMISSION FACTOR MEASUREMENTStove fuel system used Traditional single pot mud stove

5-wood species, dung-cake and 10-crop waste types

High and low power phases

Dilution samplerOptimized for aerosol stabilization

Mass of fuel, duct velocity, temperatures in combustion zone, duct and plenum recorded each minute

Pollutant measurement

PM2.5: Multi-stream cyclone sampler

OC-BC: Thermal optical transmittance (S. California Particle Centre and Supersite)SO2, NO2, ions, trace elements and absorption

Dilution sampler Burn cycle

 

 Multi-stream aerosol sampler

AIHL Cyclone

Equilibration cylinder

Inlet for air

Filter holders

Cyclone outlet pipe

Connection to Pump

Critical Orifices for flow control

OC-EC MEASUREMENTOC-EC MEASUREMENT

Front QFF for sample OC

Front Teflon filter

Back-up QFF for collection of artifact OC

PM2.5 sampler

inlet

Typical thermogram OC artifact measurement

0.0 0.2 0.4 0.6 0.8 1.0

BC emission factors (gkg-1)

Wood-LBR

Wood-HBRWoody crop stalks

Fibrous crop stalksRice straw

Dried cattle manureKerosene

LPG

Bio

fue

l ty

pe

RES: EMISSION FACTORS FROM COOKINGRES: EMISSION FACTORS FROM COOKING

RES: BLACK CARBON EMISSIONSRES: BLACK CARBON EMISSIONS

Wood Crop waste

81%

17%2% 0.34%

Dried cattle manure

Kerosene LPG

Black carbon: 174 (Ggy-1)

(Ggy-1)

Fuelwood: 281x103 (Ggy-1)

(Ggy-1) 0.2 0.3 0.5 0.6 0.7 0.8(Gg y-1)

BC (old estimate): 270

(Ggy-1)

•Activity data statistics : - cooking (various fuel categories) ~45-85%

•Emission factors :- cooking (traditional stoves / various fuel categories) ~20-100%

•Needs : - water heating / space heating ~unknown- emission factors for mixed fuel use ”- emissions from improved cooking technologies ”- outdoor penetration of residential emissions ”

RES: UNCERTAINTIESRES: UNCERTAINTIES

OPEN BURNINGOPEN BURNING

MODIS vegetation map Forest / grassland / shrubland:E = A . AFL . CE . EF•uncertain A (area under different land types).•uncertain AFL, CE.•seasonal / interannual variability.•shifting cultivation practices (Jhum)

Agricultural waste burning in croplands:E = CP . RPR. F. DM. CE. EF• uncertain F, CE.•systematic spatial/temporal variation.

OB: INTERANNUAL-SEASONAL VARIABILITYOB: INTERANNUAL-SEASONAL VARIABILITY

12month

1monthmonth) jj, (ii,

MODIS_ffr

month) jj, (ii,MODIS_ffr x

j) (i,EM

month) j, (i,

EM

ij

ii

jj

Integrating bottom-up and top-down approaches:• biomass burnt from bottom up methods.• temporal / spatial distribution from active fires and vegetation data.• high resolution, co-location.

Annual Emissions per grid (bottom-up estimate)

OB: AG WASTE MASS BALANCEOB: AG WASTE MASS BALANCE

Fodder kg/animal/day

% Crop waste in fodder

Statewise crop production(FAI 1999-2000)

Statewise cattle and livestock population (BAHS, 1999)

Statewise major crop waste

generated (Tgy-1)

Statewise annual crop waste use as fodder

(Tgy-1)

Statewise unutilized crop waste (TgyStatewise unutilized crop waste (Tgy-1-1) = generation - uses) = generation - usesStatewise unutilized crop waste (TgyStatewise unutilized crop waste (Tgy-1-1) = generation - uses) = generation - uses

National % of CW use: industrial fuel

Statewise major crop waste (cereals, pulses, oilseed and fibre crop) used as fodder, domestic fuel, thatching and industrial fuel (Tgy-1)

Statewise crop waste use for cooking (Tgy-1)

National use of rice straw for thatching (% of generation)

Crop waste /grain ratio

OB: UNUTILIZED AG WASTEOB: UNUTILIZED AG WASTE

0 5 10 15 20 25 30 35 40

Unutilized crop waste (Tgy-1)

Total unutilized crop waste 116 (60-217) Tg y-1

0

5

10

15

20

25

J F M A M J J A S O N DMonth

% o

f em

issi

ons/

biom

ass

0

5

10

15

20

25

% o

f em

issi

ons/

biom

assPM2.5 - 492

BC - 80

OM - 214

Biomass 81x10^3

Total (Gg/y)

AG WASTE: SEASONAL EMISSIONSAG WASTE: SEASONAL EMISSIONS

704 114 306 116 x 103

AG WASTE: AG WASTE: SPATIAL-TEMPORAL VARIABILITYSPATIAL-TEMPORAL VARIABILITY

MODIS fire map: Oct 10, 2002

INDIA BC EMISSIONS SUMMARY (Gg yINDIA BC EMISSIONS SUMMARY (Gg y-1-1))This work Mayol-

Bracero et al., 2002

Dickerson et al., 2002

Reddy and Venkataraman, 2002; Reddy et

al., 2002

Streets et al., 2003

Bond et al., 2004

Base year 2000 2000-01

2000-2001 1999-00 2000 1996

Total all sources

426 (156-1365)a

(200%)

503 414 600 597[671]

Biofuel: 174 (86-360)(100%)b

399 420 167 420(350%)

330[351]d

OpenBrn: crop wst

Forest

114 (45-295)(160%)

38 (5-305)(700%)

- 29c 147 87(700%)

 87

Fossil fuel:

65 54 90 (20-405)(350%)

97(350%)

233

Amean and range; buncertainty at 95% CI; conly for forest fire; dUpgraded for current base year using rural population as proxy.

Simulations of the INDOEX Simulations of the INDOEX ““intensive field intensive field phasephase”” in the LMDZ-GCM in the LMDZ-GCM

Introduction of multi-component aerosols: sulphate, black carbon, organic matter, fly-ash, dust (<1µm; 1-10µm) and sea-salt (8 size bins).

India emissions at 0.25ox0.25o with ground level and elevated sources (MSR/CV 2002a, b). Asia emissions (Streets et al. 2003).

Seasonal/inter-annual BC emissions from biomass open burning distributed using ATSR fire counts.

Nudged to ECMWF winds from Nov 1998 to March 1999.

Parameterisation for carbonaceous aerosol growth from hydrophobic to hydrophilic state.

Wavelength depended aerosol optical properties at different relative humidity.

Evaluation with measurements:Evaluation with measurements:Surface concentrations at Goa (Surface concentrations at Goa (15.5N, 73.8E)15.5N, 73.8E)

Days of February-March

BC

con

cen

trat

ion

(

g m

-3)

Measurements: Chazette, 2003

Ron Brown cruise track

Evaluation with measurements:Evaluation with measurements:Surface concentrations over ocean (Ron Brown)Surface concentrations over ocean (Ron Brown)

Sensitivity to Asian BC emissions: Sensitivity to Asian BC emissions: Surface concentrations at KCO Surface concentrations at KCO 4.9N, 73.5E

(Chowdhury et al., 2001)

Sensitivity to Asian BC emissions: AODSensitivity to Asian BC emissions: AOD

1.1

1.08

1.05

1.04

1.02

1.01

1.8

1.5

1.4

1.3

1.2

1.1

Sensitivity to Asian BC emissions: SSASensitivity to Asian BC emissions: SSA

15.5N, 73.8E 4.9N, 73.5E

TOA

SUR

ATM

CONTROL 2 x BC POLDER

Sensitivity to Asian BC emissions: Radiative forcingSensitivity to Asian BC emissions: Radiative forcing

SUMMARY / NEEDSSUMMARY / NEEDS

• Uncertainties in S. Asian BC emissions can be constrained within factor of 2 (or 100%). - emission factors ~ vehicle classes (on-road emissions), industrial plants, brick kilns, etc.- open burning amounts and seasonal / interannual variability. - missing sources ~ agro-industries e.g. spice / tea drying, small-scale industries - restaurants/ confectioners, glass and bangle making, crematoriums.

• Present model estimates show systematic under-prediction. - restricted to winter monsoon – check seasonal / interannual variability. - wind data that drive model.- assimilate satellite derived AOD.

• Ambient measurements.- network of aethalometer measured “BC” being set up.- need ambient measurements.

RELATED PAPERSRELATED PAPERS

• M.S. Reddy and C. Venkataraman (2002). Inventory of Aerosol and Sulphur Dioxide Emissions from India: I – Fossil Fuel Combustion, Atmospheric Environment, 36 (4), 677-697.

• M.S. Reddy and C. Venkataraman (2002). Inventory of Aerosol and Sulphur Dioxide Emissions from India: II – Biomass Combustion, Atmospheric Environment, 36 (4), 699-712.

• G. Habib, C. Venkataraman, M. Shrivastava, R. Banerji, J. Stehr and R. Dickerson (2004). New methodology for estimating biofuel consumption for cooking: Atmospheric emissions of black carbon and sulfur dioxide from India, Global Biogeochemical Cycles, 18, GB3007, doi:10.1029/2003GB002157.

• M.S. Reddy, O. Boucher, C. Venkataraman, S. Verma, N. Bellouin and M. Pham (2004). GCM estimates of aerosol transport and radiative forcing during INDOEX, Journal of Geophysical Research, 109, D16205, doi:10.1029/2004JD004557.

• C. Venkataraman, G. Habib, A. Eiguren-Fernandez, A.H. Miguel and S.K. Friedlander (2004). Carbonaceous aerosol emissions from residential biofuel combustion in S. Asia and climate implications, submitted.

• G. Habib, C. Venkataraman, T.C. Bond and J.J. Schauer, A. Eiguren-Fernandez, A.H. Miguel, S.K. Friedlander (2004). Primary particle emissions biofuel combustion: Chemical composition, size distribution and optical properties, in preparation.