Indoor aerosol modelings Joonas Koivisto, National Research centre for the Working Environment; jok@nrcwe.dk NanoSafety Cluster WG4, 11.4.2017
Outline Exposure modelings 1. Dispersion models 2. Source characterization 3. Engineered emission controls and personal protective equipment Modeling Examples • Laser printer emission rates • Indoor air cleaner cleaning efficiency • Exposure modelings in a paint factory • Exposure modelings during nanodiamond handling
1. Dispersion models
Single compartment model Compartment Emission concentration source 𝑒𝐷(𝑢) = 𝜇𝑄𝐷 0 (𝑢) + 𝑇 𝐷 (𝑢) − 𝜇 + 𝛿 + 𝜕 𝐷(𝑢) 𝑒𝑢 𝑊 Background Particle removal particles from by ventilation, deposition, ventilation air and coagulation Assumptions: • Particles are fully mixed at all Terms and parameters: times m -3 C ( t ) Indoor aerosol concentration C o ( t ) m -3 Outdoor aerosol concentration λ s -1 Ventilation rate P - Particle penetration factor Usually C o = 0, and γ « λ and ω « λ : S ( t ) # s -1 Indoor particle source 𝑒𝐷(𝑢) = 𝑇 𝐷 (𝑢) m 3 Compartment volume V − 𝜇𝐷(𝑢) γ s -1 Particle deposition rate 𝑒𝑢 𝑊 ω s -1 Particle coagulation rate Q m 3 s -1 Ventilation flow
Two compartment model (incl. Deposition and coagulation) A Near Field/Far Field (NF/FF) model: 𝑒𝐷 𝐺𝐺 (𝑢) 𝑊 𝐺𝐺 = 𝛾𝐷 𝐺𝐺 𝑢 − 𝛾 + 𝑅 𝐷 𝐺𝐺 𝑢 − 𝜇 𝑒,𝑂𝐺 𝐷 𝐺𝐺 𝑢 + 𝐾 𝑑𝑝𝑏,𝐺𝐺 𝑒𝑢 𝑒𝐷 𝑂𝐺 (𝑢) 𝑊 𝑂𝐺 = 𝐹𝑆 𝑗 (𝑢) + 𝛾𝐷 𝐺𝐺 𝑢 − 𝛾𝐷 𝑂𝐺 𝑢 − 𝜇 𝑒,𝑂𝐺 𝐷 𝑂𝐺 𝑢 + 𝐾 𝑑𝑝𝑏,𝑂𝐺 𝑒𝑢 Assumptions: • Particles are fully mixed at all times in the NF and FF • No significant cross-draft • Limited air exchange between NF and FF volumes (3 m 3 min -1 < β < 30 m 3 min -1 ; Cherrie, 1999) • Particle losses via ventilation and deposition Zhang et al ., (2009) describes the NF/FF model in detail.
2. Source characterization
Particle emission source (# s -1 ) characterization Powder dustiness measurements Chamber measurements Di m , [mg kg -1 ] 𝑒𝐷(𝑢) = 0 ; 1. In steady-state ( 𝑒𝑢 C o = 0, and γ « λ and ω « λ ): 𝑇 𝑑 = 𝐸𝐽 𝑑 ∙ 𝐼 ∙ 𝑒𝑁 𝑒𝑢 𝑒𝐷(𝑢) 𝑇 𝐷 (𝑢) = − 𝜇𝐷 𝑢 = 0 𝑒𝑢 𝑊 𝑇 𝐷 (𝑢) = 𝜇𝑊𝐷(𝑢) Terms and parameters: • 2. Non-steady-state: The # kg -1 Dustiness index DI c • H - The handling energy factor mathematical difference 𝑒𝑁 • between measured and kg s -1 𝑒𝑢 The powder use rate simulated concentration (Hussein et al . 2006) Schneider and Jensen (2008) 3. Convolution theorem (Schripp et al. 2008)
Particle emission source library • Process and material specific emission source library : • Quantitative material releases from articles containing manufactured nanomaterials (Koivisto et al. 2017) • Dustiness library 100000 2 of 66 "NM" above 25000 mg/Kg • Collation of data as Need for Very High? part of the EU FP7 SUN 10000 DI(Inh) [mg/Kg] High project Moderate 1000 Low Very Low 100 0 10 20 30 40 50 60 70 Ranked Sample Order Dustiness indices from small rotating drum method
Library Format Statistical analysis grouping and read-across rules Study information Traceability, transparency Process information: e.g. sanding and details, Usability, link to reality, grouping Materials: matrix, NM con- centration ,… Fragments information: Tier 1 information for EHS assessment density, description of (fraction of bioavailable NM, shape, released fragments elemental compositions ,…) Material Removal Rate or Quantitative release rate for use rate, Size-resolved dispersion models. MRR is used to release rates: S m (µg/sec), estimate surface contamination and S SA (µm2/sec), S N (1/sec) environmental release. Dynamic tranformation Tier 2 information for EHS assessment Hazard library Dose-response
Study Gomez et al. (2014) Reference/contact information N Process Sanding Metabo FSR200 hand-held sanding machine where the dust collector cartridge was removed (user modified) and equipped with a sanding paper with a grit size of 120. Sn include sanding Process details machine electric engine emissions. Matrix Acryl binder Matrix details Paint matrix: water, propylene glycol, Uradil AZ XP 601Z44, and others NM TiO2 NM vendor N/A NM product name Nano Amor and NaBond TiO2 rutile NM concentration (wt.%) 36 NM state Embedded MN PP size (nm) NaBond 80 nm and Nano Amor 50 nm Other information for materials and NaBond specific surface area of 28.2 m2/g and Nano Amor specific surface area of 139.1 methods m2/g Process rate (mg/sec) N/A Sanding of the paints produced airborne particles mainly > 1µm in size. Free TiO 2 particles Released fragments were not observed. Fragments density (g/cm3) 2.1628 Notes Density calculated as 36 % (TiO2 rutile) of 4.23 g/cm3 and 64 % density of 1 g/cm3 Mass emission Sm (ug/sec) 6800 GMDm (um) 3.6 GSDm 2 Sanding machine electric engine emissions are neglected from the emission rate. Sm is for Notes particles <10 um in diameter. Surface area emission Ssa (ug/sec) NaN GMDsa (um) NaN GSDsa NaN Notes N Number emission Sn (1/sec) 2.50E+09 GMD (um) 0.56 GSD 1.9 Notes Sanding machine electric engine emissions are excluded Dp1 (um) 9.89E-03 Dp2 2.13E-02 Dp3 3.95E-02 Dp4 7.29E-02
3. Emission controls and PPE
Emission and exposure controls • Engineered Control equipment ( EC ) and air circulation trough filter: 𝑒𝐷(𝑢) 𝑝 𝐷 𝑝 𝑢 + 𝑇 𝐷 𝑢 = 𝜇𝑄 ∙ 𝑄 𝐹𝐷 − 𝜇 + 𝛿 + 𝜕 𝐷 𝑢 + 𝜀(𝑄 𝑔 − 1)𝐷 𝑢 𝑒𝑢 𝑊 where 0< P <1 (0 = perfect filtration, 1 = no filtration) • Personal protective equipment ( PPE ): 𝐷 𝑗𝑜ℎ (𝑢) = 𝐷(𝑢) ∙ 𝑄 𝑄𝑄𝐹 Air cleaner study (Mølgaard et al . 2014) Respirator study (Koivisto et al . 2015)
Emission Control Efficiency Library for nanomaterials (ECEL) by TNO Study Control measure Activity/process NM PSD (device/range) N AvEffect SD Median 5th Min Max % perc all data (5−560 nm) Containment Loading in compounding process, Nanoalumina, nanoclay, 15 96,1 12,8 99,9919 80,9 50,3 99,9992 1,2 harvesting graphene platelets 5−560 nm Containment (high level) 8 99,890 0,26 99,48 99,3 99,9992 Containment (medium level) 200nm 7 91,7 18,4 63,4 50,3 99,9989 3,4 Fixed capturing hood Reactor clean-out, welding Mostly metals/oxides all data (<20µm) 26 94,1 7,1 95,8 79,5 74,6 100,000 45-216 nm 1 87,5 300-500 nm 3 97,85 3,7 <1µm 11 95,1 6,7 85,0 85,0 100,00 1µm-10µm 12 94,3 6,7 83,9 74,6 100,00 >10µm 3 90,3 11,2 Movable capturing hood Synthesis/handling; welding Mostly metals/oxides all data (<10µm) 5 88,2 8,5 90,4 77,5 75,9 97,3 3,4 Welding fume 45-216 nm 1 97,3 5.6−560 nm 1 Enclosing hood (process Reactor, tank cleaning Graphene platelets 1 74,6 blower) 5,6 Fume cupboard (without glove Reactor set-up & handling; TiO 2 , CNT all data (<1µm) 3 81,1 20,8 91,0 60,5 57,1 95,0 bags) sanding TiO 2 30-50 nm 1 95,0 7 Wetting at point of release Cutting band saw Base carbon all data (0.5-20µm) 3 87,9 17,3 96,4 70,8 67,9 99,3 5.6 to 560 nm 1 67,9 10 Unidirectional room airflow Sanding (hand) CNT 20-10000nm 2 91,7 7,5 89,0 88,7 94,8 systems (spray rooms) 20-300nm 1 88,7 8 Unidirectional room airflow Solution spraying CNT 14-630 nm 1 84,0 (52*1473nm → 56 - systems (not specified) 1760nm) 12 Glove boxes Handling powders, polishing, CNT 20-1000nm 2 97,2 scratching, drying … … … … … … … … … … … … See Fransman et al. ( 2008 ) ECEL for dusts
Multi-compartment Indoor aerosol modeling X X 𝜺 , P f FF P o C o NF; S c 𝑒𝐷 𝐺𝐺 (𝑢) 𝑊 𝐺𝐺 = 𝛾𝐷 𝐺𝐺 𝑢 − 𝛾 + 𝑅 𝐷 𝐺𝐺 𝑢 − 𝜇 𝑒,𝑂𝐺 𝐷 𝐺𝐺 𝑢 + 𝐾 𝑑𝑝𝑏,𝐺𝐺 + 𝜀(𝑄 𝑔 − 1)𝐷 𝑢 𝑒𝑢 𝑒𝐷 𝑂𝐺 (𝑢) 𝑊 𝑂𝐺 = 𝐹𝑆 𝑗 (𝑢) + 𝛾𝐷 𝐺𝐺 𝑢 − 𝛾𝐷 𝑂𝐺 𝑢 − 𝜇 𝑒,𝑂𝐺 𝐷 𝑂𝐺 𝑢 + 𝐾 𝑑𝑝𝑏,𝑂𝐺 𝑒𝑢
Modeling examples: A. Laser printer emission rates B. Indoor air cleaner cleaning efficiency C. Exposure modelings in a paint factory D. Exposure modelings during nanodiamond handling
A) Particle emissions from laser printers (Koivisto et al. 2010) 1. Put a source into a well controlled environment
Concentrations and indoor aerosol modelings 2. Measure concentrations 3. Define deposition rates when 𝑇 𝑂 𝑗 𝑢 = 0, 𝑂 𝑗 𝑢 ≪ 10 4 cm -3 𝑒𝑂 𝑗 (𝑢) ≈ − 𝜇 + 𝛿 𝑗 𝑂 𝑗 (𝑢) ( 𝜇 is known) 𝑒𝑢 4. Solve emission rates Lai and Nazaroff (2000) 𝑻 𝑶 𝒋 (𝒖) 𝑒𝑂 𝑗 (𝑢) = − 𝜇 + 𝛿 𝑗 + 𝜕 𝑗 𝑂 𝑗 (𝑢) 𝑒𝑢 𝑊 Measured Korhonen Solve et al., (2004) Particle concentrations during pre-operation and printing phases
Time and size resolved emission rates from laser printers
Modeled particle concentrations in a office 𝑇 𝑂 𝑗 (𝑢) 𝑒𝑶 𝒋 (𝒖) = 𝜇𝑄𝐷 0 𝑢 + − 𝜇 + 𝛿 𝑗 + 𝜕 𝑗 𝑶 𝒋 (𝒖) 𝑒𝑢 𝑊
B) Air clearer test (Mølgaard et al. 2014) Blue parameters are known
Recommend
More recommend
