Development of an improved product based approach for the calculation of

NMVOC emissions from solvent use in Germany and uncertainty analysis

Jochen ThelokeUniversity of Stuttgart

San Diego/JT/30/04/03

Contents

1. Introduction

2. Definitions

3. Method

4. Results

5. Uncertainties

6. Conclusion

San Diego/JT/30/04/03

Solvent use was the largest NMVOC emissionsource group in Germany in 1998

Sum: 2 436 kt

combustion sources from industry and

power plants 1%

small combustion plants/households

3%

traffic27%

solvent use36%

production processes8%

biogenic 27%

San Diego/JT/30/04/03

VOC - Volatile Organic Compound

“Volatile organic compound shall mean any organic compound having at 293,15 K a vapour pressure of 0,01 kPa or more, or having a corresponding volatility under the particular condition of use.”

San Diego/JT/30/04/03

Solvent Use• Application of solvents

• Application of solvent containing products

• Production of solvent containing products

• Propellants (Propane/Butane)

• Extraction of fat, edible and non edible oil (n-hexane)

• Cooling agents (HFHC`s, HFCHC`s, alcohols)

• Softening agents (Phtalates)

• Concrete additives

• Application of thinners

San Diego/JT/30/04/03

Method for the calculation of emissions from solvent use

importsolvent x

exportsolvent x

productionsolvent x

inlandconsumption

solvent x

manufacturechem.

product y

importchem.

product y

exportchem.

product y

inland consumptionchem.

product y

emissionreduction techniques

other pathways

Solvent Balance Manufacture Consumption

emissions

San Diego/JT/30/04/03

870 Gg NMVOC emitted from solvent use in Germany 2000

thinner without specific

application26%

Degreasing, dry cleaning and electronics

4%

Products manufacturing or

processing5%

Preservation of wood

4%

Paint application25%

Domestic solvent use (other than

paint application)13%

other14%

printing industry9%

San Diego/JT/30/04/03

Paint application (sum 223 Gg)

San Diego/JT/30/04/03

Manufacture of automobiles

6%

Synthetics4%

Plate emballages

3%

other paint applications

16%

Wood16%

constructions and buildings-professional

14%Car repairing

10%

Corrosion-inhibiting

primer8%

Mechanical engineering

8%

Do It Yourself 6%

white goods3%

Boat building2%

Prefabricated units

2% road marking varnishes

2%

The domestic consumption of thinners increasedcontinously in the last years

100

150

200

250

300

1992 1994 1996 1998 2000 2002Year

Th

inn

er C

on

sum

pti

on

[G

g/y

]

San Diego/JT/30/04/03

NMVOC emissions from domestic solvent use (sum 105 Gg)

San Diego/JT/30/04/03

coolant and fire extinguisher

2%aftershave, hygienic water and perfume

19%

hairspray19%

personal hygiene products

22%

antifreeze for cars24%

detergents14%

*) pure Ethanol

NMVOC emissions from domestic solvent use

San Diego/JT/30/04/03

Solvent consumption

Emission factor

NMVOC emissions

Mg/y % Mg/y

Antifreeze for cars 51 100 50 25 550

Detergent for dishwasher 9 900 3 300

Detergent for car washing 10 900 3 330

Detergent for washing-machine 12 660 3 380

Alcohol consumption 750 000* 1 7 500

Hairsprays 19 350 95 18 400

Personal hygiene products 24 400 95 23 150

Toilet water 6 100 95 5 800

Aftershave 5 600 95 5 300

Perfume 2 500 95 2 400

Deodorants 1 180 95 1 120

NMVOC emissions from printing processes (sum 80 Gg)

Rotogravure printing14,8%

Offset printing39,4%Offset-Heatset-

printing21,9%

Letterpress printing

0,6%

Package printing20,1%

Screen printing3,1%

San Diego/JT/30/04/03

Speciation of the NMVOC emissions from solvent use to substance classes

glycol derivates8,5%

alcohols26,3%

aromatics21,7%aliphatics

20,8%

ketones6,2%

halogenated hydrocarbons

1,5%other2,5%

terpenes1,2%

aldehydes0,02%

organic acids0,1%

ethers1,5%

esters9,8%

San Diego/JT/30/04/03

Verification - Plausibility check

Substance classes

Domestic consumption [t]

(solvent industry)

Domestic consumption

[t](own

estimation)

Deviation[%]

Aliphatics 300 000 280 000 - 6,7Aromatics 250 000 286 000 +14,4Terpenes 10 000Halogenic hydrocarbons 35 000 35 000Alcohols 360 000 316 000 -12,2-Ethanol 50 000 85 000-Isopropanol 250 000 144 000-n-Propanol 40 000 200-n/i-Butanol 10 000 56 000Glycolderivates 75 000 78 000 +4,0Esters 75 000 76 000 +1,3Ketones 60 000 61 000 +1,7Ethers 30 000 29 000 -3,3Aldehydes 200Organic Acids 800Softening agents 225 000Other VOC`s 65 000Sum (Sum of bold faced substance classes)

1 185 000 1 161 000 -2,0

San Diego/JT/30/04/03

Assessment and calculation of uncertainties of NMVOC emissions from solvent use - Methods

Uncertainty analysis

ØQualitative assessmentØ Semiquantitative methods

Quantitative methods

Ø Error propagationØMonte-Carlo simulation

Verification

Ø Plausibility checksØ Evaluation experiments

San Diego/JT/30/04/03

Semi-quantitative methods

San Diego/JT/30/04/03

Uncertainty bandwith

Source groups

Yearly

emissions

Spatial

resolution

Tem

poral resolution

VO

C

speciation

Paint application 2-3 3 3-4 3 Domestic solvent use 3 2 5 2 Printing processes 2 2 3 2 Synthetics processing 4 4 4 3 Metal degreasing 4 4 4 4 Other source groups 3 3 3-4 4 Solvent use Complete assessment

3 3 4 3

Monte-Carlo simulations for emissions from paintapplication - Preliminary results

San Diego/JT/30/04/03

0

1

2

3

4

180 200 220 240 260 280 300Emissions in Gg per year

Val

ues

in

10^

-5

Mean value: 231 Gg

Calculated : 222 Gg (without uncertainty

analysis)209 255

90 % of the calculated values between 209 Gg

and 255 Gg

Germany 2000

Verification - Results of city experiments

0,1 1 10

2,3-Dimethylpentane

Ethane

Ethine

n-Heptane

3-Methylhexane

2-Methylhexane

Ethene

3-Methylpentane

Benzene

2-Methylpentane

Propane

Propene

Ethylbenzene

n-Octane

i-Pentane

EVA

ESQUIF

(HCi/CO)modelled / (HCi/CO)measured

0,5 2

San Diego/JT/30/04/03

0,1 1 10 100

m/p-Xylene

i-Butane

n-Pentane

Toluene

1,2,4-TMB

o-Xylene

n-Butane

n-Hexane

1,3,5-TMB

Propylbenzene

n-Nonane

n-Decane

(HCi/CO)modelled / (HCi/CO)measured

EVA

ESQUIF

0,5 2

Verification - Results of city experiments

San Diego/JT/30/04/03

Conclusion• Emissions from solvent use was the largest NMVOC emission source

group in Germany in 1998

• The developed improved product based approach for the calculation of NMVOC emissions from solvent use in Germany was further developed and applied for 2000

• The method can in principle also be used for other OECD countries

• Nearly 75% of emissions from solvent use are caused by paint application, application of thinners, domestic solvent use and printing processes

• Nearly 50% of emitted solvents are oxygen containing NMVOC`s

• A plausibility check between a solvent based approach and a product based approach yielded a good agreement

• The results of city experiments in Augsburg and Paris indicates thatfurther experiments especially with measurements of OVOC are required

• Monte-Carlo simulation appears to be a promising approach for thequantification of uncertainties from solvent use emissions

San Diego/JT/30/04/03